diff --git a/Book/_build/.doctrees/05.02-ttest.doctree b/Book/_build/.doctrees/05.02-ttest.doctree index 084987c0..3562f774 100644 Binary files a/Book/_build/.doctrees/05.02-ttest.doctree and b/Book/_build/.doctrees/05.02-ttest.doctree differ diff --git a/Book/_build/.doctrees/environment.pickle b/Book/_build/.doctrees/environment.pickle index fc818816..4a2dc048 100644 Binary files a/Book/_build/.doctrees/environment.pickle and b/Book/_build/.doctrees/environment.pickle differ diff --git a/Book/_build/.doctrees/glue_cache.json b/Book/_build/.doctrees/glue_cache.json index db3f51fc..b24044e0 100644 --- a/Book/_build/.doctrees/glue_cache.json +++ b/Book/_build/.doctrees/glue_cache.json @@ -1,46 +1,42 @@ { "01.01-intro": { - "berkely-fig": { + "berkley-table": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "berkely-fig", + "name": "berkley-table", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": " Department Male Applicants Male Percent Admitted Female Applicants \\\n0 A 825 62% 108 \n1 B 560 63% 25 \n2 C 325 37% 593 \n3 D 417 33% 375 \n4 E 191 28% 393 \n5 F 272 6% 341 \n\n Female Percent Admitted \n0 82% \n1 68% \n2 34% \n3 35% \n4 24% \n5 7% ", + "application/papermill.record/text/html": "
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
DepartmentMale ApplicantsMale Percent AdmittedFemale ApplicantsFemale Percent Admitted
0A82562%10882%
1B56063%2568%
2C32537%59334%
3D41733%37535%
4E19128%39324%
5F2726%3417%
\n
" } }, - "berkley-table": { + "berkely-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "berkley-table", + "name": "berkely-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": " Department Male Applicants Male Percent Admitted Female Applicants \\\n0 A 825 62% 108 \n1 B 560 63% 25 \n2 C 325 37% 593 \n3 D 417 33% 375 \n4 E 191 28% 393 \n5 F 272 6% 341 \n\n Female Percent Admitted \n0 82% \n1 68% \n2 34% \n3 35% \n4 24% \n5 7% ", - "application/papermill.record/text/html": "
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
DepartmentMale ApplicantsMale Percent AdmittedFemale ApplicantsFemale Percent Admitted
0A82562%10882%
1B56063%2568%
2C32537%59334%
3D41733%37535%
4E19128%39324%
5F2726%3417%
\n
" + "application/papermill.record/text/plain": "" } } }, "04.02-probability": { - "sdnorm_fig2": { + "two-normals_fig2": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "sdnorm_fig2", + "name": "two-normals_fig2", "mime_prefix": "application/papermill.record/" - }, - "filenames": { - "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/04.02-probability_34_0.png" } }, "data": { - "application/papermill.record/text/plain": "
", - "application/papermill.record/image/png": null + "application/papermill.record/text/plain": "" } }, "normal_fig": { @@ -55,44 +51,40 @@ "application/papermill.record/text/plain": "" } }, - "two-normals_fig2": { + "tdist-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "two-normals_fig2", + "name": "tdist-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "Fdist-fig": { + "pants-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "Fdist-fig", + "name": "pants-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "sdnorm_fig": { + "skulls-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "sdnorm_fig", + "name": "skulls-fig", "mime_prefix": "application/papermill.record/" - }, - "filenames": { - "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/04.02-probability_32_0.png" } }, "data": { - "application/papermill.record/text/plain": "
", - "application/papermill.record/image/png": null + "application/papermill.record/text/plain": "" } }, "fig-binomial": { @@ -111,11 +103,11 @@ "application/papermill.record/image/png": null } }, - "tdist-fig": { + "Fdist-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "tdist-fig", + "name": "Fdist-fig", "mime_prefix": "application/papermill.record/" } }, @@ -123,95 +115,102 @@ "application/papermill.record/text/plain": "" } }, - "skulls-fig": { + "sdnorm_fig2": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "skulls-fig", + "name": "sdnorm_fig2", "mime_prefix": "application/papermill.record/" + }, + "filenames": { + "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/04.02-probability_34_0.png" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "
", + "application/papermill.record/image/png": null } }, - "pants-fig": { + "sdnorm_fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "pants-fig", + "name": "sdnorm_fig", "mime_prefix": "application/papermill.record/" + }, + "filenames": { + "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/04.02-probability_32_0.png" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "
", + "application/papermill.record/image/png": null } }, - "frequentist_probability_fig": { + "chi2-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "frequentist_probability_fig", + "name": "chi2-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "chi2-fig": { + "frequentist_probability_fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "chi2-fig", + "name": "frequentist_probability_fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } } }, "03.01-descriptives": { - "afl_fig": { + "sleep_scatter-fig1": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "afl_fig", + "name": "sleep_scatter-fig1", "mime_prefix": "application/papermill.record/" + }, + "filenames": { + "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/03.01-descriptives_122_0.png" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "
", + "application/papermill.record/image/png": null } }, - "sleep_scatter-fig1": { + "kurtosis_fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "sleep_scatter-fig1", + "name": "kurtosis_fig", "mime_prefix": "application/papermill.record/" - }, - "filenames": { - "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/03.01-descriptives_122_0.png" } }, "data": { - "application/papermill.record/text/plain": "
", - "application/papermill.record/image/png": null + "application/papermill.record/text/plain": "" } }, - "corr-interpretation-table": { + "afl_fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "corr-interpretation-table", + "name": "afl_fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": " Correlation Strength Direction\n0 -1.0 to -0.9 Very strong Negative\n1 -0.9 to -0.7 Strong Negative\n2 -0.7 to -0.4 Moderate Negative\n3 -0.4 to -0.2 Weak Negative\n4 -0.2 to 0 Negligible Negative\n5 0 to 0.2 Negligible Positive\n6 0.2 to 0.4 Weak Positive\n7 .4 to 0.7 Moderate Positive\n8 0.7 to 0.9 Strong Positive\n9 0.9 to 1.0 Very strong Positive", - "application/papermill.record/text/html": "
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
CorrelationStrengthDirection
0-1.0 to -0.9Very strongNegative
1-0.9 to -0.7StrongNegative
2-0.7 to -0.4ModerateNegative
3-0.4 to -0.2WeakNegative
4-0.2 to 0NegligibleNegative
50 to 0.2NegligiblePositive
60.2 to 0.4WeakPositive
7.4 to 0.7ModeratePositive
80.7 to 0.9StrongPositive
90.9 to 1.0Very strongPositive
\n
" + "application/papermill.record/text/plain": "" } }, "sleep_scatter-fig2": { @@ -230,66 +229,67 @@ "application/papermill.record/image/png": null } }, - "skew_fig": { + "corr-interpretation-table": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "skew_fig", + "name": "corr-interpretation-table", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": " Correlation Strength Direction\n0 -1.0 to -0.9 Very strong Negative\n1 -0.9 to -0.7 Strong Negative\n2 -0.7 to -0.4 Moderate Negative\n3 -0.4 to -0.2 Weak Negative\n4 -0.2 to 0 Negligible Negative\n5 0 to 0.2 Negligible Positive\n6 0.2 to 0.4 Weak Positive\n7 .4 to 0.7 Moderate Positive\n8 0.7 to 0.9 Strong Positive\n9 0.9 to 1.0 Very strong Positive", + "application/papermill.record/text/html": "
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
CorrelationStrengthDirection
0-1.0 to -0.9Very strongNegative
1-0.9 to -0.7StrongNegative
2-0.7 to -0.4ModerateNegative
3-0.4 to -0.2WeakNegative
4-0.2 to 0NegligibleNegative
50 to 0.2NegligiblePositive
60.2 to 0.4WeakPositive
7.4 to 0.7ModeratePositive
80.7 to 0.9StrongPositive
90.9 to 1.0Very strongPositive
\n
" } }, - "kurtosis_fig": { + "skew_fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "kurtosis_fig", + "name": "skew_fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } } }, "04.04-hypothesis-testing": { - "esp-alternative-fig2": { + "powerfunctionsample-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "esp-alternative-fig2", + "name": "powerfunctionsample-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "esp-alternative-fig": { + "espcritical-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "esp-alternative-fig", + "name": "espcritical-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "estimation-fig": { + "espcritical-onesided-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "estimation-fig", + "name": "espcritical-onesided-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, "powerfunction-fig": { @@ -304,49 +304,49 @@ "application/papermill.record/text/plain": "" } }, - "espcritical-onesided-fig": { + "esp-alternative-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "espcritical-onesided-fig", + "name": "esp-alternative-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "powerfunctionsample-fig": { + "estimation-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "powerfunctionsample-fig", + "name": "estimation-fig", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } }, - "espcritical-fig": { + "esp-alternative-fig2": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "espcritical-fig", + "name": "esp-alternative-fig2", "mime_prefix": "application/papermill.record/" } }, "data": { - "application/papermill.record/text/plain": "" + "application/papermill.record/text/plain": "" } } }, "05.02-ttest": { - "zeppo-fig": { + "ztest-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "zeppo-fig", + "name": "ztest-fig", "mime_prefix": "application/papermill.record/" } }, @@ -354,11 +354,23 @@ "application/papermill.record/text/plain": "" } }, - "ttesthyp_onesample-fig": { + "qqheavy_fig": { + "data": { + "application/papermill.record/text/plain": "" + }, + "metadata": { + "scrapbook": { + "mime_prefix": "application/papermill.record/", + "name": "qqheavy_fig" + } + }, + "output_type": "display_data" + }, + "zeppo-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "ttesthyp_onesample-fig", + "name": "zeppo-fig", "mime_prefix": "application/papermill.record/" } }, @@ -366,11 +378,23 @@ "application/papermill.record/text/plain": "" } }, - "ztest-fig": { + "qq_fig": { + "data": { + "application/papermill.record/text/plain": "" + }, + "metadata": { + "scrapbook": { + "mime_prefix": "application/papermill.record/", + "name": "qq_fig" + } + }, + "output_type": "display_data" + }, + "ttesthyp_onesample-fig": { "output_type": "display_data", "metadata": { "scrapbook": { - "name": "ztest-fig", + "name": "ttesthyp_onesample-fig", "mime_prefix": "application/papermill.record/" } }, diff --git a/Book/_build/html/05.02-ttest.html b/Book/_build/html/05.02-ttest.html index 33ff7b84..a0fcff92 100644 --- a/Book/_build/html/05.02-ttest.html +++ b/Book/_build/html/05.02-ttest.html @@ -1487,7 +1487,7 @@

2.3.2. Introducing the test -
<matplotlib.lines.Line2D at 0x134b4bac0>
+
<matplotlib.lines.Line2D at 0x135505be0>
 

 

 _images/05.02-ttest_61_1.png
@@ -1702,7 +1702,7 @@ 
2.3.7. Positive and negative
-
<matplotlib.lines.Line2D at 0x134bf6f40>
+
<matplotlib.lines.Line2D at 0x1355b60a0>
 

 

 _images/05.02-ttest_76_1.png
@@ -2330,76 +2330,100 @@ 
2.7.4. Cohen’s 
 
2.8. Checking the normality of a sample

-
All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section @ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is actually an average of lots of different things, there’s a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using \(t\)-tests. However, life doesn’t come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there’s a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.^[This is a massive oversimplification.] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test.

+
All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section @ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is actually an average of lots of different things, there’s a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using \(t\)-tests. However, life doesn’t come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there’s a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.13 This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test.

 

 
2.8.1. QQ plots

-
```{r qq1a, fig.cap=”Histogram of normal.data, a normally distributed sample with 100 observations.”, echo=FALSE}
-width <- 5
-height <- 5

-
plotOne <- function( data, title ) {

-
hist( x = data ,
-      xlab = "Value", # x-axis label
-      main = title,                     
-      border="white", 
-      col=ifelse(colour,emphColLight,emphGrey),
-      font.main = 1
-)         
-
-require(psych)
-cat(title,"\n")
-cat( "skew=",skew(data),"\n")
-cat( "kurtosis=",kurtosi(data),"\n")
-print( shapiro.test(data) )
-

-

-
}

-
normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers
-plotOne( normal.data, “Normally Distributed Data” )

-

-```{r qq1b, fig.cap="Normal QQ plot of `normal.data`, a normally distributed sample with 100 observations.", echo=FALSE}
-  plotTwo <- function( data ) {   
-    qqnorm( y = data,
-            pch=19,
-            font.main = 1
-    )        # draw the QQ plot
-  
-  }
+
One way to check whether a sample violates the normality assumption is to draw a “quantile-quantile” plot (QQ plot). This allows you to visually check whether you’re seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot, using probplot from scipy. We can compare this with a histogram of the data as well:

+

+

+
import numpy as np
+import seaborn as sns
+from scipy.stats import probplot
+
+np.random.seed(42)
+normal_data = np.random.normal(size=100)
+
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
+
+qq = probplot(normal_data, dist="norm", plot = ax1)
+hist = sns.histplot(normal_data, axes=ax2)
 
-  plotTwo( normal.data )
+glue("qq_fig", ax, display=False)
 

 

-
The Shapiro-Wilk statistic associated with the data in Figures @ref(fig:qq1a) and @ref(fig:qq1b) is \(W = .99\), indicating that no significant departures from normality were detected (\(p = .73\)). As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the two data sets shown in Figures @ref(fig:qq2a), @ref(fig:qq2b), @ref(fig:qq2c), @ref(fig:qq2d). Figures @ref(fig:qq2a) and @ref(fig:qq2b) show the histogram and a QQ plot for a data set that is highly skewed: the QQ plot curves upwards. Figures @ref(fig:qq2c) and @ref(fig:qq2d) show the same plots for a heavy tailed (i.e., high kurtosis) data set: in this case, the QQ plot flattens in the middle and curves sharply at either end.

-
```{r qq2a, fig.cap=”A histogram of the 100 observations in a skewed.data set”, echo=FALSE}
-skewed.data <- rgamma( n = 100,1,2 )
-plotOne( skewed.data, “Skewed Data” )

-

-```{r qq2b, fig.cap="A normal QQ plot of the 100 observations in a `skewed.data` set", echo=FALSE}
-  plotTwo( skewed.data )
+

+

+_images/05.02-ttest_126_1.png
+

+

+

+

+
<AxesSubplot:xlabel='grades', ylabel='Count'>
 

 

-
The skewness of the data in Figures @ref(fig:qq2a) and @ref(fig:qq2b) is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is \(W=.80\), reflecting a significant departure from normality (\(p<.001\)).

-
```{r qq2c, fig.cap=”A histogram of the 100 observations in a heavy tailed` data set, again consisting of 100 observations. “, echo=FALSE}
-heavy.tailed.data <- c(rnorm( n = 80 ) , rnorm( 20, sd=5))
-plotOne( heavy.tailed.data, “Heavy-Tailed Data” )

-

-```{r qq2d, fig.cap="A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. ", echo=FALSE}
-  plotTwo( heavy.tailed.data )
+

+
Fig. 2.24 QQ plot (left) and histogram (right) of normal_data, a normally distributed sample with 100 observations. The Shapiro-Wilk statistic associated with these data is \(W = .99\), indicating that no significant departures from normality were detected (\(p = .73\)).

+

+
And the results are shown in {numref}(fig-qq), above.

+
As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the data shown in {numref}(qqskew) and {numref}(qqheavy), which show the histogram and a QQ plot for a data sets that are highly skewed and have a heavy tail (i.e., high kurtosis), respectively.

+

+

+
df = pd.read_csv("https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/skewed_data.csv")
+
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
+
+qq = probplot(df['data'], dist="norm", plot = ax1)
+hist = sns.histplot(df['data'], axes=ax2)
+
+glue("qqskew_fig", ax, display=False)
 

 

-
Figures @ref(fig:qq2c) and @ref(fig:qq2d) shows the same plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is \(W = .93\), again reflecting significant non-normality (\(p < .001\)).

-
One way to check whether a sample violates the normality assumption is to draw a “quantile-quantile” plot (QQ plot). This allows you to visually check whether you’re seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot using the R function qqnorm(). The qqnorm() function has a few arguments, but the only one we really need to care about here is y, a vector specifying the data whose normality we’re interested in checking. Here’s the R commands:

+

+

+_images/05.02-ttest_129_0.png
+

+

+

+
Fig. 2.25 The skewness of these data of 100 observations is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is \(W=.80\), reflecting a significant departure from normality (\(p<.001\)).

+

+

+

+
df = pd.read_csv("https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/heavy_tailed_data.csv")
+
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
+
+qq = probplot(df['data'], dist="norm", plot = ax1)
+hist = sns.histplot(df['data'], axes=ax2)
+
+glue("qqheavy_fig", ax, display=False)
+

+

+

+

+_images/05.02-ttest_131_1.png
+

+

+

+

+
<AxesSubplot:xlabel='grades', ylabel='Count'>
+

+

+

+
Fig. 2.26 Plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is \(W = .93\), again reflecting significant non-normality (\(p < .001\)).

+

 

 

 
2.8.2. Shapiro-Wilk tests

-
Although QQ plots provide a nice way to informally check the normality of your data, sometimes you’ll want to do something a bit more formal. And when that moment comes, the Shapiro-Wilk test [@Shapiro1965] is probably what you’re looking for.^[Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I’ve read seem to suggest Shapiro-Wilk is the better test of normality; although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests; in R it’s implemented via the ks.test() function.] As you’d expect, the null hypothesis being tested is that a set of \(N\) observations is normally distributed. The test statistic that it calculates is conventionally denoted as \(W\), and it’s calculated as follows. First, we sort the observations in order of increasing size, and let \(X_1\) be the smallest value in the sample, \(X_2\) be the second smallest and so on. Then the value of \(W\) is given by
-$\(
+
Although QQ plots provide a nice way to informally check the normality of your data, sometimes you’ll want to do something a bit more formal. And when that moment comes, the Shapiro-Wilk test [@Shapiro1965] is probably what you’re looking for.14 As you’d expect, the null hypothesis being tested is that a set of \(N\) observations is normally distributed. The test statistic that it calculates is conventionally denoted as \(W\), and it’s calculated as follows. First, we sort the observations in order of increasing size, and let \(X_1\) be the smallest value in the sample, \(X_2\) be the second smallest and so on. Then the value of \(W\) is given by

+

+\[
 W = \frac{ \left( \sum_{i = 1}^N a_i X_i \right)^2 }{ \sum_{i = 1}^N (X_i - \bar{X})^2}
-\)\(
-where \)\bar{X}\( is the mean of the observations, and the \)a_i$ values are … mumble, mumble … something complicated that is a bit beyond the scope of an introductory text.

+\]

+
where \(\bar{X}\) is the mean of the observations, and the \(a_i\) values are … mumble, mumble … something complicated that is a bit beyond the scope of an introductory text.

 
Because it’s a little hard to explain the maths behind the \(W\) statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we’ll encounter in this book, it’s actually small values of \(W\) that indicated departure from normality. The \(W\) statistic has a maximum value of 1, which arises when the data look “perfectly normal”. The smaller the value of \(W\), the less normal the data are. However, the sampling distribution for \(W\)  – which is not one of the standard ones that I discussed in Chapter @ref(probability) and is in fact a complete pain in the arse to work with – does depend on the sample size \(N\). To give you a feel for what these sampling distributions look like, I’ve plotted three of them in Figure @ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near \(W=1\), and as a consequence, for larger samples \(W\) doesn’t have to be very much smaller than 1 in order for the test to be significant.

-
knitr::include_graphics(file.path(projecthome, "img/ttest2/shapirowilkdist.png"))
-
-

+

+_images/shapirowilkdist.png
+
Fig. 2.27 Sampling distribution of the Shapiro-Wilk \(W\) statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that small values of \(W\) indicate departure from normality.

 

 
To run the test in R, we use the shapiro.test() function. It has only a single argument x, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our normal.data, we get the following:

 
So, not surprisingly, we have no evidence that these data depart from normality. When reporting the results for a Shapiro-Wilk test, you should (as usual) make sure to include the test statistic \(W\) and the \(p\) value, though given that the sampling distribution depends so heavily on \(N\) it would probably be a politeness to include \(N\) as well.

@@ -2484,6 +2508,12 @@ 
2.10. Summary12
 
This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. I’ve never taught an intro stats class with less than 350 students.

 

+
13

+
This is a massive oversimplification.

+

+
14

+
Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I’ve read seem to suggest Shapiro-Wilk is the better test of normality, although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests.

+

 
 

 

diff --git a/Book/_build/html/_images/05.02-ttest_126_0.png b/Book/_build/html/_images/05.02-ttest_126_0.png
new file mode 100644
index 00000000..e52ee6ef
Binary files /dev/null and b/Book/_build/html/_images/05.02-ttest_126_0.png differ
diff --git a/Book/_build/html/_images/05.02-ttest_126_1.png b/Book/_build/html/_images/05.02-ttest_126_1.png
new file mode 100644
index 00000000..e52ee6ef
Binary files /dev/null and b/Book/_build/html/_images/05.02-ttest_126_1.png differ
diff --git a/Book/_build/html/_images/05.02-ttest_129_0.png b/Book/_build/html/_images/05.02-ttest_129_0.png
new file mode 100644
index 00000000..2e981f7a
Binary files /dev/null and b/Book/_build/html/_images/05.02-ttest_129_0.png differ
diff --git a/Book/_build/html/_images/05.02-ttest_131_0.png b/Book/_build/html/_images/05.02-ttest_131_0.png
new file mode 100644
index 00000000..f79b40c9
Binary files /dev/null and b/Book/_build/html/_images/05.02-ttest_131_0.png differ
diff --git a/Book/_build/html/_images/05.02-ttest_131_1.png b/Book/_build/html/_images/05.02-ttest_131_1.png
new file mode 100644
index 00000000..f79b40c9
Binary files /dev/null and b/Book/_build/html/_images/05.02-ttest_131_1.png differ
diff --git a/Book/_build/html/_images/05.02-ttest_58_0.png b/Book/_build/html/_images/05.02-ttest_58_0.png
index f2ce2894..4bdb39da 100644
Binary files a/Book/_build/html/_images/05.02-ttest_58_0.png and b/Book/_build/html/_images/05.02-ttest_58_0.png differ
diff --git a/Book/_build/html/_images/05.02-ttest_89_0.png b/Book/_build/html/_images/05.02-ttest_89_0.png
index a09d07cc..5e2eae0a 100644
Binary files a/Book/_build/html/_images/05.02-ttest_89_0.png and b/Book/_build/html/_images/05.02-ttest_89_0.png differ
diff --git a/Book/_build/html/_images/shapirowilkdist.png b/Book/_build/html/_images/shapirowilkdist.png
new file mode 100644
index 00000000..c47849c1
Binary files /dev/null and b/Book/_build/html/_images/shapirowilkdist.png differ
diff --git a/Book/_build/html/_sources/05.02-ttest.ipynb b/Book/_build/html/_sources/05.02-ttest.ipynb
index 2d792269..14c16221 100644
--- a/Book/_build/html/_sources/05.02-ttest.ipynb
+++ b/Book/_build/html/_sources/05.02-ttest.ipynb
@@ -230,7 +230,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "hired-boulder",
+   "id": "surface-watershed",
    "metadata": {},
    "source": [
     "### Constructing the hypothesis test\n",
@@ -255,7 +255,7 @@
   {
    "cell_type": "code",
    "execution_count": 253,
-   "id": "secret-feedback",
+   "id": "tired-secondary",
    "metadata": {
     "tags": [
      "hide-input"
@@ -327,7 +327,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "instant-adult",
+   "id": "balanced-gamma",
    "metadata": {},
    "source": [
     "```{glue:figure} ztesthyp-fig\n",
@@ -344,7 +344,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "limiting-administration",
+   "id": "dedicated-portable",
    "metadata": {},
    "source": [
     "The next step is to figure out what we would be a good choice for a diagnostic test statistic; something that would help us discriminate between $H_0$ and $H_1$. Given that the hypotheses all refer to the population mean $\\mu$, you'd feel pretty confident that the sample mean $\\bar{X}$ would be a pretty useful place to start. What we could do, is look at the difference between the sample mean $\\bar{X}$ and the value that the null hypothesis predicts for the population mean. In our example, that would mean we calculate $\\bar{X} - 67.5$. More generally, if we let $\\mu_0$ refer to the value that the null hypothesis claims is our population mean, then we'd want to calculate\n",
@@ -399,7 +399,7 @@
   {
    "cell_type": "code",
    "execution_count": 254,
-   "id": "timely-place",
+   "id": "reflected-flood",
    "metadata": {
     "tags": [
      "hide-input"
@@ -478,7 +478,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "united-module",
+   "id": "interested-plate",
    "metadata": {},
    "source": [
     "```{glue:figure} ztest-fig\n",
@@ -495,7 +495,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "parallel-trustee",
+   "id": "fatal-audience",
    "metadata": {},
    "source": [
     "And what this meant, way back in the days where people did all their statistics by hand, is that someone could publish a table like this:"
@@ -503,7 +503,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "minute-stone",
+   "id": "cordless-parts",
    "metadata": {},
    "source": [
     "|                   || critical z value             |\n",
@@ -517,7 +517,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "printable-forest",
+   "id": "arctic-handy",
    "metadata": {},
    "source": [
     "which in turn meant that researchers could calculate their $z$-statistic by hand, and then look up the critical value in a text book.  That was an incredibly handy thing to be able to do back then, but it's kind of unnecessary these days, since it's trivially easy to do it with software like Python."
@@ -525,7 +525,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "elder-empire",
+   "id": "ruled-marijuana",
    "metadata": {},
    "source": [
     "### A worked example using Python\n",
@@ -536,7 +536,7 @@
   {
    "cell_type": "code",
    "execution_count": 262,
-   "id": "secondary-transsexual",
+   "id": "isolated-edinburgh",
    "metadata": {},
    "outputs": [
     {
@@ -558,7 +558,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "trained-position",
+   "id": "talented-correspondence",
    "metadata": {},
    "source": [
     "Then, I create variables corresponding to known population standard deviation ($\\sigma = 9.5$), and the value of the population mean that the null hypothesis specifies ($\\mu_0 = 67.5$):"
@@ -567,7 +567,7 @@
   {
    "cell_type": "code",
    "execution_count": 261,
-   "id": "executed-irish",
+   "id": "generous-cherry",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -577,7 +577,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "controlled-thought",
+   "id": "sticky-nancy",
    "metadata": {},
    "source": [
     "Let's also create a variable for the sample size. We could count up the number of observations ourselves, and type `N = 20` at the command prompt, but counting is tedious and repetitive. Let's get Python to do the tedious repetitive bit by using the `len()` function, which tells us how many elements there are in a vector:"
@@ -586,7 +586,7 @@
   {
    "cell_type": "code",
    "execution_count": 263,
-   "id": "lightweight-kennedy",
+   "id": "according-violence",
    "metadata": {},
    "outputs": [
     {
@@ -607,7 +607,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continuous-briefing",
+   "id": "applied-terminology",
    "metadata": {},
    "source": [
     "Next, let's calculate the (true) standard error of the mean:"
@@ -616,7 +616,7 @@
   {
    "cell_type": "code",
    "execution_count": 266,
-   "id": "banned-isaac",
+   "id": "premium-animation",
    "metadata": {},
    "outputs": [
     {
@@ -638,7 +638,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fifth-basket",
+   "id": "amended-politics",
    "metadata": {},
    "source": [
     "And finally, we calculate our $z$-score:"
@@ -647,7 +647,7 @@
   {
    "cell_type": "code",
    "execution_count": 268,
-   "id": "caroline-uzbekistan",
+   "id": "liked-faith",
    "metadata": {},
    "outputs": [
     {
@@ -668,7 +668,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "killing-fifth",
+   "id": "compound-wednesday",
    "metadata": {},
    "source": [
     "At this point, we would traditionally look up the value 2.26 in our table of critical values. Our original hypothesis was two-sided (we didn't really have any theory about whether psych students would be better or worse at statistics than other students) so our hypothesis test is two-sided (or two-tailed) also. Looking at the little table that I showed earlier, we can see that 2.26 is bigger than the critical value of 1.96 that would be required to be significant at $\\alpha = .05$, but smaller than the value of 2.58 that would be required to be significant at a level of $\\alpha = .01$. Therefore, we can conclude that we have a significant effect, which we might write up by saying something like this:\n",
@@ -681,7 +681,7 @@
   {
    "cell_type": "code",
    "execution_count": 286,
-   "id": "economic-october",
+   "id": "everyday-intermediate",
    "metadata": {},
    "outputs": [
     {
@@ -703,7 +703,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "northern-haven",
+   "id": "peripheral-extraction",
    "metadata": {},
    "source": [
     "`NormalDist().cdf()` calculates the \"cumulative density function\" for a normal distribution. Translated to something slightly less opaque, this means that  `NormalDist().cdf()` gives us the probability that a random variable X will be less than or equal to a given value. In our case, the given value for the lower tail of the distribution was our z-score, $2.259$. So `NormalDist().cdf(-z_score)` gives us the probability that a random value draw from a normal distribution would be less than or equal to $-2.259$.\n",
@@ -716,7 +716,7 @@
   {
    "cell_type": "code",
    "execution_count": 288,
-   "id": "sixth-somerset",
+   "id": "molecular-tobacco",
    "metadata": {},
    "outputs": [
     {
@@ -739,7 +739,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "static-nursery",
+   "id": "invisible-biology",
    "metadata": {},
    "source": [
     "(zassumptions)=\n",
@@ -756,7 +756,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "popular-portugal",
+   "id": "owned-coordinator",
    "metadata": {},
    "source": [
     "(onesamplettest)=\n",
@@ -768,7 +768,7 @@
   {
    "cell_type": "code",
    "execution_count": 289,
-   "id": "discrete-blank",
+   "id": "eastern-manchester",
    "metadata": {},
    "outputs": [
     {
@@ -789,7 +789,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "labeled-probability",
+   "id": "national-batman",
    "metadata": {},
    "source": [
     "In other words, while I can't say that I know that $\\sigma = 9.5$, I *can* say that $\\hat\\sigma = 9.52$. \n",
@@ -800,7 +800,7 @@
   {
    "cell_type": "code",
    "execution_count": 291,
-   "id": "earlier-transportation",
+   "id": "latest-alcohol",
    "metadata": {
     "tags": [
      "hide-input"
@@ -873,7 +873,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "uniform-scroll",
+   "id": "detected-robinson",
    "metadata": {},
    "source": [
     "\n",
@@ -893,7 +893,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "bronze-pocket",
+   "id": "personalized-surface",
    "metadata": {},
    "source": [
     "### Introducing the $t$-test\n",
@@ -910,7 +910,7 @@
   {
    "cell_type": "code",
    "execution_count": 325,
-   "id": "soviet-calendar",
+   "id": "significant-peoples",
    "metadata": {
     "tags": [
      "hide-input"
@@ -966,7 +966,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continuing-oxford",
+   "id": "cubic-glenn",
    "metadata": {},
    "source": [
     "```{glue:figure} ttestdist-fig\n",
@@ -982,7 +982,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "imposed-herald",
+   "id": "aging-assumption",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -995,7 +995,7 @@
   {
    "cell_type": "code",
    "execution_count": 359,
-   "id": "dimensional-desperate",
+   "id": "eastern-grave",
    "metadata": {},
    "outputs": [
     {
@@ -1016,7 +1016,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "affected-bailey",
+   "id": "capable-inventory",
    "metadata": {},
    "source": [
     "So that seems straightforward enough. Our calculation resulted in a $t$-statistic of 2.54, and a $p$-value of 0.36. Now what do we *do* with this output? Well, since we're pretending that we actually care about my toy example, we're overjoyed to discover that the result is statistically significant (i.e. $p$ value below .05), and we will probably want to report our result. We could report the result by saying something like this:\n",
@@ -1033,7 +1033,7 @@
   {
    "cell_type": "code",
    "execution_count": 352,
-   "id": "steady-pollution",
+   "id": "conventional-technical",
    "metadata": {},
    "outputs": [
     {
@@ -1055,7 +1055,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "accepted-mozambique",
+   "id": "comic-offer",
    "metadata": {},
    "source": [
     "Now at least we have the bare minimum of what is necessary to report our results. Still, it would be sweet if we could get those confidence intervals as well. `scipy` actually has all the tools we need, and why these are not just built into the `ttest_1samp()` method is beyond me. To find the confidence interval, we need to:\n",
@@ -1071,7 +1071,7 @@
   {
    "cell_type": "code",
    "execution_count": 356,
-   "id": "regulation-biography",
+   "id": "turkish-killer",
    "metadata": {},
    "outputs": [
     {
@@ -1099,7 +1099,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "collective-solomon",
+   "id": "cardiac-database",
    "metadata": {},
    "source": [
     "Whew. Now at least we have everything we need for a full report of our results.\n",
@@ -1113,7 +1113,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "cellular-drinking",
+   "id": "present-wagon",
    "metadata": {},
    "source": [
     "(ttestoneassumptions)=\n",
@@ -1129,7 +1129,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "unlimited-facing",
+   "id": "ignored-conditions",
    "metadata": {},
    "source": [
     "(studentttest)=\n",
@@ -1140,7 +1140,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "adjacent-buying",
+   "id": "expanded-preserve",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1151,7 +1151,7 @@
   {
    "cell_type": "code",
    "execution_count": 360,
-   "id": "eight-opening",
+   "id": "tight-career",
    "metadata": {},
    "outputs": [
     {
@@ -1232,7 +1232,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "varying-liverpool",
+   "id": "verified-edgar",
    "metadata": {},
    "source": [
     "As we can see, there's a single data frame with two variables, `grade` and `tutor`. The `grade` variable is a numeric vector, containing the grades for all $N = 33$ students taking Dr Harpo's class; the `tutor` variable is a factor that indicates who each student's tutor was. The first five observations in this data set are shown above, and below is a nice little table with some summary statistics:"
@@ -1241,7 +1241,7 @@
   {
    "cell_type": "code",
    "execution_count": 437,
-   "id": "forty-florida",
+   "id": "present-purpose",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1320,7 +1320,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "acoustic-swaziland",
+   "id": "conscious-sampling",
    "metadata": {},
    "source": [
     "To give you a more detailed sense of what's going on here, I've plotted histograms showing the distribution of grades for both tutors {numref}`fig-harpohist`. Inspection of these histograms suggests that the students in Anastasia's class may be getting slightly better grades on average, though they also seem a little more variable."
@@ -1329,7 +1329,7 @@
   {
    "cell_type": "code",
    "execution_count": 408,
-   "id": "normal-jewel",
+   "id": "assumed-ancient",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1371,7 +1371,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "million-enterprise",
+   "id": "direct-laugh",
    "metadata": {},
    "source": [
     " ```{glue:figure} harpohist_fig\n",
@@ -1384,7 +1384,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "acceptable-cause",
+   "id": "color-demographic",
    "metadata": {},
    "source": [
     "{numref}`fig-ttestci` is a simpler plot showing the means and corresponding confidence intervals for both groups of students."
@@ -1393,7 +1393,7 @@
   {
    "cell_type": "code",
    "execution_count": 413,
-   "id": "danish-conjunction",
+   "id": "assured-treaty",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1430,7 +1430,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "hazardous-semiconductor",
+   "id": "increased-pursuit",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttestci-fig\n",
@@ -1444,7 +1444,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "worse-triumph",
+   "id": "extreme-thunder",
    "metadata": {},
    "source": [
     "### Introducing the test\n",
@@ -1464,7 +1464,7 @@
   {
    "cell_type": "code",
    "execution_count": 436,
-   "id": "daily-poverty",
+   "id": "amazing-soviet",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1528,7 +1528,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "selective-prince",
+   "id": "improving-pittsburgh",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp_fig\n",
@@ -1541,7 +1541,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "younger-anthropology",
+   "id": "noticed-sound",
    "metadata": {},
    "source": [
     "To construct a hypothesis test that handles this scenario, we start by noting that if the null hypothesis is true, then the difference between the population means is *exactly* zero, \n",
@@ -1563,7 +1563,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "anticipated-compression",
+   "id": "forty-occasions",
    "metadata": {},
    "source": [
     "### A \"pooled estimate\" of the standard deviation\n",
@@ -1594,7 +1594,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "inclusive-andrews",
+   "id": "printable-paraguay",
    "metadata": {},
    "source": [
     "### The same pooled estimate, described differently\n",
@@ -1622,7 +1622,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "informed-syria",
+   "id": "horizontal-latino",
    "metadata": {},
    "source": [
     "(indsamplesttest_formula)=\n",
@@ -1645,7 +1645,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "binary-habitat",
+   "id": "healthy-disabled",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1658,7 +1658,7 @@
   {
    "cell_type": "code",
    "execution_count": 454,
-   "id": "tracked-harvest",
+   "id": "connected-alexander",
    "metadata": {},
    "outputs": [
     {
@@ -1684,7 +1684,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fourth-touch",
+   "id": "likely-sensitivity",
    "metadata": {},
    "source": [
     "This is fairly straightforward, and just as it was for the one-sample $t$-test, `scipy` does very little to format its results or give you any information over the bare minimum. You get a $t$-statistic and a $p$-value and that's that. Luckily, as was the case with the one-sampel $t$-test, getting the other elements we need to report our results isn't too bad. We will need the $t$-statistic, the $p$-value, the mean of each group, and the degrees of freedom. The first two we already have, and the last two are easy to get. As we discussed [above](indsamplesttest_formula), the degrees of freedom for an independent samples $t$-test is $N-2$, so..."
@@ -1693,7 +1693,7 @@
   {
    "cell_type": "code",
    "execution_count": 465,
-   "id": "dependent-circumstances",
+   "id": "structural-tragedy",
    "metadata": {},
    "outputs": [
     {
@@ -1717,7 +1717,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "exciting-growth",
+   "id": "registered-siemens",
    "metadata": {},
    "source": [
     "You probably noticed that in addition to telling `ttest_ind` which means I wanted to compare, I also added the argument `equal_var = True` to the command. This wasn't strictly necessary in this case, because by default this argument is set to `True`. But I made it explicit anyway, because we will be using this argument again later. By saying `equal_var = True`, what we're really doing is telling Python to use the *Student* independent samples $t$-test. More on this later."
@@ -1725,7 +1725,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "square-scott",
+   "id": "south-supervision",
    "metadata": {},
    "source": [
     "In any case, the difference between the two groups is significant (just barely), so we might write up the result using text like this:\n",
@@ -1735,7 +1735,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "bright-maker",
+   "id": "appointed-shipping",
    "metadata": {},
    "source": [
     " \n",
@@ -1767,7 +1767,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "damaged-peter",
+   "id": "organizational-drill",
    "metadata": {},
    "source": [
     "(studentassumptions)= \n",
@@ -1783,7 +1783,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "binding-struggle",
+   "id": "distributed-lawrence",
    "metadata": {},
    "source": [
     "(welchttest)=\n",
@@ -1817,7 +1817,7 @@
   {
    "cell_type": "code",
    "execution_count": 530,
-   "id": "sublime-hospital",
+   "id": "functioning-andrews",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1888,7 +1888,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "standing-crawford",
+   "id": "serial-offset",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp2_fig\n",
@@ -1901,7 +1901,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "intelligent-profile",
+   "id": "inappropriate-knife",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1912,7 +1912,7 @@
   {
    "cell_type": "code",
    "execution_count": 531,
-   "id": "established-bryan",
+   "id": "simple-lloyd",
    "metadata": {},
    "outputs": [
     {
@@ -1933,7 +1933,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fixed-elder",
+   "id": "moving-cabin",
    "metadata": {},
    "source": [
     "Not too difficult, right? Not surprisingly, the output has exactly the same format as it did last time too: a test statistic $t$, and a $p$-value. So that's all pretty easy. \n",
@@ -1948,7 +1948,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "lesbian-wright",
+   "id": "ceramic-favor",
    "metadata": {},
    "source": [
     "(pairedsamplesttest)=\n",
@@ -1959,7 +1959,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "central-interim",
+   "id": "transsexual-armor",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1970,7 +1970,7 @@
   {
    "cell_type": "code",
    "execution_count": 563,
-   "id": "graphic-wellington",
+   "id": "better-depth",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1980,7 +1980,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "failing-split",
+   "id": "prepared-sword",
    "metadata": {},
    "source": [
     "The data frame `chico` contains three variables: an `id` variable that identifies each student in the class, the `grade_test1` variable that records the student grade for the first test, and the `grade_test2` variable that has the grades for the second test. Here's the first five students:"
@@ -1989,7 +1989,7 @@
   {
    "cell_type": "code",
    "execution_count": 534,
-   "id": "higher-prefix",
+   "id": "chinese-missouri",
    "metadata": {},
    "outputs": [
     {
@@ -2073,7 +2073,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "aquatic-attention",
+   "id": "divided-deadline",
    "metadata": {},
    "source": [
     "At a glance, it does seem like the class is a hard one (most grades are between 50\\% and 60\\%), but it does look like there's an improvement from the first test to the second one. If we take a quick look at the descriptive statistics"
@@ -2082,7 +2082,7 @@
   {
    "cell_type": "code",
    "execution_count": 535,
-   "id": "apparent-partition",
+   "id": "endangered-theme",
    "metadata": {},
    "outputs": [
     {
@@ -2178,7 +2178,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "professional-mounting",
+   "id": "reserved-procedure",
    "metadata": {},
    "source": [
     "we see that this impression seems to be supported. Across all 20 students[^note12] the mean grade for the first test is 57\\%, but this rises to 58\\% for the second test. Although, given that the standard deviations are 6.6\\% and 6.4\\% respectively, it's starting to feel like maybe the improvement is just illusory; maybe just random variation. This impression is reinforced when you see the means and confidence intervals plotted in {numref}`pairedta` panel A. If we were to rely on this plot alone, we'd come to the same conclusion that we got from looking at the descriptive statistics that the `describe()` method produced. Looking at how wide those confidence intervals are, we'd be tempted to think that the apparent improvement in student performance is pure chance."
@@ -2187,7 +2187,7 @@
   {
    "cell_type": "code",
    "execution_count": 595,
-   "id": "empty-backup",
+   "id": "compact-employer",
    "metadata": {
     "tags": [
      "hide-input"
@@ -2237,7 +2237,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "virtual-humor",
+   "id": "falling-language",
    "metadata": {},
    "source": [
     "```{glue:figure} pairedta_fig\n",
@@ -2251,7 +2251,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continued-temperature",
+   "id": "bearing-allowance",
    "metadata": {},
    "source": [
     "Nevertheless, this impression is wrong. To see why, take a look at the scatterplot of the grades for test 1 against the grades for test 2.  shown in {numref}`fig-pairedta` panel B. \n",
@@ -2264,7 +2264,7 @@
   {
    "cell_type": "code",
    "execution_count": 596,
-   "id": "median-pharmacy",
+   "id": "earlier-director",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2273,7 +2273,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "pharmaceutical-pickup",
+   "id": "white-sweden",
    "metadata": {},
    "source": [
     "Notice that I assigned the output to a variable called `df['improvement]`. That has the effect of creating a new column called `improvement` inside the `chico` data frame. Now that we've created and stored this `improvement` variable, we can draw a histogram showing the distribution of these improvement scores, shown in {numref}`fig-pairedta` panel C. \n",
@@ -2285,7 +2285,7 @@
   {
    "cell_type": "code",
    "execution_count": 598,
-   "id": "ceramic-spice",
+   "id": "plain-massage",
    "metadata": {},
    "outputs": [
     {
@@ -2310,7 +2310,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "reasonable-slide",
+   "id": "fixed-concentrate",
    "metadata": {},
    "source": [
     "we see that it is 95\\% certain that the true (population-wide) average improvement would lie between 0.95\\% and 1.86\\%. So you can see, qualitatively, what's going on: there is a real \"within student\" improvement (everyone improves by about 1\\%), but it is very small when set against the quite large \"between student\" differences (student grades vary by about 20\\% or so). "
@@ -2318,7 +2318,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "maritime-harvest",
+   "id": "dramatic-injury",
    "metadata": {},
    "source": [
     "### What is the paired samples $t$-test?\n",
@@ -2357,7 +2357,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "annual-inspector",
+   "id": "prompt-parade",
    "metadata": {},
    "source": [
     "### Doing the test in Python \n",
@@ -2368,7 +2368,7 @@
   {
    "cell_type": "code",
    "execution_count": 600,
-   "id": "about-restoration",
+   "id": "attached-catalyst",
    "metadata": {},
    "outputs": [
     {
@@ -2389,7 +2389,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "listed-bones",
+   "id": "incident-thailand",
    "metadata": {},
    "source": [
     "However, suppose you're lazy and you don't want to go to all the effort of creating a new variable. Or perhaps you just want to keep the difference between one-sample and paired-samples tests clear in your head. In that case, `scipy` also has a built-in method for conducting paired $t$-tests called `ttest_rel` (the `_rel` part is for \"related\"). Using this method, we get:"
@@ -2398,7 +2398,7 @@
   {
    "cell_type": "code",
    "execution_count": 599,
-   "id": "english-china",
+   "id": "handled-produce",
    "metadata": {},
    "outputs": [
     {
@@ -2420,7 +2420,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "eleven-governor",
+   "id": "joined-skiing",
    "metadata": {},
    "source": [
     "Either way, the result is exactly the same, which is strangely comforting, actually.  Not only that, but the result confirms our intuition. There’s an average improvement of 1.4% from test 1 to test 2, and this is significantly different from 0 ($t$(19) = 6.48, $p$ < .001). In fact, $p$ is quite a bit less than one, since the $p$-value has been given in scientific notation. The exact $p$-value is $3.32^{-06}$, that is, $p$ = 0.0000032."
@@ -2428,7 +2428,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "rational-carolina",
+   "id": "integral-intervention",
    "metadata": {},
    "source": [
     "## One sided tests\n",
@@ -2439,7 +2439,7 @@
   {
    "cell_type": "code",
    "execution_count": 606,
-   "id": "toxic-boating",
+   "id": "elect-oasis",
    "metadata": {},
    "outputs": [
     {
@@ -2464,7 +2464,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "possible-precipitation",
+   "id": "scientific-portfolio",
    "metadata": {},
    "source": [
     "The $t$-statistics are exactly the same, which makes sense, if you think about it, because the calculation of the $t$ is based on the mean and standard deviation, and these do not change. The $p$-value, on the other hand, is lower for the one-sided test. The only thing that changes between the two tests is the _expectation_ that we bring to data. The way that the $p$-value is calculated depends on those expectations, and they are the reason for choosing one test over the other. It should go without saying, but maybe is worth saying anyway, that our reasons for choosing one test over the other should be theoretical, and not based on which test is more likely to give us the $p$-value we want!\n",
@@ -2476,7 +2476,7 @@
   {
    "cell_type": "code",
    "execution_count": 613,
-   "id": "living-auditor",
+   "id": "covered-thanks",
    "metadata": {},
    "outputs": [
     {
@@ -2510,7 +2510,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "american-chrome",
+   "id": "frank-neutral",
    "metadata": {},
    "source": [
     "What about the paired samples $t$-test? Suppose we wanted to test the hypothesis that grades go *up* from test 1 to test 2 in Dr. Chico's class, and are not prepared to consider the idea that the grades go down. Again, we can use the `alternative` argument to specify the one-sided test, and it works the same way it does for the independent samples $t$-test. Since we are comparing test 1 to test 2 by substracting one from the other, it makes a difference whether we subract test 1 from test 2, or test 2 from test 1. So, to test the hypothesis that grades for test 2 are higher than test 2, we will need to enter the grades from test 2 first; otherwise we are testing the opposite hypothesis: "
@@ -2519,7 +2519,7 @@
   {
    "cell_type": "code",
    "execution_count": 616,
-   "id": "lovely-cylinder",
+   "id": "constitutional-newfoundland",
    "metadata": {},
    "outputs": [
     {
@@ -2545,7 +2545,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sporting-wednesday",
+   "id": "downtown-float",
    "metadata": {},
    "source": [
     "(cohensd)=\n",
@@ -2564,7 +2564,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "played-church",
+   "id": "awful-annual",
    "metadata": {},
    "source": [
     "(dinterpretation)=\n",
@@ -2574,7 +2574,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "treated-quebec",
+   "id": "corrected-reconstruction",
    "metadata": {},
    "source": [
     "| d-value   | rough interpretation |\n",
@@ -2586,7 +2586,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "advised-oasis",
+   "id": "precious-scholarship",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from one sample\n",
@@ -2601,7 +2601,7 @@
   {
    "cell_type": "code",
    "execution_count": 624,
-   "id": "featured-somerset",
+   "id": "rural-ratio",
    "metadata": {},
    "outputs": [
     {
@@ -2628,7 +2628,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "square-culture",
+   "id": "cooperative-final",
    "metadata": {},
    "source": [
     "What does this effect size mean? Overall, then, the psychology students in Dr Zeppo's class are achieving grades (mean = 72.3\\%) that are about .5 standard deviations higher than the level that you'd expect (67.5\\%) if they were performing at the same level as other students. Judged against Cohen's rough guide, this is a moderate effect size."
@@ -2636,7 +2636,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "administrative-chile",
+   "id": "noted-timeline",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Student $t$ test\n",
@@ -2658,7 +2658,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "foster-correction",
+   "id": "reliable-trigger",
    "metadata": {},
    "source": [
     "However, there are other possibilities, which I'll briefly describe. Firstly, you may have reason to want to use only one of the two groups as the basis for calculating the standard deviation. This approach (often called Glass' $\\Delta$) only makes most sense when you have good reason to treat one of the two groups as a purer reflection of \"natural variation\" than the other. This can happen if, for instance, one of the two groups is a control group. Secondly, recall that in the usual calculation of the pooled standard deviation we divide by $N-2$ to correct for the bias in the sample variance; in one version of Cohen's $d$ this correction is omitted. Instead, we divide by $N$. This version makes sense primarily when you're trying to calculate the effect size in the sample; rather than estimating an effect size in the population. Finally, there is a version based on @Hedges1985, who point out there is a small bias in the usual (pooled) estimation for Cohen's $d$. Thus they introduce a small correction, by multiplying the usual value of $d$ by $(N-3)/(N-2.25)$. \n",
@@ -2669,7 +2669,7 @@
   {
    "cell_type": "code",
    "execution_count": 654,
-   "id": "western-document",
+   "id": "suffering-technical",
    "metadata": {},
    "outputs": [
     {
@@ -2719,7 +2719,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "aboriginal-calcium",
+   "id": "present-kennedy",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Welch test\n",
@@ -2749,7 +2749,7 @@
   {
    "cell_type": "code",
    "execution_count": 657,
-   "id": "advance-cleanup",
+   "id": "romantic-reality",
    "metadata": {},
    "outputs": [
     {
@@ -2784,7 +2784,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "above-yeast",
+   "id": "diagnostic-processor",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a paired-samples test\n",
@@ -2802,7 +2802,7 @@
   {
    "cell_type": "code",
    "execution_count": 663,
-   "id": "legendary-crown",
+   "id": "classical-juice",
    "metadata": {},
    "outputs": [
     {
@@ -2831,7 +2831,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "altered-example",
+   "id": "animated-brunei",
    "metadata": {},
    "source": [
     "The only wrinkle is figuring out whether this is the measure you want or not. To the extent that you care about the practical consequences of your research, you often want to measure the effect size relative to the *original* variables, not the *difference* scores (e.g., the 1\\% improvement in Dr Chico's class is pretty small when measured against the amount of between-student variation in grades), in which case you use the same versions of Cohen's $d$ that you would use for a Student or Welch test. For instance, when we do that for Dr Chico's class, "
@@ -2840,7 +2840,7 @@
   {
    "cell_type": "code",
    "execution_count": 666,
-   "id": "noble-history",
+   "id": "structured-colon",
    "metadata": {},
    "outputs": [
     {
@@ -2874,7 +2874,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "northern-appeal",
+   "id": "functioning-circus",
    "metadata": {},
    "source": [
     "what we see is that the overall effect size is quite small, when assessed on the scale of the original variables."
@@ -2882,107 +2882,246 @@
   },
   {
    "cell_type": "markdown",
-   "id": "frequent-market",
+   "id": "economic-medicine",
    "metadata": {},
    "source": [
     "(shapiro)=\n",
     "## Checking the normality of a sample\n",
     "\n",
-    "All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section \\@ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is *actually* an average of lots of different things, there's a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using $t$-tests. However, life doesn't come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there's a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.^[This is a massive oversimplification.] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test.\n",
-    "\n",
+    "All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section \\@ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is *actually* an average of lots of different things, there's a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using $t$-tests. However, life doesn't come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there's a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.[^note13] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "northern-pressing",
+   "metadata": {},
+   "source": [
     "### QQ plots\n",
     "\n",
-    "```{r qq1a, fig.cap=\"Histogram of `normal.data`, a normally distributed sample with 100 observations.\", echo=FALSE}\n",
-    "  width <- 5\n",
-    "  height <- 5\n",
-    "  \n",
-    "  plotOne <- function( data, title ) {\n",
-    "    \n",
-    "\n",
-    "    hist( x = data ,\n",
-    "          xlab = \"Value\", # x-axis label\n",
-    "          main = title,                     \n",
-    "          border=\"white\", \n",
-    "          col=ifelse(colour,emphColLight,emphGrey),\n",
-    "          font.main = 1\n",
-    "    )         \n",
-    "    \n",
-    "    require(psych)\n",
-    "    cat(title,\"\\n\")\n",
-    "    cat( \"skew=\",skew(data),\"\\n\")\n",
-    "    cat( \"kurtosis=\",kurtosi(data),\"\\n\")\n",
-    "    print( shapiro.test(data) )\n",
-    "    \n",
-    "    \n",
-    "  }\n",
-    "  \n",
-    "  normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers\n",
-    "  plotOne( normal.data, \"Normally Distributed Data\" )\n",
-    "```\n",
+    "One way to check whether a sample violates the normality assumption is to draw a **_\"quantile-quantile\" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot, using `probplot` from `scipy`. We can compare this with a histogram of the data as well:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 706,
+   "id": "seeing-consideration",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "application/papermill.record/text/plain": ""
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qq_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "import numpy as np\n",
+    "import seaborn as sns\n",
+    "from scipy.stats import probplot\n",
     "\n",
-    "```{r qq1b, fig.cap=\"Normal QQ plot of `normal.data`, a normally distributed sample with 100 observations.\", echo=FALSE}\n",
-    "  plotTwo <- function( data ) {   \n",
-    "    qqnorm( y = data,\n",
-    "            pch=19,\n",
-    "            font.main = 1\n",
-    "    )        # draw the QQ plot\n",
-    "  \n",
-    "  }\n",
+    "np.random.seed(42)\n",
+    "normal_data = np.random.normal(size=100)\n",
     "\n",
-    "  plotTwo( normal.data )\n",
-    "```\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
     "\n",
-    "The Shapiro-Wilk statistic associated with the data in Figures \\@ref(fig:qq1a) and \\@ref(fig:qq1b) is $W = .99$, indicating that no significant departures from normality were detected ($p = .73$). As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the two data sets shown in Figures \\@ref(fig:qq2a), \\@ref(fig:qq2b), \\@ref(fig:qq2c), \\@ref(fig:qq2d). Figures \\@ref(fig:qq2a) and \\@ref(fig:qq2b) show the histogram and a QQ plot for a data set that is highly skewed: the QQ plot curves upwards. Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) show the same plots for a heavy tailed (i.e., high kurtosis) data set: in this case, the QQ plot flattens in the middle and curves sharply at either end.\n",
+    "qq = probplot(normal_data, dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(normal_data, axes=ax2)\n",
     "\n",
-    "```{r qq2a, fig.cap=\"A histogram of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  skewed.data <- rgamma( n = 100,1,2 ) \n",
-    "  plotOne( skewed.data, \"Skewed Data\" )\n",
-    "```\n",
+    "glue(\"qq_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "variable-cedar",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qq_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qq\n",
     "\n",
-    "```{r qq2b, fig.cap=\"A normal QQ plot of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  plotTwo( skewed.data )\n",
-    "```\n",
+    "QQ plot (left) and histogram (right) of `normal_data`, a normally distributed sample with 100 observations. The Shapiro-Wilk statistic associated with these data is $W = .99$, indicating that no significant departures from normality were detected ($p = .73$).\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "hired-classic",
+   "metadata": {},
+   "source": [
+    "And the results are shown in {numref}(`fig-qq`), above.\n",
+    "\n",
+    "As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the data shown in {numref}(qqskew) and {numref}(qqheavy), which show the histogram and a QQ plot for a data sets that are highly skewed and have a heavy tail (i.e., high kurtosis), respectively."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 704,
+   "id": "protected-sunday",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/skewed_data.csv\")\n",
     "\n",
-    "The skewness of the data in Figures \\@ref(fig:qq2a) and \\@ref(fig:qq2b) is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
     "\n",
-    "```{r qq2c, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  heavy.tailed.data <- c(rnorm( n = 80 ) , rnorm( 20, sd=5))\n",
-    "  plotOne( heavy.tailed.data, \"Heavy-Tailed Data\" )\n",
-    "```\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
     "\n",
-    "```{r qq2d, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  plotTwo( heavy.tailed.data )\n",
-    "```\n",
+    "glue(\"qqskew_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "improving-husband",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqskew_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqskew\n",
     "\n",
-    "Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) shows the same plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
+    "The skewness of these data of 100 observations is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
+    "\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 707,
+   "id": "understood-michigan",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "application/papermill.record/text/plain": ""
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qqheavy_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/heavy_tailed_data.csv\")\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
+    "\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
+    "\n",
+    "glue(\"qqheavy_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "precious-sally",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqheavy_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqheavy\n",
+    "\n",
+    "Plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
     "\n",
-    "One way to check whether a sample violates the normality assumption is to draw a **_\"quantile-quantile\" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot using the R function `qqnorm()`. The `qqnorm()` function has a few arguments, but the only one we really need to care about here is `y`, a vector specifying the data whose normality we're interested in checking. Here's the R commands:\n",
-    "```{r}\n",
-    "normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers\n",
-    "hist( x = normal.data )          # draw a histogram of these numbers\n",
-    "qqnorm( y = normal.data )        # draw the QQ plot\n",
     "```"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "established-system",
+   "id": "silent-desire",
    "metadata": {},
    "source": [
     "### Shapiro-Wilk tests\n",
     "\n",
-    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.^[Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality; although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests; in R it's implemented via the `ks.test()` function.] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.[^note14] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "\n",
     "$$\n",
     "W = \\frac{ \\left( \\sum_{i = 1}^N a_i X_i \\right)^2 }{ \\sum_{i = 1}^N (X_i - \\bar{X})^2}\n",
     "$$\n",
+    "\n",
     "where $\\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. \n",
     "\n",
-    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. \n",
+    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "stuck-prison",
+   "metadata": {},
+   "source": [
+    "```{figure} ../img/ttest2/shapirowilkdist.png\n",
+    ":name: fig-shapirowilk\n",
+    ":width: 600px\n",
+    ":align: center\n",
     "\n",
-    "```{r swdist, fig.cap=\"Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\", echo=FALSE}\n",
-    "knitr::include_graphics(file.path(projecthome, \"img/ttest2/shapirowilkdist.png\"))\n",
+    "Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\n",
+    "\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "lesbian-cooperation",
+   "metadata": {},
+   "source": [
     "\n",
-    "```\n",
     "\n",
     "To run the test in R, we use the `shapiro.test()` function. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal.data`, we get the following:\n",
     "```{r}\n",
@@ -3099,7 +3238,11 @@
     "\n",
     "[^note11]: At this point we have Drs Harpo, Chico and Zeppo. No prizes for guessing who Dr Groucho is.\n",
     "\n",
-    "[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students."
+    "[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students.\n",
+    "\n",
+    "[^note13]: This is a massive oversimplification.\n",
+    "\n",
+    "[^note14]: Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality, although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests."
    ]
   }
  ],
diff --git a/Book/_build/html/objects.inv b/Book/_build/html/objects.inv
index 46dad800..86d7735e 100644
Binary files a/Book/_build/html/objects.inv and b/Book/_build/html/objects.inv differ
diff --git a/Book/_build/html/searchindex.js b/Book/_build/html/searchindex.js
index 67e463b7..6cc9bef7 100644
--- a/Book/_build/html/searchindex.js
+++ b/Book/_build/html/searchindex.js
@@ -1 +1 @@
-Search.setIndex({docnames:["01.01-intro","01.02-studydesign","02.01-getting_started_with_python","02.02-more_python_concepts","03.01-descriptives","03.02-drawing_graphs","03.03-pragmatic_matters","03.04-basic_programming","04.01-intro-to-probability","04.02-probability","04.03-estimation","04.04-hypothesis-testing","05.01-chisquare","05.02-ttest","05.03-anova","05.04-regression","05.05-anova2","06.01-bayes","06.02-epilogue","bibliography","landingpage"],envversion:{"sphinx.domains.c":2,"sphinx.domains.changeset":1,"sphinx.domains.citation":1,"sphinx.domains.cpp":3,"sphinx.domains.index":1,"sphinx.domains.javascript":2,"sphinx.domains.math":2,"sphinx.domains.python":2,"sphinx.domains.rst":2,"sphinx.domains.std":2,"sphinx.ext.intersphinx":1,"sphinxcontrib.bibtex":7,sphinx:56},filenames:["01.01-intro.ipynb","01.02-studydesign.ipynb","02.01-getting_started_with_python.ipynb","02.02-more_python_concepts.ipynb","03.01-descriptives.ipynb","03.02-drawing_graphs.ipynb","03.03-pragmatic_matters.ipynb","03.04-basic_programming.ipynb","04.01-intro-to-probability.ipynb","04.02-probability.ipynb","04.03-estimation.ipynb","04.04-hypothesis-testing.ipynb","05.01-chisquare.ipynb","05.02-ttest.ipynb","05.03-anova.ipynb","05.04-regression.ipynb","05.05-anova2.ipynb","06.01-bayes.ipynb","06.02-epilogue.ipynb","bibliography.md","landingpage.md"],objects:{},objnames:{},objtypes:{},terms:{"000":[0,4,10,11,13],"00000":4,"000000":[4,5,10,13],"0000000":[],"00000000":[],"000000000":[],"000000000000000000000000136":11,"0000000001":11,"0000032":13,"001":[9,11,13],"00135199382707":[],"002":[],"0024":1,"00248896":[],"00404299296417":[],"005":9,"005833":4,"005833399":[],"006":4,"00601151900518":[],"007":[],"007499":[],"010434":4,"011492":4,"01152043":[],"011922871882469877":13,"013":[],"01379741":[],"015884":4,"016":4,"016644":[],"017532279773745":[],"01807260939072272":13,"01929769888727":[],"019586288325765":[],"019d":[],"01eff593ab66":[],"020409":4,"02077226":[],"020978735677851718":11,"021":11,"02126474":13,"02183303":[],"02342152288259":10,"023666733139155":[],"023845743764939753":13,"024978":10,"025":[10,13],"02514322152811":[],"02745433":[],"02807813330601":[],"02949514355197":[],"029c62871301":[],"030000":4,"031":1,"0326992":[],"03418661":13,"036":[],"03608366":[],"03614521878144544":13,"03643339":10,"03915839942863":10,"03921221":[],"03935379301905":[],"03970372":[],"041595":[],"04182658946993":[],"0424d1d3":[],"04252949":13,"043":4,"04499342":[],"04539b4d4399":[],"046821":4,"0471a470108d":[],"048173":[],"049":[11,13],"049200":4,"04967":4,"049904":[],"050000":13,"051":[11,13],"05360997":13,"054":13,"05555555555556":13,"055556":13,"05686539981934":[],"057":11,"05705482891125":[],"05751208":[],"05822858861367":[],"058309":4,"058309485":[],"05830949":[],"05922473578477":10,"06054205":[],"060789":[],"06132891":[],"061329":4,"062":11,"0642512":[],"0643495578651616":4,"06434955786516161":4,"06816586":[],"068166":4,"06e1":10,"070000":4,"0722941558371":[],"07344737":[],"073636":[4,5],"073636359108274":4,"074232":4,"07436796218789":[],"074844":[],"076479":4,"076796665":[],"07679667":[],"076797":4,"0769629":[],"0787":[],"078784":[],"079791106564333":10,"080842":[],"08301893":[],"08323040323924":[],"0877058441629":[],"090232":13,"09083859":[],"09110059":[],"09148816719804":10,"09374202":[],"09526109998637":[],"09690401":[],"09817021869561":[],"09825660326588":[],"098408":4,"09dd6645":[],"09de7a5f":[],"09e25dbb":[],"0b25dde20c3":[],"0b6b":[],"0b6e4918":[],"0b70":[],"0ca4c902e18":[],"0dbb3e59":[],"0e79":[],"0edf":[],"0f20":[],"0th":6,"0x104f641c0":[],"0x105028e80":[],"0x105bd6580":[],"0x107adf640":[],"0x107d8aee0":[],"0x108936a60":[],"0x10b7ff730":[],"0x10bf73400":[],"0x10bf79130":[],"0x10db54490":[],"0x10dcb70a0":[],"0x10e730f40":[],"0x10f47eb20":[],"0x1111bb340":[],"0x111a2c580":[],"0x111cd6d60":[],"0x1141189d0":[],"0x1145ce1f0":[],"0x11539afa0":[],"0x116b7cb50":[],"0x116e25d60":4,"0x118c67940":[],"0x119810790":[],"0x119bf8d30":[],"0x11ab0ef40":[],"0x11ba1d6a0":[],"0x11c9d64f0":[],"0x12df8a3d0":[],"0x12efa5a00":[],"0x12f5196a0":[],"0x12f79f2b0":[],"0x12f883d90":[],"0x12fc07dc0":[],"0x130acaf70":[],"0x130efda30":[],"0x131878bb0":[],"0x131faf610":[],"0x1320ecd00":[],"0x132356d60":[],"0x1324848b0":[],"0x132780fa0":[],"0x132890f70":[],"0x1328b4ca0":[],"0x132c0f8e0":[],"0x133e60f40":[],"0x133ea0d90":[],"0x1341032b0":[],"0x13421a460":[],"0x134424220":[],"0x1344887f0":[],"0x1347150a0":[],"0x1348a5430":[],"0x134a1cfa0":[],"0x134b42730":[],"0x134b4bac0":13,"0x134b8f340":[],"0x134bf6f40":13,"0x134c4cb20":[],"0x134fcf0d0":[],"0x135151df0":[],"0x13515f460":[],"0x135c8feb0":10,"0x135d13220":[],"0x135f21310":[],"0x1362803a0":[],"0x1362da3a0":[],"0x1362e02e0":[],"0x1363a16a0":[],"0x136ecc3a0":[],"0x137054610":[],"0x13706d430":[],"0x1370dd250":[],"0x137613f10":[],"0x137bb82b0":[],"0x137c10b80":[],"0x137c217f0":[],"0x137c904f0":[],"0x137fef040":[],"0x1380475b0":[],"0x13804b280":[],"0x1380f2b20":[],"0x138112f70":[],"0x1381a4520":[],"0x138449af0":[],"0x1384ffee0":[],"0x13869d8b0":[],"0x13871f130":[],"0x138731ac0":[],"0x1387b3070":[],"0x138913190":[],"0x138a28280":[],"0x138bbc880":[],"0x138bcd370":[],"0x138dc27f0":[],"0x138e1c520":[],"0x138eb2310":[],"0x139315bb0":[],"0x1395244c0":[],"0x13998ae50":[],"0x139b96f40":[],"0x139c90340":[],"0x13a17fe20":[],"0x13a775430":[],"0x13a7df280":[],"0x13a91f460":[],"0x13ad5f550":[],"0x13b0f4fa0":[],"0x13b2ee6d0":[],"0x13b563cd0":[],"0x13b6a9250":[],"0x13b6dcf40":[],"0x13ba01af0":[],"0x13ba16280":[],"0x13bbb7df0":[],"0x13c0aac70":[],"0x13c407ee0":[],"0x13c5bd6a0":[],"0x13c683d60":[],"0x13c7b74c0":[],"0x13c9b6070":[],"0x13ca8ca60":[],"0x13cd8ad60":[],"0x13d1e5520":[],"0x13d369d60":[],"0x13d3eda90":[],"0x13d4ab070":[],"0x13d5873d0":[],"0x13d6987f0":[],"0x13d90e190":[],"0x13dab3340":[],"0x13e0c4880":[],"0x13e1214f0":[],"0x13e6088e0":[],"0x13e6516d0":[],"0x13efe3c70":[],"0x13f018760":[],"0x13f8859a0":[],"0x13fc0da60":[],"0x13ff8d280":[],"0x13ff9a9a0":[],"0x140383340":[],"0x140d0bd00":[],"0x1417afe50":[],"0x141af5f70":[],"0x141ca25e0":[],"0x7fb554040320":3,"0x7fd3d2cdfc20":[],"0x7fe46813ba90":[],"100":[0,1,3,4,6,9,10,11,13],"1000":[1,4,9,10],"10000":[9,10,11],"100000":[4,13],"1000000":10,"10000000":9,"100cc72e6535":[],"100th":6,"101":10,"10109718805638757":4,"102":10,"1024":11,"103":[10,19],"104":[],"105":10,"105084137429943":10,"105084137429944":10,"105248":[],"10577300630605":[],"106":10,"107":10,"108":[0,5],"10837250197972":[],"109":10,"10th":[4,8],"110":[4,10],"111":10,"111686f274e2":[],"112":10,"112154266955443":[],"1123667414013":[],"113":10,"1135102194217":[],"11352343709996":[],"114":[],"114494":4,"115":[10,19],"1151":19,"11543239":13,"116":[4,5],"11646993":[],"117":[4,10],"1172":19,"118":10,"119":10,"11a110c312b2":[],"120":[5,10],"121":10,"122":10,"123":4,"1231055273238":[],"124":[],"124228":[],"1242645786248002":13,"125":10,"126":[],"127":[],"128":[],"129":[],"129391846315059":[],"12ae68273d06":10,"12f6a9b78854":[],"130":[],"13066559719219":[],"131":[],"132":[3,4],"133":10,"13486720826272":[],"135":[],"136":4,"137833":[],"138":[],"13814737543848":[],"139":[],"13th":8,"140":10,"1415":3,"14513279173728":[],"145558":4,"1455581":[],"14555814":[],"147":[],"14778526288103":[],"148318398823882":[],"14967680216347":[],"14a9":[],"150":[1,4],"15131168":[],"152174":4,"15237303113537":[],"15264631":[],"15293958344104":10,"154f":[],"155418":[],"15864518":[],"159":4,"15978061608057":[],"15bd4ff623a9":[],"15c588d0eaae":2,"15th":10,"16420029223272":[],"1645":[],"16496580927726":4,"165":4,"1664897805783":[],"167":9,"1672629":[],"16965322610622":[],"171":4,"1713":10,"171322032470227":[],"172":4,"173":4,"173127":10,"17385161646514":[],"174":4,"17467105":[],"175":4,"17535285392665":[],"176":[4,5],"177":5,"18018727663605":[],"181503":[],"1825667":[],"183":4,"18332":[],"1854":5,"186":5,"187":19,"18725684911405":[],"18728588770138":[],"18fab7820478":[],"1902517605674":[],"1907":1,"1908":13,"191":0,"1911":19,"191144":[],"1917431a980f":[],"1920":1,"1920872":[],"1922":[11,19],"1937":4,"1938":[1,4],"1945":4,"1946":[0,19],"1955803":[],"1963":19,"1967":19,"1968":2,"19689447267619":[],"1971":[],"1973":[0,19],"1975":[6,19],"197569402522694":[],"1976":11,"1977":[],"19804714297581":[],"1983":19,"1984":19,"1986":7,"1987":[4,5],"1988":19,"19c4":[],"1d79":[],"1e3a6be03e39":[],"1e4eab12":[],"1f9614a14a60":[],"1st":[1,2,8],"200":4,"20000":[],"200000":[4,10],"2002":[1,4,19],"2003":1,"2004":[5,19],"2005":[5,19],"2006":19,"2007":0,"2008":1,"2010":[1,4,5,9,19],"2011":[1,5,19],"2014":19,"2016":0,"2018":1,"2022036":9,"20224782":[],"2048":9,"20842152145724":[],"20th":[9,11],"21061879":[],"212737":[],"21548225636099":[],"21752682":[],"219ddd80":[],"21st":0,"2213698410715347":13,"222":19,"22213677386856":[],"22235022":[],"222672621395798":[],"22659769":[],"228001905654434":[],"22f8":[],"230":9,"233454":[],"234":3,"2345":[],"23579970776728":[],"237":19,"23802375287762":[],"2382c84d04fd":[],"238842":[],"23908691097166":[],"24159f7b86d4":[],"24191232":[],"24381557":[],"2446":[],"2450":2,"245021001876143":[],"24536718309507":[],"24657209":[],"24930821":9,"24b97ad0d3d7":[],"25000":4,"250000":4,"251":19,"25471286700693":13,"25480038":[],"255":[],"255188":[],"25753056":[],"258331":[],"259":13,"259605535157681":13,"25th":[4,5],"26062329":[],"26215716":10,"265855":[],"2661476":9,"268202":[],"271319":[],"27173353050394":[],"272":0,"2733d20e1661":[],"2765451":[],"27661201499386":[],"27739475":[],"27841143860383":[],"279523":[],"281552":13,"282385719772392":[],"285000":4,"28685666":[],"28686088":[],"2870":[],"290527":13,"292500":4,"29288462":[],"293833":[],"295":19,"2_p":13,"2a09":[],"2c15d7a0":[],"2d6bbc3f":[],"2eacbe14":[],"2eb4759cd4c":[],"2f68":[],"2f9b":[],"2nd":[1,19],"300":[1,5],"3000":2,"300000":[4,13],"30107606819503":[],"301136":[4,5],"30113636363637":4,"3021480":4,"30247052":[],"306":19,"30622a0f322f":[],"309":[3,19],"30c2ad06":[],"310189":[],"3114fd44":[],"31192405433491":[],"314cb8c72701":10,"315019":[],"3151aa31a887":[],"316655":[],"31775495":[],"318654":[],"31ec45c6":[],"320670056126948e":13,"322544":[],"3227dad5f5a0":[],"324":4,"325":0,"326348":13,"32642464":[],"327":5,"327322":[],"3279":10,"328":19,"328028":[],"32b0":[],"32b38c54":[],"32c1e318a0b5":[],"331":19,"33162749802028":[],"33174961835198":[],"333":[2,4,6],"334":19,"33rd":6,"3404275653649":[],"341":0,"342":[],"34201924105093":[],"344":8,"345":19,"34648913":9,"346834":[],"3477":2,"34841741999587":[],"34fe5eabced4":[],"350":[2,13],"353":3,"3538706638441":[],"35615280000385":10,"36324512":[],"36569752995766":[],"367872":[],"368":19,"36826646949605":[],"37497756":[],"375":0,"37594501":[],"376":4,"37707609":[],"3824c47e7b89":[],"38498210704823":[],"385000":13,"38887014":[],"38965611715034":[],"390295":[],"391":[],"39106267174839":[],"393":0,"3938":[],"39398848099482":[],"395876":[],"395945":[],"39768396673382":[],"398":19,"39878609":[],"39909089":[],"3bd0":[],"3c11":[],"3c5c0d4116c9":[],"3cf564f2":[],"3de11ed2a147":[],"3e593bc2":[],"3ecc4ad645bb":[],"3rd":[1,19],"400":[4,5,8],"40000":4,"400000":[],"4003":[],"401":[],"402c":[],"403819175526497":[],"40384719999614":10,"404":19,"405":4,"405612":13,"4056320378121":[],"406136267820491":[],"40763325859871":[],"4084":[],"40d6":[],"40f6b37c":10,"4112":[],"411c":[],"411d":[],"4124":[],"41322915":[],"413f":[],"417":0,"417ca995":[],"41891313174203":[],"4192":[],"41cf":[],"41dc":[],"41f2":[],"42137238":[],"422b":10,"4249832892949":[],"425000":4,"42663639956145":[],"4291":5,"4292":5,"4293":5,"4294":5,"4295":5,"42db":10,"42fb":[],"43076225311911":[],"430d":[],"431":3,"431f":[],"4321":0,"4323":[],"4346":[],"434fe2c9":[],"4358":[],"43781029194585":[],"438009":[],"4385":[],"4397":[],"43a7":[],"43c3":[],"44042383577023":[],"44158856139617":[],"441c":[],"44218970805414":[],"444141":[],"444575":[],"44684489":[],"44893732825162":[],"44923774688088":[],"4494":[],"44b5":[],"4567":[],"4570":[],"457273":[],"4575312":9,"457c":[],"4595724346351":[],"45bb":[],"45d1":[],"45da":[],"460000":4,"4610025":9,"4610795":9,"46177141138634":[],"46473890001363":[],"464b":[],"465292":[],"46656561":[],"4687b280":[],"46bd":[],"46e0":10,"471008":[],"471c5998e834":[],"475436088339379":13,"475f":[],"4785":[],"47985760094015":[],"47a9617c07f0":[],"47c5":[],"47d0":[],"47d8":[],"47e4":[],"47f2":[],"4809":[],"48197624712238":[],"4826":4,"4827":[],"485127007903577":[],"4855669223382983":13,"488088":[],"48906353198835":[],"48bd":[],"48ec":[],"48fd":10,"49077598":[],"4913":[],"497619":[],"4979":[],"49b3":[],"49e0":[],"49e3":[],"49ee":[],"4a44":[],"4a81":[],"4a813ce2e90a":[],"4ad1":[],"4b00":[],"4b54a7b31a35":[],"4b86":[],"4b9f":[],"4ba3":10,"4bac":[],"4bcd":[],"4be0":[],"4be2":[],"4c15":[],"4cd2":[],"4ce5":[],"4ce6":[],"4cf1":[],"4d07":[],"4d095a9bf3d1":[],"4d27":[],"4d56":[],"4d61":[],"4d7a":[],"4dd9b0dca8c6":[],"4e2a":[],"4e5428219d97":[],"4e9c":[],"4ec1":[],"4ee2":10,"4ef5":[],"4efec19b21b5":[],"4f02":[],"4f2a":[],"4f42d043":[],"4f65b33001b2":[],"4f7d":[],"4f87":[],"4f8d":[],"4fd8":[],"4fe3":[],"4th":2,"500":[4,5,11],"50000":10,"500000":[4,5],"50070230111274":[],"500a":[],"501852":[],"504":[],"5041691240370938":13,"50499343":9,"508880dfa229":[],"50th":4,"5104":[],"514097":[],"51575876538577":[],"516059467297646":10,"51726195":[],"51840029644295":[],"51b4c8f8":[],"51e324fc7254":[],"520614752375915":13,"52086912":[],"521":[],"52197825786087":[],"521c3ce2":[],"52579322606445":[],"525b8b1eaeea":[],"533333":13,"53333333333333":13,"533854":4,"533c3bcdb859":[],"53501671070511":[],"53644252":[],"539539450000001":3,"53992f56":[],"53f3d79e401b":[],"54296008":[],"545":[],"54974968":[],"550":2,"552":3,"554":[],"555103":[],"55fbbb17":[],"560":0,"56064620698095":[],"5619":10,"56195285702418":[],"562fe19a":[],"565964":4,"56648022":[],"567802669":[],"56780267":[],"567803":4,"570814":[],"572706":10,"573214":[],"575495":[],"575829":13,"57822143":[],"57826317":9,"580":11,"58064012543642":[],"58245324413445":[],"5849b013":[],"58656066":[],"5865607":[],"586561":4,"590":[],"593":0,"59da466cd70f":[],"5a56":[],"5ade8840c87a":[],"5baafabf4e64":[],"5be0c505":[],"5c86":[],"5ff9d08b":[],"5th":10,"600":11,"60000":[],"600000":[10,13],"600px":11,"601312462441985":[],"60137":[],"60220688933644":[],"60453660558522":[],"6061293361559":[],"607289":[],"609":9,"60994147639813":[],"60d1731417b7":10,"610":9,"610000":4,"613248":[],"614723":4,"61472303":[],"61485677847188":[],"615192":[],"616137":13,"617321676258056":[],"61cf3d61":[],"62034303842519":[],"6213":4,"624":3,"62454125":9,"625":2,"6250671":9,"62783751403471":[],"627949":4,"6282":[],"6287":[],"62891818":[],"62940661917412":[],"63132613":[],"634":3,"63409284352663":[],"635":3,"635000":4,"63674390464717":[],"63868541943094":[],"639498":4,"63949845":[],"6394985":[],"64324782236976":[],"644854":13,"645512":[],"64762696886464":[],"649363":[],"65281f7c":[],"65579917":[],"655825":[],"65689746":[],"65965696157481":[],"660335028063474e":13,"662":3,"66243843":[],"666666":[],"667":6,"66th":6,"672807771136599":[],"675":4,"6769080804625":[],"677":19,"67866353":[],"679":4,"680":19,"683321":[],"68557758":[],"686729":[],"691012":[],"6927102":[],"69274315088595":[],"693064":[],"696153":[],"696f1210f32f":[],"697":19,"698084151486567":[],"6a39a87b":[],"6af2d884":[],"6b6d":[],"6d27a6aa":[],"6d76":[],"6dce":[],"700000":13,"701":19,"70452372014537":[],"71000":4,"7183":3,"7189413":[],"72131458056907":[],"72231603326618":[],"72309191953751":[],"730527":[],"73294549":[],"7337826a4685":[],"737963401065596":[],"73923279574828":[],"73935987456358":[],"740000":4,"74076720425171":[],"74108686825797":[],"7486295728919304":13,"750":4,"750000":[4,5,13],"751886920458144":[],"754921071175007":[],"75578486208585":13,"75595700703582":[],"756":[],"75th":[4,5],"7608":[],"76084160057138":10,"7640042619421049":[],"76431b569704":[],"766bd3cf":[],"77032319783652":[],"771594":[],"772600263582405":[],"774918":13,"77668129338701":[],"779712104491898":[],"77d39d61":[],"7804075289401982":4,"783555899725448":[],"785000":4,"785299ab":[],"791a5ddf":[],"79278116248614":[],"79341":[],"795":9,"7a5eb02c":[],"7ac79054":[],"7b0f24b5":[],"7bf862e18266":[],"7c840be9d48d":[],"7cc6":[],"80000":[],"800000":4,"80014239905985":[],"80019238829465":10,"800723274871205":[],"8009":[],"803261864666974":[],"803408":[],"8044":[],"80647656290003":[],"809191273277598":[],"812312089763958":[],"813457":[],"816":4,"8162365060002428":4,"8162867394895984":4,"81642051634484":4,"8165214368885028":4,"816e":[],"816f0097":[],"817":4,"817956":[],"81c5":[],"820431":10,"82077526421523":10,"825":0,"82606624234876":[],"8290da4f":[],"8291":[],"82933440280782":[],"8307d73c":[],"83157847854275":[],"83247670748433":[],"834512987013":4,"83576592197558":[],"835814885142842":[],"839394488381325":[],"839742":[],"839e":[],"840000":4,"840608":[],"8420":[],"844":[],"8442":0,"84421513791415":13,"84423407802443":[],"844a":[],"84590715647336":[],"8461":[],"84675":[],"8482816486408":[],"84896776":[],"8492":[],"84928176192457":[],"84974823943259":[],"84a8":[],"850000":4,"85050485644803":[],"851574":4,"85422699369396":[],"854708":[],"854854723088971":[],"85786322613146":[],"8591313907397005":13,"85946705720211":[],"85th":10,"860011":[],"87155738":[],"8744":[],"8778708198222":[],"88000072":[],"881944541256573":[],"883333":4,"883954":[],"8845":[],"88721883751387":[],"88839966861514":[],"888659358775472":[],"89336360043855":[],"89470711547601":[],"8949158625700564":10,"894915862570057":10,"89887521":[],"898a0657":[],"89951468":[],"899515":4,"89dd":[],"8a13e497":10,"8a53":[],"8a5a":[],"8aa8":[],"8ad32b31":[],"8b18bf777617":[],"8b1feb08ff56":[],"8b4b":[],"8b92":[],"8bc0bee751cd":[],"8bc15872bd39":[],"8bff":[],"8c61":[],"8c7826c2":[],"8c83":[],"8c88":[],"8d863ff96bb1":13,"8ef1":[],"8f9f":[],"900000":13,"90165863":[],"903384":4,"9033840374657273":4,"903442":4,"90344244":[],"903442442":[],"90588738":[],"90664850568817":[],"908712":[],"9094019658612525":4,"90944411144406":[],"91071823807542":[],"91074326085924":4,"91148659677621":[],"911521":[],"9118":[],"91427124":[],"91430247004234":[],"91648970627955":[],"916b":[],"91802174213913":[],"919674372409807":[],"91fc":[],"92214ba0":[],"925":3,"92849522":[],"928e":[],"92af":[],"92cb7f49":[],"933d9d08126b":[],"93634803022955":[],"936367228048367":[],"938b":[],"9396542167696":[],"93991042213851":[],"93f6003e1b4":[],"94183567":9,"942":6,"94282227653469":[],"9429":[],"9457":[],"950":4,"950000":4,"9508686092602991":13,"9508925471332":[],"95297701":[],"954451":[],"958188":[],"95996398":10,"959964":13,"96020126":10,"965200":4,"966636":10,"96b7":[],"96c7dabb":10,"96f0":[],"96f1":[],"97034388284968":[],"970716532238185":[],"970c":[],"9718168904958":4,"974ee0c2":[],"975":[10,11],"97547627985463":[],"976923":4,"977535679139253":[],"97873526":13,"97892393180497":[],"980000":13,"981176628376893":[],"9818297813044":[],"98200487":[],"98236724494748":[],"98343998090161":[],"984159075050805":[],"98597539":9,"98612011347937":[],"9880771":[],"989422804014327":9,"98ae8990":[],"990351575034936":[],"99093646801164":[],"990c0101f8fd":[],"9913048":[],"9957":[],"998942":13,"999":6,"9998":9,"99998":9,"9999983396649719":13,"9999999999999999":4,"99ef":[],"9ae08854":10,"9bda":[],"9bf47972c6dd":[],"9c0c":[],"9d40":[],"9d4cea9d":[],"9db7f240":[],"9de9":[],"9eddbab0c553":[],"9f95":[],"9fe7cbb7":[],"9th":[],"\u0105":11,"\u03b1":[],"\u03b8":[9,11],"\u03bc":9,"\u03c3":9,"\u3164\u3164":4,"\ufb01le":[],"\ufb01nd":[],"\ufb01nding":[],"\ufb01rst":[],"abstract":[1,2,3,4,10,11],"break":[1,4,5,6,9,13],"case":[0,1,2,3,5,6,8,9,10,11,13],"catch":[1,3,11],"class":[0,1,2,3,4,8,9,10,13],"default":[1,3,4,5,10,13],"di\ufb00er":[],"export":[3,5],"final":[0,1,3,4,5,9,10,11,13],"float":[2,3,4],"function":[0,1,2,3,4,5,7,9,10,13,20],"import":[0,1,2,4,5,6,8,9,10,11,13],"int":[3,10],"k\u00fchberger":19,"long":[1,2,3,4,5,9,10,11,13],"new":[0,1,2,3,4,5,6,7,8,9,10,11,13],"null":[1,13],"public":[1,5,9,10,19],"return":[1,2,3,4,5,9,10,11],"short":[0,1,3,4,5,6,7,9,10,11],"speci\ufb01":[],"speci\ufb01c":[],"stu\ufb00":[],"super":7,"switch":[4,5,8,9,13],"throw":[5,9],"true":[0,1,2,3,4,5,6,7,8,9,10,11,13],"try":[0,1,2,3,4,5,6,7,8,9,10,11,13],"typeof":[],"var":[4,10,13],"while":[0,1,2,3,4,5,6,8,9,11,13],AND:3,Adding:9,Age:1,And:[0,1,2,3,4,5,6,7,8,10,11,13],Being:1,Bus:1,But:[0,1,2,3,4,5,6,7,8,9,10,11,13],DOS:[],Doing:[7,10],Drs:13,Eyes:1,FOR:[],For:[0,1,2,3,4,5,6,7,8,9,10,11,13],Going:[5,8],His:11,IDE:2,IQs:10,IVs:1,Its:13,Los:1,Not:[0,1,2,3,4,9,10,11,13],One:[1,2,3,4,5,6,9,10,11,19],SAS:[],Such:5,That:[0,1,2,3,4,5,6,7,8,9,10,11,13],The:[2,3,5,6,7,8,19,20],Their:8,Then:[1,3,4,5,7,13],There:[1,2,3,4,5,6,7,8,9,10,11,13],These:[0,1,2,3,4,5,6,9,10,13],USE:[],Use:4,Used:4,Useful:[],Using:[0,5,10,11,13],WITH:[],Will:1,With:[0,3,4,5,6,7,9,10,11,13],Yes:[3,8,10,11,13],_95:[],__init__:3,_between:[],_blank:[],_class_:[],_cohen:[],_data:[],_degre:[],_factors_:[],_formula_:[],_gener:[],_independ:[],_lists_:[],_method:[],_method_:[],_mode_:[],_nonparametr:[],_one:[],_pair:[],_pool:[],_rel:13,_remove_:[],_shapiro:[],_standard:[],_student:[],_two:[],_type_:[],_welch:[],_within:[],_work:[],_workspace_:[],a13ba159:[],a15:[],a187:[],a18d91877a5a:[],a1ac068357bc:[],a1b3a402:[],a1c7:[],a1eb:[],a21faee3aca1:3,a232:[],a28c:[],a2a7:[],a2d93521:[],a372e0c08cdb:[],a3e3:[],a451:[],a460:[],a491:[],a4bb:[],a4f388fc:10,a54b:[],a632:10,a656:[],a66:[],a6d1:[],a736:[],a74f0709e759:[],a849:[],a8d84704:[],a910:[],a98172db74f9:[],a9e3:[],a9ed06cbf49a:[],a_i:13,aa5fdd172618:[],aa69:[],aaaand:3,aad:4,ab13:[],ab34:[],ab9b70e376da:[],abandon:10,abbrevi:[3,4,5],abc:[0,4],abe1:[],abil:[1,3,7,10,11],abl:[0,1,2,3,4,5,7,8,10,11,13],ablin:[],about:[0,2,4,5,6,8,9,10,11,13],abov:[0,1,2,3,4,5,6,8,9,10,11,13],abroad:11,abs:[4,6,11],absenc:[],absolut:[1,2,3,6,11,13],absurdli:2,abx:[],ac00:[],ac717453c36f:[],academ:[9,10],accept:[1,3,4,8,9,10,11],access:[2,3,10,20],accident:[1,9,13],accommod:13,accompani:[],accomplish:7,accord:[4,11,13],account:[0,1,10,11,13],accur:[1,2,4,10],accuraci:13,accus:1,achiev:[3,4,5,6,9,10,13],acknowledg:1,acronym:[4,10],across:[0,1,2,4,8,9,10,11,13],act:[1,2,10,13],action:[4,11],activ:1,actual:[0,1,2,3,4,5,6,7,8,9,10,11,13],ad4328bb:[],ad65:[],ad6858b801e8:[],ada84:[1,19],ada:[],adair1984:[],adair:19,adapt:[7,13,20],add:[1,2,3,4,5,7,9,10,13],addarrow:[],adddistplot:[],added:[1,2,3,5,11,13],addict:0,adding:[3,13],addit:[1,2,3,4,13],addition:13,address:[0,1,3,4,10,11],adelaid:[4,5,9,10],adjac:11,adjust:[5,10,13],administ:[1,4,10],administr:9,admiss:[0,19],admit:[0,8,11,13],admittedli:[9,10],adopt:[1,6,7,9,10,11,13],adult:[1,6],advanc:[0,1,2,7,9,11,13],advantag:[1,3,4,9,10,11],advers:1,adversari:11,advic:[3,5,6,11,13],advis:6,advisedli:3,ae1d:[],ae2b:[],ae370d91032a:[],aesthet:5,af05d5f5cd6f:[],af86:[],af9a:[],afd1:[],affair:[0,1,10],affect:[4,11],afford:[0,4,9],afl24:[],afl2smal:5,afl:[4,5,13],afl_fig:4,afl_finalist:[4,5],afl_margin:[4,5],aflsmal:[],aflsmall2:5,afraid:[1,8],after:[0,1,2,3,4,5,7,8,9,10,11,13],afterward:[1,11],again:[1,2,3,4,5,7,8,9,10,11,13],against:[0,4,5,6,8,13],age:[1,3,6,10],agent:9,ages:[7,10],agg:9,aggreg:[0,1],ago:2,agpp:[],agre:[0,1,4,6,8,9,11],agreement:[0,6],agricultur:13,aha:5,ahead:[1,4,10,11],aim:[1,8,9],aka:3,ala:5,alan:1,albeit:7,alert:11,alex:[],algebra:10,alic:1,alien:9,align:[],aliv:[4,10],all:[0,1,2,3,4,5,6,7,8,9,10,11,13],alleg:1,allevi:[5,10],alloc:13,allow:[0,1,2,3,4,5,6,9,10,11,13],allud:[],almost:[0,1,2,3,4,5,6,9,10,11,13],alon:[0,13],along:[0,2,4,5,9,11,13,20],alongsid:3,alp:9,alpha:[9,10,11,13],alphabet:4,alreadi:[1,2,3,4,5,6,7,8,9,10,11,13],alright:5,also:[0,1,2,3,4,5,6,7,9,10,11,13,20],alter:[1,9,10,11],altern:[1,2,10,13],although:[1,2,3,4,5,6,7,9,10,11,13],alwai:[0,1,2,3,4,5,9,10,11,13],amass:10,amaz:[0,8,9],amazingli:11,ambigu:[4,11,13],american:19,among:[0,1,9,10,11,13],amount:[0,1,4,5,10,11,13],amstat:10,anaconda:[2,3],analog:[5,11,13],analogu:4,analys:[0,1,2,4,5,6,7,9,11],analysi:[0,1,2,3,4,5,6,7,8,9,10,11,13,19],analyst:[4,5,7],analyz:2,anastasia:13,anchovi:1,andrei:9,andrew:1,anecdot:0,angel:1,anger:2,angri:0,ani:[0,1,2,3,4,5,6,7,8,9,11,13],anim:19,annihil:5,annoi:[1,2,4,5,7,10,11,13],annot:11,annoyingli:1,anoth:[0,1,3,4,5,6,7,8,9,10,11,13],anova:[6,13],ans73:[4,19],anscomb:[4,19],anscombe1973:[],anscombe1973graph:[],anscombesquartet:[],answer:[0,1,2,3,4,5,8,9,10,11,13],ant:[],anti:1,anxieti:[1,13],anxifre:4,anymor:[3,4,9,13],anyon:[1,4,7,10,11],anyth:[0,1,2,3,4,5,6,7,8,9,10,11,13],anywai:[2,5,9,13],anywher:[2,3,4,6,10],apart:[1,4],api:10,apolog:11,appar:[0,1,5,13],appeal:[1,5,11],appear:[1,2,3,4,6,9,10,11,13],append:[4,10,11],appl:[3,4,13],appli:[0,1,4,5,6,7,9,10,13,19],applic:[0,2,5,9,10,11],appreci:2,apprehens:1,approach:[0,1,4,5,7,8,9,10,11,13,20],appropri:[4,9,10,11,13],approxim:[0,1,2,4,6,9,10,11,13],arang:[9,11,13],arbitrari:[0,1,4,9,13],arduino:9,area:[0,1,5,9,11,13],aren:[0,1,2,3,4,5,6,7,8,9,10,11,13],arg:[],argblock:[],argu:[1,2,3,4,10,11,13],arguabl:[2,10],arguement:5,argument:[0,1,4,5,6,9,11,13],aris:[0,1,4,6,9,10,13],arisen:1,arithmet:2,arm:[4,10],around:[1,3,4,5,7,8,9,10,11,13],arrai:[1,4,6,10,11,13],arrang:5,arrest:9,arriv:[1,9,10,11],arrog:13,arrow:[],arrowprop:11,arrowstyl:11,ars:13,arsen:9,art:[2,7,8],arthur:8,articl:[0,1,4,9],artifact:19,artifactu:1,artifici:2,artist:5,asbesto:1,ascend:4,ascii:[],ashlei:11,asia:19,asid:[0,1,10,11],ask:[0,1,2,3,4,6,7,8,9,10,11,13],aspect:[1,4,13],assembl:2,assert:[],assess:[4,11,13],asset:4,assign:[0,1,2,3,4,6,9,11,13],assist:3,associ:[2,6,8,9,10,11,13],associationtest:[],assum:[1,2,3,4,5,8,9,10,11,13],assumpt:[1,4,5,6,9,10,11],astyp:10,asymmetr:4,asymmetri:[4,11],ate:[],athough:4,atroc:4,attach:[4,6,9,10,11],attempt:[0,1,2,4,5,7,9,10,11,13],attent:[1,3,4,5,9,10,11],attest:11,attitud:1,attract:[4,5],attribut:[3,9,11],auckland:4,auden:0,audienc:[4,5],australia:[4,10,11],australian:[0,1,4,5,9,10,13],author:[0,1,3,5,13],authori:1,autism:1,autocorrect:2,autom:2,automat:[1,5,10],automaton:[2,4,5],auxiliari:1,avail:[2,3,4,5,8,10,11],availab:2,avenu:3,averag:[1,4,9,10,11,13],averagesamplemean:10,averagesamplesd:10,avoid:[1,4,7,8,11,13],awai:[1,4,5,6,7,8,9,10,11,13],awar:[1,2,3,4,5,9,10,11,13],awesom:[1,5,6,9,13],awesome2:13,awful:1,awkward:[1,4],awri:1,ax0:13,ax1:[4,5,13],ax1a:[],ax2:[4,5,9,10,13],ax2a:[],ax3:[4,5,13],ax3a:[],axes:[4,5,9,10,13],axessubplot:[4,5,9,10,11,13],axez:[],axhlin:[10,13],axi:[4,6,9,10,13],axisgrid:4,axvlin:[10,13],b0080e30277c:[],b025cd0c:[],b0e0:[],b158:[],b19c:[],b2bba8a2:[],b2e1:[],b38bcb52:[],b3a6b9f4780c:[],b458:[],b529:10,b5ac:10,b5d2:[],b62e:[],b65115ca7d06:[],b671:[],b6c2:[],b6cf:[],b6f8e681:[],b75b:[],b78a100a:[],b810:10,b85001a53d3d:[],b869:[],b926:10,b967:[],b970:[],ba17a6df509a:[],babbl:[],babi:[1,4,5],baby_sleep:4,back:[1,2,3,4,5,6,7,9,10,11,13],background:[2,3,5,9,10,13],backslash:3,backward:3,bad:[0,1,2,3,4,5,6,8,9,10,11,13],badli:11,bag:10,balanc:[1,4],banana:3,bank:4,bar:[4,9,10,11,13],bare:[10,11,13],bargraph:[],barplot:[5,9],barston:19,bartlebi:8,base:[0,1,2,3,4,5,6,9,10,11,13,19],baselin:0,basi:[0,1,8,9,10,11,13],basic:[0,1,2,3,4,5,6,10,11,13],basicprob:[],bastard:[],batch:8,bathroom:1,battlefield:11,bay:10,bayesian:10,bb7f:[],bba44a4097c2:[],bbox_inch:5,bd44:[],bd5f:[],bd60:[],bd6739b1e5e8:10,bd7d:[],bd95:[],bear:[3,10,11],beast:[3,4],becam:1,becaus:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],becauw:13,becom:[0,1,2,3,4,5,6,8,9,10,11,13],bed2:[],bedma:2,been:[1,2,3,4,5,6,7,8,9,10,11,13,20],befor:[0,1,2,3,4,5,6,7,8,9,10,11,13],begin:[1,3,5,7,9,10,11,13],beginn:7,behav:[1,3,4,9,13],behavior:19,behaviour:[0,1,2,4,5,9,10,13],behind:[1,2,3,4,5,7,11,13],being:[0,1,2,3,4,5,7,8,9,10,11,13],beings:[1,9,10],belief:[1,8,9,10,11,19],believ:[0,1,4,8,9,10,11,13],bell:[4,9,10],belong:[4,10],below:[0,1,2,3,4,5,9,10,11,13],ben:1,benchmark:4,benefit:7,bereft:13,berk:0,berkelei:0,berkeley2:0,berkelyplot:0,berklei:0,bernadett:13,bernoulli:10,besid:[1,4,9,10,13],best:[0,1,2,3,4,5,6,7,9,10,11,13],bet:[4,8,9,11],beta:[10,11],better:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],between:[0,1,3,4,5,6,7,8,9,10,13,19],beyond:[0,1,4,5,9,10,11,13],bezo:5,bf41:[],bf5c:[],bfae:[],bfbb:[],bfe1:[],bhoconnell75:[0,19],bia:[4,10,11,13,19],bianca:1,bias:[0,8,10,11,13],bickel1975:[],bickel1975sex:[],bickel:19,bicker:3,bicycl:1,big:[0,1,2,3,4,5,7,8,9,10,11,13],bigger:[1,4,9,10,11,13],biggest:[0,1,4,11,13],bill:[4,10],billion:1,bin:[5,6,9,11],binari:[10,13],binom:[9,11],binom_test:11,binomi:11,binomial1:[],binomial2a:[],binomial2b:[],binomplot:[],binwidth:[4,9,10,11,13],biolog:[0,10],biologi:1,birth:1,birthdai:[],bit:[0,1,3,4,5,6,7,8,9,10,11,13,20],bitter:8,bizarr:[0,1,10],black:[4,5,7,9,10,11,13],blackston:11,blah:10,blame:2,blank:9,blatantli:0,blind:1,blindingli:[1,10],blindli:13,block2:[],block:[2,8,13],blood:13,bloodi:[0,1,10,13],blowout:4,blue:[1,5,9,10,13],bluntli:11,blur:1,board:9,bob:4,bodi:4,boi:11,boil:3,bold:[0,1,4],bolt:[],bonu:2,book:[0,1,4,5,6,7,8,9,10,11,13,20],bookdown:20,booksal:3,booksales2:[],booksalescsv:[],bookshelf:9,booktab:[],boost:11,border:[10,13],borderlin:11,bore:[0,1,4,5,6,11],boredom:1,boringli:9,born:[1,10],boston:19,both:[0,1,2,3,4,5,6,8,9,10,11,13],bother:[0,4,7,11,13],bottom:[4,5,9,10,13],bought:13,bound:1,boundari:[6,11],box1987:13,box:[4,10,11],boxprop:5,brace:3,bracket:[2,3,4],brag:5,brain:[1,11],brainer:[],branch:9,bread:1,breakdown:[0,9,10],breath:5,breweri:13,bridg:0,brief:[8,9,11,13],briefli:[1,3,4,11,13],bright:[1,11],bring:[1,3,13],brisban:[4,5],broad:[1,5,6,9,13],broader:[0,1,5,10,13],broadli:[3,4],broken:[4,11],brooklyn:0,brown:1,brows:[],browser:3,brs:10,brush:[5,13],bug:[],bui:2,build:[3,10],built:[1,2,3,5,9,10,11,13],bulldog:[4,5],bump:5,bunch:[3,4,7,9,10,13],bundl:[],bureaucrat:11,buri:[0,4],burn:[4,13],burt:1,bus:[1,9],busi:[1,2,4,5],butter:1,button:[2,9],bypass:[],c39bad49718f:[],c6e5fe9a:[],c8e34490:[],cake:[],calcul:[3,5,6,9,11,13],calculu:9,california:0,call:[0,1,2,3,4,5,6,7,9,10,11,13],calm:1,cambridg:19,came:[0,3,4,5,8,9,10,11],cameron:1,campbel:19,campbell1963:[],campbell2015experiment:[],can:[0,1,2,3,4,5,6,7,8,9,10,11,13],cancer:1,candid:[0,5],cannot:[2,3,4,5,8,9,10,11],canva:5,cap:[9,10,13],capabl:0,capac:[10,13],capprop:5,caption:[2,4],captur:[1,2,4,5,10,11],car:1,card:[9,11],care:[0,1,3,4,6,7,9,10,11,13],carefulli:[0,1,4,13],carlton:[4,5],carri:1,carriag:8,carrol:8,carv:6,casual:5,cat:[3,13],categor:[1,9,13],categori:[0,1,4,9,10,11],categoris:6,caught:11,caus:[0,1,4,5,7,10,11,13],causal:[1,5,11],caution:[4,6],caveat:[],cba:4,cbc482d2:[],cbt:4,cc834daf:[],cc8ea8bc:[],ccc:8,cd43:[],cdf:[11,13],cell:1,celsiu:1,censor:1,cent:9,center:[4,6,11,13],centr:[4,5,6,9],central:[11,13],centraltend:[],centuri:[0,9,11],certain:[0,2,4,5,6,8,9,10,11,13],certainli:[0,1,4,5,6,7,9,10,11,13],certainti:[10,11],certif:1,cetera:6,cex:[],cfa4aaec:[],challeng:13,chanc:[0,1,3,4,5,7,8,9,10,11,13],chang:[0,1,2,3,5,8,9,10,11,13],chapek9:[],chapter:[1,2,3,4,5,6,7,8,9,10,11,13],charact:[3,4,6,11,13],character:2,characteris:1,characterist:[0,1,4,9,10,11],charg:0,chart:[5,9],chdir:[0,4,5],cheap:1,cheapli:2,cheat:1,cheater:1,check:[1,4,8,9,11],checkmark:[1,13],chemic:1,chemist:13,chemistri:0,chi2:9,chi:[9,11,13,20],chi_square_20:9,chi_square_3:9,chi_square_data:9,chicago:1,chick:0,chico2:[],chico3:[],chico:13,chief:4,child:[1,4,6],childish:1,children:1,chill:4,chip:10,chisq:[],chisqdist:[],chisqreport:11,chisquar:[9,13],chitchat:1,choic:[0,1,3,4,5,6,8,9,11,13],cholera:5,choos:[0,1,2,3,4,5,6,11,13],chop:7,chose:[1,10,11],chosen:[1,2,4,5,10,11,13],chromosom:1,chunk:[5,9],ci_1:10,ci_2:10,ci_:[],ci_int:10,ci_upp:[],cigarett:[0,1],cimean:13,circ:1,circl:5,circumst:[4,11],cirep:10,citat:1,cite:[],citi:[4,9],claim:[1,7,9,10,11,13],clairvoy:11,clarifi:4,clariti:[4,9],classic:[0,1,4,5,19],classif:[1,6],clean:[1,4,5,7],cleaner:11,cleanli:13,cleanup:[],clear:[0,1,2,3,4,5,6,9,10,11,13],clearer:[4,5,13],clearli:[1,2,3,4,5,9,10,13],clever:[1,4,7,9,11,19],click:[3,9],climat:1,clinic:[0,1,4,10],clinical_trial_data:4,clinicaltri:[],clinicaltrial_old:[],clipboard:[],close:[0,1,2,4,5,6,9,10,11,13],closer:[1,2,3,4,10,11,13],cloud:[2,11],clt:[10,13],cltdemo:10,cltdemoa:10,cltdemob:10,cltdemoc:10,club:11,clue:4,clump:13,cluster:10,clutter:5,coars:6,coast:[4,5],coat:[1,10],code:[3,4,5,6,9,11,13,20],coder:5,coef:[],coerc:6,coercion:6,coffe:[],coffer:10,cognit:[1,10,19],coh88:[11,19],cohen1988:13,cohen:19,cohensd:13,cohensdinterpret:[],coher:[3,5,6],coin:[9,11],coin_flip:9,coincid:[0,4,11],col:[4,13],col_0:6,colab:2,cold:[4,10],colleagu:7,collect:[0,1,2,3,4,5,6,9,10,11],colleg:0,collingwood:[4,5],color:[4,9,10,11,13],colour:[1,10,11,13],coltot:6,column1:4,column2:4,column3:4,column:[3,4,5,6,13],com:[0,3,4,8,13],comb:7,combin:[3,4,5],come:[0,1,2,3,4,5,6,7,9,10,11,13,20],comfort:[1,2,4,5,6,9,10,11,13],comma:3,command:[3,4,5,6,7,9,11,13],commenc:0,comment:[7,13,19],commentari:[0,9],commerci:2,commiss:9,commission:9,commit:[0,20],commmand:[],common:[0,1,2,3,4,5,6,8,9,10,13],commonli:[3,4,9,13],commonplac:13,commonwealth:4,commsec:4,commun:[1,4],commut:[1,10],compact:[4,5,13],compani:[2,9,10,19],companion:8,compar:[0,4,9,10,11],comparison:[0,4,9,10,13],compel:0,compens:0,compet:0,competit:[4,8,11],complain:[],complement:[1,4],complet:[0,1,2,3,4,5,9,10,11],complex:[0,3,4,10],complianc:0,complic:[0,2,3,4,5,6,7,10,11,13],compon:1,comprehens:[1,3,4,6,10],compress:13,compromis:[],comput:[1,2,3,4,5,6,7,9,10,11,13],con:[],concat:9,conceal:5,conced:11,conceiv:11,concentr:20,concept:[0,1,2,4,5,6,7,9,10,11],conceptu:[1,2,9,10],concern:[0,1,2,3,4,10,11],concert:3,concis:10,conclud:[1,9,13],conclus:[0,1,8,10,13],concret:[1,2,4,10,11],condens:4,condescend:10,condit:[1,6,11,13],conduct:[1,9,10,11,13],confess:[],confid:[1,5,9,11,13],confidence_interv:13,confidence_level:13,configur:5,confin:6,confintrepl:10,confirm:[1,10,13],conflat:[11,13],conflict:19,confront:[0,1],confus:[1,2,3,4,5,9,10,11],congress:1,conjunct:7,connect:[1,3,9],connel:19,conscienti:1,conscious:2,consensu:9,consent:10,consequ:[0,1,2,3,4,5,8,9,10,11,13],conserv:2,consid:[1,4,5,6,7,9,10,13],consider:[],consist:[0,1,4,5,8,10,11,13],consol:[4,7],constant:[1,4,9,13],constitut:[0,5],constraint:[5,9,13],construct:[0,3,4,6,9,10,11],consul:5,consult:[1,11],consum:1,contact:[10,11],contain:[2,3,4,5,6,9,10,11,13],contempl:11,content:[6,7,9,11],context:[1,2,4,5,6,9,11,13],continu:[2,3,4,5,6,8,9,10,13],continuousdiscret:[],contradict:0,contradictori:[5,11],contrari:[0,8],contrast:[1,4,5,9,10,11,13],contribut:[4,19],contriv:11,control:[0,1,3,5,11,13],controversi:9,convei:[1,4,5,11],conveni:[3,5,6,10,11,13],convent:[1,3,4,5,10,11],convention:13,converg:9,convers:[7,8,9,11],convert:[2,4,5,10,11,13],convict:11,convinc:[0,1,9,10,11],cool:[4,11,13],coolest:4,cope:4,copi:[2,3,4,7,13],copiou:5,cor:4,core:[0,2,3,4],corner:[],corr:4,correct:[0,1,2,4,7,8,9,10,11,13],correctli:[1,4,11],correl:[5,11,13,19],correspond:[1,2,3,4,5,6,8,9,10,11,13],cosmet:[],could:[1,2,3,4,5,6,7,8,9,10,11,13],couldn:[1,4],counsel:[],count:[1,2,3,4,5,7,9,10,11,13],counterintuit:[0,9,10,11],countri:[2,4,10,11],countrysid:8,coupl:[1,4,6,10],courier:[],cours:[0,1,2,3,4,5,7,8,9,10,11,13,20],courtesi:10,cov:4,covari:[1,4],cover:[0,1,3,4,5,6,9,10,11,13],coverag:9,craft:5,craig:4,cram:[5,13],cramp:5,cran:[],cranni:0,crap:1,crash:9,crazi:[0,9],creat:[1,2,3,4,5,7,9,10,13,20],creation:19,creativ:[5,11],credibl:10,credit:[1,4],crime:[1,11],crimin:[1,11],crippl:2,crisi:11,crit1:[],crit2:[],crit3:[],crit4:[],crit:13,crithi:11,critic:[0,1,3,8,9,10,13],criticis:[1,10],critiqu:10,critlo:11,croker:[8,10],cromul:10,crop:13,cross:[4,13],crosstab:6,crucial:[3,10],crude:[1,10],cruel:0,cryptic:4,cs63:[1,19],csg63:[1,19],csv:[0,4,5,7,13],cube:[2,4],cubism:5,cue:4,cuebal:3,cultur:0,cumbersom:3,cumsum:9,cumul:13,cunningli:0,cup:9,cure:[1,13],curios:13,curiou:[0,1,4,6,13],curli:[3,9],current:[1,2,3,4,6,10,13],cursori:11,curv:[4,5,9,10,13],curvi:5,custom:[5,10],customari:5,customis:5,cut:[0,4,5,9],cute:1,cwc:8,cwd:[0,4,5],cynic:0,cyril:1,d009972d:[],d13d:[],d36a6fc3adb7:[],d5b41eeb:[],d5ee87d58c65:[],d69586c7f512:[],d9846b90f6de:[],d997af956b7d:[],dabbl:1,dae4:[],dai:[0,1,2,4,5,8,9,10,11,13],daisi:6,damag:1,damn:0,dampen:9,dan:[1,3,4,5,9],dan_grump:4,dan_sleep:4,danger:[1,2,10,11],daniel:[1,20],danrbook:3,dark:5,darken:11,dart:9,dash:[4,9,10],dashdot:10,data:[0,1,2,7,8,9,10,19],databas:[],datafram:[0,4,5,6,9,10,13],datahandl:[],datapoint:10,dataset2:4,dataset3:4,dataset:[3,4,10],datastructur:6,date:[1,2,4,11],datum:13,david:[1,8],dbb77e8ee3ad:[],dbinom:[],dchisq:[],ddition:2,ddof:4,dead:[1,10],deal:[1,2,3,4,5,10,13],dealt:[1,3],dear:0,death:[4,5],debug:[],decad:[9,11],deceiv:1,decemb:1,decent:[5,7],decid:[0,2,3,5,9,10,11,13],decim:[2,3],decis:[0,5,6,9],decisionmak:[],deck:9,declin:1,decreas:[1,4,10,13],dedic:3,deduct:[0,8],deem:[4,11],deep:[1,2,9,11],deeper:[0,1,11],deepli:[0,1,4,5,9],def:[9,10],defeat:[],defenc:13,defend:11,deficit:10,defin:[2,3,4,9,11,13],definit:[1,4,5,9,10,11,13],degfre:[9,13],degfree1:9,degfree2:9,degre:[1,5,9,10,11,13],degrees_freedom:13,del:3,deleg:6,delet:5,deliber:[0,1,10,11],delici:13,deliv:0,delta:13,delv:2,demand:[0,10],demographia:4,demolit:5,demoni:4,demonin:4,demonstr:[9,10,11],deni:4,denomin:13,denot:[1,2,4,9,10,11,13],dens:[],densiti:[5,13,20],depart:[0,13],departur:13,depend:[0,1,2,3,4,5,8,9,10,11,13],depict:[4,5,9,10,13],deprec:[],depress:1,depressingli:3,depth:[9,10],deriv:[1,4,9],describ:[1,3,5,6,8,9,10,11],describeimg:[],descript:[1,2,5,9,10,11,13],descriptor:1,descrstatsw:10,design:[0,3,5,9,10,11,13,19],desir:[1,5,6,7,9,10,11,13],desktop:[3,5],desper:[0,11],despin:[5,9,13],despit:[1,8,11],destin:10,destroi:[1,9],detach:[],detail:[0,1,2,3,4,5,7,9,10,11,13],detect:[1,5,13],determin:[0,1,4,5,6,10,11,13],determinist:9,dev:[4,13],devat:10,develop:[2,3,4,10],development:1,deviat:[2,9],devis:13,devoid:9,devot:[0,3,9],dewei:9,df0e8d462b68:[],df1:[],df2:[5,13],df_1:9,df_2:9,df_berk:0,df_clintrial:4,df_kurtosi:4,df_skew:4,df_wide:[],diagnosi:[13,19],diagnost:13,diagon:13,dialog:8,dialogu:[6,8],diamond:5,dice:9,dict:[4,11],dictat:0,dictionari:3,did:[0,1,2,3,4,5,7,8,9,10,11,13],didn:[1,2,4,8,9,11,13],die:9,died:1,diet:1,diff_mean:[],differ:[0,1,2,3,5,6,7,8,10,19],difference_:[],differentialattrit:1,difficult:[1,2,5,6,7,9,10,11,13],dig:[0,1,5,9,13],digit:5,digress:[3,9,11],dim:13,dinner:10,dinosaur:6,dip:5,dir:3,direct:[2,5,6,10,13],directli:[1,2,3,4,5,7,11],directori:3,directory_:[],disadvantag:[9,10,11],disaggreg:[0,1],disagr:[1,11],disagre:[0,1,6,11],disappear:[],discard:4,disciplin:[7,9],discov:[1,2,5,9,11,13],discredit:1,discrep:4,discret:[6,9,10,11],discrimin:[0,8,11,13],discuss:[0,1,2,3,4,5,6,7,8,9,10,11,13],diseas:1,dishonesti:1,disingenu:1,disk:3,dismal:2,dismiss:10,disobedi:1,disorganis:[],dispatch:[],dispatch_:[],displai:[0,1,3,4,5,9,10,11,13],displaystyl:13,displot:4,disput:11,disregard:5,dissect:8,dist:9,distanc:4,distant:5,distformula:[],distinct:[1,3,4,5,9,10,11,13],distinguish:[1,4,6,10],distort:[9,10],distract:11,distractor:0,distribut:[0,2,4,5,6,13,19,20],disturb:9,div:6,divd:3,dive:[5,10],divers:8,divid:[1,3,4,5,6,9,10,11,13],divin:11,divis:[1,2,10],dnorm:[],doc:3,doctrin:9,document:[3,4,5,7,9,10],doe:[0,1,2,4,5,6,7,8,11,13],doesn:[0,1,2,3,4,5,6,8,9,10,11,13],dog:[],dogma:11,dogmat:11,doing:[0,1,2,3,4,5,6,7,8,9,10,11,13],dollar:[1,2,4,9],domain:[11,13],domest:4,domin:[1,5,9,11],don:[0,1,2,3,4,5,6,7,8,9,11,13],donald:19,donat:2,done:[0,1,2,3,4,5,6,7,8,9,11,13],doomsay:4,dot:[4,5,11,13],doubl:1,doubt:[2,4,11],down:[1,2,3,4,5,7,9,10,11,13],download:2,downsid:5,dozen:[1,4,7,9],drag:4,dramat:[4,9],drastic:10,draw:[1,3,4,8,9,10,13],drawback:[6,7,11,13],drawn:[1,5,9,10,13],dread:4,dream:7,dreari:13,drew:[9,10],dri:4,drill:4,drive:[3,8],drop:[1,4,11,13],dropdown:[],dropna:4,drown:0,drug:[1,2,4,13],dtype:[3,4],dubiou:1,due:[1,4,9,10,11,13],dug:2,duke:8,dull:4,dumb:[1,2,4],dure:[1,2,3,4,5,6,7],duti:1,dwell:3,dwr_all:[],e170:[],e1e3cb50:[],e1e8:[],e225286a3f01:[],e272:[],e30f90a78005:[],e3c489536945:[],e3cf:[],e592:[],e635:[],e811:[],e835:[],e87501de0303:[],e8d3c3d8:[],ea51:[],each:[0,1,2,3,4,5,6,7,8,9,10,11,13],eaerlier:4,earli:[3,11,13],earlier:[0,1,2,4,5,9,10,11,13],earn:4,earth:13,earthquak:1,eas:5,easi:[0,1,2,3,4,5,6,7,8,10,11,13],easier:[1,2,3,4,5,6,10,11,13],easiest:[2,4,5,9,11,13],easili:[1,4,5,10,13],ebp83:[0,19],ecb93e1df467:[],echo:13,ecologist:10,econom:[0,10],economist:[4,10],ecosystem:[2,3],edg:[3,6,11],edgecolor:[5,13],edit:[5,19,20],editor:[],editori:[],educ:[0,1,2,10,11],edward:5,eecaa5d710b8:[],ef46:[],ef63631a:[],effect:[0,2,3,4,5,7,9,10,19],effects:[],effici:10,effort:[1,3,4,9,11,13],ehp11:[9,19],eight:4,either:[0,1,2,3,4,5,6,8,9,10,11,13],elbow:5,elect:9,elector:9,electrodynam:0,electron:0,eleg:[4,5],elegantli:4,element:[3,4,6,10,11,13],elementari:[4,9],elimin:[1,5,8,11],ell02:[4,19],ell10:[11,19],elli:19,ellis2010:[],ellman2002:[],ellman2002soviet:[],ellman:[4,19],els:[0,1,3,4,5,6,7,8,9,10,11,13],elsewher:2,email:[7,13],embarrass:1,emerg:11,emili:20,emphasis:[11,13],emphat:11,emphcol:[],emphcollight:13,emphgrei:13,emploi:1,employe:13,empti:5,enamor:3,enclos:2,encod:[],encompass:11,encount:[0,1,3,4,8,9,10,11,13],encourag:[10,13,20],end:[0,1,2,3,4,5,7,8,9,10,11,13],endlessli:3,endors:[1,9],endpoint:6,enemi:5,energi:1,enforc:2,engag:[0,1,7],engin:[0,1,4],english:[0,7,9,11],enjoi:5,enorm:9,enough:[0,1,2,3,4,7,8,9,10,11,13],enrol:[0,1,9,10],enrypt:2,ensconc:11,ensur:[1,4,9,11,13],entail:9,enter:[2,4,13],enterpris:9,entertain:[9,13],entir:[0,1,3,4,5,6,8,9,10,11,13],entireti:11,entiti:[1,9,10],entri:4,enumer:[4,10],envir:[],enviro:3,environ:[1,2,3,10,11],eol:[],eps:[],equal:[0,2,4,5,6,8,9,10,11,13],equal_var:13,equat:[2,4,9,10,13],equip:[],equival:[1,2,4,6,9,11,13],eras:1,erich:19,erkelei:19,erlbaum:19,error:[0,1,2,3,5,10,13],errortyp:[],escap:[8,11],escapechar:[],ese:[4,13],esp:11,espcrit:11,especi:[0,1,2,3,4,5,6,10,11,13],espestim:[],essai:1,essenc:[1,10,11],essendon:[4,5],essenti:[6,7,10,11,13,19],est:[],establish:[1,11],estim:[1,4,5,8,11],estimatorbia:10,etc:[1,3,4,10],ethan:[3,4,5,20],ethanwe:[4,13],ethic:1,europ:[3,19],european:13,eval:[],evalu:[0,2,4],evan:19,evans1983:[],evans1983conflict:[],evans2000:[],evas:11,even:[0,1,2,3,4,5,6,7,8,9,10,11,13],event:[1,4,9,10,11,13],eventnam:9,eventu:[1,2,9,10,11],ever:[0,1,2,3,8,9,11],everi:[0,1,3,4,5,8,9,10,11,13],everybodi:9,everydai:[1,4,7,8,9,10,11],everyon:[0,1,2,4,9,10,11,13],everyth:[0,1,3,4,5,6,7,8,9,10,13],everywher:[0,7,9,13],evid:[0,1,5,8,10,11,13],evidentiari:[1,10,11],evil:[0,1],evolut:0,exact:[1,3,4,5,9,10,11,13],exactli:[0,1,2,3,4,5,6,7,9,10,11,13],exagger:[4,13],exam:[1,4,11],examin:[1,4],exampl:[0,1,2,3,5,6,8,9,10,11],exce:4,excel:[2,3,4,9],except:[1,2,4,7,8,10,11,13],excit:[3,4,20],exclud:5,exclus:[10,11],excus:4,execut:[2,3,4,7],exercis:[1,5,7,10,13],exhaust:[9,10],exist:[0,1,3,4,5,6,8,9,11,13],exorbit:2,exot:13,exp:9,expand:[1,4,5,11,13],expans:[],expect:[0,1,2,3,4,5,8,9,10,11,13],expens:[0,2,4,10,13],experi:[0,1,2,3,4,5,7,8,9,10,11],experienc:1,experiment:[0,9,10,11,13,19],experimet:13,expert:[2,8],explain:[0,1,2,3,4,5,7,8,9,10,11,13],explan:[1,6,9,11],explanatori:3,explicit:[4,6,10,11,13],explicitli:[1,4,10,11],exploit:9,explor:[4,7,13],exploratori:5,exponenti:9,expos:[],express:[1,4,5,6,10],expt:[],exrem:11,extend:[2,3,5,8,9,10],extens:[2,3,11],extent:[0,1,2,4,5,6,13],extra:[2,3,4,10],extract:[2,3,4,6,9],extraordinarili:10,extrasensori:11,extravers:4,extravert:4,extrem:[1,2,4,5,9,20],eye:[1,3,5],eyebal:4,eyes:[5,6,10],eyewit:1,eyman:19,f26b7c7c1738:[],f29618a6:[],f40c1ed9:[],f45fd3136913:[],f50fea121ed3:[],f533bacecaf1:[],f63645209886:[],f636:10,f66e:[],f6eba0bb:[],f6f1d9db:[],f70d:[],f7430430:[],f795f0294715:[],f796:[],f7b7794b1a5e:10,f7f2c52b62e9:[],f927:[],f_3_20:9,fa58:[],fabric:1,face:[4,5,9,11],facecolor:5,facet_kw:4,facetgrid:4,fact:[0,1,2,4,5,6,8,9,10,11,13],facto:5,factor:[1,3,5,11,13],factori:[1,9],factual:8,fail:[0,1,11,13],failur:1,faint:5,fair:[0,4,5,8,9,11,13],fairi:13,fairli:[0,1,2,3,4,5,6,10,11,13],faith:[0,1,2,4],fake:10,fall:[0,1,4,5,9,10,11,13],fallaci:1,fals:[0,1,4,5,9,10,11,13,19],falsifi:1,falsification:11,falsificationist:11,familar:3,famili:4,familiar:[1,2,4,13],famou:[1,5,10],famous:[10,11],fan:20,fanci:[1,3,4,8],fancier:[],fancyp:[],fantast:20,far:[0,2,3,4,5,6,7,8,9,10,11,13,20],fashion:[1,2,3,4,5,10,11],fast:[1,3,4],faster:[1,2],father:6,favor:5,favorit:4,favour:[0,2,4,8,11],favourit:[0,1,13],fb50:[],fba6:[],fc9b:[],fd6e:[],fdd8cff7b4db:[],fdist:9,fe89e91d93ec:[],fear:[0,7],feasibl:11,feat:[0,1],featur:[1,5,9],feb6a0b4c657:[],februari:1,feder:9,fee:[0,2],feed:[3,4,13],feedback:1,feel:[0,1,2,4,5,8,9,10,11,13],fell:[1,11],felt:[0,3,13],femal:[0,1,10,11],few:[0,1,2,3,4,5,7,8,9,10,11,13],fewer:[1,3,5,11],ff01:[],ff2f:[],fibonacci:5,fictiti:[4,10],field:[0,1,2,4,8,9,11],fig1:4,fig2:[4,11],fig:[0,4,5,9,10,11,13],fig_powerfunct:[],figsiz:[4,5,9,10,13],figur:[0,1,2,3,4,5,6,7,9,10,11,13,20],figwidth:11,file1:4,file2:4,file:[2,4,6,7,10,13],filenam:[],fileopen:[],filepanel:[],filepath:[],filesystem:[],fill:[1,4,9,13],fill_between:[9,13],finalist:[4,5],find:[0,1,2,4,5,6,7,8,9,10,11,13,19],finder:[],fine:[10,11],finish:[1,3,4,5,7,10,11],finit:10,finnegan:1,firmli:[1,4,5],first:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],firstli:[0,1,2,4,5,6,8,9,10,11,13],fis22:[9,19],fischer:4,fisher1922b:[],fisher:[1,4,9,13,19],fist:2,fit:[3,5,9],five:[1,3,4,5,9,10,13],fix:[3,4,5,7,10,11,13],fixat:0,flat:[4,10],flatten:13,flavor:20,flavour:9,flaw:[1,2,4,11],flexibl:[3,11],flick:6,flip:[6,9,11],flood:1,floor:3,floppi:[],flourish:5,fluctuat:[1,9],fluke:[9,13],focu:[0,1,3,4,5,7,9,11,13],focus:[0,1,2,4,7,8,9,10,13],folder:3,folk:[4,10,11],follow:[0,1,2,3,4,5,6,9,10,11,13],fond:11,font:13,font_scal:5,fool:7,foolish:13,foot:1,footbal:[4,5],footi:[4,5],footnot:4,forbid:9,forc:[1,2,4,5,7,10,11],forego:20,foreign:[],forev:[3,10],forget:[0,1,2,3,4,5,9],forgiv:10,forgiven:0,forgot:4,forgotten:[3,13],form:[1,2,3,4,5,7,8,9,10,11,13],formal:[1,11,13],formalis:[9,10,13],format:[3,4,5,6,7,9,13],former:[],formula1:[],formula2:[],formula3:[],formula4:[],formula:[3,4,9,13],formulat:[],forth:7,fortuit:10,fortun:[4,5,9,10,11,13],forward:3,found:[0,1,3,9,10,11],foundat:[3,9,10,11,19],founder:10,four:[1,2,3,4,6,9,10,11],fourth:[1,3,4,8,9],fox2011:[],frac:[4,9,10,13],fraction:11,frailti:0,frame:[6,10,13],frame_:[],framework:11,franci:[],francisco:4,frankli:[0,4,5],fraudul:1,freak:[1,4],free:[2,4,8,11,20],freedom:[9,10,11,13],freedom_:[],freeli:2,freez:1,fremantl:[4,5],french:11,freq:4,frequenc:[4,5,6,9,10,11],frequent:[0,4],frequentist:10,frequentist_prob:[],frequentist_probability_fig:9,frequentistprob:[],friedman:10,friend:[1,4,5,7,9,13],friendli:4,frill:[],fritz:19,from:[0,1,2,4,5,6,7,8,9,11,19,20],front:13,fruit:13,frustrat:[2,4,9,11,13],fulfil:1,full:[2,3,4,5,11,13],fulli:0,fun:[3,4,20],function_:[],fund:0,fundament:[0,1,4,9,11,13],funni:[4,5,10,13],further:[0,1,2,5,6,7,9,11],futur:[5,10,11],fuzzi:1,fyi:2,gage:19,gain:[1,4,9],gambl:[8,9],gambler:8,game:[4,5,8,9,10],garbag:[],garden:[6,13],gather:[6,9],gaussian:9,gave:[1,2,3,4,13],gdata:[],geelong:[4,5],gelman2006:[],gelman:19,gender:[0,1],gener:[0,1,2,3,4,5,6,9,10,11,13],generalis:[0,1,4,10],gentli:1,genuin:[1,2,4,5,6,9,11,13],georg:11,geq:11,ges:3,get:[0,1,3,5,6,7,8,9,10,11,13,20],get_x:11,get_xaxi:13,get_xlim:13,get_yaxi:13,get_ylim:13,getcwd:[0,4,5],getwd:[],gibberish:[1,4,10,11],gibson:7,gist:[4,9],github:[3,4,5],githubusercont:[4,13],give:[1,2,3,4,5,6,8,9,10,11,13],giveawai:1,given:[0,1,2,3,4,5,8,9,10,11,13],glad:4,glanc:[5,13],glass:13,glib:0,global:[],globalenv:[],gloriou:11,glorious:0,gloss:[4,10],glossi:2,glue:[0,2,4,9,11,13],goal:[2,3,4,5,6,7,8,9,10,11,13],god:[9,11],goe:[1,2,4,8,9,10,11,13],going:[0,1,2,3,4,5,7,8,9,10,11,13],gold:10,gone:[1,3,13,20],good:[0,1,2,3,5,6,7,8,9,10,11,13],goodman:8,googl:[1,2],gosset:13,got:[0,1,2,3,4,7,10,11,13],gotten:[4,11,20],govern:[0,1,9,10],gplot:[],grab:5,grade:[1,4,13],grade_test1:13,grade_test2:13,grade_time1:[],grade_time2:[],graduat:19,grain:6,grammat:1,grant:[9,11],graph:[4,9,10,19],graphic:[4,10,13],grappl:4,grasp:[6,9],graviti:4,great:[1,3,4,5,9,10,13],greater:[0,1,9,11,13],greatli:5,greek:[4,11],green:1,grei:[9,11],grid:[],grossli:1,groucho:13,ground:[0,1,3,4,5,9,11],groundwork:[],group:[0,1,3,4,6,10,11,13],groupdescript:[],grow:[1,9,10],grown:10,grump:[4,5],grump_fig:4,grumpi:[4,5],grumpier:4,gs06:[11,19],guarante:[1,4,10,11,13],guess:[0,1,3,4,5,8,9,10,11,13],gui:[0,13],guid:[1,4,10,11,13,19],guidelin:1,guilti:11,guin:13,gulag:4,gullibl:1,gunpoint:10,gut:[0,5,11],gzcon:[],gzfile:[],h_0:[11,13],h_1:[11,13],habit:[2,4,5],had:[0,1,2,3,4,5,6,8,9,10,11,13],haha:4,half:[1,3,4,5,8,9,10,11,13],hallmark:8,hammel:19,hammer:1,han:[1,19],hand:[0,1,2,4,5,7,9,10,11,13],handful:3,handi:[1,2,3,9,10,11,13],handili:4,handl:[0,1,2,5,7,10,13],hang:3,hangman:6,happen:[0,1,2,3,4,5,6,8,9,10,11,13],happi:[0,5,9,11,13],happiest:5,happili:11,hard:[0,1,2,3,4,5,7,9,10,11,13,20],harder:[0,1,2,13],hardest:[5,6,10],hardli:[0,11],harken:5,harm:3,harmon:11,harpo2:[],harpo:13,harpo_summari:13,harpohist:[],harpohistanastasia:[],harpohistbernadett:[],harri:2,harvard:0,harvardmagazin:0,has:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],hasn:[7,10],hast:19,hat:[4,10,13],hatch:13,hate:11,have:[0,1,2,3,4,5,6,7,8,9,11,13,20],haven:[3,4,5,6,8,9,10,11],hawthorn:[1,4,5,19],head:[3,4,5,9,10,11,13],headach:5,header1:[],header2:[],header3:[],header:[],headlin:1,hear:[1,4,13],heart:[4,9,10],heat:1,heavi:[0,13],heavier:[9,13],heavili:[0,1,5,10,13],heck:[0,4,13],hedg:13,hedges1981:13,hedges1985:13,hei:[5,10],height:[3,4,5,9,13],held:0,hello:3,help:[0,1,2,3,4,5,6,7,9,10,11,13,20],henc:[],henri:19,her:[1,4,9,11,13],herald:9,here:[0,1,2,3,4,5,6,9,10,11,13],heresi:11,heterogen:1,heurist:[4,13],hidden:[0,1,10,11],hide:[1,4,5,9,10],high:[0,1,2,4,5,9,10,11,13],higher:[0,1,4,9,11,13],highest:10,highli:[1,2,3,4,5,6,10,11,13],highlight:10,hilari:1,hill:[8,10,19],him:[1,8],himself:[10,13],hint:[4,11],hirtum:0,his:[1,4,8,9,10,11,13],hist:13,histdata:[],histgram:10,histogram1:[],histogram:[4,9,10,13],histor:11,histori:[0,4,7,9,11,13,19],histplot:[4,5,9,10,11,13],hit:[2,9],hline:13,hoax:1,hoc:1,hold:[1,5,7,9,10],hole:2,holm:8,holt:19,home:[1,3,4,5,6,10],homework:[],homogen:[1,13],homoscedast:13,honest:[0,1,4,5,9,11,13],honesti:[1,4],honestli:[6,9,13],honour:9,hood:[4,5],hope:[0,8,9,10,11,13],hopefulli:[0,3],horizont:[5,9],horribl:[0,1,5],hors:[1,19],hot04:[1,19],hot:[1,9],hothersal:19,hothersall2004:[],houghton:19,hour:[1,4,5,7],hourli:2,hous:[4,10,13],household:4,housingpriceexampl:0,how:[0,1,2,4,5,6,7,8,11,13,20],howev:[0,1,2,3,4,5,6,7,9,10,11,13],howl:2,href:[],htm:4,html:[0,3,8,10],http:[0,1,3,4,8,10,13],hue:[0,9,10],huei:9,huge:[0,4,5,10,13,20],human:[0,1,2,3,4,7,8,9,10,11,13,19],humbl:[1,3,5],hume:8,hunch:1,hundr:[1,7,13],hurt:[9,10],hwang:1,hybrid:11,hypothei:11,hypothes:[1,10,13],hypothesi:[1,4,8,10],hypothesis:[1,13],hypothesistest:[],hypothet:[0,4,9],icon:[],idea:[0,1,2,3,4,5,6,7,8,9,10,11,13],ideal:[1,4,7,10,11],ident:[0,3,4,5,6,9,10,11,13],identif:1,identifi:[1,2,11,13],ideolog:[9,11],idiosyncrat:[0,1],idiot:4,ifels:13,ignor:[1,2,3,4,6,7,9,11,13],ill:3,illeg:5,illus:1,illusori:[10,13],illustr:[1,3,4,5,6,9,10,11,13],imag:7,imagin:[1,3,4,5,9,10,11,13],imbal:1,img:[10,13],immedi:[0,3,4,10],immun:1,immut:11,impact:[0,9],imparti:[0,1],impenetr:9,implaus:[10,11],implement:[11,13],impli:[2,4,9,10,11,13],implicit:8,implicitli:1,importantli:[0,9,10,11],impos:[5,7],imposs:[1,7,8,9,10,11],imprecis:10,impress:[5,7,9,11,13],impressionist:5,improv:[0,1,4,13],inabl:8,inaccur:10,inadvert:5,inappropri:11,incid:[1,5],includ:[0,1,3,4,5,7,9,10,11,13],include_graph:[10,13],includng:[],incom:4,incomplet:10,incomprehens:4,inconsist:[8,11],inconveni:1,incorrect:[0,4,9,10],incorrectli:[0,11],increas:[1,4,5,9,10,13],increasingli:[9,11],incred:[0,1,5,13],inde:[2,4,5,8,10,13],indenpend:[],indent:2,independ:[1,9,10,11],independentsamplesttest:[],index:[4,5,6],indian:1,indic:[1,2,3,4,5,6,10,11,13],indici:5,individu:[0,1,2,4,9,10,13],induct:[0,8,10,11],indulg:11,industri:10,industrialis:10,inert:1,inevit:[7,9],inexpens:0,inf:[],infam:4,infant:4,infer:[0,4,8,9,10],inferenti:[4,9,10,11],infin:[9,10],infinit:[9,10,13],inflex:2,influenc:[0,1,2,4,11],info:[3,5],inform:[0,1,2,3,4,5,6,8,9,10,11,13],infti:[9,10],infuriatingli:0,inher:6,initi:[4,5],injur:9,inlin:9,innoc:11,inplac:5,input:[2,3,4,6,9,13],inquisitori:11,insan:[1,10],insert:10,insid:[1,2,3,4,5,6,7,10,11,13],insight:4,insist:[5,8],insofar:[1,10],inspect:[1,4,5,9,10,13],inspir:5,instal:[],instanc:[0,1,2,3,4,5,6,9,10,11,13],instead:[0,2,3,4,5,6,7,9,10,11,13],instinct:[0,10],instruct:[1,2,7,10],instructor:2,insult:[1,10],insur:5,int64:[3,4],int_a:9,integ:[2,3,5,10],integr:[2,9],intellect:11,intellectu:9,intellig:[1,2,4,9,10,11],intend:[1,4,5,9,10,13],intent:[1,4,10,11],inter:1,interact:[1,2,3,7,11],intercept:[],interest:[0,1,2,3,4,5,6,7,9,10,11,13],interestd:13,interfac:2,interim:2,intern:4,internalis:1,internet:[0,1,5,11],interpol:[],interpret:[0,1,2,5,6,7,9,11,13,19],interquartil:5,intertwin:[0,1],interv:[4,5,6,13],interval_:[],intimid:[],intoler:11,intro:[13,20],introduc:[1,2,3,4,5,6,10,11],introducd:4,introduct:[3,4,6,8],introducton:5,introductori:[0,3,9,10,11,13],intror:[],intrus:10,intuit:[0,1,5,9,10,11,13],intut:10,invalid:[0,1],invas:1,invent:[1,2,4,11],invest:1,investig:[1,4],invis:[],invok:[],involv:[0,1,2,4,5,7,9,10,11,13],ioa05:[1,19],ioannidi:19,ioannidis2005:[],ioff:[],ipynb:3,ipython:[2,3,13],iq1:10,iq2:10,iq3:10,iq4:10,iq5:10,iq_10000:10,iq_100:10,iq_10:10,iq_1:10,iq_2:10,iqdist:10,iqdista:10,iqr:4,iqsampa:10,iqsampb:10,iqsampc:10,iron:[2,10],irrelev:[0,5,9],irrespect:[9,10],irritatingli:4,isn:[0,1,2,3,4,5,6,8,9,10,11,13],isol:[1,7],issu:[0,1,3,4,5,6,8,10],italicis:1,item:[1,3,4],iteself:13,its:[0,1,2,3,4,5,6,7,8,9,10,11,13],itself:[0,1,2,3,4,5,6,9,10,11,13,19],ivis:2,ivori:11,jacob:10,jame:4,jamovi:20,jane:4,januari:[1,7],jargon:[11,13],jasp:20,jean:9,jerk:[8,10,13],jerzi:11,jitter:5,job:[0,3,4,5,9,11,20],joe:[],john:[0,5,9,10,19],joke:[1,13],josef:4,journal:[1,11,19],journalist:[0,4],joyzepam:[],jpg:[],jse:10,judg:[4,11,13],judgement:0,judgment:5,julian:19,jump:[0,4,13],jungl:9,junk:[],jupyt:[2,3,5,7],jurist:11,just:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],justic:1,justif:[8,10,11],justifi:[8,10,11],kabl:[],kage:3,kahneman1973:[],kahneman:19,kardinia:5,karl:11,karyotyp:1,kate:4,kde:10,keep:[0,1,2,3,4,5,6,7,8,9,10,11,13],keeper:3,kei:[1,2,3,4,5,6,9,11,13],kept:[1,8,9,11],kernal:5,keyboard:2,keyn:10,keynes1923:10,keyword:[],kid:[1,4],kilda:[4,5],kill:[1,3,5,13],kind:[0,1,2,3,4,5,6,7,9,10,11,13,20],kinder:13,kindli:[3,5],kingdom:13,klinfelt:1,knew:[1,10,13],knitr:[10,13],know:[0,1,2,4,5,6,7,8,9,10,11,13,20],knowledg:[0,1,4,5,9,10,13],known:[0,1,4,5,7,8,9,10,11,13],kolmogorov:[9,13],kolomogorov:13,koth:20,kt73:[1,19],kuhberger2014:[],kuhbergerfs14:[1,19],kurtosi:13,kurtosis_fig:4,kurtosisdata:4,kurtosisdata_ncurv:4,lab:[1,10],label:[1,2,4,6,9,10,13],labelbottom:9,labelleft:9,labelright:9,labeltop:9,labor:9,labori:[4,10],lack:[1,5],lai:[],laid:6,land:9,languag:[0,2,3,4,5,7,9,11,13],larg:[0,1,3,4,5,9,11,13],larger:[1,4,5,9,10,11,13],largest:[0,1,4,5,10],larri:9,last:[0,1,2,3,4,5,7,8,9,10,11,13],late:1,later:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],latter:5,launceston:5,law:[0,1,9,11],lawlargenumb:10,lawrenc:19,lawyer:11,lazi:[2,3,9,10,13],laziest:4,lccc:[],lead:[1,3,4,9,10,11,13],leagu:[4,5],learn:[1,2,3,4,5,7,9,10,11,13],learningstatisticswithr:[],least:[0,1,2,3,4,5,7,9,10,11,13],leav:[0,1,2,4,5,7,9,10,11,13],lectur:[0,1,2,4,10,13],led:[4,8],lee:19,left:[1,2,3,4,5,8,9,10,11,13],leftmost:5,legal:[0,11],legitim:[2,3,4,5,7,8,9,10,11],leh11:[11,19],lehmann2011:[],lehmann:19,len:[3,4,10,13],lender:4,length:[1,4,10],lengthi:[],leprechaun:13,leptokurt:4,leq:[10,11],less:[0,1,3,4,5,7,8,9,10,11,13],lesser:[],lesson:[0,4],let:[0,1,2,3,4,5,6,7,8,9,10,11,13],letter:[4,10,11],level:[0,2,4,5,7,8,9,10,11,13],leven:13,lewi:8,liar:9,liber:2,librari:[2,4,5,6,19],licenc:2,licens:[2,20],lie:[1,4,5,9,10,13],lied:[4,9,10],lies:[4,9,10,11],life:[1,2,5,6,8,9,10,11,13],lifetim:4,light:[1,2,4,11,13],lightgrei:11,lightli:6,like:[0,1,2,3,4,5,6,7,8,9,10,11,13],likert:[1,6,9],limit:[1,2,3,4,9,13],line2d:[10,13],line:[0,1,2,3,4,5,6,7,9,10,11,13,20],linear:[4,6,10],lineplot:[4,9,10,11,13],linestyl:[9,10,13],link:[2,3,4,11],linspac:[4,9,10,13],linux:3,lion:9,list:[1,2,4,5,6,9,10,11,13],listen:[2,11,13],liter:[2,4,6,9,11],literatur:[0,4,8,9,11],littl:[0,1,2,3,4,6,7,8,9,10,11,13],live:[0,1,2,4,10,11,13],liz:[],lme4:[],lmplot:0,lmtest:[],load:[4,5,7,13],load_dataset:10,loadingcsv:[],loan:4,loc:[3,4,5,6,10,13],local:10,locat:[1,3,4,5,6,9],lock:9,log:[7,11],logic:[0,1,4,11,19],logico:11,lone:4,longer:[1,2,3,4,5,8,11,13],longest:6,longtowid:[],look:[0,1,2,4,5,6,7,8,9,10,11,13],loop:[3,10],loos:3,lopsid:5,lose:[1,4,5,6,8,9],loss:[4,5,8,10],lost:[1,4,8],lot:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],lotteri:9,loui:3,louie:9,love:[0,1,2,3,4,5,9,11,20],lover:3,low:[0,1,4,5,10,11,13],lower:[0,4,10,11,13],lower_area:13,lowest:[4,6],lsquo:[],lsr:[3,20],lty:[],luck:[1,8],lucki:1,luckili:[3,4,5,13],ludwig:11,lunch:[8,10],lung:1,luxuri:0,lwd:[],lying:[0,4],lyre:0,mac:3,machin:[2,3,7,8,9],maco:3,mad:[0,1,4,11],made:[0,1,2,3,4,5,6,8,9,10,11,13],magic:[0,1,3,4,7,10,11,13],magnifi:11,magnitud:[4,11],mai:[1,2,3,4,5,6,7,9,10,11,13],maiden:9,mail:1,main:[1,2,3,4,5,9,10,13],mainli:[],major:[1,4,5,9,10,11,13],make:[0,1,2,3,4,5,6,7,9,10,13],makeov:5,maker:1,makka:6,male:[0,1,10,11],malici:1,man:[1,4],manag:[1,2,3,9],mani:[0,1,2,3,4,5,6,8,9,10,11,13],manipul:[1,6],mann:13,manner:1,manual:[4,5,13],map:[5,9,11],margin:[4,5,6,11,13],margin_titl:4,mark:[1,3,5,13],marker:[],market:4,mash:11,mask:[],mass:11,massacr:5,massiv:[0,2,4,5,13],match:[1,4,9,13],materi:[1,5,9,10],math:[1,2,3,4,9,10,13],mathbf:4,mathemat:[1,2,4,7,9,10,11,13,19],mathematician:[9,10],mathes:[],mathfunc:[],matlab:5,matplotlib:[3,4,5,9,10,11,13],matrix:4,matter:[0,1,2,3,4,9,11,13],max:[4,5,10,13],maximis:11,maximum:[4,9,10,13],mayb:[1,2,3,4,5,6,7,8,9,11,13],maynard:10,mbox:[4,9,10,11,13],mcar:4,mcgrath2006:13,mcgraw:19,mcnemar:13,mean:[0,2,3,5,6,8,11],mean_diff:13,mean_numb:[],meaning:[1,4,5,6,9,11,13],meaningfulli:[4,11],meaningless:[1,4,6,9,11],meanmedian:[],meant:[1,2,11,13],meantim:1,measur:[2,10,11,13,19],mechan:[3,6,9,11,13],med:19,median:5,medianprop:5,medic:13,mee67:[9,19],meehl1967:[],meehl:[9,19],meet:[5,11],melbourn:[4,5],member:10,membership:[],memori:[10,13,19],men:10,mental:[3,13],mention:[0,1,2,3,4,5,9,10,13],menu:[],mere:[1,4,8,10],merg:6,merit:[0,1,13],merri:9,mersenn:[],mesokurt:4,mess:[0,5,11],messag:[2,3,7],messi:[0,2,4,6,11],messier:[1,4,13],met:[5,13],meta:19,metaphor:5,meteorologist:9,method:[1,2,3,4,5,6,9,10,11,13],methodolog:[0,1,10,11,19],mfcol:[],mfg:[],mfrow:[],mgp:[],mice:11,michael:19,microsoft:2,mid:[4,6],middl:[0,1,4,5,6,9,13],midpoint:6,mifflin:19,might:[0,1,2,3,4,5,6,8,9,10,11,13],milan:9,million:[4,10],millisecond:[],min:[4,5,9,10,13],mind:[0,1,3,4,5,7,9,10,11],mindless:[2,4,5,9],mindlessli:2,mine:[1,4,5,8,9,13],mini:5,minim:3,minimis:[1,11,13],minimum:[4,10,13],minitab:[],minor:[4,9,10,11],minu:[4,13],minut:[1,5],miracl:10,miscdatahandl:6,miscellan:6,misdesign:1,mishmash:11,mislead:[1,4,5,10,11],misrepres:1,misrepresent:1,miss:[1,5,6,10],mistak:[0,1,4,5,7,9],mistaken:11,mister:9,mix:3,mmr:1,mock:4,modal:4,mode:[],model:[2,3,5,9,10,11],moder:[4,13],modern:[1,11],modest:[5,10],modul:[2,3,4,9,13],modulenotfounderror:[],modulu:2,moe:9,moment:[0,1,2,3,4,6,7,9,10,11,13],mondai:10,monei:[0,2,10,11],monoton:4,month:[1,3,7],mood:[1,4],moral:[1,5],more:[1,2,4,7,8,9,10,11,13],moreov:[1,10,11],morn:[4,9],moron:1,mortem:1,mortgag:4,most:[0,1,2,3,4,5,6,8,9,11,13,19,20],mostli:[0,1,3,4,5,7,9,13],motiv:[6,8,10,11,13],mous:[],move:[1,3,4,6,9,11,13],movement:1,mu0:[],mu1:13,mu2:13,mu_0:13,mu_1:13,mu_2:13,mu_d:13,mu_nul:13,mu_o:13,much:[0,1,2,3,4,5,6,7,8,9,11,13,20],multcomp:[],multi:4,multipl:[1,2,3,4,9,10,11,13],multipli:[1,3,4,9,10,13],multitud:4,multivariate_norm:4,mum:1,mumbl:13,murder:1,murki:0,music:[11,13],must:[1,2,3,5,9,10,11,13],my_column:3,my_fil:3,my_row:3,mydataanalysi:7,myfil:[],mynewdatanalaysi:7,myrawdata:7,myself:[1,2,4,5,6,7,9,11,13],myst_nb:[0,4,9,11,13],mysteri:[4,10,13],n_1:13,n_2:13,n_h:9,nabig:[],naisbitt:0,naiv:[0,1],name:[0,1,2,3,4,5,9,10,11,13],nameerror:[3,13],namespac:[],nan:4,nana:[],nancorr:4,nanmax:4,nanmean:4,nanmedian:4,nanmin:4,nanpercentil:4,nanstd:4,nansum:4,narrow:[1,5,9,10,11],narrowest:10,nasmal:[],nasti:[3,7],nastier:13,nathaniel:19,nation:[0,4,10],nationwid:4,nativ:13,natur:[1,2,3,4,6,9,10,11,13],naturalist:1,nav3:[],navarro:20,navig:3,navigationr:[],nearli:[0,6,7,10],neat:[1,6],neater:4,neatli:4,necess:11,necessari:[1,2,8,9,13],necessarili:[0,1,4,8,9,11,13],need:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],neg:[1,3,4,6,9],neglect:11,neglig:[4,13],neither:[1,4,11],nelson:8,neq:[9,11,13],nerd:[4,7],nerdstuff:[],nervou:1,net:4,network:10,neuropsycholog:1,neutral:[0,11],never:[1,2,4,5,6,7,8,9,10,11,13],nevertheless:[1,4,5,9,10,11,13],newborn:1,newcom:3,newer:[],newspap:[0,1,9],next:[0,1,2,3,4,5,7,8,9,10,11,13],nhst:11,nhstimg:[],nhstmess:[],nice:[0,1,2,4,5,7,9,10,11,13],nicer:3,nicknam:1,niggl:4,night:6,nightgarden2:[],nightgarden:[],nilli:[3,6],nine:3,nitpick:13,nlme:[],no_mean:10,nobodi:4,nomean_low:10,nomean_up:10,nomin:[4,5,13],non:[4,5,6,7,9,10,11],none:[0,1,4,5,8,9,10,13],nonparametr:13,nonsens:[2,5,13],nook:0,nor:[0,1,9,11],norm:[4,9,10,13],normal:[1,3,4,5,6,10],normal_a:9,normal_b:9,normal_c:9,normal_d:9,normal_fig:9,normaldist:13,normals2:[],normals_fig2:9,normals_fig:9,normdist:[],normmean:[],normsd:[],norri:4,north:[4,5],notabl:[],notat:[4,9,10,11,13],note1:[],note3:[],note4:[],note5:[],note8:[],note:[0,1,3,4,5,6,7,9,10,13],notebook:[2,3,5,7],noth:[1,2,5,7,8,9,10,11,13],notic:[0,1,2,3,4,5,6,7,9,10,11,13],notion:[1,4,5,9],notori:2,notwithstand:9,novel:4,novemb:9,novic:[],now:[0,1,2,3,4,6,7,8,9,10,11,13,20],nsw:9,number:[0,1,3,4,5,6,7,9,11,13],number_of_flip:9,number_of_head:9,numer:[1,3,4,10,13],numpi:[2,3,4,6,9,10,11,13],numr:[],numref:9,nut:[],nutrient:13,nutshel:9,obedi:[1,2],obei:9,object:[0,4,9,10,11],oblig:1,oblivi:4,obnoxi:5,obs:[],obscur:[5,11],observ:[1,4,5,6,7,8,9,10,11,13],obsess:[0,11],obsolet:[],obstetrician:1,obtain:[1,4,6,9,10,11,13],obviou:[1,2,3,4,5,8,9,10,11,13],obvious:[1,2,3,8,9,10,11,13,20],occas:[1,4,5],occasion:[1,2,4],occult:11,occur:[1,3,4,9,11,13],ocean:10,oct:9,odd:[1,3,4,7,11,13],odditi:11,oddli:[0,3,9],off:[1,2,3,4,5,8,9,10,13,20],offer:[0,4,9,10,11],offic:1,offici:1,offspr:9,often:[0,1,2,3,4,5,6,7,8,9,10,11,13],oil:5,okaaaai:[],okai:[0,1,2,3,4,8,9,10,11,13],old:[1,2,3,4,6,7,9,10,11],old_par:[],older:[1,6],oldest:1,omit:[4,9,11,13],onc:[1,2,3,4,5,6,7,8,9,11,13],one:[0,1,2,3,4,5,6,7,8,9,10],onecorplot:[],ones:[0,1,3,4,8,9,10,11,13],onesamplettest:13,onesid:11,onesidedtest:[],onk:6,onli:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],onlin:[2,3],onscreen:[],onto:[5,9,11],ontolog:11,opaqu:13,open:[1,2,3,4,6,10,20],openoffic:[],openscreen:[],oper:[1,2,3,4,10],operation:[10,11],operationalis:[4,9,10,11],opinion:[0,1,5,6,9,10,11],opinion_direct:6,opinion_strength:6,oppon:9,opportun:4,oppos:[4,5,9,13],opposit:[1,5,11,13],option:[1,3,5,6,10,13],orang:[4,5,13],order:[0,1,3,4,5,6,7,9,10,11,13],orderedfactor:[],ordin:[4,13],ordinari:[4,13],ore:10,org:[1,3,10],organ:[1,5,10],organis:[3,4,6,8,13],orient:[1,5],origin:[0,2,3,4,5,6,7,9,10,11,13],orthodox:[9,11],osten:19,ostens:1,other:[0,2,3,4,5,6,7,8,10,11,13],otherdist:[],otherwis:[2,3,10,11,13],ought:[0,1,2,3,5,7,10],our:[0,1,2,3,4,5,6,8,9,10,11,13],ourselv:[0,1,3,4,5,9,10,11,13],out:[0,1,2,4,5,6,7,8,9,10,11,13,20],outbreak:5,outcom:[3,5,6,8,9,10,11,13],outdat:1,outier:5,outlier:[4,5],outlin:[4,9,10,13],output:[2,3,4,5,6,7,11,13],outright:1,outset:[],outsid:[1,5,7,9,11],outward:9,outweigh:[2,11],over:[0,1,3,4,5,7,8,9,10,11,13],overal:[0,5,9,11,13],overarch:5,overflow:10,overgeneralis:13,overjoi:13,overlai:[5,9,20],overlap:[5,11],overli:1,overlook:0,overrid:5,overrul:2,oversampl:10,oversimplif:[4,13],oversimplifi:[],overview:[1,13],overwhelm:[5,11],overwrit:2,own:[0,1,2,3,4,5,6,8,9,10,11,13],p_lower:13,p_upper:13,p_valu:13,packag:[2,3,4,10,11],packageinstal:[],packet:4,page:[1,4],pai:[1,2,4,9,11],paid:8,pain:[5,13],paint:5,painter:5,pair:[4,9,10],pairedsamplesttest:13,pairedta:13,pairedtb:[],pairedtc:[],pakka:6,panda:[0,2,3,4,5,6,9,10,13],pandas_mad:4,panel:[4,5,9,10,13],panic:13,pant:9,pantsprob:[],paper:[0,1,2,5,7,9,10,11],par:[],paradox:[1,19],paragraph:[1,4],parallel:[1,10],paralys:[],paramet:[4,9,11,13],paranoid:4,parent:[0,1,4,5,13],parenthes:3,parenthood2:4,parenthood:[4,5],park:5,part:[0,1,2,3,4,5,8,9,10,11,13],partial:[1,4,5,13],particip:[1,2,3,5,10,11,13],particp:3,particular:[0,1,2,3,4,5,6,8,9,10,11,13],particularli:[5,11],partit:[],partli:11,pass:[1,3,4,7,9],passag:10,passeng:9,passion:[],past:[2,3,4,10],patch:11,path:[0,4,5,9,10,13],pathet:0,pathin:[4,5],pathlib:[0,4,5],patholog:[11,13],patienc:9,patient:[7,13],pattern:[1,4,9,11],paul:9,payoff:[5,9],pbinom:[],pch:13,pchisq:[],pdf:[3,4,9,10,13],pdistnam:[],peacock:19,peak:[3,5,9],pear:3,pearson:[4,11],pearsonr:4,pecentag:0,peculiar:3,pedagog:[2,7],pedant:10,peek:[3,4,5],pencil:2,peopl:[0,1,2,3,4,5,6,7,8,9,10,11,13],per:[1,2,5,9],perceiv:[0,7],percent:0,percentag:[0,4,9],percentil:[4,5,6,10],percept:11,perfect:[0,1,4,5,10,11,13],perfectli:[0,1,3,4,8,9,10,11,13],perform:[1,2,5,11,13],perhap:[1,3,4,5,9,10,11,13],period:[4,5],person:[0,1,2,3,4,5,7,8,9,10,11,13],perspect:[0,1,2,3,4,7,9,10,11],pertain:[5,10],pervas:1,pervers:0,peski:4,pfu11:[1,19],pfungst1911:[],pfungst:19,phase:5,phenomenon:[1,10],philosoph:[6,8,19],philosophi:[8,11,19],philosphicu:11,phone:10,phosphoru:13,phrase:[1,4,7,10,11,13],physic:[0,1,3,9,19],physicist:[0,1],physiolog:1,pick:[0,1,4,5,9,10,11,13],picki:[4,5],pictur:[4,5,8,9,10,11],piec:[1,2,3,4,5,7,9,10,11,13],piltdown:1,pip:6,pirat:1,piri:10,pitch:8,place:[0,1,2,3,5,9,11,13],placebo:4,plai:[0,1,3,4,5,8,9,10],plain:[3,5],planet:0,plant:13,platform:5,platykurt:4,plausibl:[0,8,9,10,13],player:5,pleas:[5,10,13],pleasant:9,plo:19,plot:[0,4,6,7,9,11],plot_:[],plothist:[],plotmean:[],ploton:13,plotsampl:10,plottwo:13,plt:[4,5,9,10,11,13],plu:[1,9,13],png:[5,10,13],pnorm:[],pocket:9,poem:0,poet:0,point:[0,1,2,3,4,5,6,7,8,9,10,11,13],pointestim:10,pointi:4,pointless:1,pointplot:[10,13],poison:[10,13],polic:1,polici:1,polit:13,politician:1,poll:[9,10],pollard:19,pollut:10,polygon:[],pool:[],pooled_s:[],poor:[1,4,11,13],poorli:5,pop:[8,9,10,11,13],popmean:13,popper:11,popul:[1,4,9,11,13],populac:1,popular:[1,4,5],population_mean:10,population_vari:4,port:[4,5,10],portabl:9,portion:3,portrai:11,pos:[],pose:1,posit:[1,2,3,4,6,10,11],possess:[1,10],possibl:[1,2,3,4,5,7,8,9,10,11,13],post:[0,1],poster:5,postgradu:[0,13],potato:1,potent:9,potenti:[5,9],potsdam:4,potter:2,poulat:10,pow:2,power:[0,1,3,4,5,6,9,10,13,19],powerfunct:11,powerfunction_fig:[],powerfunctionsampl:11,ppf:[10,11],practic:[0,1,2,3,4,5,6,9,10,13],practition:11,pragmat:[0,1,6,9],pragmatist:9,pre:[0,1],preambl:5,precis:[1,2,3,4,9,10,11,13],pred1:[],pred2:[],pred:[],predict:[4,8,9,11,13,19],predictor:[6,13],prefer:[0,1,2,3,4,6,7,8,9,11,13],prefix:[],prelud:10,prematur:1,premier:9,premis:0,prepar:[5,13],present:[0,1,2,4,5,10],preserv:4,presidenti:1,press:[2,19],pressur:[1,13],presum:[2,4,5],presumpt:11,pretend:[1,4,5,9,13],pretti:[0,1,2,3,4,5,6,7,8,9,10,11,13],prevent:[1,7,9,11],preview:[],previou:[1,2,3,4,9,10,11,13],previous:[4,6,13],price:[2,4],primari:[2,9],primarili:[3,8,13],prime:13,principl:[0,1,11,13],print:[2,3,4,9,10,11,13],printout:[],prior:0,prioriti:[0,2],privaci:0,privat:11,prize:[4,13],prob:[9,11],prob_low:11,probabail:10,probabilist:[9,10],probabl:[0,1,2,3,4,5,6,8,10,13,20],probablil:11,problem:[0,1,2,3,4,5,6,7,8,9,10,11,20],probmean:10,probrul:[],probstat:[],proce:9,procedur:[6,9,10,11],process:[0,1,2,3,4,5,7,10,11,13],proclaim:11,produc:[0,1,2,4,5,6,9,10,11,13],product:[1,4],profession:[0,1],professor:11,profil:1,profit:[],program:[2,3,5,9],progress:9,project:[0,1,6,10],projecthom:[10,13],promin:[3,11],promis:[9,10],promot:11,prompt:[7,13],pronoun:1,proof:[10,11],prop:[],proper:[1,3,4,8,10,11],properli:[1,5,9,11,20],properti:[2,4,5,9,10],prophesi:1,proport:[0,1,4,9,11],proportion_head:9,proposit:6,proprietari:2,prosecutor:11,protect:11,prove:[1,10,11],provid:[0,1,3,4,5,6,8,9,10,11,13],proviso:11,proxi:1,pseudo:4,pseudonym:13,pseudorandom:9,psych:[0,1,13],psychic:11,psycholog:[2,3,4,5,8,9,10,11,13,19],psychologist:[0,1,8,9,10,11,13],psychotherapi:1,psychotrop:2,publish:[1,2,6,9,11,13,19],pull:[1,4,10,13],pump:5,punish:11,pure:[0,1,4,7,9,10,11,13],purer:13,purpos:[1,2,3,4,5,6,9,10,11,13],purview:0,push:2,put:[0,1,2,3,4,5,6,9,10,11,20],puzzl:8,pvaltabl:[],pvalu:13,pvarianc:4,pygmalion:1,pyplot:[4,5,9,10,11,13],pythob:[],python3:3,python:[6,7,11],pythonbook:[3,4,5,13],qbinom:11,qchisq:[],qcut:6,qnorm:[],qq1a:13,qq1b:13,qq2a:13,qq2b:13,qq2c:13,qq2d:13,qqnorm:13,qualifi:[0,5],qualit:[1,4,9,11,13],qualiti:[0,5,9],quantifi:[4,10,11],quantil:[4,6,10,13],quantit:[4,5],quantiti:[4,9,11,13],quantum:0,quarter:4,quarterli:[],quartet:4,quartil:[4,5],quasi:[1,19],queensland:1,quest:2,question:[0,1,2,4,5,6,9,10,11,13],questionnair:[0,1,4,9],quick:[1,3,4,5,11,13],quickli:[2,3,5,9],quiet:11,quietli:9,quirk:[2,3,11],quit:[0,1,2,3,4,5,6,7,9,10,11,13],quizz:0,quot:[0,4,5,6,7,8,10,11],quotat:[3,5],race:1,racket:2,rain:9,rais:[1,2,4,10],rake:9,ralph:4,rambl:10,ramp:10,ran:[0,10,11,13],random:[1,4,5,7,9,11,13,20],randomis:[1,11],randomli:[1,4,9,10,13],rang:[2,5,6,9,10,11],rank:[0,4],rankcorr:4,rankcorrp:[],ranked_grad:4,ranked_hour:4,rant:10,rapidli:[1,7],rare:[0,1,3,4,10,11,13],rate:[0,1,4,11,13],rater:1,rather:[0,1,2,3,4,5,6,7,9,10,11,13],ratio:[4,5,9,13,20],ration:9,rational:6,ravish:9,raw:[0,1,3,4,5,6,7,10,11,13],rbind:[],rbinom:[],rbook:3,rchisq:[],rcl:13,rda:[],rdata:13,rdb:[],rdx:[],reach:[0,10],reactionari:0,read:[0,1,2,3,4,5,6,7,9,10,11,13],read_csv:[0,3,4,5,13],readabl:[3,4,5],reader:[1,4,5,9,10,11,13],readership:[4,9],readonli:[],readrd:[],real:[0,1,2,5,6,7,8,9,10,11,13],realis:[1,2,5,7,8,11,13],realist:[1,13],realiti:6,realiz:[3,13,20],realli:[0,1,2,3,4,5,6,7,8,9,10,11,13],realm:10,reanalys:1,reang:4,reason:[0,1,2,3,4,5,6,7,9,10,11,13,19],reassign:2,recal:[1,4,10,13],recalcul:10,recap:11,reccomend:2,receiv:[4,5,13],recent:[0,3,5,8,9,13],reckon:8,recogn:3,recognis:[1,7,9,10,11,13],recognit:2,recommend:[3,13],reconsider:19,record:[1,3,4,7,8,10,13],recreat:7,rectangular:[],red:10,redefin:13,redescrib:[9,13],redraw:5,reduc:[1,5,9,11,13],redund:2,ref:[0,1,6,10,11,13],refer:[1,3,4,5,6,7,8,9,10,11,13,20],reffig:[],refineri:10,reflect:[0,1,4,6,9,10,11,13],refsmmat:0,refus:[0,9],regard:[1,11],regardless:[1,3,4,5,6,10,11,13],region:13,regist:1,regplot:[4,5,10],regress:[6,10,13],regular:[3,4,13],rehash:10,reinforc:[1,13],reinvent:3,reject:[0,4,8,11,13],rel:[1,2,3,4,6,9,10,13],rela:4,relat:[0,1,2,3,4,9,10,11,13],relatedli:5,relationship:[1,3,5,6,9,10,11,13],relax:9,releas:[],relev:[1,3,4,9,10,11],reli:[0,1,4,5,6,7,8,9,10,11,13],reliabl:[4,10],relianc:10,reliant:[0,6],reload:[],remain:[1,4,5,8,9,11],remark:[0,3,8,9,10],remedi:[4,13],rememb:[0,1,2,3,4,5,7,9,10,11,13],remind:[1,5],remov:[0,4,5,11,13],renam:[3,4,5,6],rent:4,rep:[],repeat:[9,10,13],repeatedli:9,repetit:[2,9,13],replac:[5,10,13],repli:1,replic:[10,11],report:[0,1,4,5,9,10,13],repres:[1,3,5,6,9,10],represent:4,repress:[4,19],reproduc:7,reput:11,request:[],requir:[0,1,2,4,6,7,8,9,10,11,13],res:[],research:[2,4,5,6,9,10,13,19],resent:10,reserv:[],reset:3,reset_index:5,reshap:6,resolv:[2,13],resort:3,resourc:5,respect:[1,4,5,8,9,10,11,13],respond:[1,3,11],respons:[2,3,4,6,9,10,11,13],rest:[4,5,6,8,10,11,13],restor:[],restrict:[10,11,13],result:[0,1,2,3,4,5,6,9,10,13,19],retain:[11,13],retent:11,retest:1,reus:7,reveal:[4,10],revel:[],revenu:2,revers:[0,1,11,13],review:19,revis:8,revisit:[1,3,4,9,11],reward:[4,9],rewrit:10,rewritten:10,rgamma:13,rgb:[],rhistori:[],rho:4,rich:[2,3],richmond:[4,5],rid:3,riddl:[5,8,10],ride:8,ridicul:[1,13],rig:[9,10,13],right:[0,1,3,4,5,6,7,8,10,11,13],rightarrow:[9,10],rigor:11,rigour:1,ring:10,rise:[1,4,9,11,13],risk:1,rmdnote:[],rmsd:4,rneg:4,rnorm:[10,13],robot:[2,4,9],robust:4,rock:[],roger:[2,6],role:[0,2,10],roll:9,ronald:[1,9,11,13],ronald_fish:1,room:[5,11],root:[3,4,9,10,13],rosenthal1966:1,rotat:5,rough:[1,4,13],roughli:[6,9,10],round:[5,8],roundabout:1,row1:4,row2:4,row:[3,4,5,6,7,8,9,11],rowtot:6,royal:19,royalti:2,rseek:[],rsquo:[],rstudio:[],rtfm:[],rubbish:4,rude:[],rule:[1,2,4,5,8,9,10,11,13],run1:9,run2:9,run3:9,run4:9,run:[0,1,2,3,4,5,6,7,8,9,10,13],rush:11,sabotag:0,sack:1,sad:9,sadder:9,sadli:13,safe:[0,1,2,3,5,11,13],safeguard:0,safer:[2,4],safest:4,sai:[0,1,2,3,4,5,6,8,9,10,11,13,20],said:[0,1,2,3,4,5,8,9,10,11,13],sake:[3,4,5,13],salari:1,sale:[2,3],salem:[],same:[0,1,2,3,4,5,6,8,9,10,11],sampdistmax:10,sampdistmean:10,sampdistsd:10,sampel:13,sampl:[1,4,5,8,9,19,20],sample_max:10,sample_mean:[10,13],sample_mean_1:10,sample_sd:10,sample_standard_error:13,sample_vari:4,samplemean:10,samplesandclt:10,samplesd:10,samplingdist:[],san:4,sane:9,saniti:6,satisfi:[6,9,10,11,13],saturdai:4,save1:[],save:[2,4,6,7],savefig:5,savvi:[],saw:[1,4,5,8,9,13],scale:[4,5,6,9,10,13,19,20],scaled_chi_square_20:9,scaled_chi_square_3:9,scaled_chi_square_data:9,scalescont:[],scan:1,scare:[1,5],scari:[],scat:13,scatterplot:[4,13],scenario:[1,3,4,9,11,13],scene:2,scheme:[1,4,10],scherndl:19,schizophren:10,schizophrenia:10,school:[1,2,3,10,11],sci:[],scienc:[0,1,2,10,11,19],scientif:[0,1,4,5,9,10,11,13],scientist:[0,1,4,10,11,13],scipi:[2,3,4,9,10,11,13],sciplot:[],scope:[1,4,5,9,10,13],score:[1,3,6,8,10,11,13],score_data:3,scorecard:8,scotsman:1,scratch:6,screen:[2,3,4],screenshot:[],scribbl:[4,7],script:[3,5],scriptstyl:4,scroll:3,sd0:[],sd1:[],sd_diff:13,sd_true:13,sdnorm1:[],sdnorm2:[],sdnorm:9,sdnorm_fig2:9,sdnorm_fig:9,se_est:13,seaborn:[0,2,3,4,5,9,10,11,13],seali:13,search:[1,3,5,11],searchabl:[],season:[4,5,10],second:[0,1,2,3,4,5,6,8,9,10,11,13],secondari:11,secondli:[0,1,5,6,8,9,10,13],secret:13,section:[0,1,3,4,5,6,7,9,10,11,13],secur:[4,5],see:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],seek:[1,11],seeker:3,seem:[0,1,2,3,4,5,6,7,8,9,10,11,13],seen:[1,2,3,4,5,6,9,11],select:[0,4,5,6,10],self:[0,2,3,10],sell:[0,2],sem:[10,13],sem_tru:13,semest:[4,13],send:[2,5,7],senior:4,sens:[0,1,2,4,5,6,7,8,9,10,11,13],sensibl:[0,1,5,7,8,9,10,11,13],sensit:[2,4],sent:13,sep:[],separ:[3,4,5,6,9,10,11,13],septemb:[4,7],seq:[],seq_along:[],sequenc:[8,9,10],seri:[2,3,4,5],serial:[],seriou:[4,6,11],serious:[1,2,5,9,13],serv:[1,3,4,13],session:[1,4],set:[0,1,2,3,4,5,6,7,8,9,10,11,13],set_capt:[],set_color:11,set_context:5,set_frame_on:13,set_linestyl:4,set_text:13,set_titl:[4,5,9,10,11,13],set_vis:[4,5,10,13],set_xlabel:4,set_xlim:[4,13],set_ylabel:4,set_ylim:[4,13],settl:[2,4,9],setupplot:[],setwd:[],seven:5,seventh:8,sever:[1,2,3,4,5,6,7,9,10,11,13],sex:[10,19],sexist:[0,10],shade:[4,5,9,11],shake:10,shall:[10,13],shalt:0,shame:[5,7,11],shape:[4,5,9,10],shapiro1965:13,shapiro:[],shapirowilkdist:13,share:[0,1,5,7,9,10],sharei:[4,9,10,13],sharex:10,sharpli:[11,13],she:[1,2,10,11,13],shed:4,sheer:13,shelf:10,shell:2,sherlock:8,shift:[2,5,9,11],shini:6,shirt:10,shock:[1,4,13],shockingli:[4,7],shoe:[1,10],shook:10,shop:3,shortag:0,shortcom:4,shorten:[4,11],shorter:[1,10],shorthand:13,shortli:[4,7,9,11],shot:[1,10,11],should:[0,1,2,3,4,5,6,7,8,9,10,11,13],shoulder:3,shouldn:[0,1,4,5,8,9,10,13],show:[0,1,2,3,4,5,6,7,9,10,11,13],shown:[0,1,2,4,7,9,10,11,13],shrink:9,shrunk:[6,11],shuffl:9,side:[0,4,5,7,8,9,10],sidewai:4,sig:11,sigex:[],sigh:[1,4],sigma1:13,sigma2:13,sigma:[4,9,10,13],sigma_0:13,sigma_1:13,sigma_2:13,sigma_d:13,sigma_p:13,sign:[2,3,4,6,13],signif:[],signific:[0,13,19],significantli:[11,13],silli:[1,2,5,10,11,13],sim:[9,11,13],simdata:10,similar:[1,2,4,5,7,9,10,11,13],similarli:[1,4,9,11,13],simpl:[0,1,3,4,5,6,7,9,11,13],simpler:[2,4,5,9,13],simplest:[1,2,3,4,5,9,10,11,13],simpli:[1,2,3,4,5,6,9,10,13],simplic:[5,13],simplif:[10,11],simplifi:[5,10],simpson:1,simpsont:[],simul:[9,10],sinc:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],sinclair:1,singl:[0,2,3,5,6,7,9,10,11,13],sink:[],sir:[1,9,11,13],sit:[0,1,3,4,7,9,10,11],site:[5,10],situat:[0,2,4,5,6,8,9,10,11,13],six:[0,4,7,8,9],sixth:4,size:[0,1,3,6,9,10,19],skew:[9,13],skew_fig:4,skewdata:4,skill:[1,2,3,7,8,10],skim:9,skip:[4,13],skipna:4,skull:9,skulls_fig:[],sky:13,slant:1,slash:3,sleep:[4,5],sleep_scatt:4,sleepi:4,sleepier:4,slept:4,slice:[4,7],slightli:[0,1,2,4,5,6,7,9,10,11,13],slope:5,sloppi:10,slower:[0,20],slowli:[4,11],small:[0,4,5,6,7,9,10,11,13],smaller:[4,9,10,13],smallest:[4,9,11,13],smallish:[6,7],smart:[0,1,2,7,13],smarter:[0,1],smelter:10,smirnov:13,smoke:1,smoker:[1,10],smooth:9,smoothi:13,sms:10,sneaki:[1,10,13],snippet:3,snow:5,snowbal:10,sns:[0,3,4,5,9,10,11,13],soccer:9,social:[0,1,2,6,10],societi:19,sociolog:0,soft:11,soften:11,softwar:[1,2,3,4,11,13],sokal:1,sole:[4,5],solid:[1,4,9,10,13],solut:[1,2,4,5,9,13],solv:[0,1,2,3,4,5,7,11,13],some:[0,2,3,4,5,6,7,8,9,10,13,19],somebodi:1,somehow:[10,11],someon:[0,1,2,4,5,6,8,9,10,11,13],someth:[0,1,2,3,4,5,7,8,9,10,11,13],sometim:[0,1,2,3,4,5,6,7,8,9,10,11,13],somewhat:[0,1,4,6,9,10,11,13],somewher:[3,4,9,11],son:[4,10],song:11,soon:2,sooner:5,sophist:[1,4],sorri:8,sort:[0,1,2,4,5,6,7,9,10,11,13],sort_valu:4,sorted_margin:4,sortfram:[],sound:[0,1,2,3,4,5,8,9,10,11],sourc:[1,2,4,5,8,10,11,20],south:10,soviet:[4,9,19],space:[5,9],span:[4,5,6],speak:[2,3,4,5,7,8,9,10,11,13],speaker:6,spearman:4,spearmanr:4,speci:0,special:[1,2,3,4,7,8,9,10,13],specialis:[5,7],specif:[0,1,2,3,4,5,9,10,11,13,20],specifi:[1,4,5,6,9,10,11,13],specifii:[4,10],spectacularli:4,speech:2,speed:11,spell:1,spend:[0,1,2,3,4,5,6,7,8,9,10,11],spent:[1,4,7,9],spine:[4,5,10],spirit:[3,4],spit:2,split:4,sport:4,spot:9,spous:9,spread:[4,5],spreadsheet:[2,5],spring:9,springer:19,spss:11,sqrt:[4,9,10,13],squar:[2,3,4,9,10,11,13,20],squared_error:4,squat:5,srs1:10,srs2:[],srs:10,stabl:2,stackoverflow:5,stage:[3,5,6,8,10,11],staggeringli:2,stai:[0,5,6,9],stalin:4,stalinist:4,stamp:9,stand:[1,3,4,9],standard:[1,2,3,5,6,9,11],standardis:[4,13],standardnorm:[],stanlei:19,star:[5,11],stare:5,start:[0,1,3,4,5,6,7,8,9,10,11,13,20],starv:0,stat:[0,1,2,4,9,10,11,13,20],stata:[],state:[1,2,9,10,11,13],statement:[0,1,2,3,4,9,10,11],statist:[1,2,3,5,6,7,13,19],statistic_:[],statistician:[0,1,4,8,9,10,11,13,19],statsmodel:[3,4,10],statsmodels_mad:4,std1:[],std2:[],std:[4,5,10,13],stdev:[4,10,13],ste46:[1,19],steadili:9,steel:10,steep:5,stem:1,stencil:5,step:[1,2,3,4,6,8,11,13],stereotyp:1,steril:9,stern:19,steven:19,stevens1946:[],stick:[3,4,6],stigler1986:10,still:[0,1,2,3,4,5,8,9,10,11,13],stomach:4,stone:13,stop:[1,2,4,5,7,9,13],store:[3,4,5,6,7,9,13],stori:[0,1,2,4,6,9,10,11,13],storm:10,str:[0,3,4,5,10],straight:[2,4,11,13],straightforward:[2,3,4,5,9,10,13],strang:[3,4,6,13],strata:10,strategi:[2,8],stratifi:10,streak:[5,8],street:5,strength:[2,5,6,10,13],strengthen:11,stress:[0,4,5,10],strict:[1,2],strictli:[1,4,11,13],strike:0,string:[3,7],stringsasfactor:[],strip:[5,7],stripplot:5,stroke:5,strong:[0,1,4,5,6,8,9,11,13],stronger:[4,11],strongli:[1,2,6],stroppi:[],structur:[0,1,3,5,11,13],struggl:5,strung:5,stuck:2,student10:[],student1908:13,student1:13,student2:13,student3:13,student4:13,student5:13,student:[0,1,2,4,10,11,20],studentassumpt:13,studentttest:13,studi:[0,2,4,5,6,10,13,19],studydesign:10,stuf:1,stuff:[0,1,2,3,4,10,11],stupid:[1,2,4,5,9,10,11,13],style:[1,5,7,10,11],stylis:5,subfield:4,subfold:[],subgroup:1,subject:[1,4,9,10,13],subject_:[],subjectivist:9,submit:1,subplot:[4,5,9,10,13],subpopul:10,subract:13,subsampl:1,subscrib:9,subscript:[4,13],subsequ:[1,5],subset:[6,10],subsetdatafram:6,substant:11,substanti:[1,4,13],substitut:[0,13],substract:13,subtl:[0,1,2,3,4,5,9,10],subtleti:[4,5,9],subtli:[1,11,13],subtract:[1,6],succeed:[],success:[0,4,9,11],successfulli:11,succumb:11,suck:2,suddenli:7,sued:0,suffer:[0,11],suffic:3,suffici:[0,1,2,4,10,11,13],suggest:[0,1,2,3,4,5,9,11,13],suit:[3,5,9,13],suitabl:[],suk:1,sullivan:3,sum:[2,3,4,6,9,13],sum_:[4,10,13],sum_numb:[],summar:5,summari:[],summaris:[1,4,7,9,10,11],summat:4,summon:4,sun:[10,11],superior:1,supplement:[],suppli:[0,5,10],support:[3,11,13],suppos:[0,1,2,3,4,5,6,7,8,9,10,11,13],suppress:11,suptitl:[4,9,10],sure:[1,2,3,4,5,7,8,9,10,11,13],surfac:6,surpris:[0,1,2,3,4,5,9,11,13],surprisingli:[0,1,2,4,9,10,11,13],surround:11,survei:[1,2,4,9,10],surviv:0,suscept:0,suspect:[1,4,5,9,10,13],suspici:[4,5,8,11,13],suspicion:5,swai:0,swath:4,swdist:13,sweet:[2,3,13],swing:1,sydnei:[4,5,9],syllogist:19,symbol:[1,2,3,4,5,10],symmetr:[4,6,13],syndrom:1,synonym:4,syntax:5,syntaxerror:2,systat:[],system:[2,3,5,10,11],systemat:[0,1,4,10,13],t_3:9,t_stat:[],tab:[6,10],tabl:[0,1,2,3,4,9,10,11,13],tabul:13,tack:4,tackl:[3,10],tag:2,tail:[4,5,9,11,13],take:[0,1,3,4,5,6,8,9,10,11,13,20],taken:[0,2,3,4,6,7,9,10,11,13],talk:[0,1,2,3,4,5,6,8,9,10,11,13],tall:1,taller:1,talli:9,target:[],task:[1,2,3,4,5,6,9,10],taught:[2,10,13],tautolog:11,tconfint_mean:10,tdist:9,tdist_3:[],teach:[0,1,7,13,20],team:[4,5,9,10],teas:1,techin:5,technic:[1,4,9,10,11,13],techniqu:[10,11],tediou:[1,2,4,13],teensi:0,tell:[1,2,3,4,5,6,7,9,10,11,13],temperatur:[1,9],tempestu:10,temporarili:1,tempt:[4,5,11,13],temptat:[0,11],ten:11,tend:[0,1,4,5,6,7,9,10,11,13],tendenc:10,tension:11,tenur:11,term:[0,1,2,3,4,5,6,8,9,10,11,13],terminolog:[1,2,9,10,11],terplot:13,terrestri:10,terribl:[1,4,9,11,13],territori:9,ters:11,test1:[],test2:[],test:[0,3,4,6,8,9,10,19],test_1samp:13,test_:[],testabl:11,tests_:[],testsm:13,teststatist:[],text:[2,3,4,5,6,7,9,10,11,13,20],textbook:[1,2,4,9,11],textedit:[],textprocess:6,textual:20,than:[1,2,3,4,5,6,7,8,9,10,11,13],thank:[0,9,20],thankfulli:0,theatr:1,thei:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],theirs:1,them:[0,1,2,3,4,5,6,7,9,10,11,13],themselv:[0,1,4,5,6,7,9,10,13],theorem:[8,13],theoret:[1,8,10,11,13,19],theori:[0,4,10,11,13,19],therapi:4,therebi:[1,4],therefor:[0,1,2,3,8,10,11,13],thermomet:9,theta:[9,11],theta_0:11,thi:[0,1,4,5,6,7,8,9,10,11,13,20],thick:1,thin:5,thing:[0,1,2,4,5,6,7,8,9,10,11,13,20],think:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],third:[1,2,3,4,5,6,9,10,11,13],thirdli:[0,5],thoma:19,thorni:10,thorough:10,thoroughli:11,those:[0,1,2,3,4,5,6,7,8,9,10,11,13],thou:0,though:[0,1,2,3,4,5,6,8,9,10,11,13],thought:[0,3,4,9,10,11,13],thoughtless:11,thousand:[1,2,3,9,11],threaten:1,three:[0,1,2,4,5,6,8,9,10,11,13],through:[1,3,4,5,6,7,8,9,10,11,13,20],throughout:[1,4,5,8,9,11],thu:[3,4,5,9,13],thumb:4,tibbl:[],tick:[1,5,13],tick_param:[9,10],tidi:6,tidli:5,tie:1,tied:0,tier:11,ties:[2,13],tiger:11,tight:5,tightli:[1,9,10,13],tild:[],time:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],tini:[4,7,9,10,11,13],tionship:4,tip:[5,10],tire:[1,4],titl:[1,4,11,13],tmp:[],to_csv:3,to_fram:5,todai:[1,9,11],togeth:[1,3,4,5,6,7,9,10,13],toi:[3,4,6,11,13],told:[0,3,4,5,8,9,11],toler:[1,11],tolist:3,tombliboo:6,tomorrow:[9,11],too:[0,1,2,3,4,5,6,7,8,9,10,11,13],too_high:10,too_low:10,took:[0,1,3,4,10,13],tool:[0,1,2,4,5,7,8,9,10,11,13],top:[0,1,3,4,5,9,10,13,20],topic:[0,1,3,4,6,9,10,11,13],total:[0,4,6,9,10,11,13],total_bil:10,touch:[4,5,10],tough:9,toward:[1,4,9,13],tower:[9,11],town:10,traceback:[3,13],track:9,tracksuit:9,tract:0,tractatu:11,traction:9,tractor:6,trade:[4,13],tradit:[1,4,13],tradition:[4,10,13],traditionalist:1,traffic:[],tragedi:4,train:[0,1,9],tran:10,transact:19,transcrib:6,transfer:1,transferr:1,transform:[4,10],transgend:[1,10],transit:[],translat:[4,5,7,9,10,11,13,20],transluc:5,transport:1,trap:[0,1,2],travel:[1,4],treat:[1,3,4,11,13],treatment:[1,11],trend:0,tri:[1,2,3,4,5,6,7,11],trial:[1,4,9,11],triangular:10,tribbl:[],trick:[1,2,6,7,9,13],tricki:[1,4,9,11],trickier:[4,9,11,13],trigger:13,trillion:11,trim_mean:4,trite:9,trivial:[6,13],troubl:[1,4],truism:10,truli:[4,9],trust:[0,1,7,9,13],trustworthi:[0,1,4,13],truth:[0,1,2,4,8,9,10],ttest2:13,ttest:[],ttest_1samp:13,ttest_1sampresult:13,ttest_ind:13,ttest_indresult:13,ttest_rel:13,ttest_relresult:13,ttestci:[],ttestdist:[],ttestfunct:[],ttesthyp2:13,ttesthyp:13,ttesthyp_onesampl:13,ttestimg:[],ttestoneassumpt:13,tuck:[],tuft:5,tupl:3,turf:5,turn:[0,1,3,4,5,8,9,10,11,13],tutor1:13,tutor2:13,tutor:13,tutori:[3,13],tverski:19,tweak:[5,7,10],twice:[1,10],twilight:9,twini:9,twinx:[4,9,10],twist:10,twister:[],two:[1,2,3,4,5,6,7,8,9,10,20],twofold:5,type:[0,1,4,5,6,7,9,10,13],typic:[0,1,3,4,9,10,13],typo:[4,5,7],ubiquit:11,ubtract:2,ugh:4,ugli:13,ultim:[8,11,20],ultipl:2,unabl:11,unaccept:[],unacknowledg:1,unambigu:[4,9,13],unavoid:[1,8],unawar:11,unbeliev:0,unbias:[0,4,10],unbiased:10,uncertain:10,uncertainti:[1,5,10,13],uncomfort:9,uncommon:[1,4,6,13],uncontrol:1,undefin:[],under:[0,2,4,5,9,10,11,13,20],undergradu:[0,1,9,10],underli:[5,6,9,11],underpin:[8,9,11],underpow:5,underst:4,understand:[0,1,2,3,4,5,6,7,9,10,11,13],understood:[4,9],undesir:9,undo:[3,5],unduli:4,unenlighten:0,unequ:[6,13],uneth:[1,10],unexpect:3,unfair:1,unfairli:0,unfalsifi:11,unfamiliar:[],unfold:9,unfortun:[1,3,4,5,9,10,11],ungrammat:4,unhelp:[4,10],unicorn:13,uniform:9,unimagin:[],uninform:1,unintent:1,unintention:1,uninterpret:4,uniqu:[4,5,10,11],unit:[4,6],univers:[0,1,9,10,13,19],unix:3,unknow:5,unknown:[3,4,9,11,13],unlabel:[],unless:[2,3,4,5,6,7,10,11],unlik:[1,3,4,5,9,11,13],unlink:[],unlucki:11,unmeasur:1,unnatur:1,unnecessari:[4,13],unpopular:9,unproblemat:10,unproduct:11,unrealist:[1,4],unreason:[4,8,13],unrel:4,unreli:1,unremark:9,unrepent:11,unrepres:1,unsatisfi:[],unseri:[],unsur:2,unsurpris:11,until:[0,2,3,4,5,7,9,10,13],untitl:[],untrain:1,untru:4,unus:3,unusu:[4,5],unwari:[0,11],updat:[1,20],updatepackag:[],upload:2,upon:[0,1,5,10,11],upper:[1,4,10,11,13],upper_area:13,upset:2,upsi:6,uptight:2,upton:1,upward:[11,13],urgent:0,url:10,usabl:2,usag:[4,13],use:[0,1,2,5,6,8,9,10,11,13,20],used:[0,1,2,3,4,5,7,9,10,11,13],useful:[0,1,2,3,4,5,6,7,10,11,13],useless:[0,4,5,10,13],user:[2,3,4,5],uses:[1,2,3,4,6,10,13],using:[0,1,3,4,5,6,7,8,9,10,11,20],usingfunct:[],usual:[0,1,2,3,4,5,6,9,10,11,13],usus:5,utf:[],utter:6,v10n3:10,vaccin:1,vagu:[1,10,11],val:[4,10],valid:0,valign:[],valu:[0,1,2,3,5,6,9,10],valuabl:1,value_count:[4,5],van:0,vancouv:4,var1:4,var2:4,var3:4,vari:[1,3,4,9,10,11,13],variabl:[5,7,9,10,13],varianc:[9,10,13],variance_:[],variant:1,variat:[1,2,4,10,13],variaterel:[],varieti:[2,5,13],variou:[0,2,3,7,9,11],vast:[0,1,4,5,9,10,11,13],vector:[4,6,9,10,13],vehement:9,verbal:[0,1],verbalis:1,verbos:[],veri:[0,1,3,4,5,6,7,9,10,11,13,20],verifi:[2,4],version:[0,1,2,4,5,6,11,13,20],versitil:3,versu:[9,13],vertic:[5,9],vertigo:7,vest:4,via:[7,13],viabl:9,victori:[8,9],view:[0,1,5,6,10,13],viewer:5,violat:[1,6,13],violinplot:5,virtual:3,visibl:[],vision:2,visual:[1,4,5,9,10,13],visualis:[4,5,9],vline:10,vocabulari:1,voic:1,volum:1,von:19,vote:[9,10],voter:[9,10],vulner:[1,4],w_1:13,w_2:13,waaaai:5,wacki:11,wage:[2,4],wai:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],wait:[0,4,5],wake:[1,13],wakefield:1,walk:[9,11],wall:11,wander:10,want:[0,1,2,3,4,5,6,7,8,9,10,11,13],war:[4,8],wardrob:9,wari:11,warn:[0,1,3],warrant:[4,13],wasn:[1,2,5,9,11,13],wast:[5,9],watch:[1,4,5,6,8,9],water:[1,5],watercolour:5,wcc:8,weak:[0,2,4,5,11,13],weaken:13,weaker:[10,11],wealthi:10,wealthiest:4,wear:9,weather:9,web:2,webcom:3,websit:[0,4,10],wed:[],weed:20,week:[1,3,4,5],weigh:13,weight:[1,13],weird:[0,1,4,5,8,9,10,11],weirdli:10,welch1947:13,welch:[],welchttest:13,welcom:3,well:[0,1,2,3,4,5,6,7,8,9,10,11,13],welles:[8,10],welleslei:[8,10],wellington:8,wellselei:10,went:[0,1,4,5,9],were:[0,1,2,3,4,5,7,8,9,10,11,13],weren:[5,10],west:[4,5],western:[4,5],what:[0,1,2,5,6,7,8,10,11,20],whatev:[1,2,3,4,9,11],whatsoev:[4,13],wheel:3,when:[0,1,2,3,4,5,6,7,8,9,10,11,13],whenev:[1,6,9,10,11,13],where:[0,1,2,3,4,5,6,8,9,10,11,13],wherea:[0,1,4,6,9,10,11,13],whether:[0,1,2,3,4,5,6,8,9,11,13],whew:13,which:[0,1,2,3,5,6,7,8,9,10,11,13,20],whichev:11,whing:2,whisker:5,whiskerprop:5,white:[1,5,10,11,13],whitnei:13,who:[1,2,3,4,5,7,8,10,11,13],whole:[0,1,2,3,4,8,9,10,11,13],whose:[1,2,4,5,9,10,11,13],why:[1,2,3,4,5,6,8,9,10,11,13,19],why_you_gotta_be_so:3,whyalla:10,whyfor:[],wide:[2,5,10,13],widen:9,wider:[1,9,10],widetolong:[],width:[5,9,13],wife:11,wiki:1,wikiquot:1,wilcox1:13,wilcox2:13,wilcox:[],wilcoxon:[],wild:1,wildli:[1,9],wilei:19,wilfri:0,wilk:[],willi:[3,6],william:[7,11,13],willing:[1,8,9,11,13],win:[4,5,8,9,10,11,13],winc:2,window:[3,11],wipe:4,wise:[0,6],wish:4,wit:1,within:[0,1,2,4,5,6,7,9,10,11,13],without:[0,1,2,3,4,5,9,10,11,13],witter:10,wittgenstein:11,wmc:[],woman:4,women:0,won:[1,2,3,4,5,7,8,9,10,13,20],wonder:[4,9,10,11,13],woo:1,word:[0,1,2,3,4,6,7,8,9,10,11,13],work:[0,1,2,3,4,5,6,7,8,10,11,20],workaround:[],worker:1,workspac:7,workspace2:[],workspacepanel2:[],workspacepanel:[],world:[0,1,2,3,4,5,6,7,8,9,10,11,13],worri:[0,1,4,6,9,10,11,13],wors:[0,1,4,11,13],worst:4,worth:[1,2,3,4,5,7,9,10,11,13],would:[0,1,2,3,4,5,6,7,8,9,10,11,13],wouldn:[0,1,2,4,5,6,7,9,10,11,13],wrangl:[3,5],wrap:9,wrapper:[],wrinkl:13,write:[0,1,2,3,4,5,7,9,10,11,13],writeup:[],written:[1,2,3,4,5,9,10,11,13],wrong:[0,1,2,4,5,7,8,9,10,11,13],wrongfulli:11,wrongli:1,wrote:[1,2,3,6,7,10,13],wwc:8,www:[0,4,8,10],x_100:9,x_1:[1,4,9,13],x_20:9,x_2:[1,4,9,13],x_3:[4,9],x_4:[4,9],x_5:[4,9],x_bar:4,x_estim:10,x_fill1:9,x_fill2:9,x_i:[4,10,13],x_n:4,xkcd:3,xlab:13,xlabel:[0,4,5,9,10,11,13],xlim:[9,11],xlsx:[],xmax:13,xmin:13,xponent:2,xtick:9,xval:[],xxx:[],xxy:1,xytext:11,y_fill1:9,y_fill2:9,y_fill:[],y_i:4,y_norm:13,y_t:13,ya1:[],ya2:[],yeah:[8,9],year:[0,1,3,4,5,6,8,9,10,13],yep:[],yes:[1,4,8,9,10,11,13],yesterdai:1,yet:[0,1,2,3,4,5,8,9,10,11,13],yield:11,yike:[3,4],ylab:[],ylabel:[0,4,5,9,10,11,13],ylim:[10,13],ymax:10,ymin:10,ynull:[],york:4,you:[0,1,2,3,4,5,6,7,8,9,11,13],young:[1,6],younger:6,your:[0,2,3,4,5,6,7,8,9,10,13],yourself:[0,1,2,3,4,5,7,9,13],yticklabel:10,yum:5,yval:[],yyi:8,z_i:4,z_score:13,zag:11,zassumpt:13,zealous:2,zelazni:[2,6],zeppo:13,zero:[1,3,4,5,6,7,9,13],zeroeth:3,zig:11,zip:[],zscore:4,ztest1:13,ztest2:13,ztest:13,ztesthyp:13},titles:["1. Why do we learn statistics?","2. A brief introduction to research design","1. Getting Started with Python","3. More Python Concepts","1. Descriptive statistics","2. Drawing Graphs","3. Data Wrangling","4. Basic Programming","1. Statistical theory","2. Introduction to Probability","3. Estimating unknown quantities from a sample","4. Hypothesis Testing","1. Categorical data analysis","2. Comparing Two Means","3. Comparing several means","4. Linear regression","5. Factorial ANOVA","1. Bayesian Statistics","2. Epilogue","3. References","Learning Statistics with Python"],titleterms:{"case":4,"function":[6,11],"import":3,"null":11,Adding:2,And:9,Doing:[2,13],One:13,The:[0,1,4,9,10,11,13],There:0,Useful:3,Using:[2,3],about:[1,3],absolut:4,activ:3,address:13,altern:[5,11],analysi:12,ani:10,anova:16,artifact:1,assess:1,assumpt:[8,13],attrit:1,avoid:2,axi:5,bar:5,basic:[7,9],basicprob:[],bayesian:[9,11,17],belief:0,between:11,bia:[0,1],bibliographi:[],binomi:9,bit:2,book:[2,3],box:5,boxplot:5,brief:1,calcul:[2,4,10],care:2,categor:12,categori:6,caution:5,cautionari:0,central:[4,10],check:[3,13],code:2,coeffici:4,cohen:13,cohensd:[],color:5,command:2,comment:3,common:11,compar:[13,14],complet:13,complex:[1,5],concept:3,confid:10,confound:1,consid:11,consol:2,construct:[1,13],content:3,continu:1,convert:6,correl:4,count:6,creat:6,critic:11,cross:6,csv:3,curs:0,cut:6,data:[3,4,5,6,11,12,13],datafram:3,decept:1,decis:11,decisionmak:[],defin:[1,10],demand:1,densiti:9,describ:[4,13],descript:4,design:1,deviat:[4,10,13],differ:[4,9,11,13],differenti:1,differentialattrit:[],direct:4,discret:1,distribut:[9,10,11],divid:2,document:[],doe:[3,9,10],don:10,draw:5,ecolog:1,effect:[1,11,13],effects:[],enough:5,epilogu:[4,18],error:11,errortyp:[],estim:[10,13],everydai:0,exampl:[4,13],exist:10,experiment:1,extern:1,extrem:11,face:1,factor:[],factori:16,few:6,file:[3,5],filesystem:3,find:3,fisher:11,flexibl:2,font:5,formula:10,frame:[3,4],fraud:1,frequentist:[9,11],from:[3,10,13],gener:[],get:[2,4],good:4,graph:5,graphic:5,handl:4,have:10,help:[],histogram:5,histori:1,how:[3,9,10],hypothes:11,hypothesi:[11,13],hypothesistest:[],imag:5,increas:11,independ:13,index:3,infer:13,inform:[],instal:[2,3],intern:1,interpret:[4,10],interquartil:4,interv:[1,10],introduc:[9,13],introduct:[1,5,9],issu:11,ivdv:[],know:3,kurtosi:4,label:5,larg:10,law:10,learn:[0,8,20],length:3,level:[],librari:3,life:[0,4],limit:[8,10],linear:15,list:3,littl:5,load:3,loadingcsv:[],logic:8,look:3,loop:7,make:[8,11],mathemat:6,matter:[5,10],matur:1,mean:[1,4,9,10,13,14],measur:[1,4],median:4,memori:3,menageri:11,method:0,miss:4,mistak:[10,11],mode:4,more:[0,3,5,6],most:10,move:[],much:10,multipl:5,multipli:2,myth:8,neg:13,neyman:11,nhstmess:[],nomin:1,non:[1,13],normal:[9,13],note:11,now:5,number:[2,10],numer:6,object:3,one:[11,13],onesamplettest:[],onesidedtest:[],open:5,oper:6,operationalis:1,order:2,ordin:1,other:[1,9],otherdist:[],out:3,outcom:1,overal:4,overview:5,packageinstal:[],pair:13,pairedsamplesttest:[],pairwis:4,paradox:0,paramet:10,part:[],path:3,placebo:1,plot:[5,10,13],pool:13,popul:10,posit:13,power:[2,11],practic:11,pragmat:[],predictor:1,probabl:[9,11],problem:13,probmean:[],probstat:[],program:7,proport:6,propos:11,psycholog:[0,1],pull:3,pvalu:[],python:[2,3,4,5,9,10,13,20],quantiti:10,random:10,rang:4,rank:[],ratio:1,reactiv:1,read:[],real:4,reason:8,recod:6,refer:19,region:11,regress:[1,15],relat:5,relationship:4,reliabl:1,remov:3,repeat:1,report:11,research:[0,1,11],resourc:[],respons:1,result:11,right:[2,9],road:3,role:1,run:11,same:13,sampl:[10,11,13],save:[3,5],scale:1,scatterplot:5,score:4,script:7,select:1,self:1,sever:14,shapiro:13,side:[11,13],signific:11,simpl:[2,10],simpson:0,singl:4,situat:1,size:[5,11,13],skew:4,slice:3,slight:10,softer:11,solut:11,some:[1,11],space:2,spearman:[],standard:[4,10,13],start:2,statist:[0,4,8,9,10,11,17,20],store:2,strength:4,student:13,studentassumpt:[],studentttest:[],studi:[1,11],subpopul:1,subtract:2,summari:[1,3,4,9,10,11,13],tabl:6,tabul:6,take:2,tale:0,teeni:2,tendenc:4,test:[1,11,13],teststatist:[],than:0,theorem:10,theori:[8,9],thi:[2,3],thing:3,thought:1,threat:1,three:3,tip:3,titl:5,transform:6,trap:11,trim:4,ttest:[],ttestfunct:[],ttestoneassumpt:[],two:[11,13],type:[2,3,11],typo:2,unknown:10,use:[3,4,7],useful:9,using:[2,13],valid:1,valu:[4,11,13],variabl:[1,2,3,4,6],varianc:4,veri:2,versu:[1,11],view:[9,11],violin:5,welch:13,welchttest:[],what:[3,4,9,13],which:4,who:9,why:[0,7],wilcox:13,wilcoxon:13,wilk:13,without:8,word:5,work:[9,13],workspac:3,wrangl:6,writeup:[],you:10,your:[1,11],zassumpt:[]}})
\ No newline at end of file
+Search.setIndex({docnames:["01.01-intro","01.02-studydesign","02.01-getting_started_with_python","02.02-more_python_concepts","03.01-descriptives","03.02-drawing_graphs","03.03-pragmatic_matters","03.04-basic_programming","04.01-intro-to-probability","04.02-probability","04.03-estimation","04.04-hypothesis-testing","05.01-chisquare","05.02-ttest","05.03-anova","05.04-regression","05.05-anova2","06.01-bayes","06.02-epilogue","bibliography","landingpage"],envversion:{"sphinx.domains.c":2,"sphinx.domains.changeset":1,"sphinx.domains.citation":1,"sphinx.domains.cpp":3,"sphinx.domains.index":1,"sphinx.domains.javascript":2,"sphinx.domains.math":2,"sphinx.domains.python":2,"sphinx.domains.rst":2,"sphinx.domains.std":2,"sphinx.ext.intersphinx":1,"sphinxcontrib.bibtex":7,sphinx:56},filenames:["01.01-intro.ipynb","01.02-studydesign.ipynb","02.01-getting_started_with_python.ipynb","02.02-more_python_concepts.ipynb","03.01-descriptives.ipynb","03.02-drawing_graphs.ipynb","03.03-pragmatic_matters.ipynb","03.04-basic_programming.ipynb","04.01-intro-to-probability.ipynb","04.02-probability.ipynb","04.03-estimation.ipynb","04.04-hypothesis-testing.ipynb","05.01-chisquare.ipynb","05.02-ttest.ipynb","05.03-anova.ipynb","05.04-regression.ipynb","05.05-anova2.ipynb","06.01-bayes.ipynb","06.02-epilogue.ipynb","bibliography.md","landingpage.md"],objects:{},objnames:{},objtypes:{},terms:{"000":[0,4,10,11,13],"00000":4,"000000":[4,5,10,13],"0000000":[],"00000000":[],"000000000":[],"000000000000000000000000136":11,"0000000001":11,"0000032":13,"001":[9,11,13],"00135199382707":[],"002":[],"0024":1,"00248896":[],"00404299296417":[],"005":9,"005833":4,"005833399":[],"006":4,"00601151900518":[],"007":[],"007499":[],"010434":4,"011492":4,"01152043":[],"011922871882469877":13,"013":[],"01379741":[],"015884":4,"016":4,"016644":[],"017532279773745":[],"01807260939072272":13,"01929769888727":[],"019586288325765":[],"019d":[],"01eff593ab66":[],"020409":4,"02077226":[],"020978735677851718":11,"021":11,"02126474":13,"02183303":[],"02342152288259":10,"023666733139155":[],"023845743764939753":13,"024978":10,"025":[10,13],"02514322152811":[],"02745433":[],"02807813330601":[],"02949514355197":[],"029c62871301":[],"030000":4,"031":1,"0326992":[],"03418661":13,"036":[],"03608366":[],"03614521878144544":13,"03643339":10,"03915839942863":10,"03921221":[],"03935379301905":[],"03970372":[],"041595":[],"04182658946993":[],"0424d1d3":[],"04252949":13,"043":4,"04499342":[],"04539b4d4399":[],"046821":4,"0471a470108d":[],"048173":[],"049":[11,13],"049200":4,"04967":4,"049904":[],"050000":13,"051":[11,13],"05360997":13,"054":13,"05555555555556":13,"055556":13,"05686539981934":[],"057":11,"05705482891125":[],"05751208":[],"05822858861367":[],"058309":4,"058309485":[],"05830949":[],"05922473578477":10,"06054205":[],"060789":[],"06132891":[],"061329":4,"062":11,"0642512":[],"0643495578651616":4,"06434955786516161":4,"06816586":[],"068166":4,"06e1":10,"070000":4,"0722941558371":[],"07344737":[],"073636":[4,5],"073636359108274":4,"074232":4,"07436796218789":[],"074844":[],"076479":4,"076796665":[],"07679667":[],"076797":4,"0769629":[],"0787":[],"078784":[],"079791106564333":10,"080842":[],"08301893":[],"08323040323924":[],"0877058441629":[],"090232":13,"09083859":[],"09110059":[],"09148816719804":10,"09374202":[],"09526109998637":[],"09690401":[],"09817021869561":[],"09825660326588":[],"098408":4,"09dd6645":[],"09de7a5f":[],"09e25dbb":[],"0b25dde20c3":[],"0b6b":[],"0b6e4918":[],"0b70":[],"0ca4c902e18":[],"0dbb3e59":[],"0e79":[],"0edf":[],"0f20":[],"0th":6,"0x104f641c0":[],"0x105028e80":[],"0x105bd6580":[],"0x107adf640":[],"0x107d8aee0":[],"0x108936a60":[],"0x10b7ff730":[],"0x10bf73400":[],"0x10bf79130":[],"0x10db54490":[],"0x10dcb70a0":[],"0x10e730f40":[],"0x10f47eb20":[],"0x1111bb340":[],"0x111a2c580":[],"0x111cd6d60":[],"0x1141189d0":[],"0x1145ce1f0":[],"0x11539afa0":[],"0x116b7cb50":[],"0x116e25d60":4,"0x118c67940":[],"0x119810790":[],"0x119bf8d30":[],"0x11ab0ef40":[],"0x11ba1d6a0":[],"0x11c9d64f0":[],"0x12df8a3d0":[],"0x12efa5a00":[],"0x12f5196a0":[],"0x12f79f2b0":[],"0x12f883d90":[],"0x12fc07dc0":[],"0x130acaf70":[],"0x130efda30":[],"0x131878bb0":[],"0x131faf610":[],"0x1320ecd00":[],"0x132356d60":[],"0x1324848b0":[],"0x132780fa0":[],"0x132890f70":[],"0x1328b4ca0":[],"0x132c0f8e0":[],"0x133e60f40":[],"0x133ea0d90":[],"0x1341032b0":[],"0x13421a460":[],"0x134424220":[],"0x1344887f0":[],"0x1347150a0":[],"0x1348a5430":[],"0x134a1cfa0":[],"0x134b42730":[],"0x134b4bac0":[],"0x134b8f340":[],"0x134bf6f40":[],"0x134c4cb20":[],"0x134fcf0d0":[],"0x135151df0":[],"0x13515f460":[],"0x135505be0":13,"0x1355b60a0":13,"0x135c8feb0":10,"0x135d13220":[],"0x135f21310":[],"0x1362803a0":[],"0x1362da3a0":[],"0x1362e02e0":[],"0x1363a16a0":[],"0x136766760":[],"0x136a7ef70":[],"0x136ecc3a0":[],"0x137054610":[],"0x13706d430":[],"0x1370dd250":[],"0x137613f10":[],"0x137bb82b0":[],"0x137c10b80":[],"0x137c217f0":[],"0x137c904f0":[],"0x137fef040":[],"0x1380475b0":[],"0x13804b280":[],"0x1380f2b20":[],"0x138112f70":[],"0x1381a4520":[],"0x138449af0":[],"0x1384ffee0":[],"0x13869d8b0":[],"0x13871f130":[],"0x138731ac0":[],"0x1387b3070":[],"0x13881c940":[],"0x1388e1220":[],"0x138913190":[],"0x138a28280":[],"0x138bbc880":[],"0x138bcd370":[],"0x138dc27f0":[],"0x138e1c520":[],"0x138eb2310":[],"0x139315bb0":[],"0x1395244c0":[],"0x13998ae50":[],"0x139b96f40":[],"0x139c90340":[],"0x13a17fe20":[],"0x13a775430":[],"0x13a7df280":[],"0x13a91f460":[],"0x13ad5f550":[],"0x13b0f4fa0":[],"0x13b2ee6d0":[],"0x13b563cd0":[],"0x13b6a9250":[],"0x13b6dcf40":[],"0x13ba01af0":[],"0x13ba16280":[],"0x13bbb7df0":[],"0x13c0aac70":[],"0x13c407ee0":[],"0x13c5bd6a0":[],"0x13c683d60":[],"0x13c7b74c0":[],"0x13c9b6070":[],"0x13ca8ca60":[],"0x13cc53730":[],"0x13cd06490":[],"0x13cd8ad60":[],"0x13d1e5520":[],"0x13d369d60":[],"0x13d3eda90":[],"0x13d4ab070":[],"0x13d5873d0":[],"0x13d6987f0":[],"0x13d90e190":[],"0x13dab3340":[],"0x13e0c4880":[],"0x13e1214f0":[],"0x13e6088e0":[],"0x13e6516d0":[],"0x13efe3c70":[],"0x13f018760":[],"0x13f8859a0":[],"0x13fc0da60":[],"0x13ff8d280":[],"0x13ff9a9a0":[],"0x140383340":[],"0x140d0bd00":[],"0x1417afe50":[],"0x141af5f70":[],"0x141ca25e0":[],"0x7fb554040320":3,"0x7fd3d2cdfc20":[],"0x7fe46813ba90":[],"100":[0,1,3,4,6,9,10,11,13],"1000":[1,4,9,10],"10000":[9,10,11],"100000":[4,13],"1000000":10,"10000000":9,"100cc72e6535":[],"100th":6,"101":10,"10109718805638757":4,"102":10,"1024":11,"103":[10,19],"104":[],"105":10,"105084137429943":10,"105084137429944":10,"105248":[],"10577300630605":[],"106":10,"107":10,"108":[0,5],"10837250197972":[],"109":10,"10th":[4,8],"110":[4,10],"111":10,"111686f274e2":[],"112":10,"112154266955443":[],"1123667414013":[],"113":10,"1135102194217":[],"11352343709996":[],"114":[],"114494":4,"115":[10,19],"1151":19,"11543239":13,"116":[4,5],"11646993":[],"117":[4,10],"1172":19,"118":10,"119":10,"11a110c312b2":[],"120":[5,10],"121":10,"122":10,"123":4,"1231055273238":[],"124":[],"124228":[],"1242645786248002":13,"125":10,"126":[],"127":[],"128":[],"129":[],"129391846315059":[],"12ae68273d06":10,"12f6a9b78854":[],"130":[],"13066559719219":[],"131":[],"132":[3,4],"133":10,"13486720826272":[],"135":[],"136":4,"137833":[],"138":[],"13814737543848":[],"139":[],"13th":8,"140":10,"1415":3,"14513279173728":[],"145558":4,"1455581":[],"14555814":[],"147":[],"14778526288103":[],"148318398823882":[],"14967680216347":[],"14a9":[],"150":[1,4],"15131168":[],"152174":4,"15237303113537":[],"15264631":[],"15293958344104":10,"154f":[],"155418":[],"15864518":[],"159":4,"15978061608057":[],"15bd4ff623a9":[],"15c588d0eaae":2,"15th":10,"16420029223272":[],"1645":[],"16496580927726":4,"165":4,"1664897805783":[],"167":9,"1672629":[],"16965322610622":[],"171":4,"1713":10,"171322032470227":[],"172":4,"173":4,"173127":10,"17385161646514":[],"174":4,"17467105":[],"175":4,"17535285392665":[],"176":[4,5],"177":5,"18018727663605":[],"181503":[],"1825667":[],"183":4,"18332":[],"1854":5,"186":5,"187":19,"18725684911405":[],"18728588770138":[],"18fab7820478":[],"1902517605674":[],"1907":1,"1908":13,"191":0,"1911":19,"191144":[],"1917431a980f":[],"1920":1,"1920872":[],"1922":[11,19],"1937":4,"1938":[1,4],"1945":4,"1946":[0,19],"1955803":[],"1963":19,"1967":19,"1968":2,"19689447267619":[],"1971":[],"1973":[0,19],"1975":[6,19],"197569402522694":[],"1976":11,"1977":[],"19804714297581":[],"1983":19,"1984":19,"1986":7,"1987":[4,5],"1988":19,"19c4":[],"1d79":[],"1e3a6be03e39":[],"1e4eab12":[],"1f9614a14a60":[],"1st":[1,2,8],"200":4,"20000":[],"200000":[4,10],"2002":[1,4,19],"2003":1,"2004":[5,19],"2005":[5,19],"2006":19,"2007":0,"2008":1,"2010":[1,4,5,9,19],"2011":[1,5,19],"2014":19,"2016":0,"2018":1,"2022036":9,"20224782":[],"2048":9,"20842152145724":[],"20th":[9,11],"21061879":[],"212737":[],"21548225636099":[],"21752682":[],"219ddd80":[],"21st":0,"2213698410715347":13,"222":19,"22213677386856":[],"22235022":[],"222672621395798":[],"22659769":[],"228001905654434":[],"22f8":[],"230":9,"233454":[],"234":3,"2345":[],"23579970776728":[],"237":19,"23802375287762":[],"2382c84d04fd":[],"238842":[],"23908691097166":[],"24159f7b86d4":[],"24191232":[],"24381557":[],"2446":[],"2450":2,"245021001876143":[],"24536718309507":[],"24657209":[],"24930821":9,"24b97ad0d3d7":[],"25000":4,"250000":4,"251":19,"25471286700693":13,"25480038":[],"255":[],"255188":[],"25753056":[],"258331":[],"259":13,"259605535157681":13,"25th":[4,5],"26062329":[],"26215716":10,"265855":[],"2661476":9,"268202":[],"271319":[],"27173353050394":[],"272":0,"2733d20e1661":[],"2765451":[],"27661201499386":[],"27739475":[],"27841143860383":[],"279523":[],"281552":13,"282385719772392":[],"285000":4,"28685666":[],"28686088":[],"2870":[],"290527":13,"292500":4,"29288462":[],"293833":[],"295":19,"2_p":13,"2a09":[],"2c15d7a0":[],"2d6bbc3f":[],"2eacbe14":[],"2eb4759cd4c":[],"2f68":[],"2f9b":[],"2nd":[1,19],"300":[1,5],"3000":2,"300000":[4,13],"30107606819503":[],"301136":[4,5],"30113636363637":4,"3021480":4,"30247052":[],"306":19,"30622a0f322f":[],"309":[3,19],"30c2ad06":[],"310189":[],"3114fd44":[],"31192405433491":[],"314cb8c72701":10,"315019":[],"3151aa31a887":[],"316655":[],"31775495":[],"318654":[],"31ec45c6":[],"320670056126948e":13,"322544":[],"3227dad5f5a0":[],"324":4,"325":0,"326348":13,"32642464":[],"327":5,"327322":[],"3279":10,"328":19,"328028":[],"32b0":[],"32b38c54":[],"32c1e318a0b5":[],"331":19,"33162749802028":[],"33174961835198":[],"333":[2,4,6],"334":19,"33rd":6,"3404275653649":[],"341":0,"342":[],"34201924105093":[],"344":8,"345":19,"34648913":9,"346834":[],"3477":2,"34841741999587":[],"34fe5eabced4":[],"350":[2,13],"353":3,"3538706638441":[],"35615280000385":10,"36324512":[],"36569752995766":[],"367872":[],"368":19,"36826646949605":[],"37497756":[],"375":0,"37594501":[],"376":4,"37707609":[],"3824c47e7b89":[],"38498210704823":[],"385000":13,"38887014":[],"38965611715034":[],"390295":[],"391":[],"39106267174839":[],"393":0,"3938":[],"39398848099482":[],"395876":[],"395945":[],"39768396673382":[],"398":19,"39878609":[],"39909089":[],"3bd0":[],"3c11":[],"3c5c0d4116c9":[],"3cf564f2":[],"3de11ed2a147":[],"3e593bc2":[],"3ecc4ad645bb":[],"3rd":[1,19],"400":[4,5,8],"40000":4,"400000":[],"4003":[],"401":[],"402c":[],"403819175526497":[],"40384719999614":10,"404":19,"405":4,"405612":13,"4056320378121":[],"406136267820491":[],"40763325859871":[],"4084":[],"40d6":[],"40f6b37c":10,"4112":[],"411c":[],"411d":[],"4124":[],"41322915":[],"413f":[],"417":0,"417ca995":[],"41891313174203":[],"4192":[],"41cf":[],"41dc":[],"41f2":[],"42137238":[],"422b":10,"4249832892949":[],"425000":4,"42663639956145":[],"4291":5,"4292":5,"4293":5,"4294":5,"4295":5,"42db":10,"42fb":[],"43076225311911":[],"430d":[],"431":3,"431f":[],"4321":0,"4323":[],"4346":[],"434fe2c9":[],"4358":[],"43781029194585":[],"438009":[],"4385":[],"4397":[],"43a7":[],"43c3":[],"44042383577023":[],"44158856139617":[],"441c":[],"44218970805414":[],"444141":[],"444575":[],"44684489":[],"44893732825162":[],"44923774688088":[],"4494":[],"44b5":[],"4567":[],"4570":[],"457273":[],"4575312":9,"457c":[],"4595724346351":[],"45bb":[],"45d1":[],"45da":[],"460000":4,"4610025":9,"4610795":9,"46177141138634":[],"46473890001363":[],"464b":[],"465292":[],"46656561":[],"4687b280":[],"46bd":[],"46e0":10,"471008":[],"471c5998e834":[],"475436088339379":13,"475f":[],"4785":[],"47985760094015":[],"47a9617c07f0":[],"47c5":[],"47d0":[],"47d8":[],"47e4":[],"47f2":[],"4809":[],"48197624712238":[],"4826":4,"4827":[],"485127007903577":[],"4855669223382983":13,"488088":[],"48906353198835":[],"48bd":[],"48ec":[],"48fd":10,"49077598":[],"4913":[],"497619":[],"4979":[],"49b3":[],"49e0":[],"49e3":[],"49ee":[],"4a44":[],"4a81":[],"4a813ce2e90a":[],"4ad1":[],"4b00":[],"4b54a7b31a35":[],"4b86":[],"4b9f":[],"4ba3":10,"4bac":[],"4bcd":[],"4be0":[],"4be2":[],"4c15":[],"4cd2":[],"4ce5":[],"4ce6":[],"4cf1":[],"4d07":[],"4d095a9bf3d1":[],"4d27":[],"4d56":[],"4d61":[],"4d7a":[],"4dd9b0dca8c6":[],"4e2a":[],"4e5428219d97":[],"4e9c":[],"4ec1":[],"4ee2":10,"4ef5":[],"4efec19b21b5":[],"4f02":[],"4f2a":[],"4f42d043":[],"4f65b33001b2":[],"4f7d":[],"4f87":[],"4f8d":[],"4fd8":[],"4fe3":[],"4th":2,"500":[4,5,11],"50000":10,"500000":[4,5],"50070230111274":[],"500a":[],"501852":[],"504":[],"5041691240370938":13,"50499343":9,"508880dfa229":[],"50th":4,"5104":[],"514097":[],"51575876538577":[],"516059467297646":10,"51726195":[],"51840029644295":[],"51b4c8f8":[],"51e324fc7254":[],"520614752375915":13,"52086912":[],"521":[],"52197825786087":[],"521c3ce2":[],"52579322606445":[],"525b8b1eaeea":[],"533333":13,"53333333333333":13,"533854":4,"533c3bcdb859":[],"53501671070511":[],"53644252":[],"539539450000001":3,"53992f56":[],"53f3d79e401b":[],"54296008":[],"545":[],"54974968":[],"550":2,"552":3,"554":[],"555103":[],"55fbbb17":[],"560":0,"56064620698095":[],"5619":10,"56195285702418":[],"562fe19a":[],"565964":4,"56648022":[],"567802669":[],"56780267":[],"567803":4,"570814":[],"572706":10,"573214":[],"575495":[],"575829":13,"57822143":[],"57826317":9,"580":11,"58064012543642":[],"58245324413445":[],"5849b013":[],"58656066":[],"5865607":[],"586561":4,"590":[],"593":0,"59da466cd70f":[],"5a56":[],"5ade8840c87a":[],"5baafabf4e64":[],"5be0c505":[],"5c86":[],"5ff9d08b":[],"5th":10,"600":11,"60000":[],"600000":[10,13],"600px":11,"601312462441985":[],"60137":[],"60220688933644":[],"60453660558522":[],"6061293361559":[],"607289":[],"609":9,"60994147639813":[],"60d1731417b7":10,"610":9,"610000":4,"613248":[],"614723":4,"61472303":[],"61485677847188":[],"615192":[],"616137":13,"617321676258056":[],"61cf3d61":[],"62034303842519":[],"6213":4,"624":3,"62454125":9,"625":2,"6250671":9,"62783751403471":[],"627949":4,"6282":[],"6287":[],"62891818":[],"62940661917412":[],"63132613":[],"634":3,"63409284352663":[],"635":3,"635000":4,"63674390464717":[],"63868541943094":[],"639498":4,"63949845":[],"6394985":[],"64324782236976":[],"644854":13,"645512":[],"64762696886464":[],"649363":[],"65281f7c":[],"65579917":[],"655825":[],"65689746":[],"65965696157481":[],"660335028063474e":13,"662":3,"66243843":[],"666666":[],"667":6,"66th":6,"672807771136599":[],"675":4,"6769080804625":[],"677":19,"67866353":[],"679":4,"680":19,"683321":[],"68557758":[],"686729":[],"691012":[],"6927102":[],"69274315088595":[],"693064":[],"696153":[],"696f1210f32f":[],"697":19,"698084151486567":[],"6a39a87b":[],"6af2d884":[],"6b6d":[],"6d27a6aa":[],"6d76":[],"6dce":[],"700000":13,"701":19,"70452372014537":[],"71000":4,"7183":3,"7189413":[],"72131458056907":[],"72231603326618":[],"72309191953751":[],"730527":[],"73294549":[],"7337826a4685":[],"737963401065596":[],"73923279574828":[],"73935987456358":[],"740000":4,"74076720425171":[],"74108686825797":[],"7486295728919304":13,"750":4,"750000":[4,5,13],"751886920458144":[],"754921071175007":[],"75578486208585":13,"75595700703582":[],"756":[],"75th":[4,5],"7608":[],"76084160057138":10,"7640042619421049":[],"76431b569704":[],"766bd3cf":[],"77032319783652":[],"771594":[],"772600263582405":[],"774918":13,"77668129338701":[],"779712104491898":[],"77d39d61":[],"7804075289401982":4,"783555899725448":[],"785000":4,"785299ab":[],"791a5ddf":[],"79278116248614":[],"79341":[],"795":9,"7a5eb02c":[],"7ac79054":[],"7b0f24b5":[],"7bf862e18266":[],"7c840be9d48d":[],"7cc6":[],"80000":[],"800000":4,"80014239905985":[],"80019238829465":10,"800723274871205":[],"8009":[],"803261864666974":[],"803408":[],"8044":[],"80647656290003":[],"809191273277598":[],"812312089763958":[],"813457":[],"816":4,"8162365060002428":4,"8162867394895984":4,"81642051634484":4,"8165214368885028":4,"816e":[],"816f0097":[],"817":4,"817956":[],"81c5":[],"820431":10,"82077526421523":10,"825":0,"82606624234876":[],"8290da4f":[],"8291":[],"82933440280782":[],"8307d73c":[],"83157847854275":[],"83247670748433":[],"834512987013":4,"83576592197558":[],"835814885142842":[],"839394488381325":[],"839742":[],"839e":[],"840000":4,"840608":[],"8420":[],"844":[],"8442":0,"84421513791415":13,"84423407802443":[],"844a":[],"84590715647336":[],"8461":[],"84675":[],"8482816486408":[],"84896776":[],"8492":[],"84928176192457":[],"84974823943259":[],"84a8":[],"850000":4,"85050485644803":[],"851574":4,"85422699369396":[],"854708":[],"854854723088971":[],"85786322613146":[],"8591313907397005":13,"85946705720211":[],"85th":10,"860011":[],"87155738":[],"8744":[],"8778708198222":[],"88000072":[],"881944541256573":[],"883333":4,"883954":[],"8845":[],"88721883751387":[],"88839966861514":[],"888659358775472":[],"89336360043855":[],"89470711547601":[],"8949158625700564":10,"894915862570057":10,"89887521":[],"898a0657":[],"89951468":[],"899515":4,"89dd":[],"8a13e497":10,"8a53":[],"8a5a":[],"8aa8":[],"8ad32b31":[],"8b18bf777617":[],"8b1feb08ff56":[],"8b4b":[],"8b92":[],"8bc0bee751cd":[],"8bc15872bd39":[],"8bff":[],"8c61":[],"8c7826c2":[],"8c83":[],"8c88":[],"8d863ff96bb1":13,"8ef1":[],"8f9f":[],"900000":13,"90165863":[],"903384":4,"9033840374657273":4,"903442":4,"90344244":[],"903442442":[],"90588738":[],"90664850568817":[],"908712":[],"9094019658612525":4,"90944411144406":[],"91071823807542":[],"91074326085924":4,"91148659677621":[],"911521":[],"9118":[],"91427124":[],"91430247004234":[],"91648970627955":[],"916b":[],"91802174213913":[],"919674372409807":[],"91fc":[],"92214ba0":[],"925":3,"92849522":[],"928e":[],"92af":[],"92cb7f49":[],"933d9d08126b":[],"93634803022955":[],"936367228048367":[],"938b":[],"9396542167696":[],"93991042213851":[],"93f6003e1b4":[],"94183567":9,"942":6,"94282227653469":[],"9429":[],"9457":[],"950":4,"950000":4,"9508686092602991":13,"9508925471332":[],"95297701":[],"954451":[],"958188":[],"95996398":10,"959964":13,"96020126":10,"965200":4,"966636":10,"96b7":[],"96c7dabb":10,"96f0":[],"96f1":[],"97034388284968":[],"970716532238185":[],"970c":[],"9718168904958":4,"974ee0c2":[],"975":[10,11],"97547627985463":[],"976923":4,"977535679139253":[],"97873526":13,"97892393180497":[],"980000":13,"981176628376893":[],"9818297813044":[],"98200487":[],"98236724494748":[],"98343998090161":[],"984159075050805":[],"98597539":9,"98612011347937":[],"9880771":[],"989422804014327":9,"98ae8990":[],"990351575034936":[],"99093646801164":[],"990c0101f8fd":[],"9913048":[],"9957":[],"998942":13,"999":6,"9998":9,"99998":9,"9999983396649719":13,"9999999999999999":4,"99ef":[],"9ae08854":10,"9bda":[],"9bf47972c6dd":[],"9c0c":[],"9d40":[],"9d4cea9d":[],"9db7f240":[],"9de9":[],"9eddbab0c553":[],"9f95":[],"9fe7cbb7":[],"9th":[],"\u0105":11,"\u03b1":[],"\u03b8":[9,11],"\u03bc":9,"\u03c3":9,"\u3164\u3164":4,"\ufb01le":[],"\ufb01nd":[],"\ufb01nding":[],"\ufb01rst":[],"abstract":[1,2,3,4,10,11],"break":[1,4,5,6,9,13],"case":[0,1,2,3,5,6,8,9,10,11,13],"catch":[1,3,11],"class":[0,1,2,3,4,8,9,10,13],"default":[1,3,4,5,10,13],"di\ufb00er":[],"export":[3,5],"final":[0,1,3,4,5,9,10,11,13],"float":[2,3,4],"function":[0,1,2,3,4,5,7,9,10,13,20],"import":[0,1,2,4,5,6,8,9,10,11,13],"int":[3,10],"k\u00fchberger":19,"long":[1,2,3,4,5,9,10,11,13],"new":[0,1,2,3,4,5,6,7,8,9,10,11,13],"null":[1,13],"public":[1,5,9,10,19],"return":[1,2,3,4,5,9,10,11],"short":[0,1,3,4,5,6,7,9,10,11],"speci\ufb01":[],"speci\ufb01c":[],"stu\ufb00":[],"super":7,"switch":[4,5,8,9,13],"throw":[5,9],"true":[0,1,2,3,4,5,6,7,8,9,10,11,13],"try":[0,1,2,3,4,5,6,7,8,9,10,11,13],"typeof":[],"var":[4,10,13],"while":[0,1,2,3,4,5,6,8,9,11,13],AND:3,Adding:9,Age:1,And:[0,1,2,3,4,5,6,7,8,10,11,13],Being:1,Bus:1,But:[0,1,2,3,4,5,6,7,8,9,10,11,13],DOS:[],Doing:[7,10],Drs:13,Eyes:1,FOR:[],For:[0,1,2,3,4,5,6,7,8,9,10,11,13],Going:[5,8],His:11,IDE:2,IQs:10,IVs:1,Its:13,Los:1,Not:[0,1,2,3,4,9,10,11,13],One:[1,2,3,4,5,6,9,10,11,19],SAS:[],Such:5,That:[0,1,2,3,4,5,6,7,8,9,10,11,13],The:[2,3,5,6,7,8,19,20],Their:8,Then:[1,3,4,5,7,13],There:[1,2,3,4,5,6,7,8,9,10,11,13],These:[0,1,2,3,4,5,6,9,10,13],USE:[],Use:4,Used:4,Useful:[],Using:[0,5,10,11,13],WITH:[],Will:1,With:[0,3,4,5,6,7,9,10,11,13],Yes:[3,8,10,11,13],_95:[],__init__:3,_between:[],_blank:[],_class_:[],_cohen:[],_data:[],_degre:[],_factors_:[],_formula_:[],_gener:[],_independ:[],_lists_:[],_method:[],_method_:[],_mode_:[],_nonparametr:[],_one:[],_pair:[],_pool:[],_rel:13,_remove_:[],_shapiro:[],_standard:[],_student:[],_two:[],_type_:[],_welch:[],_within:[],_work:[],_workspace_:[],a13ba159:[],a15:[],a187:[],a18d91877a5a:[],a1ac068357bc:[],a1b3a402:[],a1c7:[],a1eb:[],a21faee3aca1:3,a232:[],a28c:[],a2a7:[],a2d93521:[],a372e0c08cdb:[],a3e3:[],a451:[],a460:[],a491:[],a4bb:[],a4f388fc:10,a54b:[],a632:10,a656:[],a66:[],a6d1:[],a736:[],a74f0709e759:[],a849:[],a8d84704:[],a910:[],a98172db74f9:[],a9e3:[],a9ed06cbf49a:[],a_i:13,aa5fdd172618:[],aa69:[],aaaand:3,aad:4,ab13:[],ab34:[],ab9b70e376da:[],abandon:10,abbrevi:[3,4,5],abc:[0,4],abe1:[],abil:[1,3,7,10,11],abl:[0,1,2,3,4,5,7,8,10,11,13],ablin:[],about:[0,2,4,5,6,8,9,10,11,13],abov:[0,1,2,3,4,5,6,8,9,10,11,13],abroad:11,abs:[4,6,11],absenc:[],absolut:[1,2,3,6,11,13],absurdli:2,abx:[],ac00:[],ac717453c36f:[],academ:[9,10],accept:[1,3,4,8,9,10,11],access:[2,3,10,20],accident:[1,9,13],accommod:13,accompani:[],accomplish:7,accord:[4,11,13],account:[0,1,10,11,13],accur:[1,2,4,10],accuraci:13,accus:1,achiev:[3,4,5,6,9,10,13],acknowledg:1,acronym:[4,10],across:[0,1,2,4,8,9,10,11,13],act:[1,2,10,13],action:[4,11],activ:1,actual:[0,1,2,3,4,5,6,7,8,9,10,11,13],ad4328bb:[],ad65:[],ad6858b801e8:[],ada84:[1,19],ada:[],adair1984:[],adair:19,adapt:[7,13,20],add:[1,2,3,4,5,7,9,10,13],addarrow:[],adddistplot:[],added:[1,2,3,5,11,13],addict:0,adding:[3,13],addit:[1,2,3,4,13],addition:13,address:[0,1,3,4,10,11],adelaid:[4,5,9,10],adjac:11,adjust:[5,10,13],administ:[1,4,10],administr:9,admiss:[0,19],admit:[0,8,11,13],admittedli:[9,10],adopt:[1,6,7,9,10,11,13],adult:[1,6],advanc:[0,1,2,7,9,11,13],advantag:[1,3,4,9,10,11],advers:1,adversari:11,advic:[3,5,6,11,13],advis:6,advisedli:3,ae1d:[],ae2b:[],ae370d91032a:[],aesthet:5,af05d5f5cd6f:[],af86:[],af9a:[],afd1:[],affair:[0,1,10],affect:[4,11],afford:[0,4,9],afl24:[],afl2smal:5,afl:[4,5,13],afl_fig:4,afl_finalist:[4,5],afl_margin:[4,5],aflsmal:[],aflsmall2:5,afraid:[1,8],after:[0,1,2,3,4,5,7,8,9,10,11,13],afterward:[1,11],again:[1,2,3,4,5,7,8,9,10,11,13],against:[0,4,5,6,8,13],age:[1,3,6,10],agent:9,ages:[7,10],agg:9,aggreg:[0,1],ago:2,agpp:[],agre:[0,1,4,6,8,9,11],agreement:[0,6],agricultur:13,aha:5,ahead:[1,4,10,11],aim:[1,8,9],aka:3,ala:5,alan:1,albeit:7,alert:11,alex:[],algebra:10,alic:1,alien:9,align:[],aliv:[4,10],all:[0,1,2,3,4,5,6,7,8,9,10,11,13],alleg:1,allevi:[5,10],alloc:13,allow:[0,1,2,3,4,5,6,9,10,11,13],allud:[],almost:[0,1,2,3,4,5,6,9,10,11,13],alon:[0,13],along:[0,2,4,5,9,11,13,20],alongsid:3,alp:9,alpha:[9,10,11,13],alphabet:4,alreadi:[1,2,3,4,5,6,7,8,9,10,11,13],alright:5,also:[0,1,2,3,4,5,6,7,9,10,11,13,20],alter:[1,9,10,11],altern:[1,2,10,13],although:[1,2,3,4,5,6,7,9,10,11,13],alwai:[0,1,2,3,4,5,9,10,11,13],amass:10,amaz:[0,8,9],amazingli:11,ambigu:[4,11,13],american:19,among:[0,1,9,10,11,13],amount:[0,1,4,5,10,11,13],amstat:10,anaconda:[2,3],analog:[5,11,13],analogu:4,analys:[0,1,2,4,5,6,7,9,11],analysi:[0,1,2,3,4,5,6,7,8,9,10,11,13,19],analyst:[4,5,7],analyz:2,anastasia:13,anchovi:1,andrei:9,andrew:1,anecdot:0,angel:1,anger:2,angri:0,ani:[0,1,2,3,4,5,6,7,8,9,11,13],anim:19,annihil:5,annoi:[1,2,4,5,7,10,11,13],annot:11,annoyingli:1,anoth:[0,1,3,4,5,6,7,8,9,10,11,13],anova:[6,13],ans73:[4,19],anscomb:[4,19],anscombe1973:[],anscombe1973graph:[],anscombesquartet:[],answer:[0,1,2,3,4,5,8,9,10,11,13],ant:[],anti:1,anxieti:[1,13],anxifre:4,anymor:[3,4,9,13],anyon:[1,4,7,10,11],anyth:[0,1,2,3,4,5,6,7,8,9,10,11,13],anywai:[2,5,9,13],anywher:[2,3,4,6,10],apart:[1,4],api:10,apolog:11,appar:[0,1,5,13],appeal:[1,5,11],appear:[1,2,3,4,6,9,10,11,13],append:[4,10,11],appl:[3,4,13],appli:[0,1,4,5,6,7,9,10,13,19],applic:[0,2,5,9,10,11],appreci:2,apprehens:1,approach:[0,1,4,5,7,8,9,10,11,13,20],appropri:[4,9,10,11,13],approxim:[0,1,2,4,6,9,10,11,13],arang:[9,11,13],arbitrari:[0,1,4,9,13],arduino:9,area:[0,1,5,9,11,13],aren:[0,1,2,3,4,5,6,7,8,9,10,11,13],arg:[],argblock:[],argu:[1,2,3,4,10,11,13],arguabl:[2,10],arguement:5,argument:[0,1,4,5,6,9,11,13],aris:[0,1,4,6,9,10,13],arisen:1,arithmet:2,arm:[4,10],around:[1,3,4,5,7,8,9,10,11,13],arrai:[1,4,6,10,11,13],arrang:5,arrest:9,arriv:[1,9,10,11],arrog:13,arrow:[],arrowprop:11,arrowstyl:11,ars:13,arsen:9,art:[2,7,8],arthur:8,articl:[0,1,4,9],artifact:19,artifactu:1,artifici:2,artist:5,asbesto:1,ascend:4,ascii:[],ashlei:11,asia:19,asid:[0,1,10,11],ask:[0,1,2,3,4,6,7,8,9,10,11,13],aspect:[1,4,13],assembl:2,assert:[],assess:[4,11,13],asset:4,assign:[0,1,2,3,4,6,9,11,13],assist:3,associ:[2,6,8,9,10,11,13],associationtest:[],assum:[1,2,3,4,5,8,9,10,11,13],assumpt:[1,4,5,6,9,10,11],astyp:10,asymmetr:4,asymmetri:[4,11],ate:[],athough:4,atroc:4,attach:[4,6,9,10,11],attempt:[0,1,2,4,5,7,9,10,11,13],attent:[1,3,4,5,9,10,11],attest:11,attitud:1,attract:[4,5],attribut:[3,9,11],auckland:4,auden:0,audienc:[4,5],australia:[4,10,11],australian:[0,1,4,5,9,10,13],author:[0,1,3,5,13],authori:1,autism:1,autocorrect:2,autom:2,automat:[1,5,10],automaton:[2,4,5],auxiliari:1,avail:[2,3,4,5,8,10,11],availab:2,avenu:3,averag:[1,4,9,10,11,13],averagesamplemean:10,averagesamplesd:10,avoid:[1,4,7,8,11,13],awai:[1,4,5,6,7,8,9,10,11,13],awar:[1,2,3,4,5,9,10,11,13],awesom:[1,5,6,9,13],awesome2:13,awful:1,awkward:[1,4],awri:1,ax0:13,ax1:[4,5,13],ax1a:[],ax2:[4,5,9,10,13],ax2a:[],ax3:[4,5,13],ax3a:[],axes:[4,5,9,10,13],axessubplot:[4,5,9,10,11,13],axez:[],axhlin:[10,13],axi:[4,6,9,10],axisgrid:4,axvlin:[10,13],b0080e30277c:[],b025cd0c:[],b0e0:[],b158:[],b19c:[],b2bba8a2:[],b2e1:[],b38bcb52:[],b3a6b9f4780c:[],b458:[],b529:10,b5ac:10,b5d2:[],b62e:[],b65115ca7d06:[],b671:[],b6c2:[],b6cf:[],b6f8e681:[],b75b:[],b78a100a:[],b810:10,b85001a53d3d:[],b869:[],b926:10,b967:[],b970:[],ba17a6df509a:[],babbl:[],babi:[1,4,5],baby_sleep:4,back:[1,2,3,4,5,6,7,9,10,11,13],background:[2,3,5,9,10,13],backslash:3,backward:3,bad:[0,1,2,3,4,5,6,8,9,10,11,13],badli:11,bag:10,balanc:[1,4],banana:3,bank:4,bar:[4,9,10,11,13],bare:[10,11,13],bargraph:[],barplot:[5,9],barston:19,bartlebi:8,base:[0,1,2,3,4,5,6,9,10,11,13,19],baselin:0,basi:[0,1,8,9,10,11,13],basic:[0,1,2,3,4,5,6,10,11,13],basicprob:[],bastard:[],batch:8,bathroom:1,battlefield:11,bay:10,bayesian:10,bb7f:[],bba44a4097c2:[],bbox_inch:5,bd44:[],bd5f:[],bd60:[],bd6739b1e5e8:10,bd7d:[],bd95:[],bear:[3,10,11],beast:[3,4],becam:1,becaus:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],becauw:13,becom:[0,1,2,3,4,5,6,8,9,10,11,13],bed2:[],bedma:2,been:[1,2,3,4,5,6,7,8,9,10,11,13,20],befor:[0,1,2,3,4,5,6,7,8,9,10,11,13],begin:[1,3,5,7,9,10,11,13],beginn:7,behav:[1,3,4,9,13],behavior:19,behaviour:[0,1,2,4,5,9,10,13],behind:[1,2,3,4,5,7,11,13],being:[0,1,2,3,4,5,7,8,9,10,11,13],beings:[1,9,10],belief:[1,8,9,10,11,19],believ:[0,1,4,8,9,10,11,13],bell:[4,9,10],belong:[4,10],below:[0,1,2,3,4,5,9,10,11,13],ben:1,benchmark:4,benefit:7,bereft:13,berk:0,berkelei:0,berkeley2:0,berkelyplot:0,berklei:0,bernadett:13,bernoulli:10,besid:[1,4,9,10,13],best:[0,1,2,3,4,5,6,7,9,10,11,13],bet:[4,8,9,11],beta:[10,11],better:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],between:[0,1,3,4,5,6,7,8,9,10,13,19],beyond:[0,1,4,5,9,10,11,13],bezo:5,bf41:[],bf5c:[],bfae:[],bfbb:[],bfe1:[],bhoconnell75:[0,19],bia:[4,10,11,13,19],bianca:1,bias:[0,8,10,11,13],bickel1975:[],bickel1975sex:[],bickel:19,bicker:3,bicycl:1,big:[0,1,2,3,4,5,7,8,9,10,11,13],bigger:[1,4,9,10,11,13],biggest:[0,1,4,11,13],bill:[4,10],billion:1,bin:[5,6,9,11],binari:[10,13],binom:[9,11],binom_test:11,binomi:11,binomial1:[],binomial2a:[],binomial2b:[],binomplot:[],binwidth:[4,9,10,11,13],biolog:[0,10],biologi:1,birth:1,birthdai:[],bit:[0,1,3,4,5,6,7,8,9,10,11,13,20],bitter:8,bizarr:[0,1,10],black:[4,5,7,9,10,11,13],blackston:11,blah:10,blame:2,blank:9,blatantli:0,blind:1,blindingli:[1,10],blindli:13,block2:[],block:[2,8,13],blood:13,bloodi:[0,1,10,13],blowout:4,blue:[1,5,9,10,13],bluntli:11,blur:1,board:9,bob:4,bodi:4,boi:11,boil:3,bold:[0,1,4],bolt:[],bonu:2,book:[0,1,4,5,6,7,8,9,10,11,13,20],bookdown:20,booksal:3,booksales2:[],booksalescsv:[],bookshelf:9,booktab:[],boost:11,border:10,borderlin:11,bore:[0,1,4,5,6,11],boredom:1,boringli:9,born:[1,10],boston:19,both:[0,1,2,3,4,5,6,8,9,10,11,13],bother:[0,4,7,11,13],bottom:[4,5,9,10,13],bought:13,bound:1,boundari:[6,11],box1987:13,box:[4,10,11],boxprop:5,brace:3,bracket:[2,3,4],brag:5,brain:[1,11],brainer:[],branch:9,bread:1,breakdown:[0,9,10],breath:5,breweri:13,bridg:0,brief:[8,9,11,13],briefli:[1,3,4,11,13],bright:[1,11],bring:[1,3,13],brisban:[4,5],broad:[1,5,6,9,13],broader:[0,1,5,10,13],broadli:[3,4],broken:[4,11],brooklyn:0,brown:1,brows:[],browser:3,brs:10,brush:[5,13],bug:[],bui:2,build:[3,10],built:[1,2,3,5,9,10,11,13],bulldog:[4,5],bump:5,bunch:[3,4,7,9,10,13],bundl:[],bureaucrat:11,buri:[0,4],burn:[4,13],burt:1,bus:[1,9],busi:[1,2,4,5],butter:1,button:[2,9],bypass:[],c39bad49718f:[],c6e5fe9a:[],c8e34490:[],cake:[],calcul:[3,5,6,9,11,13],calculu:9,california:0,call:[0,1,2,3,4,5,6,7,9,10,11,13],calm:1,cambridg:19,came:[0,3,4,5,8,9,10,11],cameron:1,campbel:19,campbell1963:[],campbell2015experiment:[],can:[0,1,2,3,4,5,6,7,8,9,10,11,13],cancer:1,candid:[0,5],cannot:[2,3,4,5,8,9,10,11],canva:5,cap:[9,10],capabl:0,capac:[10,13],capprop:5,caption:[2,4],captur:[1,2,4,5,10,11],car:1,card:[9,11],care:[0,1,3,4,6,7,9,10,11,13],carefulli:[0,1,4,13],carlton:[4,5],carri:1,carriag:8,carrol:8,carv:6,casual:5,cat:3,categor:[1,9,13],categori:[0,1,4,9,10,11],categoris:6,caught:11,caus:[0,1,4,5,7,10,11,13],causal:[1,5,11],caution:[4,6],caveat:[],cba:4,cbc482d2:[],cbt:4,cc834daf:[],cc8ea8bc:[],ccc:8,cd43:[],cdf:[11,13],cell:1,celsiu:1,censor:1,cent:9,center:[4,6,11,13],centr:[4,5,6,9],central:[11,13],centraltend:[],centuri:[0,9,11],certain:[0,2,4,5,6,8,9,10,11,13],certainli:[0,1,4,5,6,7,9,10,11,13],certainti:[10,11],certif:1,cetera:6,cex:[],cfa4aaec:[],challeng:13,chanc:[0,1,3,4,5,7,8,9,10,11,13],chang:[0,1,2,3,5,8,9,10,11,13],chapek9:[],chapter:[1,2,3,4,5,6,7,8,9,10,11,13],charact:[3,4,6,11,13],character:2,characteris:1,characterist:[0,1,4,9,10,11],charg:0,chart:[5,9],chdir:[0,4,5],cheap:1,cheapli:2,cheat:1,cheater:1,check:[1,4,8,9,11],checkmark:[1,13],chemic:1,chemist:13,chemistri:0,chi2:9,chi:[9,11,13,20],chi_square_20:9,chi_square_3:9,chi_square_data:9,chicago:1,chick:0,chico2:[],chico3:[],chico:13,chief:4,child:[1,4,6],childish:1,children:1,chill:4,chip:10,chisq:[],chisqdist:[],chisqreport:11,chisquar:[9,13],chitchat:1,choic:[0,1,3,4,5,6,8,9,11,13],cholera:5,choos:[0,1,2,3,4,5,6,11,13],chop:7,chose:[1,10,11],chosen:[1,2,4,5,10,11,13],chromosom:1,chunk:[5,9],ci_1:10,ci_2:10,ci_:[],ci_int:10,ci_upp:[],cigarett:[0,1],cimean:13,circ:1,circl:5,circumst:[4,11],cirep:10,citat:1,cite:[],citi:[4,9],claim:[1,7,9,10,11,13],clairvoy:11,clarifi:4,clariti:[4,9],classic:[0,1,4,5,19],classif:[1,6],clean:[1,4,5,7],cleaner:11,cleanli:13,cleanup:[],clear:[0,1,2,3,4,5,6,9,10,11,13],clearer:[4,5,13],clearli:[1,2,3,4,5,9,10,13],clever:[1,4,7,9,11,19],click:[3,9],climat:1,clinic:[0,1,4,10],clinical_trial_data:4,clinicaltri:[],clinicaltrial_old:[],clipboard:[],close:[0,1,2,4,5,6,9,10,11,13],closer:[1,2,3,4,10,11,13],cloud:[2,11],clt:[10,13],cltdemo:10,cltdemoa:10,cltdemob:10,cltdemoc:10,club:11,clue:4,clump:13,cluster:10,clutter:5,coars:6,coast:[4,5],coat:[1,10],code:[3,4,5,6,9,11,13,20],coder:5,coef:[],coerc:6,coercion:6,coffe:[],coffer:10,cognit:[1,10,19],coh88:[11,19],cohen1988:13,cohen:19,cohensd:13,cohensdinterpret:[],coher:[3,5,6],coin:[9,11],coin_flip:9,coincid:[0,4,11],col:4,col_0:6,colab:2,cold:[4,10],colleagu:7,collect:[0,1,2,3,4,5,6,9,10,11],colleg:0,collingwood:[4,5],color:[4,9,10,11,13],colour:[1,10,11],coltot:6,column1:4,column2:4,column3:4,column:[3,4,5,6,13],com:[0,3,4,8,13],comb:7,combin:[3,4,5],come:[0,1,2,3,4,5,6,7,9,10,11,13,20],comfort:[1,2,4,5,6,9,10,11,13],comma:3,command:[3,4,5,6,7,9,11,13],commenc:0,comment:[7,13,19],commentari:[0,9],commerci:2,commiss:9,commission:9,commit:[0,20],commmand:[],common:[0,1,2,3,4,5,6,8,9,10,13],commonli:[3,4,9,13],commonplac:13,commonwealth:4,commsec:4,commun:[1,4],commut:[1,10],compact:[4,5,13],compani:[2,9,10,19],companion:8,compar:[0,4,9,10,11],comparison:[0,4,9,10,13],compel:0,compens:0,compet:0,competit:[4,8,11],complain:[],complement:[1,4],complet:[0,1,2,3,4,5,9,10,11],complex:[0,3,4,10],complianc:0,complic:[0,2,3,4,5,6,7,10,11,13],compon:1,comprehens:[1,3,4,6,10],compress:13,compromis:[],comput:[1,2,3,4,5,6,7,9,10,11,13],con:[],concat:9,conceal:5,conced:11,conceiv:11,concentr:20,concept:[0,1,2,4,5,6,7,9,10,11],conceptu:[1,2,9,10],concern:[0,1,2,3,4,10,11],concert:3,concis:10,conclud:[1,9,13],conclus:[0,1,8,10,13],concret:[1,2,4,10,11],condens:4,condescend:10,condit:[1,6,11,13],conduct:[1,9,10,11,13],confess:[],confid:[1,5,9,11,13],confidence_interv:13,confidence_level:13,configur:5,confin:6,confintrepl:10,confirm:[1,10,13],conflat:[11,13],conflict:19,confront:[0,1],confus:[1,2,3,4,5,9,10,11],congress:1,conjunct:7,connect:[1,3,9],connel:19,conscienti:1,conscious:2,consensu:9,consent:10,consequ:[0,1,2,3,4,5,8,9,10,11,13],conserv:2,consid:[1,4,5,6,7,9,10,13],consider:[],consist:[0,1,4,5,8,10,11,13],consol:[4,7],constant:[1,4,9,13],constitut:[0,5],constraint:[5,9,13],construct:[0,3,4,6,9,10,11],consul:5,consult:[1,11],consum:1,contact:[10,11],contain:[2,3,4,5,6,9,10,11,13],contempl:11,content:[6,7,9,11],context:[1,2,4,5,6,9,11,13],continu:[2,3,4,5,6,8,9,10,13],continuousdiscret:[],contradict:0,contradictori:[5,11],contrari:[0,8],contrast:[1,4,5,9,10,11,13],contribut:[4,19],contriv:11,control:[0,1,3,5,11,13],controversi:9,convei:[1,4,5,11],conveni:[3,5,6,10,11,13],convent:[1,3,4,5,10,11],convention:13,converg:9,convers:[7,8,9,11],convert:[2,4,5,10,11,13],convict:11,convinc:[0,1,9,10,11],cool:[4,11,13],coolest:4,cope:4,copi:[2,3,4,7,13],copiou:5,cor:4,core:[0,2,3,4],corner:[],corr:4,correct:[0,1,2,4,7,8,9,10,11,13],correctli:[1,4,11],correl:[5,11,13,19],correspond:[1,2,3,4,5,6,8,9,10,11,13],cosmet:[],could:[1,2,3,4,5,6,7,8,9,10,11,13],couldn:[1,4],counsel:[],count:[1,2,3,4,5,7,9,10,11,13],counterintuit:[0,9,10,11],countri:[2,4,10,11],countrysid:8,coupl:[1,4,6,10],courier:[],cours:[0,1,2,3,4,5,7,8,9,10,11,13,20],courtesi:10,cov:4,covari:[1,4],cover:[0,1,3,4,5,6,9,10,11,13],coverag:9,craft:5,craig:4,cram:[5,13],cramp:5,cran:[],cranni:0,crap:1,crash:9,crazi:[0,9],creat:[1,2,3,4,5,7,9,10,13,20],creation:19,creativ:[5,11],credibl:10,credit:[1,4],crime:[1,11],crimin:[1,11],crippl:2,crisi:11,crit1:[],crit2:[],crit3:[],crit4:[],crit:13,crithi:11,critic:[0,1,3,8,9,10,13],criticis:[1,10],critiqu:10,critlo:11,croker:[8,10],cromul:10,crop:13,cross:[4,13],crosstab:6,crucial:[3,10],crude:[1,10],cruel:0,cryptic:4,cs63:[1,19],csg63:[1,19],csv:[0,4,5,7,13],cube:[2,4],cubism:5,cue:4,cuebal:3,cultur:0,cumbersom:3,cumsum:9,cumul:13,cunningli:0,cup:9,cure:[1,13],curios:13,curiou:[0,1,4,6,13],curli:[3,9],current:[1,2,3,4,6,10,13],cursori:11,curv:[4,5,9,10,13],curvi:5,custom:[5,10],customari:5,customis:5,cut:[0,4,5,9],cute:1,cwc:8,cwd:[0,4,5],cynic:0,cyril:1,d009972d:[],d13d:[],d36a6fc3adb7:[],d5b41eeb:[],d5ee87d58c65:[],d69586c7f512:[],d9846b90f6de:[],d997af956b7d:[],dabbl:1,dae4:[],dai:[0,1,2,4,5,8,9,10,11,13],daisi:6,damag:1,damn:0,dampen:9,dan:[1,3,4,5,9],dan_grump:4,dan_sleep:4,danger:[1,2,10,11],daniel:[1,20],danrbook:3,dark:5,darken:11,dart:9,dash:[4,9,10],dashdot:10,data:[0,1,2,7,8,9,10,19],databas:[],datafram:[0,4,5,6,9,10,13],datahandl:[],datapoint:10,dataset2:4,dataset3:4,dataset:[3,4,10],datastructur:6,date:[1,2,4,11],datum:13,david:[1,8],dbb77e8ee3ad:[],dbinom:[],dchisq:[],ddition:2,ddof:4,dead:[1,10],deal:[1,2,3,4,5,10,13],dealt:[1,3],dear:0,death:[4,5],debug:[],decad:[9,11],deceiv:1,decemb:1,decent:[5,7],decid:[0,2,3,5,9,10,11,13],decim:[2,3],decis:[0,5,6,9],decisionmak:[],deck:9,declin:1,decreas:[1,4,10,13],dedic:3,deduct:[0,8],deem:[4,11],deep:[1,2,9,11],deeper:[0,1,11],deepli:[0,1,4,5,9],def:[9,10],defeat:[],defenc:13,defend:11,deficit:10,defin:[2,3,4,9,11,13],definit:[1,4,5,9,10,11,13],degfre:[9,13],degfree1:9,degfree2:9,degre:[1,5,9,10,11,13],degrees_freedom:13,del:3,deleg:6,delet:5,deliber:[0,1,10,11],delici:13,deliv:0,delta:13,delv:2,demand:[0,10],demographia:4,demolit:5,demoni:4,demonin:4,demonstr:[9,10,11],deni:4,denomin:13,denot:[1,2,4,9,10,11,13],dens:[],densiti:[5,13,20],depart:[0,13],departur:13,depend:[0,1,2,3,4,5,8,9,10,11,13],depict:[4,5,9,10,13],deprec:[],depress:1,depressingli:3,depth:[9,10],deriv:[1,4,9],describ:[1,3,5,6,8,9,10,11],describeimg:[],descript:[1,2,5,9,10,11,13],descriptor:1,descrstatsw:10,design:[0,3,5,9,10,11,13,19],desir:[1,5,6,7,9,10,11,13],desktop:[3,5],desper:[0,11],despin:[5,9,13],despit:[1,8,11],destin:10,destroi:[1,9],detach:[],detail:[0,1,2,3,4,5,7,9,10,11,13],detect:[1,5,13],determin:[0,1,4,5,6,10,11,13],determinist:9,dev:[4,13],devat:10,develop:[2,3,4,10],development:1,deviat:[2,9],devis:13,devoid:9,devot:[0,3,9],dewei:9,df0e8d462b68:[],df1:[],df2:[5,13],df_1:9,df_2:9,df_berk:0,df_clintrial:4,df_kurtosi:4,df_skew:4,df_wide:[],diagnosi:[13,19],diagnost:13,diagon:13,dialog:8,dialogu:[6,8],diamond:5,dice:9,dict:[4,11],dictat:0,dictionari:3,did:[0,1,2,3,4,5,7,8,9,10,11,13],didn:[1,2,4,8,9,11,13],die:9,died:1,diet:1,diff_mean:[],differ:[0,1,2,3,5,6,7,8,10,19],difference_:[],differentialattrit:1,difficult:[1,2,5,6,7,9,10,11,13],dig:[0,1,5,9,13],digit:5,digress:[3,9,11],dim:13,dinner:10,dinosaur:6,dip:5,dir:3,direct:[2,5,6,10,13],directli:[1,2,3,4,5,7,11],directori:3,directory_:[],disadvantag:[9,10,11],disaggreg:[0,1],disagr:[1,11],disagre:[0,1,6,11],disappear:[],discard:4,disciplin:[7,9],discov:[1,2,5,9,11,13],discredit:1,discrep:4,discret:[6,9,10,11],discrimin:[0,8,11,13],discuss:[0,1,2,3,4,5,6,7,8,9,10,11,13],diseas:1,dishonesti:1,disingenu:1,disk:3,dismal:2,dismiss:10,disobedi:1,disorganis:[],dispatch:[],dispatch_:[],displai:[0,1,3,4,5,9,10,11,13],displaystyl:13,displot:4,disput:11,disregard:5,dissect:8,dist:[9,13],distanc:4,distant:5,distformula:[],distinct:[1,3,4,5,9,10,11,13],distinguish:[1,4,6,10],distort:[9,10],distract:11,distractor:0,distribut:[0,2,4,5,6,13,19,20],disturb:9,div:6,divd:3,dive:[5,10],divers:8,divid:[1,3,4,5,6,9,10,11,13],divin:11,divis:[1,2,10],dnorm:[],doc:3,doctrin:9,document:[3,4,5,7,9,10],doe:[0,1,2,4,5,6,7,8,11,13],doesn:[0,1,2,3,4,5,6,8,9,10,11,13],dog:[],dogma:11,dogmat:11,doing:[0,1,2,3,4,5,6,7,8,9,10,11,13],dollar:[1,2,4,9],domain:[11,13],domest:4,domin:[1,5,9,11],don:[0,1,2,3,4,5,6,7,8,9,11,13],donald:19,donat:2,done:[0,1,2,3,4,5,6,7,8,9,11,13],doomsay:4,dot:[4,5,11,13],doubl:1,doubt:[2,4,11],down:[1,2,3,4,5,7,9,10,11,13],download:2,downsid:5,dozen:[1,4,7,9],drag:4,dramat:[4,9],drastic:10,draw:[1,3,4,8,9,10,13],drawback:[6,7,11,13],drawn:[1,5,9,10,13],dread:4,dream:7,dreari:13,drew:[9,10],dri:4,drill:4,drive:[3,8],drop:[1,4,11,13],dropdown:[],dropna:4,drown:0,drug:[1,2,4,13],dtype:[3,4],dubiou:1,due:[1,4,9,10,11,13],dug:2,duke:8,dull:4,dumb:[1,2,4],dure:[1,2,3,4,5,6,7],duti:1,dwell:3,dwr_all:[],e170:[],e1e3cb50:[],e1e8:[],e225286a3f01:[],e272:[],e30f90a78005:[],e3c489536945:[],e3cf:[],e592:[],e635:[],e811:[],e835:[],e87501de0303:[],e8d3c3d8:[],ea51:[],each:[0,1,2,3,4,5,6,7,8,9,10,11,13],eaerlier:4,earli:[3,11,13],earlier:[0,1,2,4,5,9,10,11,13],earn:4,earth:13,earthquak:1,eas:5,easi:[0,1,2,3,4,5,6,7,8,10,11,13],easier:[1,2,3,4,5,6,10,11,13],easiest:[2,4,5,9,11,13],easili:[1,4,5,10,13],ebp83:[0,19],ecb93e1df467:[],echo:[],ecologist:10,econom:[0,10],economist:[4,10],ecosystem:[2,3],edg:[3,6,11],edgecolor:[5,13],edit:[5,19,20],editor:[],editori:[],educ:[0,1,2,10,11],edward:5,eecaa5d710b8:[],ef46:[],ef63631a:[],effect:[0,2,3,4,5,7,9,10,19],effects:[],effici:10,effort:[1,3,4,9,11,13],ehp11:[9,19],eight:4,either:[0,1,2,3,4,5,6,8,9,10,11,13],elbow:5,elect:9,elector:9,electrodynam:0,electron:0,eleg:[4,5],elegantli:4,element:[3,4,6,10,11,13],elementari:[4,9],elimin:[1,5,8,11],ell02:[4,19],ell10:[11,19],elli:19,ellis2010:[],ellman2002:[],ellman2002soviet:[],ellman:[4,19],els:[0,1,3,4,5,6,7,8,9,10,11,13],elsewher:2,email:[7,13],embarrass:1,emerg:11,emili:20,emphasis:[11,13],emphat:11,emphcol:[],emphcollight:[],emphgrei:[],emploi:1,employe:13,empti:5,enamor:3,enclos:2,encod:[],encompass:11,encount:[0,1,3,4,8,9,10,11,13],encourag:[10,13,20],end:[0,1,2,3,4,5,7,8,9,10,11,13],endlessli:3,endors:[1,9],endpoint:6,enemi:5,energi:1,enforc:2,engag:[0,1,7],engin:[0,1,4],english:[0,7,9,11],enjoi:5,enorm:9,enough:[0,1,2,3,4,7,8,9,10,11,13],enrol:[0,1,9,10],enrypt:2,ensconc:11,ensur:[1,4,9,11,13],entail:9,enter:[2,4,13],enterpris:9,entertain:[9,13],entir:[0,1,3,4,5,6,8,9,10,11,13],entireti:11,entiti:[1,9,10],entri:4,enumer:[4,10],envir:[],enviro:3,environ:[1,2,3,10,11],eol:[],eps:[],equal:[0,2,4,5,6,8,9,10,11,13],equal_var:13,equat:[2,4,9,10,13],equip:[],equival:[1,2,4,6,9,11,13],eras:1,erich:19,erkelei:19,erlbaum:19,error:[0,1,2,3,5,10,13],errortyp:[],escap:[8,11],escapechar:[],ese:[4,13],esp:11,espcrit:11,especi:[0,1,2,3,4,5,6,10,11,13],espestim:[],essai:1,essenc:[1,10,11],essendon:[4,5],essenti:[6,7,10,11,13,19],est:[],establish:[1,11],estim:[1,4,5,8,11],estimatorbia:10,etc:[1,3,4,10],ethan:[3,4,5,20],ethanwe:[4,13],ethic:1,europ:[3,19],european:13,eval:[],evalu:[0,2,4],evan:19,evans1983:[],evans1983conflict:[],evans2000:[],evas:11,even:[0,1,2,3,4,5,6,7,8,9,10,11,13],event:[1,4,9,10,11,13],eventnam:9,eventu:[1,2,9,10,11],ever:[0,1,2,3,8,9,11],everi:[0,1,3,4,5,8,9,10,11,13],everybodi:9,everydai:[1,4,7,8,9,10,11],everyon:[0,1,2,4,9,10,11,13],everyth:[0,1,3,4,5,6,7,8,9,10,13],everywher:[0,7,9,13],evid:[0,1,5,8,10,11,13],evidentiari:[1,10,11],evil:[0,1],evolut:0,exact:[1,3,4,5,9,10,11,13],exactli:[0,1,2,3,4,5,6,7,9,10,11,13],exagger:[4,13],exam:[1,4,11],examin:[1,4],exampl:[0,1,2,3,5,6,8,9,10,11],exce:4,excel:[2,3,4,9],except:[1,2,4,7,8,10,11,13],excit:[3,4,20],exclud:5,exclus:[10,11],excus:4,execut:[2,3,4,7],exercis:[1,5,7,10,13],exhaust:[9,10],exist:[0,1,3,4,5,6,8,9,11,13],exorbit:2,exot:13,exp:9,expand:[1,4,5,11,13],expans:[],expect:[0,1,2,3,4,5,8,9,10,11,13],expens:[0,2,4,10,13],experi:[0,1,2,3,4,5,7,8,9,10,11],experienc:1,experiment:[0,9,10,11,13,19],experimet:13,expert:[2,8],explain:[0,1,2,3,4,5,7,8,9,10,11,13],explan:[1,6,9,11],explanatori:3,explicit:[4,6,10,11,13],explicitli:[1,4,10,11],exploit:9,explor:[4,7,13],exploratori:5,exponenti:9,expos:[],express:[1,4,5,6,10],expt:[],exrem:11,extend:[2,3,5,8,9,10],extens:[2,3,11],extent:[0,1,2,4,5,6,13],extra:[2,3,4,10],extract:[2,3,4,6,9],extraordinarili:10,extrasensori:11,extravers:4,extravert:4,extrem:[1,2,4,5,9,20],eye:[1,3,5],eyebal:4,eyes:[5,6,10],eyewit:1,eyman:19,f26b7c7c1738:[],f29618a6:[],f40c1ed9:[],f45fd3136913:[],f50fea121ed3:[],f533bacecaf1:[],f63645209886:[],f636:10,f66e:[],f6eba0bb:[],f6f1d9db:[],f70d:[],f7430430:[],f795f0294715:[],f796:[],f7b7794b1a5e:10,f7f2c52b62e9:[],f927:[],f_3_20:9,fa58:[],fabric:1,face:[4,5,9,11],facecolor:5,facet_kw:4,facetgrid:4,fact:[0,1,2,4,5,6,8,9,10,11,13],facto:5,factor:[1,3,5,11,13],factori:[1,9],factual:8,fail:[0,1,11,13],failur:1,faint:5,fair:[0,4,5,8,9,11,13],fairi:13,fairli:[0,1,2,3,4,5,6,10,11,13],faith:[0,1,2,4],fake:10,fall:[0,1,4,5,9,10,11,13],fallaci:1,fals:[0,1,4,5,9,10,11,13,19],falsifi:1,falsification:11,falsificationist:11,familar:3,famili:4,familiar:[1,2,4,13],famou:[1,5,10],famous:[10,11],fan:20,fanci:[1,3,4,8],fancier:[],fancyp:[],fantast:20,far:[0,2,3,4,5,6,7,8,9,10,11,13,20],fashion:[1,2,3,4,5,10,11],fast:[1,3,4],faster:[1,2],father:6,favor:5,favorit:4,favour:[0,2,4,8,11],favourit:[0,1,13],fb50:[],fba6:[],fc9b:[],fd6e:[],fdd8cff7b4db:[],fdist:9,fe89e91d93ec:[],fear:[0,7],feasibl:11,feat:[0,1],featur:[1,5,9],feb6a0b4c657:[],februari:1,feder:9,fee:[0,2],feed:[3,4,13],feedback:1,feel:[0,1,2,4,5,8,9,10,11,13],fell:[1,11],felt:[0,3,13],femal:[0,1,10,11],few:[0,1,2,3,4,5,7,8,9,10,11,13],fewer:[1,3,5,11],ff01:[],ff2f:[],fibonacci:5,fictiti:[4,10],field:[0,1,2,4,8,9,11],fig1:4,fig2:[4,11],fig:[0,4,5,9,10,11,13],fig_powerfunct:[],figsiz:[4,5,9,10,13],figur:[0,1,2,3,4,5,6,7,9,10,11,13,20],figwidth:11,file1:4,file2:4,file:[2,4,6,7,10,13],filenam:[],fileopen:[],filepanel:[],filepath:[],filesystem:[],fill:[1,4,9,13],fill_between:[9,13],finalist:[4,5],find:[0,1,2,4,5,6,7,8,9,10,11,13,19],finder:[],fine:[10,11],finish:[1,3,4,5,7,10,11],finit:10,finnegan:1,firmli:[1,4,5],first:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],firstli:[0,1,2,4,5,6,8,9,10,11,13],fis22:[9,19],fischer:4,fisher1922b:[],fisher:[1,4,9,13,19],fist:2,fit:[3,5,9],five:[1,3,4,5,9,10,13],fix:[3,4,5,7,10,11,13],fixat:0,flat:[4,10],flatten:13,flavor:20,flavour:9,flaw:[1,2,4,11],flexibl:[3,11],flick:6,flip:[6,9,11],flood:1,floor:3,floppi:[],flourish:5,fluctuat:[1,9],fluke:[9,13],focu:[0,1,3,4,5,7,9,11,13],focus:[0,1,2,4,7,8,9,10,13],folder:3,folk:[4,10,11],follow:[0,1,2,3,4,5,6,9,10,11,13],fond:11,font:[],font_scal:5,fool:7,foolish:13,foot:1,footbal:[4,5],footi:[4,5],footnot:4,forbid:9,forc:[1,2,4,5,7,10,11],forego:20,foreign:[],forev:[3,10],forget:[0,1,2,3,4,5,9],forgiv:10,forgiven:0,forgot:4,forgotten:[3,13],form:[1,2,3,4,5,7,8,9,10,11,13],formal:[1,11,13],formalis:[9,10,13],format:[3,4,5,6,7,9,13],former:[],formula1:[],formula2:[],formula3:[],formula4:[],formula:[3,4,9,13],formulat:[],forth:7,fortuit:10,fortun:[4,5,9,10,11,13],forward:3,found:[0,1,3,9,10,11],foundat:[3,9,10,11,19],founder:10,four:[1,2,3,4,6,9,10,11],fourth:[1,3,4,8,9],fox2011:[],frac:[4,9,10,13],fraction:11,frailti:0,frame:[6,10,13],frame_:[],framework:11,franci:[],francisco:4,frankli:[0,4,5],fraudul:1,freak:[1,4],free:[2,4,8,11,20],freedom:[9,10,11,13],freedom_:[],freeli:2,freez:1,fremantl:[4,5],french:11,freq:4,frequenc:[4,5,6,9,10,11],frequent:[0,4],frequentist:10,frequentist_prob:[],frequentist_probability_fig:9,frequentistprob:[],friedman:10,friend:[1,4,5,7,9,13],friendli:4,frill:[],fritz:19,from:[0,1,2,4,5,6,7,8,9,11,19,20],front:13,fruit:13,frustrat:[2,4,9,11,13],fulfil:1,full:[2,3,4,5,11,13],fulli:0,fun:[3,4,20],function_:[],fund:0,fundament:[0,1,4,9,11,13],funni:[4,5,10,13],further:[0,1,2,5,6,7,9,11],futur:[5,10,11],fuzzi:1,fyi:2,gage:19,gain:[1,4,9],gambl:[8,9],gambler:8,game:[4,5,8,9,10],garbag:[],garden:[6,13],gather:[6,9],gaussian:9,gave:[1,2,3,4,13],gdata:[],geelong:[4,5],gelman2006:[],gelman:19,gender:[0,1],gener:[0,1,2,3,4,5,6,9,10,11,13],generalis:[0,1,4,10],gentli:1,genuin:[1,2,4,5,6,9,11,13],georg:11,geq:11,ges:3,get:[0,1,3,5,6,7,8,9,10,11,13,20],get_x:11,get_xaxi:13,get_xlim:13,get_yaxi:13,get_ylim:13,getcwd:[0,4,5],getwd:[],gibberish:[1,4,10,11],gibson:7,gist:[4,9],github:[3,4,5],githubusercont:[4,13],give:[1,2,3,4,5,6,8,9,10,11,13],giveawai:1,given:[0,1,2,3,4,5,8,9,10,11,13],glad:4,glanc:[5,13],glass:13,glib:0,global:[],globalenv:[],gloriou:11,glorious:0,gloss:[4,10],glossi:2,glue:[0,2,4,9,11,13],goal:[2,3,4,5,6,7,8,9,10,11,13],god:[9,11],goe:[1,2,4,8,9,10,11,13],going:[0,1,2,3,4,5,7,8,9,10,11,13],gold:10,gone:[1,3,13,20],good:[0,1,2,3,5,6,7,8,9,10,11,13],goodman:8,googl:[1,2],gosset:13,got:[0,1,2,3,4,7,10,11,13],gotten:[4,11,20],govern:[0,1,9,10],gplot:[],grab:5,grade:[1,4,13],grade_test1:13,grade_test2:13,grade_time1:[],grade_time2:[],graduat:19,grain:6,grammat:1,grant:[9,11],graph:[4,9,10,19],graphic:[4,10,13],grappl:4,grasp:[6,9],graviti:4,great:[1,3,4,5,9,10,13],greater:[0,1,9,11,13],greatli:5,greek:[4,11],green:1,grei:[9,11],grid:[],grossli:1,groucho:13,ground:[0,1,3,4,5,9,11],groundwork:[],group:[0,1,3,4,6,10,11,13],groupdescript:[],grow:[1,9,10],grown:10,grump:[4,5],grump_fig:4,grumpi:[4,5],grumpier:4,gs06:[11,19],guarante:[1,4,10,11,13],guess:[0,1,3,4,5,8,9,10,11,13],gui:[0,13],guid:[1,4,10,11,13,19],guidelin:1,guilti:11,guin:13,gulag:4,gullibl:1,gunpoint:10,gut:[0,5,11],gzcon:[],gzfile:[],h_0:[11,13],h_1:[11,13],habit:[2,4,5],had:[0,1,2,3,4,5,6,8,9,10,11,13],haha:4,half:[1,3,4,5,8,9,10,11,13],hallmark:8,hammel:19,hammer:1,han:[1,19],hand:[0,1,2,4,5,7,9,10,11,13],handful:3,handi:[1,2,3,9,10,11,13],handili:4,handl:[0,1,2,5,7,10,13],hang:3,hangman:6,happen:[0,1,2,3,4,5,6,8,9,10,11,13],happi:[0,5,9,11,13],happiest:5,happili:11,hard:[0,1,2,3,4,5,7,9,10,11,13,20],harder:[0,1,2,13],hardest:[5,6,10],hardli:[0,11],harken:5,harm:3,harmon:11,harpo2:[],harpo:13,harpo_summari:13,harpohist:[],harpohistanastasia:[],harpohistbernadett:[],harri:2,harvard:0,harvardmagazin:0,has:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],hasn:[7,10],hast:19,hat:[4,10,13],hatch:13,hate:11,have:[0,1,2,3,4,5,6,7,8,9,11,13,20],haven:[3,4,5,6,8,9,10,11],hawthorn:[1,4,5,19],head:[3,4,5,9,10,11,13],headach:5,header1:[],header2:[],header3:[],header:[],headlin:1,hear:[1,4,13],heart:[4,9,10],heat:1,heavi:[0,13],heavier:[9,13],heavili:[0,1,5,10,13],heavy_tailed_data:13,heck:[0,4,13],hedg:13,hedges1981:13,hedges1985:13,hei:[5,10],height:[3,4,5,9],held:0,hello:3,help:[0,1,2,3,4,5,6,7,9,10,11,13,20],henc:[],henri:19,her:[1,4,9,11,13],herald:9,here:[0,1,2,3,4,5,6,9,10,11,13],heresi:11,heterogen:1,heurist:[4,13],hidden:[0,1,10,11],hide:[1,4,5,9,10],high:[0,1,2,4,5,9,10,11,13],higher:[0,1,4,9,11,13],highest:10,highli:[1,2,3,4,5,6,10,11,13],highlight:10,hilari:1,hill:[8,10,19],him:[1,8],himself:[10,13],hint:[4,11],hirtum:0,his:[1,4,8,9,10,11,13],hist:13,histdata:[],histgram:10,histogram1:[],histogram:[4,9,10,13],histor:11,histori:[0,4,7,9,11,13,19],histplot:[4,5,9,10,11,13],hit:[2,9],hline:13,hoax:1,hoc:1,hold:[1,5,7,9,10],hole:2,holm:8,holt:19,home:[1,3,4,5,6,10],homework:[],homogen:[1,13],homoscedast:13,honest:[0,1,4,5,9,11,13],honesti:[1,4],honestli:[6,9,13],honour:9,hood:[4,5],hope:[0,8,9,10,11,13],hopefulli:[0,3],horizont:[5,9],horribl:[0,1,5],hors:[1,19],hot04:[1,19],hot:[1,9],hothersal:19,hothersall2004:[],houghton:19,hour:[1,4,5,7],hourli:2,hous:[4,10,13],household:4,housingpriceexampl:0,how:[0,1,2,4,5,6,7,8,11,13,20],howev:[0,1,2,3,4,5,6,7,9,10,11,13],howl:2,href:[],htm:4,html:[0,3,8,10],http:[0,1,3,4,8,10,13],hue:[0,9,10],huei:9,huge:[0,4,5,10,13,20],human:[0,1,2,3,4,7,8,9,10,11,13,19],humbl:[1,3,5],hume:8,hunch:1,hundr:[1,7,13],hurt:[9,10],hwang:1,hybrid:11,hypothei:11,hypothes:[1,10,13],hypothesi:[1,4,8,10],hypothesis:[1,13],hypothesistest:[],hypothet:[0,4,9],icon:[],idea:[0,1,2,3,4,5,6,7,8,9,10,11,13],ideal:[1,4,7,10,11],ident:[0,3,4,5,6,9,10,11,13],identif:1,identifi:[1,2,11,13],ideolog:[9,11],idiosyncrat:[0,1],idiot:4,ifels:[],ignor:[1,2,3,4,6,7,9,11,13],ill:3,illeg:5,illus:1,illusori:[10,13],illustr:[1,3,4,5,6,9,10,11,13],imag:7,imagin:[1,3,4,5,9,10,11,13],imbal:1,img:[10,13],immedi:[0,3,4,10],immun:1,immut:11,impact:[0,9],imparti:[0,1],impenetr:9,implaus:[10,11],implement:11,impli:[2,4,9,10,11,13],implicit:8,implicitli:1,importantli:[0,9,10,11],impos:[5,7],imposs:[1,7,8,9,10,11],imprecis:10,impress:[5,7,9,11,13],impressionist:5,improv:[0,1,4,13],inabl:8,inaccur:10,inadvert:5,inappropri:11,incid:[1,5],includ:[0,1,3,4,5,7,9,10,11,13],include_graph:[10,13],includng:[],incom:4,incomplet:10,incomprehens:4,inconsist:[8,11],inconveni:1,incorrect:[0,4,9,10],incorrectli:[0,11],increas:[1,4,5,9,10,13],increasingli:[9,11],incred:[0,1,5,13],inde:[2,4,5,8,10,13],indenpend:[],indent:2,independ:[1,9,10,11],independentsamplesttest:[],index:[4,5,6],indian:1,indic:[1,2,3,4,5,6,10,11,13],indici:5,individu:[0,1,2,4,9,10,13],induct:[0,8,10,11],indulg:11,industri:10,industrialis:10,inert:1,inevit:[7,9],inexpens:0,inf:[],infam:4,infant:4,infer:[0,4,8,9,10],inferenti:[4,9,10,11],infin:[9,10],infinit:[9,10,13],inflex:2,influenc:[0,1,2,4,11],info:[3,5],inform:[0,1,2,3,4,5,6,8,9,10,11,13],infti:[9,10],infuriatingli:0,inher:6,initi:[4,5],injur:9,inlin:9,innoc:11,inplac:5,input:[2,3,4,6,9,13],inquisitori:11,insan:[1,10],insert:10,insid:[1,2,3,4,5,6,7,10,11,13],insight:4,insist:[5,8],insofar:[1,10],inspect:[1,4,5,9,10,13],inspir:5,instal:[],instanc:[0,1,2,3,4,5,6,9,10,11,13],instead:[0,2,3,4,5,6,7,9,10,11,13],instinct:[0,10],instruct:[1,2,7,10],instructor:2,insult:[1,10],insur:5,int64:[3,4],int_a:9,integ:[2,3,5,10],integr:[2,9],intellect:11,intellectu:9,intellig:[1,2,4,9,10,11],intend:[1,4,5,9,10,13],intent:[1,4,10,11],inter:1,interact:[1,2,3,7,11],intercept:[],interest:[0,1,2,3,4,5,6,7,9,10,11,13],interestd:13,interfac:2,interim:2,intern:4,internalis:1,internet:[0,1,5,11],interpol:[],interpret:[0,1,2,5,6,7,9,11,13,19],interquartil:5,intertwin:[0,1],interv:[4,5,6,13],interval_:[],intimid:[],intoler:11,intro:[13,20],introduc:[1,2,3,4,5,6,10,11],introducd:4,introduct:[3,4,6,8],introducton:5,introductori:[0,3,9,10,11,13],intror:[],intrus:10,intuit:[0,1,5,9,10,11,13],intut:10,invalid:[0,1],invas:1,invent:[1,2,4,11],invest:1,investig:[1,4],invis:[],invok:[],involv:[0,1,2,4,5,7,9,10,11,13],ioa05:[1,19],ioannidi:19,ioannidis2005:[],ioff:[],ipynb:3,ipython:[2,3,13],iq1:10,iq2:10,iq3:10,iq4:10,iq5:10,iq_10000:10,iq_100:10,iq_10:10,iq_1:10,iq_2:10,iqdist:10,iqdista:10,iqr:4,iqsampa:10,iqsampb:10,iqsampc:10,iron:[2,10],irrelev:[0,5,9],irrespect:[9,10],irritatingli:4,isn:[0,1,2,3,4,5,6,8,9,10,11,13],isol:[1,7],issu:[0,1,3,4,5,6,8,10],italicis:1,item:[1,3,4],iteself:13,its:[0,1,2,3,4,5,6,7,8,9,10,11,13],itself:[0,1,2,3,4,5,6,9,10,11,13,19],ivis:2,ivori:11,jacob:10,jame:4,jamovi:20,jane:4,januari:[1,7],jargon:[11,13],jasp:20,jean:9,jerk:[8,10,13],jerzi:11,jitter:5,job:[0,3,4,5,9,11,20],joe:[],john:[0,5,9,10,19],joke:[1,13],josef:4,journal:[1,11,19],journalist:[0,4],joyzepam:[],jpg:[],jse:10,judg:[4,11,13],judgement:0,judgment:5,julian:19,jump:[0,4,13],jungl:9,junk:[],jupyt:[2,3,5,7],jurist:11,just:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],justic:1,justif:[8,10,11],justifi:[8,10,11],kabl:[],kage:3,kahneman1973:[],kahneman:19,kardinia:5,karl:11,karyotyp:1,kate:4,kde:10,keep:[0,1,2,3,4,5,6,7,8,9,10,11,13],keeper:3,kei:[1,2,3,4,5,6,9,11,13],kept:[1,8,9,11],kernal:5,keyboard:2,keyn:10,keynes1923:10,keyword:[],kid:[1,4],kilda:[4,5],kill:[1,3,5,13],kind:[0,1,2,3,4,5,6,7,9,10,11,13,20],kinder:13,kindli:[3,5],kingdom:13,klinfelt:1,knew:[1,10,13],knitr:[10,13],know:[0,1,2,4,5,6,7,8,9,10,11,13,20],knowledg:[0,1,4,5,9,10,13],known:[0,1,4,5,7,8,9,10,11,13],kolmogorov:[9,13],kolomogorov:13,koth:20,kt73:[1,19],kuhberger2014:[],kuhbergerfs14:[1,19],kurtosi:13,kurtosis_fig:4,kurtosisdata:4,kurtosisdata_ncurv:4,lab:[1,10],label:[1,2,4,6,9,10],labelbottom:9,labelleft:9,labelright:9,labeltop:9,labor:9,labori:[4,10],lack:[1,5],lai:[],laid:6,land:9,languag:[0,2,3,4,5,7,9,11,13],larg:[0,1,3,4,5,9,11,13],larger:[1,4,5,9,10,11,13],largest:[0,1,4,5,10],larri:9,last:[0,1,2,3,4,5,7,8,9,10,11,13],late:1,later:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],latter:5,launceston:5,law:[0,1,9,11],lawlargenumb:10,lawrenc:19,lawyer:11,lazi:[2,3,9,10,13],laziest:4,lccc:[],lead:[1,3,4,9,10,11,13],leagu:[4,5],learn:[1,2,3,4,5,7,9,10,11,13],learningstatisticswithr:[],least:[0,1,2,3,4,5,7,9,10,11,13],leav:[0,1,2,4,5,7,9,10,11,13],lectur:[0,1,2,4,10,13],led:[4,8],lee:19,left:[1,2,3,4,5,8,9,10,11,13],leftmost:5,legal:[0,11],legitim:[2,3,4,5,7,8,9,10,11],leh11:[11,19],lehmann2011:[],lehmann:19,len:[3,4,10,13],lender:4,length:[1,4,10],lengthi:[],leprechaun:13,leptokurt:4,leq:[10,11],less:[0,1,3,4,5,7,8,9,10,11,13],lesser:[],lesson:[0,4],let:[0,1,2,3,4,5,6,7,8,9,10,11,13],letter:[4,10,11],level:[0,2,4,5,7,8,9,10,11,13],leven:13,lewi:8,liar:9,liber:2,librari:[2,4,5,6,19],licenc:2,licens:[2,20],lie:[1,4,5,9,10,13],lied:[4,9,10],lies:[4,9,10,11],life:[1,2,5,6,8,9,10,11,13],lifetim:4,light:[1,2,4,11,13],lightgrei:11,lightli:6,like:[0,1,2,3,4,5,6,7,8,9,10,11,13],likert:[1,6,9],limit:[1,2,3,4,9,13],line2d:[10,13],line:[0,1,2,3,4,5,6,7,9,10,11,13,20],linear:[4,6,10],lineplot:[4,9,10,11,13],linestyl:[9,10,13],link:[2,3,4,11],linspac:[4,9,10,13],linux:3,lion:9,list:[1,2,4,5,6,9,10,11,13],listen:[2,11,13],liter:[2,4,6,9,11],literatur:[0,4,8,9,11],littl:[0,1,2,3,4,6,7,8,9,10,11,13],live:[0,1,2,4,10,11,13],liz:[],lme4:[],lmplot:0,lmtest:[],load:[4,5,7,13],load_dataset:10,loadingcsv:[],loan:4,loc:[3,4,5,6,10,13],local:10,locat:[1,3,4,5,6,9],lock:9,log:[7,11],logic:[0,1,4,11,19],logico:11,lone:4,longer:[1,2,3,4,5,8,11,13],longest:6,longtowid:[],look:[0,1,2,4,5,6,7,8,9,10,11,13],loop:[3,10],loos:3,lopsid:5,lose:[1,4,5,6,8,9],loss:[4,5,8,10],lost:[1,4,8],lot:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],lotteri:9,loui:3,louie:9,love:[0,1,2,3,4,5,9,11,20],lover:3,low:[0,1,4,5,10,11,13],lower:[0,4,10,11,13],lower_area:13,lowest:[4,6],lsquo:[],lsr:[3,20],lty:[],luck:[1,8],lucki:1,luckili:[3,4,5,13],ludwig:11,lunch:[8,10],lung:1,luxuri:0,lwd:[],lying:[0,4],lyre:0,mac:3,machin:[2,3,7,8,9],maco:3,mad:[0,1,4,11],made:[0,1,2,3,4,5,6,8,9,10,11,13],magic:[0,1,3,4,7,10,11,13],magnifi:11,magnitud:[4,11],mai:[1,2,3,4,5,6,7,9,10,11,13],maiden:9,mail:1,main:[1,2,3,4,5,9,10,13],mainli:[],major:[1,4,5,9,10,11,13],make:[0,1,2,3,4,5,6,7,9,10,13],makeov:5,maker:1,makka:6,male:[0,1,10,11],malici:1,man:[1,4],manag:[1,2,3,9],mani:[0,1,2,3,4,5,6,8,9,10,11,13],manipul:[1,6],mann:13,manner:1,manual:[4,5,13],map:[5,9,11],margin:[4,5,6,11,13],margin_titl:4,mark:[1,3,5,13],marker:[],market:4,mash:11,mask:[],mass:11,massacr:5,massiv:[0,2,4,5,13],match:[1,4,9,13],materi:[1,5,9,10],math:[1,2,3,4,9,10,13],mathbf:4,mathemat:[1,2,4,7,9,10,11,13,19],mathematician:[9,10],mathes:[],mathfunc:[],matlab:5,matplotlib:[3,4,5,9,10,11,13],matrix:4,matter:[0,1,2,3,4,9,11,13],max:[4,5,10,13],maximis:11,maximum:[4,9,10,13],mayb:[1,2,3,4,5,6,7,8,9,11,13],maynard:10,mbox:[4,9,10,11,13],mcar:4,mcgrath2006:13,mcgraw:19,mcnemar:13,mean:[0,2,3,5,6,8,11],mean_diff:13,mean_numb:[],meaning:[1,4,5,6,9,11,13],meaningfulli:[4,11],meaningless:[1,4,6,9,11],meanmedian:[],meant:[1,2,11,13],meantim:1,measur:[2,10,11,13,19],mechan:[3,6,9,11,13],med:19,median:5,medianprop:5,medic:13,mee67:[9,19],meehl1967:[],meehl:[9,19],meet:[5,11],melbourn:[4,5],member:10,membership:[],memori:[10,13,19],men:10,mental:[3,13],mention:[0,1,2,3,4,5,9,10,13],menu:[],mere:[1,4,8,10],merg:6,merit:[0,1,13],merri:9,mersenn:[],mesokurt:4,mess:[0,5,11],messag:[2,3,7],messi:[0,2,4,6,11],messier:[1,4,13],met:[5,13],meta:19,metaphor:5,meteorologist:9,method:[1,2,3,4,5,6,9,10,11,13],methodolog:[0,1,10,11,19],mfcol:[],mfg:[],mfrow:[],mgp:[],mice:11,michael:19,microsoft:2,mid:[4,6],middl:[0,1,4,5,6,9,13],midpoint:6,mifflin:19,might:[0,1,2,3,4,5,6,8,9,10,11,13],milan:9,million:[4,10],millisecond:[],min:[4,5,9,10,13],mind:[0,1,3,4,5,7,9,10,11],mindless:[2,4,5,9],mindlessli:2,mine:[1,4,5,8,9,13],mini:5,minim:3,minimis:[1,11,13],minimum:[4,10,13],minitab:[],minor:[4,9,10,11],minu:[4,13],minut:[1,5],miracl:10,miscdatahandl:6,miscellan:6,misdesign:1,mishmash:11,mislead:[1,4,5,10,11],misrepres:1,misrepresent:1,miss:[1,5,6,10],mistak:[0,1,4,5,7,9],mistaken:11,mister:9,mix:3,mmr:1,mock:4,modal:4,mode:[],model:[2,3,5,9,10,11],moder:[4,13],modern:[1,11],modest:[5,10],modul:[2,3,4,9,13],modulenotfounderror:[],modulu:2,moe:9,moment:[0,1,2,3,4,6,7,9,10,11,13],mondai:10,monei:[0,2,10,11],monoton:4,month:[1,3,7],mood:[1,4],moral:[1,5],more:[1,2,4,7,8,9,10,11,13],moreov:[1,10,11],morn:[4,9],moron:1,mortem:1,mortgag:4,most:[0,1,2,3,4,5,6,8,9,11,13,19,20],mostli:[0,1,3,4,5,7,9,13],motiv:[6,8,10,11,13],mous:[],move:[1,3,4,6,9,11,13],movement:1,mu0:[],mu1:13,mu2:13,mu_0:13,mu_1:13,mu_2:13,mu_d:13,mu_nul:13,mu_o:13,much:[0,1,2,3,4,5,6,7,8,9,11,13,20],multcomp:[],multi:4,multipl:[1,2,3,4,9,10,11,13],multipli:[1,3,4,9,10,13],multitud:4,multivariate_norm:4,mum:1,mumbl:13,murder:1,murki:0,music:[11,13],must:[1,2,3,5,9,10,11,13],my_column:3,my_fil:3,my_row:3,mydataanalysi:7,myfil:[],mynewdatanalaysi:7,myrawdata:7,myself:[1,2,4,5,6,7,9,11,13],myst_nb:[0,4,9,11,13],mysteri:[4,10,13],n_1:13,n_2:13,n_h:9,nabig:[],naisbitt:0,naiv:[0,1],name:[0,1,2,3,4,5,9,10,11,13],nameerror:[3,13],namespac:[],nan:4,nana:[],nancorr:4,nanmax:4,nanmean:4,nanmedian:4,nanmin:4,nanpercentil:4,nanstd:4,nansum:4,narrow:[1,5,9,10,11],narrowest:10,nasmal:[],nasti:[3,7],nastier:13,nathaniel:19,nation:[0,4,10],nationwid:4,nativ:13,natur:[1,2,3,4,6,9,10,11,13],naturalist:1,nav3:[],navarro:20,navig:3,navigationr:[],nearli:[0,6,7,10],neat:[1,6],neater:4,neatli:4,necess:11,necessari:[1,2,8,9,13],necessarili:[0,1,4,8,9,11,13],need:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],neg:[1,3,4,6,9],neglect:11,neglig:[4,13],neither:[1,4,11],nelson:8,neq:[9,11,13],nerd:[4,7],nerdstuff:[],nervou:1,net:4,network:10,neuropsycholog:1,neutral:[0,11],never:[1,2,4,5,6,7,8,9,10,11,13],nevertheless:[1,4,5,9,10,11,13],newborn:1,newcom:3,newer:[],newspap:[0,1,9],next:[0,1,2,3,4,5,7,8,9,10,11,13],nhst:11,nhstimg:[],nhstmess:[],nice:[0,1,2,4,5,7,9,10,11,13],nicer:3,nicknam:1,niggl:4,night:6,nightgarden2:[],nightgarden:[],nilli:[3,6],nine:3,nitpick:13,nlme:[],no_mean:10,nobodi:4,nomean_low:10,nomean_up:10,nomin:[4,5,13],non:[4,5,6,7,9,10,11],none:[0,1,4,5,8,9,10,13],nonparametr:13,nonsens:[2,5,13],nook:0,nor:[0,1,9,11],norm:[4,9,10,13],normal:[1,3,4,5,6,10],normal_a:9,normal_b:9,normal_c:9,normal_d:9,normal_data:13,normal_fig:9,normaldist:13,normals2:[],normals_fig2:9,normals_fig:9,normdist:[],normmean:[],normsd:[],norri:4,north:[4,5],notabl:[],notat:[4,9,10,11,13],note1:[],note3:[],note4:[],note5:[],note8:[],note:[0,1,3,4,5,6,7,9,10,13],notebook:[2,3,5,7],noth:[1,2,5,7,8,9,10,11,13],notic:[0,1,2,3,4,5,6,7,9,10,11,13],notion:[1,4,5,9],notori:2,notwithstand:9,novel:4,novemb:9,novic:[],now:[0,1,2,3,4,6,7,8,9,10,11,13,20],nsw:9,number:[0,1,3,4,5,6,7,9,11,13],number_of_flip:9,number_of_head:9,numer:[1,3,4,10,13],numpi:[2,3,4,6,9,10,11,13],numr:[],numref:[9,13],nut:[],nutrient:13,nutshel:9,obedi:[1,2],obei:9,object:[0,4,9,10,11],oblig:1,oblivi:4,obnoxi:5,obs:[],obscur:[5,11],observ:[1,4,5,6,7,8,9,10,11,13],obsess:[0,11],obsolet:[],obstetrician:1,obtain:[1,4,6,9,10,11,13],obviou:[1,2,3,4,5,8,9,10,11,13],obvious:[1,2,3,8,9,10,11,13,20],occas:[1,4,5],occasion:[1,2,4],occult:11,occur:[1,3,4,9,11,13],ocean:10,oct:9,odd:[1,3,4,7,11,13],odditi:11,oddli:[0,3,9],off:[1,2,3,4,5,8,9,10,13,20],offer:[0,4,9,10,11],offic:1,offici:1,offspr:9,often:[0,1,2,3,4,5,6,7,8,9,10,11,13],oil:5,okaaaai:[],okai:[0,1,2,3,4,8,9,10,11,13],old:[1,2,3,4,6,7,9,10,11],old_par:[],older:[1,6],oldest:1,omit:[4,9,11,13],onc:[1,2,3,4,5,6,7,8,9,11,13],one:[0,1,2,3,4,5,6,7,8,9,10],onecorplot:[],ones:[0,1,3,4,8,9,10,11,13],onesamplettest:13,onesid:11,onesidedtest:[],onk:6,onli:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],onlin:[2,3],onscreen:[],onto:[5,9,11],ontolog:11,opaqu:13,open:[1,2,3,4,6,10,20],openoffic:[],openscreen:[],oper:[1,2,3,4,10],operation:[10,11],operationalis:[4,9,10,11],opinion:[0,1,5,6,9,10,11],opinion_direct:6,opinion_strength:6,oppon:9,opportun:4,oppos:[4,5,9,13],opposit:[1,5,11,13],option:[1,3,5,6,10,13],orang:[4,5,13],order:[0,1,3,4,5,6,7,9,10,11,13],orderedfactor:[],ordin:[4,13],ordinari:[4,13],ore:10,org:[1,3,10],organ:[1,5,10],organis:[3,4,6,8,13],orient:[1,5],origin:[0,2,3,4,5,6,7,9,10,11,13],orthodox:[9,11],osten:19,ostens:1,other:[0,2,3,4,5,6,7,8,10,11,13],otherdist:[],otherwis:[2,3,10,11,13],ought:[0,1,2,3,5,7,10],our:[0,1,2,3,4,5,6,8,9,10,11,13],ourselv:[0,1,3,4,5,9,10,11,13],out:[0,1,2,4,5,6,7,8,9,10,11,13,20],outbreak:5,outcom:[3,5,6,8,9,10,11,13],outdat:1,outier:5,outlier:[4,5],outlin:[4,9,10,13],output:[2,3,4,5,6,7,11,13],outright:1,outset:[],outsid:[1,5,7,9,11],outward:9,outweigh:[2,11],over:[0,1,3,4,5,7,8,9,10,11,13],overal:[0,5,9,11,13],overarch:5,overflow:10,overgeneralis:13,overjoi:13,overlai:[5,9,20],overlap:[5,11],overli:1,overlook:0,overrid:5,overrul:2,oversampl:10,oversimplif:[4,13],oversimplifi:[],overview:[1,13],overwhelm:[5,11],overwrit:2,own:[0,1,2,3,4,5,6,8,9,10,11,13],p_lower:13,p_upper:13,p_valu:13,packag:[2,3,4,10,11],packageinstal:[],packet:4,page:[1,4],pai:[1,2,4,9,11],paid:8,pain:[5,13],paint:5,painter:5,pair:[4,9,10],pairedsamplesttest:13,pairedta:13,pairedtb:[],pairedtc:[],pakka:6,panda:[0,2,3,4,5,6,9,10,13],pandas_mad:4,panel:[4,5,9,10,13],panic:13,pant:9,pantsprob:[],paper:[0,1,2,5,7,9,10,11],par:[],paradox:[1,19],paragraph:[1,4],parallel:[1,10],paralys:[],paramet:[4,9,11,13],paranoid:4,parent:[0,1,4,5,13],parenthes:3,parenthood2:4,parenthood:[4,5],park:5,part:[0,1,2,3,4,5,8,9,10,11,13],partial:[1,4,5,13],particip:[1,2,3,5,10,11,13],particp:3,particular:[0,1,2,3,4,5,6,8,9,10,11,13],particularli:[5,11],partit:[],partli:11,pass:[1,3,4,7,9],passag:10,passeng:9,passion:[],past:[2,3,4,10],patch:11,path:[0,4,5,9,10,13],pathet:0,pathin:[4,5],pathlib:[0,4,5],patholog:[11,13],patienc:9,patient:[7,13],pattern:[1,4,9,11],paul:9,payoff:[5,9],pbinom:[],pch:[],pchisq:[],pdf:[3,4,9,10,13],pdistnam:[],peacock:19,peak:[3,5,9],pear:3,pearson:[4,11],pearsonr:4,pecentag:0,peculiar:3,pedagog:[2,7],pedant:10,peek:[3,4,5],pencil:2,peopl:[0,1,2,3,4,5,6,7,8,9,10,11,13],per:[1,2,5,9],perceiv:[0,7],percent:0,percentag:[0,4,9],percentil:[4,5,6,10],percept:11,perfect:[0,1,4,5,10,11,13],perfectli:[0,1,3,4,8,9,10,11,13],perform:[1,2,5,11,13],perhap:[1,3,4,5,9,10,11,13],period:[4,5],person:[0,1,2,3,4,5,7,8,9,10,11,13],perspect:[0,1,2,3,4,7,9,10,11],pertain:[5,10],pervas:1,pervers:0,peski:4,pfu11:[1,19],pfungst1911:[],pfungst:19,phase:5,phenomenon:[1,10],philosoph:[6,8,19],philosophi:[8,11,19],philosphicu:11,phone:10,phosphoru:13,phrase:[1,4,7,10,11,13],physic:[0,1,3,9,19],physicist:[0,1],physiolog:1,pick:[0,1,4,5,9,10,11,13],picki:[4,5],pictur:[4,5,8,9,10,11],piec:[1,2,3,4,5,7,9,10,11,13],piltdown:1,pip:6,pirat:1,piri:10,pitch:8,place:[0,1,2,3,5,9,11,13],placebo:4,plai:[0,1,3,4,5,8,9,10],plain:[3,5],planet:0,plant:13,platform:5,platykurt:4,plausibl:[0,8,9,10,13],player:5,pleas:[5,10,13],pleasant:9,plo:19,plot:[0,4,6,7,9,11],plot_:[],plothist:[],plotmean:[],ploton:[],plotsampl:10,plottwo:[],plt:[4,5,9,10,11,13],plu:[1,9,13],png:[5,10,13],pnorm:[],pocket:9,poem:0,poet:0,point:[0,1,2,3,4,5,6,7,8,9,10,11,13],pointestim:10,pointi:4,pointless:1,pointplot:[10,13],poison:[10,13],polic:1,polici:1,polit:13,politician:1,poll:[9,10],pollard:19,pollut:10,polygon:[],pool:[],pooled_s:[],poor:[1,4,11,13],poorli:5,pop:[8,9,10,11,13],popmean:13,popper:11,popul:[1,4,9,11,13],populac:1,popular:[1,4,5],population_mean:10,population_vari:4,port:[4,5,10],portabl:9,portion:3,portrai:11,pos:[],pose:1,posit:[1,2,3,4,6,10,11],possess:[1,10],possibl:[1,2,3,4,5,7,8,9,10,11,13],post:[0,1],poster:5,postgradu:[0,13],potato:1,potent:9,potenti:[5,9],potsdam:4,potter:2,poulat:10,pow:2,power:[0,1,3,4,5,6,9,10,13,19],powerfunct:11,powerfunction_fig:[],powerfunctionsampl:11,ppf:[10,11],practic:[0,1,2,3,4,5,6,9,10,13],practition:11,pragmat:[0,1,6,9],pragmatist:9,pre:[0,1],preambl:5,precis:[1,2,3,4,9,10,11,13],pred1:[],pred2:[],pred:[],predict:[4,8,9,11,13,19],predictor:[6,13],prefer:[0,1,2,3,4,6,7,8,9,11,13],prefix:[],prelud:10,prematur:1,premier:9,premis:0,prepar:[5,13],present:[0,1,2,4,5,10],preserv:4,presidenti:1,press:[2,19],pressur:[1,13],presum:[2,4,5],presumpt:11,pretend:[1,4,5,9,13],pretti:[0,1,2,3,4,5,6,7,8,9,10,11,13],prevent:[1,7,9,11],preview:[],previou:[1,2,3,4,9,10,11,13],previous:[4,6,13],price:[2,4],primari:[2,9],primarili:[3,8,13],prime:13,principl:[0,1,11,13],print:[2,3,4,9,10,11,13],printout:[],prior:0,prioriti:[0,2],privaci:0,privat:11,prize:[4,13],prob:[9,11],prob_low:11,probabail:10,probabilist:[9,10],probabl:[0,1,2,3,4,5,6,8,10,13,20],probablil:11,problem:[0,1,2,3,4,5,6,7,8,9,10,11,20],probmean:10,probplot:13,probrul:[],probstat:[],proce:9,procedur:[6,9,10,11],process:[0,1,2,3,4,5,7,10,11,13],proclaim:11,produc:[0,1,2,4,5,6,9,10,11,13],product:[1,4],profession:[0,1],professor:11,profil:1,profit:[],program:[2,3,5,9],progress:9,project:[0,1,6,10],projecthom:[10,13],promin:[3,11],promis:[9,10],promot:11,prompt:[7,13],pronoun:1,proof:[10,11],prop:[],proper:[1,3,4,8,10,11],properli:[1,5,9,11,20],properti:[2,4,5,9,10],prophesi:1,proport:[0,1,4,9,11],proportion_head:9,proposit:6,proprietari:2,prosecutor:11,protect:11,prove:[1,10,11],provid:[0,1,3,4,5,6,8,9,10,11,13],proviso:11,proxi:1,pseudo:4,pseudonym:13,pseudorandom:9,psych:[0,1,13],psychic:11,psycholog:[2,3,4,5,8,9,10,11,13,19],psychologist:[0,1,8,9,10,11,13],psychotherapi:1,psychotrop:2,publish:[1,2,6,9,11,13,19],pull:[1,4,10,13],pump:5,punish:11,pure:[0,1,4,7,9,10,11,13],purer:13,purpos:[1,2,3,4,5,6,9,10,11,13],purview:0,push:2,put:[0,1,2,3,4,5,6,9,10,11,20],puzzl:8,pvaltabl:[],pvalu:13,pvarianc:4,pygmalion:1,pyplot:[4,5,9,10,11,13],pythob:[],python3:3,python:[6,7,11],pythonbook:[3,4,5,13],qbinom:11,qchisq:[],qcut:6,qnorm:[],qq1a:[],qq1b:[],qq2a:[],qq2b:[],qq2c:[],qq2d:[],qq_fig:13,qqheavi:13,qqheavy_fig:13,qqnorm:[],qqskew:13,qqskew_fig:13,qualifi:[0,5],qualit:[1,4,9,11,13],qualiti:[0,5,9],quantifi:[4,10,11],quantil:[4,6,10,13],quantit:[4,5],quantiti:[4,9,11,13],quantum:0,quarter:4,quarterli:[],quartet:4,quartil:[4,5],quasi:[1,19],queensland:1,quest:2,question:[0,1,2,4,5,6,9,10,11,13],questionnair:[0,1,4,9],quick:[1,3,4,5,11,13],quickli:[2,3,5,9],quiet:11,quietli:9,quirk:[2,3,11],quit:[0,1,2,3,4,5,6,7,9,10,11,13],quizz:0,quot:[0,4,5,6,7,8,10,11],quotat:[3,5],race:1,racket:2,rain:9,rais:[1,2,4,10],rake:9,ralph:4,rambl:10,ramp:10,ran:[0,10,11,13],random:[1,4,5,7,9,11,13,20],randomis:[1,11],randomli:[1,4,9,10,13],rang:[2,5,6,9,10,11],rank:[0,4],rankcorr:4,rankcorrp:[],ranked_grad:4,ranked_hour:4,rant:10,rapidli:[1,7],rare:[0,1,3,4,10,11,13],rate:[0,1,4,11,13],rater:1,rather:[0,1,2,3,4,5,6,7,9,10,11,13],ratio:[4,5,9,13,20],ration:9,rational:6,ravish:9,raw:[0,1,3,4,5,6,7,10,11,13],rbind:[],rbinom:[],rbook:3,rchisq:[],rcl:13,rda:[],rdata:13,rdb:[],rdx:[],reach:[0,10],reactionari:0,read:[0,1,2,3,4,5,6,7,9,10,11,13],read_csv:[0,3,4,5,13],readabl:[3,4,5],reader:[1,4,5,9,10,11,13],readership:[4,9],readonli:[],readrd:[],real:[0,1,2,5,6,7,8,9,10,11,13],realis:[1,2,5,7,8,11,13],realist:[1,13],realiti:6,realiz:[3,13,20],realli:[0,1,2,3,4,5,6,7,8,9,10,11,13],realm:10,reanalys:1,reang:4,reason:[0,1,2,3,4,5,6,7,9,10,11,13,19],reassign:2,recal:[1,4,10,13],recalcul:10,recap:11,reccomend:2,receiv:[4,5,13],recent:[0,3,5,8,9,13],reckon:8,recogn:3,recognis:[1,7,9,10,11,13],recognit:2,recommend:[3,13],reconsider:19,record:[1,3,4,7,8,10,13],recreat:7,rectangular:[],red:10,redefin:13,redescrib:[9,13],redraw:5,reduc:[1,5,9,11,13],redund:2,ref:[0,1,6,10,11,13],refer:[1,3,4,5,6,7,8,9,10,11,13,20],reffig:[],refineri:10,reflect:[0,1,4,6,9,10,11,13],refsmmat:0,refus:[0,9],regard:[1,11],regardless:[1,3,4,5,6,10,11,13],region:13,regist:1,regplot:[4,5,10],regress:[6,10,13],regular:[3,4,13],rehash:10,reinforc:[1,13],reinvent:3,reject:[0,4,8,11,13],rel:[1,2,3,4,6,9,10,13],rela:4,relat:[0,1,2,3,4,9,10,11,13],relatedli:5,relationship:[1,3,5,6,9,10,11,13],relax:9,releas:[],relev:[1,3,4,9,10,11],reli:[0,1,4,5,6,7,8,9,10,11,13],reliabl:[4,10],relianc:10,reliant:[0,6],reload:[],remain:[1,4,5,8,9,11],remark:[0,3,8,9,10],remedi:[4,13],rememb:[0,1,2,3,4,5,7,9,10,11,13],remind:[1,5],remov:[0,4,5,11,13],renam:[3,4,5,6],rent:4,rep:[],repeat:[9,10,13],repeatedli:9,repetit:[2,9,13],replac:[5,10,13],repli:1,replic:[10,11],report:[0,1,4,5,9,10,13],repres:[1,3,5,6,9,10],represent:4,repress:[4,19],reproduc:7,reput:11,request:[],requir:[0,1,2,4,6,7,8,9,10,11,13],res:[],research:[2,4,5,6,9,10,13,19],resent:10,reserv:[],reset:3,reset_index:5,reshap:6,resolv:[2,13],resort:3,resourc:5,respect:[1,4,5,8,9,10,11,13],respond:[1,3,11],respons:[2,3,4,6,9,10,11,13],rest:[4,5,6,8,10,11,13],restor:[],restrict:[10,11,13],result:[0,1,2,3,4,5,6,9,10,13,19],retain:[11,13],retent:11,retest:1,reus:7,reveal:[4,10],revel:[],revenu:2,revers:[0,1,11,13],review:19,revis:8,revisit:[1,3,4,9,11],reward:[4,9],rewrit:10,rewritten:10,rgamma:[],rgb:[],rhistori:[],rho:4,rich:[2,3],richmond:[4,5],rid:3,riddl:[5,8,10],ride:8,ridicul:[1,13],rig:[9,10,13],right:[0,1,3,4,5,6,7,8,10,11,13],rightarrow:[9,10],rigor:11,rigour:1,ring:10,rise:[1,4,9,11,13],risk:1,rmdnote:[],rmsd:4,rneg:4,rnorm:10,robot:[2,4,9],robust:4,rock:[],roger:[2,6],role:[0,2,10],roll:9,ronald:[1,9,11,13],ronald_fish:1,room:[5,11],root:[3,4,9,10,13],rosenthal1966:1,rotat:5,rough:[1,4,13],roughli:[6,9,10],round:[5,8],roundabout:1,row1:4,row2:4,row:[3,4,5,6,7,8,9,11],rowtot:6,royal:19,royalti:2,rseek:[],rsquo:[],rstudio:[],rtfm:[],rubbish:4,rude:[],rule:[1,2,4,5,8,9,10,11,13],run1:9,run2:9,run3:9,run4:9,run:[0,1,2,3,4,5,6,7,8,9,10,13],rush:11,sabotag:0,sack:1,sad:9,sadder:9,sadli:13,safe:[0,1,2,3,5,11,13],safeguard:0,safer:[2,4],safest:4,sai:[0,1,2,3,4,5,6,8,9,10,11,13,20],said:[0,1,2,3,4,5,8,9,10,11,13],sake:[3,4,5,13],salari:1,sale:[2,3],salem:[],same:[0,1,2,3,4,5,6,8,9,10,11],sampdistmax:10,sampdistmean:10,sampdistsd:10,sampel:13,sampl:[1,4,5,8,9,19,20],sample_max:10,sample_mean:[10,13],sample_mean_1:10,sample_sd:10,sample_standard_error:13,sample_vari:4,samplemean:10,samplesandclt:10,samplesd:10,samplingdist:[],san:4,sane:9,saniti:6,satisfi:[6,9,10,11,13],saturdai:4,save1:[],save:[2,4,6,7],savefig:5,savvi:[],saw:[1,4,5,8,9,13],scale:[4,5,6,9,10,13,19,20],scaled_chi_square_20:9,scaled_chi_square_3:9,scaled_chi_square_data:9,scalescont:[],scan:1,scare:[1,5],scari:[],scat:13,scatterplot:[4,13],scenario:[1,3,4,9,11,13],scene:2,scheme:[1,4,10],scherndl:19,schizophren:10,schizophrenia:10,school:[1,2,3,10,11],sci:[],scienc:[0,1,2,10,11,19],scientif:[0,1,4,5,9,10,11,13],scientist:[0,1,4,10,11,13],scipi:[2,3,4,9,10,11,13],sciplot:[],scope:[1,4,5,9,10,13],score:[1,3,6,8,10,11,13],score_data:3,scorecard:8,scotsman:1,scratch:6,screen:[2,3,4],screenshot:[],scribbl:[4,7],script:[3,5],scriptstyl:4,scroll:3,sd0:[],sd1:[],sd_diff:13,sd_true:13,sdnorm1:[],sdnorm2:[],sdnorm:9,sdnorm_fig2:9,sdnorm_fig:9,se_est:13,seaborn:[0,2,3,4,5,9,10,11,13],seali:13,search:[1,3,5,11],searchabl:[],season:[4,5,10],second:[0,1,2,3,4,5,6,8,9,10,11,13],secondari:11,secondli:[0,1,5,6,8,9,10,13],secret:13,section:[0,1,3,4,5,6,7,9,10,11,13],secur:[4,5],see:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],seed:13,seek:[1,11],seeker:3,seem:[0,1,2,3,4,5,6,7,8,9,10,11,13],seen:[1,2,3,4,5,6,9,11],select:[0,4,5,6,10],self:[0,2,3,10],sell:[0,2],sem:[10,13],sem_tru:13,semest:[4,13],send:[2,5,7],senior:4,sens:[0,1,2,4,5,6,7,8,9,10,11,13],sensibl:[0,1,5,7,8,9,10,11,13],sensit:[2,4],sent:13,sep:[],separ:[3,4,5,6,9,10,11,13],septemb:[4,7],seq:[],seq_along:[],sequenc:[8,9,10],seri:[2,3,4,5],serial:[],seriou:[4,6,11],serious:[1,2,5,9,13],serv:[1,3,4,13],session:[1,4],set:[0,1,2,3,4,5,6,7,8,9,10,11,13],set_capt:[],set_color:11,set_context:5,set_frame_on:13,set_linestyl:4,set_text:13,set_titl:[4,5,9,10,11,13],set_vis:[4,5,10,13],set_xlabel:4,set_xlim:[4,13],set_ylabel:4,set_ylim:[4,13],settl:[2,4,9],setupplot:[],setwd:[],seven:5,seventh:8,sever:[1,2,3,4,5,6,7,9,10,11,13],sex:[10,19],sexist:[0,10],shade:[4,5,9,11],shake:10,shall:[10,13],shalt:0,shame:[5,7,11],shape:[4,5,9,10],shapiro1965:13,shapiro:[],shapirowilkdist:[],share:[0,1,5,7,9,10],sharei:[4,9,10,13],sharex:10,sharpli:[11,13],she:[1,2,10,11,13],shed:4,sheer:13,shelf:10,shell:2,sherlock:8,shift:[2,5,9,11],shini:6,shirt:10,shock:[1,4,13],shockingli:[4,7],shoe:[1,10],shook:10,shop:3,shortag:0,shortcom:4,shorten:[4,11],shorter:[1,10],shorthand:13,shortli:[4,7,9,11],shot:[1,10,11],should:[0,1,2,3,4,5,6,7,8,9,10,11,13],shoulder:3,shouldn:[0,1,4,5,8,9,10,13],show:[0,1,2,3,4,5,6,7,9,10,11,13],shown:[0,1,2,4,7,9,10,11,13],shrink:9,shrunk:[6,11],shuffl:9,side:[0,4,5,7,8,9,10],sidewai:4,sig:11,sigex:[],sigh:[1,4],sigma1:13,sigma2:13,sigma:[4,9,10,13],sigma_0:13,sigma_1:13,sigma_2:13,sigma_d:13,sigma_p:13,sign:[2,3,4,6,13],signif:[],signific:[0,13,19],significantli:[11,13],silli:[1,2,5,10,11,13],sim:[9,11,13],simdata:10,similar:[1,2,4,5,7,9,10,11,13],similarli:[1,4,9,11,13],simpl:[0,1,3,4,5,6,7,9,11,13],simpler:[2,4,5,9,13],simplest:[1,2,3,4,5,9,10,11,13],simpli:[1,2,3,4,5,6,9,10,13],simplic:[5,13],simplif:[10,11],simplifi:[5,10],simpson:1,simpsont:[],simul:[9,10],sinc:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],sinclair:1,singl:[0,2,3,5,6,7,9,10,11,13],sink:[],sir:[1,9,11,13],sit:[0,1,3,4,7,9,10,11],site:[5,10],situat:[0,2,4,5,6,8,9,10,11,13],six:[0,4,7,8,9],sixth:4,size:[0,1,3,6,9,10,19],skew:[9,13],skew_fig:4,skewdata:4,skewed_data:13,skill:[1,2,3,7,8,10],skim:9,skip:[4,13],skipna:4,skull:9,skulls_fig:[],sky:13,slant:1,slash:3,sleep:[4,5],sleep_scatt:4,sleepi:4,sleepier:4,slept:4,slice:[4,7],slightli:[0,1,2,4,5,6,7,9,10,11,13],slope:5,sloppi:10,slower:[0,20],slowli:[4,11],small:[0,4,5,6,7,9,10,11,13],smaller:[4,9,10,13],smallest:[4,9,11,13],smallish:[6,7],smart:[0,1,2,7,13],smarter:[0,1],smelter:10,smirnov:13,smoke:1,smoker:[1,10],smooth:9,smoothi:13,sms:10,sneaki:[1,10,13],snippet:3,snow:5,snowbal:10,sns:[0,3,4,5,9,10,11,13],soccer:9,social:[0,1,2,6,10],societi:19,sociolog:0,soft:11,soften:11,softwar:[1,2,3,4,11,13],sokal:1,sole:[4,5],solid:[1,4,9,10,13],solut:[1,2,4,5,9,13],solv:[0,1,2,3,4,5,7,11,13],some:[0,2,3,4,5,6,7,8,9,10,13,19],somebodi:1,somehow:[10,11],someon:[0,1,2,4,5,6,8,9,10,11,13],someth:[0,1,2,3,4,5,7,8,9,10,11,13],sometim:[0,1,2,3,4,5,6,7,8,9,10,11,13],somewhat:[0,1,4,6,9,10,11,13],somewher:[3,4,9,11],son:[4,10],song:11,soon:2,sooner:5,sophist:[1,4],sorri:8,sort:[0,1,2,4,5,6,7,9,10,11,13],sort_valu:4,sorted_margin:4,sortfram:[],sound:[0,1,2,3,4,5,8,9,10,11],sourc:[1,2,4,5,8,10,11,20],south:10,soviet:[4,9,19],space:[5,9],span:[4,5,6],speak:[2,3,4,5,7,8,9,10,11,13],speaker:6,spearman:4,spearmanr:4,speci:0,special:[1,2,3,4,7,8,9,10,13],specialis:[5,7],specif:[0,1,2,3,4,5,9,10,11,13,20],specifi:[1,4,5,6,9,10,11,13],specifii:[4,10],spectacularli:4,speech:2,speed:11,spell:1,spend:[0,1,2,3,4,5,6,7,8,9,10,11],spent:[1,4,7,9],spine:[4,5,10],spirit:[3,4],spit:2,split:4,sport:4,spot:9,spous:9,spread:[4,5],spreadsheet:[2,5],spring:9,springer:19,spss:11,sqrt:[4,9,10,13],squar:[2,3,4,9,10,11,13,20],squared_error:4,squat:5,srs1:10,srs2:[],srs:10,stabl:2,stackoverflow:5,stage:[3,5,6,8,10,11],staggeringli:2,stai:[0,5,6,9],stalin:4,stalinist:4,stamp:9,stand:[1,3,4,9],standard:[1,2,3,5,6,9,11],standardis:[4,13],standardnorm:[],stanlei:19,star:[5,11],stare:5,start:[0,1,3,4,5,6,7,8,9,10,11,13,20],starv:0,stat:[0,1,2,4,9,10,11,13,20],stata:[],state:[1,2,9,10,11,13],statement:[0,1,2,3,4,9,10,11],statist:[1,2,3,5,6,7,13,19],statistic_:[],statistician:[0,1,4,8,9,10,11,13,19],statsmodel:[3,4,10],statsmodels_mad:4,std1:[],std2:[],std:[4,5,10,13],stdev:[4,10,13],ste46:[1,19],steadili:9,steel:10,steep:5,stem:1,stencil:5,step:[1,2,3,4,6,8,11,13],stereotyp:1,steril:9,stern:19,steven:19,stevens1946:[],stick:[3,4,6],stigler1986:10,still:[0,1,2,3,4,5,8,9,10,11,13],stomach:4,stone:13,stop:[1,2,4,5,7,9,13],store:[3,4,5,6,7,9,13],stori:[0,1,2,4,6,9,10,11,13],storm:10,str:[0,3,4,5,10],straight:[2,4,11,13],straightforward:[2,3,4,5,9,10,13],strang:[3,4,6,13],strata:10,strategi:[2,8],stratifi:10,streak:[5,8],street:5,strength:[2,5,6,10,13],strengthen:11,stress:[0,4,5,10],strict:[1,2],strictli:[1,4,11,13],strike:0,string:[3,7],stringsasfactor:[],strip:[5,7],stripplot:5,stroke:5,strong:[0,1,4,5,6,8,9,11,13],stronger:[4,11],strongli:[1,2,6],stroppi:[],structur:[0,1,3,5,11,13],struggl:5,strung:5,stuck:2,student10:[],student1908:13,student1:13,student2:13,student3:13,student4:13,student5:13,student:[0,1,2,4,10,11,20],studentassumpt:13,studentttest:13,studi:[0,2,4,5,6,10,13,19],studydesign:10,stuf:1,stuff:[0,1,2,3,4,10,11],stupid:[1,2,4,5,9,10,11,13],style:[1,5,7,10,11],stylis:5,subfield:4,subfold:[],subgroup:1,subject:[1,4,9,10,13],subject_:[],subjectivist:9,submit:1,subplot:[4,5,9,10,13],subpopul:10,subract:13,subsampl:1,subscrib:9,subscript:[4,13],subsequ:[1,5],subset:[6,10],subsetdatafram:6,substant:11,substanti:[1,4,13],substitut:[0,13],substract:13,subtl:[0,1,2,3,4,5,9,10],subtleti:[4,5,9],subtli:[1,11,13],subtract:[1,6],succeed:[],success:[0,4,9,11],successfulli:11,succumb:11,suck:2,suddenli:7,sued:0,suffer:[0,11],suffic:3,suffici:[0,1,2,4,10,11,13],suggest:[0,1,2,3,4,5,9,11,13],suit:[3,5,9,13],suitabl:[],suk:1,sullivan:3,sum:[2,3,4,6,9,13],sum_:[4,10,13],sum_numb:[],summar:5,summari:[],summaris:[1,4,7,9,10,11],summat:4,summon:4,sun:[10,11],superior:1,supplement:[],suppli:[0,5,10],support:[3,11,13],suppos:[0,1,2,3,4,5,6,7,8,9,10,11,13],suppress:11,suptitl:[4,9,10],sure:[1,2,3,4,5,7,8,9,10,11,13],surfac:6,surpris:[0,1,2,3,4,5,9,11,13],surprisingli:[0,1,2,4,9,10,11,13],surround:11,survei:[1,2,4,9,10],surviv:0,suscept:0,suspect:[1,4,5,9,10,13],suspici:[4,5,8,11,13],suspicion:5,swai:0,swath:4,swdist:13,sweet:[2,3,13],swing:1,sydnei:[4,5,9],syllogist:19,symbol:[1,2,3,4,5,10],symmetr:[4,6,13],syndrom:1,synonym:4,syntax:5,syntaxerror:2,systat:[],system:[2,3,5,10,11],systemat:[0,1,4,10,13],t_3:9,t_stat:[],tab:[6,10],tabl:[0,1,2,3,4,9,10,11,13],tabul:13,tack:4,tackl:[3,10],tag:2,tail:[4,5,9,11,13],take:[0,1,3,4,5,6,8,9,10,11,13,20],taken:[0,2,3,4,6,7,9,10,11,13],talk:[0,1,2,3,4,5,6,8,9,10,11,13],tall:1,taller:1,talli:9,target:[],task:[1,2,3,4,5,6,9,10],taught:[2,10,13],tautolog:11,tconfint_mean:10,tdist:9,tdist_3:[],teach:[0,1,7,13,20],team:[4,5,9,10],teas:1,techin:5,technic:[1,4,9,10,11,13],techniqu:[10,11],tediou:[1,2,4,13],teensi:0,tell:[1,2,3,4,5,6,7,9,10,11,13],temperatur:[1,9],tempestu:10,temporarili:1,tempt:[4,5,11,13],temptat:[0,11],ten:11,tend:[0,1,4,5,6,7,9,10,11,13],tendenc:10,tension:11,tenur:11,term:[0,1,2,3,4,5,6,8,9,10,11,13],terminolog:[1,2,9,10,11],terplot:13,terrestri:10,terribl:[1,4,9,11,13],territori:9,ters:11,test1:[],test2:[],test:[0,3,4,6,8,9,10,19],test_1samp:13,test_:[],testabl:11,tests_:[],testsm:13,teststatist:[],text:[2,3,4,5,6,7,9,10,11,13,20],textbook:[1,2,4,9,11],textedit:[],textprocess:6,textual:20,than:[1,2,3,4,5,6,7,8,9,10,11,13],thank:[0,9,20],thankfulli:0,theatr:1,thei:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],theirs:1,them:[0,1,2,3,4,5,6,7,9,10,11,13],themselv:[0,1,4,5,6,7,9,10,13],theorem:[8,13],theoret:[1,8,10,11,13,19],theori:[0,4,10,11,13,19],therapi:4,therebi:[1,4],therefor:[0,1,2,3,8,10,11,13],thermomet:9,theta:[9,11],theta_0:11,thi:[0,1,4,5,6,7,8,9,10,11,13,20],thick:1,thin:5,thing:[0,1,2,4,5,6,7,8,9,10,11,13,20],think:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],third:[1,2,3,4,5,6,9,10,11,13],thirdli:[0,5],thoma:19,thorni:10,thorough:10,thoroughli:11,those:[0,1,2,3,4,5,6,7,8,9,10,11,13],thou:0,though:[0,1,2,3,4,5,6,8,9,10,11,13],thought:[0,3,4,9,10,11,13],thoughtless:11,thousand:[1,2,3,9,11],threaten:1,three:[0,1,2,4,5,6,8,9,10,11,13],through:[1,3,4,5,6,7,8,9,10,11,13,20],throughout:[1,4,5,8,9,11],thu:[3,4,5,9,13],thumb:4,tibbl:[],tick:[1,5,13],tick_param:[9,10],tidi:6,tidli:5,tie:1,tied:0,tier:11,ties:[2,13],tiger:11,tight:5,tightli:[1,9,10,13],tild:[],time:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],tini:[4,7,9,10,11,13],tionship:4,tip:[5,10],tire:[1,4],titl:[1,4,11,13],tmp:[],to_csv:3,to_fram:5,todai:[1,9,11],togeth:[1,3,4,5,6,7,9,10,13],toi:[3,4,6,11,13],told:[0,3,4,5,8,9,11],toler:[1,11],tolist:3,tombliboo:6,tomorrow:[9,11],too:[0,1,2,3,4,5,6,7,8,9,10,11,13],too_high:10,too_low:10,took:[0,1,3,4,10,13],tool:[0,1,2,4,5,7,8,9,10,11,13],top:[0,1,3,4,5,9,10,13,20],topic:[0,1,3,4,6,9,10,11,13],total:[0,4,6,9,10,11,13],total_bil:10,touch:[4,5,10],tough:9,toward:[1,4,9,13],tower:[9,11],town:10,traceback:[3,13],track:9,tracksuit:9,tract:0,tractatu:11,traction:9,tractor:6,trade:[4,13],tradit:[1,4,13],tradition:[4,10,13],traditionalist:1,traffic:[],tragedi:4,train:[0,1,9],tran:10,transact:19,transcrib:6,transfer:1,transferr:1,transform:[4,10],transgend:[1,10],transit:[],translat:[4,5,7,9,10,11,13,20],transluc:5,transport:1,trap:[0,1,2],travel:[1,4],treat:[1,3,4,11,13],treatment:[1,11],trend:0,tri:[1,2,3,4,5,6,7,11],trial:[1,4,9,11],triangular:10,tribbl:[],trick:[1,2,6,7,9,13],tricki:[1,4,9,11],trickier:[4,9,11,13],trigger:13,trillion:11,trim_mean:4,trite:9,trivial:[6,13],troubl:[1,4],truism:10,truli:[4,9],trust:[0,1,7,9,13],trustworthi:[0,1,4,13],truth:[0,1,2,4,8,9,10],ttest2:13,ttest:[],ttest_1samp:13,ttest_1sampresult:13,ttest_ind:13,ttest_indresult:13,ttest_rel:13,ttest_relresult:13,ttestci:[],ttestdist:[],ttestfunct:[],ttesthyp2:13,ttesthyp:13,ttesthyp_onesampl:13,ttestimg:[],ttestoneassumpt:13,tuck:[],tuft:5,tupl:3,turf:5,turn:[0,1,3,4,5,8,9,10,11,13],tutor1:13,tutor2:13,tutor:13,tutori:[3,13],tverski:19,tweak:[5,7,10],twice:[1,10],twilight:9,twini:9,twinx:[4,9,10],twist:10,twister:[],two:[1,2,3,4,5,6,7,8,9,10,20],twofold:5,type:[0,1,4,5,6,7,9,10,13],typic:[0,1,3,4,9,10,13],typo:[4,5,7],ubiquit:11,ubtract:2,ugh:4,ugli:13,ultim:[8,11,20],ultipl:2,unabl:11,unaccept:[],unacknowledg:1,unambigu:[4,9,13],unavoid:[1,8],unawar:11,unbeliev:0,unbias:[0,4,10],unbiased:10,uncertain:10,uncertainti:[1,5,10,13],uncomfort:9,uncommon:[1,4,6,13],uncontrol:1,undefin:[],under:[0,2,4,5,9,10,11,13,20],undergradu:[0,1,9,10],underli:[5,6,9,11],underpin:[8,9,11],underpow:5,underst:4,understand:[0,1,2,3,4,5,6,7,9,10,11,13],understood:[4,9],undesir:9,undo:[3,5],unduli:4,unenlighten:0,unequ:[6,13],uneth:[1,10],unexpect:3,unfair:1,unfairli:0,unfalsifi:11,unfamiliar:[],unfold:9,unfortun:[1,3,4,5,9,10,11],ungrammat:4,unhelp:[4,10],unicorn:13,uniform:9,unimagin:[],uninform:1,unintent:1,unintention:1,uninterpret:4,uniqu:[4,5,10,11],unit:[4,6],univers:[0,1,9,10,13,19],unix:3,unknow:5,unknown:[3,4,9,11,13],unlabel:[],unless:[2,3,4,5,6,7,10,11],unlik:[1,3,4,5,9,11,13],unlink:[],unlucki:11,unmeasur:1,unnatur:1,unnecessari:[4,13],unpopular:9,unproblemat:10,unproduct:11,unrealist:[1,4],unreason:[4,8,13],unrel:4,unreli:1,unremark:9,unrepent:11,unrepres:1,unsatisfi:[],unseri:[],unsur:2,unsurpris:11,until:[0,2,3,4,5,7,9,10,13],untitl:[],untrain:1,untru:4,unus:3,unusu:[4,5],unwari:[0,11],updat:[1,20],updatepackag:[],upload:2,upon:[0,1,5,10,11],upper:[1,4,10,11,13],upper_area:13,upset:2,upsi:6,uptight:2,upton:1,upward:[11,13],urgent:0,url:10,usabl:2,usag:[4,13],use:[0,1,2,5,6,8,9,10,11,13,20],used:[0,1,2,3,4,5,7,9,10,11,13],useful:[0,1,2,3,4,5,6,7,10,11,13],useless:[0,4,5,10,13],user:[2,3,4,5],uses:[1,2,3,4,6,10,13],using:[0,1,3,4,5,6,7,8,9,10,11,20],usingfunct:[],usual:[0,1,2,3,4,5,6,9,10,11,13],usus:5,utf:[],utter:6,v10n3:10,vaccin:1,vagu:[1,10,11],val:[4,10],valid:0,valign:[],valu:[0,1,2,3,5,6,9,10],valuabl:1,value_count:[4,5],van:0,vancouv:4,var1:4,var2:4,var3:4,vari:[1,3,4,9,10,11,13],variabl:[5,7,9,10,13],varianc:[9,10,13],variance_:[],variant:1,variat:[1,2,4,10,13],variaterel:[],varieti:[2,5,13],variou:[0,2,3,7,9,11],vast:[0,1,4,5,9,10,11,13],vector:[4,6,9,10,13],vehement:9,verbal:[0,1],verbalis:1,verbos:[],veri:[0,1,3,4,5,6,7,9,10,11,13,20],verifi:[2,4],version:[0,1,2,4,5,6,11,13,20],versitil:3,versu:[9,13],vertic:[5,9],vertigo:7,vest:4,via:[7,13],viabl:9,victori:[8,9],view:[0,1,5,6,10,13],viewer:5,violat:[1,6,13],violinplot:5,virtual:3,visibl:[],vision:2,visual:[1,4,5,9,10,13],visualis:[4,5,9],vline:10,vocabulari:1,voic:1,volum:1,von:19,vote:[9,10],voter:[9,10],vulner:[1,4],w_1:13,w_2:13,waaaai:5,wacki:11,wage:[2,4],wai:[0,1,2,3,4,5,6,7,8,9,10,11,13,20],wait:[0,4,5],wake:[1,13],wakefield:1,walk:[9,11],wall:11,wander:10,want:[0,1,2,3,4,5,6,7,8,9,10,11,13],war:[4,8],wardrob:9,wari:11,warn:[0,1,3],warrant:[4,13],wasn:[1,2,5,9,11,13],wast:[5,9],watch:[1,4,5,6,8,9],water:[1,5],watercolour:5,wcc:8,weak:[0,2,4,5,11,13],weaken:13,weaker:[10,11],wealthi:10,wealthiest:4,wear:9,weather:9,web:2,webcom:3,websit:[0,4,10],wed:[],weed:20,week:[1,3,4,5],weigh:13,weight:[1,13],weird:[0,1,4,5,8,9,10,11],weirdli:10,welch1947:13,welch:[],welchttest:13,welcom:3,well:[0,1,2,3,4,5,6,7,8,9,10,11,13],welles:[8,10],welleslei:[8,10],wellington:8,wellselei:10,went:[0,1,4,5,9],were:[0,1,2,3,4,5,7,8,9,10,11,13],weren:[5,10],west:[4,5],western:[4,5],what:[0,1,2,5,6,7,8,10,11,20],whatev:[1,2,3,4,9,11],whatsoev:[4,13],wheel:3,when:[0,1,2,3,4,5,6,7,8,9,10,11,13],whenev:[1,6,9,10,11,13],where:[0,1,2,3,4,5,6,8,9,10,11,13],wherea:[0,1,4,6,9,10,11,13],whether:[0,1,2,3,4,5,6,8,9,11,13],whew:13,which:[0,1,2,3,5,6,7,8,9,10,11,13,20],whichev:11,whing:2,whisker:5,whiskerprop:5,white:[1,5,10,11],whitnei:13,who:[1,2,3,4,5,7,8,10,11,13],whole:[0,1,2,3,4,8,9,10,11,13],whose:[1,2,4,5,9,10,11,13],why:[1,2,3,4,5,6,8,9,10,11,13,19],why_you_gotta_be_so:3,whyalla:10,whyfor:[],wide:[2,5,10,13],widen:9,wider:[1,9,10],widetolong:[],width:[5,9],wife:11,wiki:1,wikiquot:1,wilcox1:13,wilcox2:13,wilcox:[],wilcoxon:[],wild:1,wildli:[1,9],wilei:19,wilfri:0,wilk:[],willi:[3,6],william:[7,11,13],willing:[1,8,9,11,13],win:[4,5,8,9,10,11,13],winc:2,window:[3,11],wipe:4,wise:[0,6],wish:4,wit:1,within:[0,1,2,4,5,6,7,9,10,11,13],without:[0,1,2,3,4,5,9,10,11,13],witter:10,wittgenstein:11,wmc:[],woman:4,women:0,won:[1,2,3,4,5,7,8,9,10,13,20],wonder:[4,9,10,11,13],woo:1,word:[0,1,2,3,4,6,7,8,9,10,11,13],work:[0,1,2,3,4,5,6,7,8,10,11,20],workaround:[],worker:1,workspac:7,workspace2:[],workspacepanel2:[],workspacepanel:[],world:[0,1,2,3,4,5,6,7,8,9,10,11,13],worri:[0,1,4,6,9,10,11,13],wors:[0,1,4,11,13],worst:4,worth:[1,2,3,4,5,7,9,10,11,13],would:[0,1,2,3,4,5,6,7,8,9,10,11,13],wouldn:[0,1,2,4,5,6,7,9,10,11,13],wrangl:[3,5],wrap:9,wrapper:[],wrinkl:13,write:[0,1,2,3,4,5,7,9,10,11,13],writeup:[],written:[1,2,3,4,5,9,10,11,13],wrong:[0,1,2,4,5,7,8,9,10,11,13],wrongfulli:11,wrongli:1,wrote:[1,2,3,6,7,10,13],wwc:8,www:[0,4,8,10],x_100:9,x_1:[1,4,9,13],x_20:9,x_2:[1,4,9,13],x_3:[4,9],x_4:[4,9],x_5:[4,9],x_bar:4,x_estim:10,x_fill1:9,x_fill2:9,x_i:[4,10,13],x_n:4,xkcd:3,xlab:[],xlabel:[0,4,5,9,10,11,13],xlim:[9,11],xlsx:[],xmax:13,xmin:13,xponent:2,xtick:9,xval:[],xxx:[],xxy:1,xytext:11,y_fill1:9,y_fill2:9,y_fill:[],y_i:4,y_norm:13,y_t:13,ya1:[],ya2:[],yeah:[8,9],year:[0,1,3,4,5,6,8,9,10,13],yep:[],yes:[1,4,8,9,10,11,13],yesterdai:1,yet:[0,1,2,3,4,5,8,9,10,11,13],yield:11,yike:[3,4],ylab:[],ylabel:[0,4,5,9,10,11,13],ylim:[10,13],ymax:10,ymin:10,ynull:[],york:4,you:[0,1,2,3,4,5,6,7,8,9,11,13],young:[1,6],younger:6,your:[0,2,3,4,5,6,7,8,9,10,13],yourself:[0,1,2,3,4,5,7,9,13],yticklabel:10,yum:5,yval:[],yyi:8,z_i:4,z_score:13,zag:11,zassumpt:13,zealous:2,zelazni:[2,6],zeppo:13,zero:[1,3,4,5,6,7,9,13],zeroeth:3,zig:11,zip:[],zscore:4,ztest1:13,ztest2:13,ztest:13,ztesthyp:13},titles:["1. Why do we learn statistics?","2. A brief introduction to research design","1. Getting Started with Python","3. More Python Concepts","1. Descriptive statistics","2. Drawing Graphs","3. Data Wrangling","4. Basic Programming","1. Statistical theory","2. Introduction to Probability","3. Estimating unknown quantities from a sample","4. Hypothesis Testing","1. Categorical data analysis","2. Comparing Two Means","3. Comparing several means","4. Linear regression","5. Factorial ANOVA","1. Bayesian Statistics","2. Epilogue","3. References","Learning Statistics with Python"],titleterms:{"case":4,"function":[6,11],"import":3,"null":11,Adding:2,And:9,Doing:[2,13],One:13,The:[0,1,4,9,10,11,13],There:0,Useful:3,Using:[2,3],about:[1,3],absolut:4,activ:3,address:13,altern:[5,11],analysi:12,ani:10,anova:16,artifact:1,assess:1,assumpt:[8,13],attrit:1,avoid:2,axi:5,bar:5,basic:[7,9],basicprob:[],bayesian:[9,11,17],belief:0,between:11,bia:[0,1],bibliographi:[],binomi:9,bit:2,book:[2,3],box:5,boxplot:5,brief:1,calcul:[2,4,10],care:2,categor:12,categori:6,caution:5,cautionari:0,central:[4,10],check:[3,13],code:2,coeffici:4,cohen:13,cohensd:[],color:5,command:2,comment:3,common:11,compar:[13,14],complet:13,complex:[1,5],concept:3,confid:10,confound:1,consid:11,consol:2,construct:[1,13],content:3,continu:1,convert:6,correl:4,count:6,creat:6,critic:11,cross:6,csv:3,curs:0,cut:6,data:[3,4,5,6,11,12,13],datafram:3,decept:1,decis:11,decisionmak:[],defin:[1,10],demand:1,densiti:9,describ:[4,13],descript:4,design:1,deviat:[4,10,13],differ:[4,9,11,13],differenti:1,differentialattrit:[],direct:4,discret:1,distribut:[9,10,11],divid:2,document:[],doe:[3,9,10],don:10,draw:5,ecolog:1,effect:[1,11,13],effects:[],enough:5,epilogu:[4,18],error:11,errortyp:[],estim:[10,13],everydai:0,exampl:[4,13],exist:10,experiment:1,extern:1,extrem:11,face:1,factor:[],factori:16,few:6,file:[3,5],filesystem:3,find:3,fisher:11,flexibl:2,font:5,formula:10,frame:[3,4],fraud:1,frequentist:[9,11],from:[3,10,13],gener:[],get:[2,4],good:4,graph:5,graphic:5,handl:4,have:10,help:[],histogram:5,histori:1,how:[3,9,10],hypothes:11,hypothesi:[11,13],hypothesistest:[],imag:5,increas:11,independ:13,index:3,infer:13,inform:[],instal:[2,3],intern:1,interpret:[4,10],interquartil:4,interv:[1,10],introduc:[9,13],introduct:[1,5,9],issu:11,ivdv:[],know:3,kurtosi:4,label:5,larg:10,law:10,learn:[0,8,20],length:3,level:[],librari:3,life:[0,4],limit:[8,10],linear:15,list:3,littl:5,load:3,loadingcsv:[],logic:8,look:3,loop:7,make:[8,11],mathemat:6,matter:[5,10],matur:1,mean:[1,4,9,10,13,14],measur:[1,4],median:4,memori:3,menageri:11,method:0,miss:4,mistak:[10,11],mode:4,more:[0,3,5,6],most:10,move:[],much:10,multipl:5,multipli:2,myth:8,neg:13,neyman:11,nhstmess:[],nomin:1,non:[1,13],normal:[9,13],note:11,now:5,number:[2,10],numer:6,object:3,one:[11,13],onesamplettest:[],onesidedtest:[],open:5,oper:6,operationalis:1,order:2,ordin:1,other:[1,9],otherdist:[],out:3,outcom:1,overal:4,overview:5,packageinstal:[],pair:13,pairedsamplesttest:[],pairwis:4,paradox:0,paramet:10,part:[],path:3,placebo:1,plot:[5,10,13],pool:13,popul:10,posit:13,power:[2,11],practic:11,pragmat:[],predictor:1,probabl:[9,11],problem:13,probmean:[],probstat:[],program:7,proport:6,propos:11,psycholog:[0,1],pull:3,pvalu:[],python:[2,3,4,5,9,10,13,20],quantiti:10,random:10,rang:4,rank:[],ratio:1,reactiv:1,read:[],real:4,reason:8,recod:6,refer:19,region:11,regress:[1,15],relat:5,relationship:4,reliabl:1,remov:3,repeat:1,report:11,research:[0,1,11],resourc:[],respons:1,result:11,right:[2,9],road:3,role:1,run:11,same:13,sampl:[10,11,13],save:[3,5],scale:1,scatterplot:5,score:4,script:7,select:1,self:1,sever:14,shapiro:13,side:[11,13],signific:11,simpl:[2,10],simpson:0,singl:4,situat:1,size:[5,11,13],skew:4,slice:3,slight:10,softer:11,solut:11,some:[1,11],space:2,spearman:[],standard:[4,10,13],start:2,statist:[0,4,8,9,10,11,17,20],store:2,strength:4,student:13,studentassumpt:[],studentttest:[],studi:[1,11],subpopul:1,subtract:2,summari:[1,3,4,9,10,11,13],tabl:6,tabul:6,take:2,tale:0,teeni:2,tendenc:4,test:[1,11,13],teststatist:[],than:0,theorem:10,theori:[8,9],thi:[2,3],thing:3,thought:1,threat:1,three:3,tip:3,titl:5,transform:6,trap:11,trim:4,ttest:[],ttestfunct:[],ttestoneassumpt:[],two:[11,13],type:[2,3,11],typo:2,unknown:10,use:[3,4,7],useful:9,using:[2,13],valid:1,valu:[4,11,13],variabl:[1,2,3,4,6],varianc:4,veri:2,versu:[1,11],view:[9,11],violin:5,welch:13,welchttest:[],what:[3,4,9,13],which:4,who:9,why:[0,7],wilcox:13,wilcoxon:13,wilk:13,without:8,word:5,work:[9,13],workspac:3,wrangl:6,writeup:[],you:10,your:[1,11],zassumpt:[]}})
\ No newline at end of file
diff --git a/Book/_build/jupyter_execute/05.02-ttest.ipynb b/Book/_build/jupyter_execute/05.02-ttest.ipynb
index 9c6028d1..a65ce8eb 100644
--- a/Book/_build/jupyter_execute/05.02-ttest.ipynb
+++ b/Book/_build/jupyter_execute/05.02-ttest.ipynb
@@ -235,7 +235,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "hired-boulder",
+   "id": "surface-watershed",
    "metadata": {},
    "source": [
     "### Constructing the hypothesis test\n",
@@ -260,7 +260,7 @@
   {
    "cell_type": "code",
    "execution_count": 4,
-   "id": "secret-feedback",
+   "id": "tired-secondary",
    "metadata": {
     "tags": [
      "hide-input"
@@ -337,7 +337,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "instant-adult",
+   "id": "balanced-gamma",
    "metadata": {},
    "source": [
     "```{glue:figure} ztesthyp-fig\n",
@@ -354,7 +354,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "limiting-administration",
+   "id": "dedicated-portable",
    "metadata": {},
    "source": [
     "The next step is to figure out what we would be a good choice for a diagnostic test statistic; something that would help us discriminate between $H_0$ and $H_1$. Given that the hypotheses all refer to the population mean $\\mu$, you'd feel pretty confident that the sample mean $\\bar{X}$ would be a pretty useful place to start. What we could do, is look at the difference between the sample mean $\\bar{X}$ and the value that the null hypothesis predicts for the population mean. In our example, that would mean we calculate $\\bar{X} - 67.5$. More generally, if we let $\\mu_0$ refer to the value that the null hypothesis claims is our population mean, then we'd want to calculate\n",
@@ -409,7 +409,7 @@
   {
    "cell_type": "code",
    "execution_count": 5,
-   "id": "timely-place",
+   "id": "reflected-flood",
    "metadata": {
     "tags": [
      "hide-input"
@@ -493,7 +493,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "united-module",
+   "id": "interested-plate",
    "metadata": {},
    "source": [
     "```{glue:figure} ztest-fig\n",
@@ -510,7 +510,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "parallel-trustee",
+   "id": "fatal-audience",
    "metadata": {},
    "source": [
     "And what this meant, way back in the days where people did all their statistics by hand, is that someone could publish a table like this:"
@@ -518,7 +518,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "minute-stone",
+   "id": "cordless-parts",
    "metadata": {},
    "source": [
     "|                   || critical z value             |\n",
@@ -532,7 +532,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "printable-forest",
+   "id": "arctic-handy",
    "metadata": {},
    "source": [
     "which in turn meant that researchers could calculate their $z$-statistic by hand, and then look up the critical value in a text book.  That was an incredibly handy thing to be able to do back then, but it's kind of unnecessary these days, since it's trivially easy to do it with software like Python."
@@ -540,7 +540,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "elder-empire",
+   "id": "ruled-marijuana",
    "metadata": {},
    "source": [
     "### A worked example using Python\n",
@@ -551,7 +551,7 @@
   {
    "cell_type": "code",
    "execution_count": 6,
-   "id": "secondary-transsexual",
+   "id": "isolated-edinburgh",
    "metadata": {},
    "outputs": [
     {
@@ -573,7 +573,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "trained-position",
+   "id": "talented-correspondence",
    "metadata": {},
    "source": [
     "Then, I create variables corresponding to known population standard deviation ($\\sigma = 9.5$), and the value of the population mean that the null hypothesis specifies ($\\mu_0 = 67.5$):"
@@ -582,7 +582,7 @@
   {
    "cell_type": "code",
    "execution_count": 7,
-   "id": "executed-irish",
+   "id": "generous-cherry",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -592,7 +592,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "controlled-thought",
+   "id": "sticky-nancy",
    "metadata": {},
    "source": [
     "Let's also create a variable for the sample size. We could count up the number of observations ourselves, and type `N = 20` at the command prompt, but counting is tedious and repetitive. Let's get Python to do the tedious repetitive bit by using the `len()` function, which tells us how many elements there are in a vector:"
@@ -601,7 +601,7 @@
   {
    "cell_type": "code",
    "execution_count": 8,
-   "id": "lightweight-kennedy",
+   "id": "according-violence",
    "metadata": {},
    "outputs": [
     {
@@ -622,7 +622,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continuous-briefing",
+   "id": "applied-terminology",
    "metadata": {},
    "source": [
     "Next, let's calculate the (true) standard error of the mean:"
@@ -631,7 +631,7 @@
   {
    "cell_type": "code",
    "execution_count": 9,
-   "id": "banned-isaac",
+   "id": "premium-animation",
    "metadata": {},
    "outputs": [
     {
@@ -653,7 +653,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fifth-basket",
+   "id": "amended-politics",
    "metadata": {},
    "source": [
     "And finally, we calculate our $z$-score:"
@@ -662,7 +662,7 @@
   {
    "cell_type": "code",
    "execution_count": 10,
-   "id": "caroline-uzbekistan",
+   "id": "liked-faith",
    "metadata": {},
    "outputs": [
     {
@@ -683,7 +683,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "killing-fifth",
+   "id": "compound-wednesday",
    "metadata": {},
    "source": [
     "At this point, we would traditionally look up the value 2.26 in our table of critical values. Our original hypothesis was two-sided (we didn't really have any theory about whether psych students would be better or worse at statistics than other students) so our hypothesis test is two-sided (or two-tailed) also. Looking at the little table that I showed earlier, we can see that 2.26 is bigger than the critical value of 1.96 that would be required to be significant at $\\alpha = .05$, but smaller than the value of 2.58 that would be required to be significant at a level of $\\alpha = .01$. Therefore, we can conclude that we have a significant effect, which we might write up by saying something like this:\n",
@@ -696,7 +696,7 @@
   {
    "cell_type": "code",
    "execution_count": 11,
-   "id": "economic-october",
+   "id": "everyday-intermediate",
    "metadata": {},
    "outputs": [
     {
@@ -718,7 +718,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "northern-haven",
+   "id": "peripheral-extraction",
    "metadata": {},
    "source": [
     "`NormalDist().cdf()` calculates the \"cumulative density function\" for a normal distribution. Translated to something slightly less opaque, this means that  `NormalDist().cdf()` gives us the probability that a random variable X will be less than or equal to a given value. In our case, the given value for the lower tail of the distribution was our z-score, $2.259$. So `NormalDist().cdf(-z_score)` gives us the probability that a random value draw from a normal distribution would be less than or equal to $-2.259$.\n",
@@ -731,7 +731,7 @@
   {
    "cell_type": "code",
    "execution_count": 12,
-   "id": "sixth-somerset",
+   "id": "molecular-tobacco",
    "metadata": {},
    "outputs": [
     {
@@ -754,7 +754,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "static-nursery",
+   "id": "invisible-biology",
    "metadata": {},
    "source": [
     "(zassumptions)=\n",
@@ -771,7 +771,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "popular-portugal",
+   "id": "owned-coordinator",
    "metadata": {},
    "source": [
     "(onesamplettest)=\n",
@@ -783,7 +783,7 @@
   {
    "cell_type": "code",
    "execution_count": 13,
-   "id": "discrete-blank",
+   "id": "eastern-manchester",
    "metadata": {},
    "outputs": [
     {
@@ -804,7 +804,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "labeled-probability",
+   "id": "national-batman",
    "metadata": {},
    "source": [
     "In other words, while I can't say that I know that $\\sigma = 9.5$, I *can* say that $\\hat\\sigma = 9.52$. \n",
@@ -815,7 +815,7 @@
   {
    "cell_type": "code",
    "execution_count": 14,
-   "id": "earlier-transportation",
+   "id": "latest-alcohol",
    "metadata": {
     "tags": [
      "hide-input"
@@ -893,7 +893,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "uniform-scroll",
+   "id": "detected-robinson",
    "metadata": {},
    "source": [
     "\n",
@@ -913,7 +913,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "bronze-pocket",
+   "id": "personalized-surface",
    "metadata": {},
    "source": [
     "### Introducing the $t$-test\n",
@@ -930,7 +930,7 @@
   {
    "cell_type": "code",
    "execution_count": 15,
-   "id": "soviet-calendar",
+   "id": "significant-peoples",
    "metadata": {
     "tags": [
      "hide-input"
@@ -989,7 +989,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continuing-oxford",
+   "id": "cubic-glenn",
    "metadata": {},
    "source": [
     "```{glue:figure} ttestdist-fig\n",
@@ -1005,7 +1005,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "imposed-herald",
+   "id": "aging-assumption",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1018,7 +1018,7 @@
   {
    "cell_type": "code",
    "execution_count": 16,
-   "id": "dimensional-desperate",
+   "id": "eastern-grave",
    "metadata": {},
    "outputs": [
     {
@@ -1039,7 +1039,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "affected-bailey",
+   "id": "capable-inventory",
    "metadata": {},
    "source": [
     "So that seems straightforward enough. Our calculation resulted in a $t$-statistic of 2.54, and a $p$-value of 0.36. Now what do we *do* with this output? Well, since we're pretending that we actually care about my toy example, we're overjoyed to discover that the result is statistically significant (i.e. $p$ value below .05), and we will probably want to report our result. We could report the result by saying something like this:\n",
@@ -1056,7 +1056,7 @@
   {
    "cell_type": "code",
    "execution_count": 17,
-   "id": "steady-pollution",
+   "id": "conventional-technical",
    "metadata": {},
    "outputs": [
     {
@@ -1078,7 +1078,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "accepted-mozambique",
+   "id": "comic-offer",
    "metadata": {},
    "source": [
     "Now at least we have the bare minimum of what is necessary to report our results. Still, it would be sweet if we could get those confidence intervals as well. `scipy` actually has all the tools we need, and why these are not just built into the `ttest_1samp()` method is beyond me. To find the confidence interval, we need to:\n",
@@ -1094,7 +1094,7 @@
   {
    "cell_type": "code",
    "execution_count": 18,
-   "id": "regulation-biography",
+   "id": "turkish-killer",
    "metadata": {},
    "outputs": [
     {
@@ -1122,7 +1122,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "collective-solomon",
+   "id": "cardiac-database",
    "metadata": {},
    "source": [
     "Whew. Now at least we have everything we need for a full report of our results.\n",
@@ -1136,7 +1136,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "cellular-drinking",
+   "id": "present-wagon",
    "metadata": {},
    "source": [
     "(ttestoneassumptions)=\n",
@@ -1152,7 +1152,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "unlimited-facing",
+   "id": "ignored-conditions",
    "metadata": {},
    "source": [
     "(studentttest)=\n",
@@ -1163,7 +1163,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "adjacent-buying",
+   "id": "expanded-preserve",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1174,7 +1174,7 @@
   {
    "cell_type": "code",
    "execution_count": 19,
-   "id": "eight-opening",
+   "id": "tight-career",
    "metadata": {},
    "outputs": [
     {
@@ -1255,7 +1255,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "varying-liverpool",
+   "id": "verified-edgar",
    "metadata": {},
    "source": [
     "As we can see, there's a single data frame with two variables, `grade` and `tutor`. The `grade` variable is a numeric vector, containing the grades for all $N = 33$ students taking Dr Harpo's class; the `tutor` variable is a factor that indicates who each student's tutor was. The first five observations in this data set are shown above, and below is a nice little table with some summary statistics:"
@@ -1264,7 +1264,7 @@
   {
    "cell_type": "code",
    "execution_count": 20,
-   "id": "forty-florida",
+   "id": "present-purpose",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1343,7 +1343,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "acoustic-swaziland",
+   "id": "conscious-sampling",
    "metadata": {},
    "source": [
     "To give you a more detailed sense of what's going on here, I've plotted histograms showing the distribution of grades for both tutors {numref}`fig-harpohist`. Inspection of these histograms suggests that the students in Anastasia's class may be getting slightly better grades on average, though they also seem a little more variable."
@@ -1352,7 +1352,7 @@
   {
    "cell_type": "code",
    "execution_count": 21,
-   "id": "normal-jewel",
+   "id": "assumed-ancient",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1397,7 +1397,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "million-enterprise",
+   "id": "direct-laugh",
    "metadata": {},
    "source": [
     " ```{glue:figure} harpohist_fig\n",
@@ -1410,7 +1410,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "acceptable-cause",
+   "id": "color-demographic",
    "metadata": {},
    "source": [
     "{numref}`fig-ttestci` is a simpler plot showing the means and corresponding confidence intervals for both groups of students."
@@ -1419,7 +1419,7 @@
   {
    "cell_type": "code",
    "execution_count": 22,
-   "id": "danish-conjunction",
+   "id": "assured-treaty",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1428,7 +1428,7 @@
    "outputs": [
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "
"
       ]
@@ -1449,7 +1449,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "hazardous-semiconductor",
+   "id": "increased-pursuit",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttestci-fig\n",
@@ -1463,7 +1463,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "worse-triumph",
+   "id": "extreme-thunder",
    "metadata": {},
    "source": [
     "### Introducing the test\n",
@@ -1483,7 +1483,7 @@
   {
    "cell_type": "code",
    "execution_count": 23,
-   "id": "daily-poverty",
+   "id": "amazing-soviet",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1493,7 +1493,7 @@
     {
      "data": {
       "text/plain": [
-       ""
+       ""
       ]
      },
      "execution_count": 23,
@@ -1550,7 +1550,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "selective-prince",
+   "id": "improving-pittsburgh",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp_fig\n",
@@ -1563,7 +1563,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "younger-anthropology",
+   "id": "noticed-sound",
    "metadata": {},
    "source": [
     "To construct a hypothesis test that handles this scenario, we start by noting that if the null hypothesis is true, then the difference between the population means is *exactly* zero, \n",
@@ -1585,7 +1585,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "anticipated-compression",
+   "id": "forty-occasions",
    "metadata": {},
    "source": [
     "### A \"pooled estimate\" of the standard deviation\n",
@@ -1616,7 +1616,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "inclusive-andrews",
+   "id": "printable-paraguay",
    "metadata": {},
    "source": [
     "### The same pooled estimate, described differently\n",
@@ -1644,7 +1644,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "informed-syria",
+   "id": "horizontal-latino",
    "metadata": {},
    "source": [
     "(indsamplesttest_formula)=\n",
@@ -1667,7 +1667,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "binary-habitat",
+   "id": "healthy-disabled",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1680,7 +1680,7 @@
   {
    "cell_type": "code",
    "execution_count": 24,
-   "id": "tracked-harvest",
+   "id": "connected-alexander",
    "metadata": {},
    "outputs": [
     {
@@ -1706,7 +1706,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fourth-touch",
+   "id": "likely-sensitivity",
    "metadata": {},
    "source": [
     "This is fairly straightforward, and just as it was for the one-sample $t$-test, `scipy` does very little to format its results or give you any information over the bare minimum. You get a $t$-statistic and a $p$-value and that's that. Luckily, as was the case with the one-sampel $t$-test, getting the other elements we need to report our results isn't too bad. We will need the $t$-statistic, the $p$-value, the mean of each group, and the degrees of freedom. The first two we already have, and the last two are easy to get. As we discussed [above](indsamplesttest_formula), the degrees of freedom for an independent samples $t$-test is $N-2$, so..."
@@ -1715,7 +1715,7 @@
   {
    "cell_type": "code",
    "execution_count": 25,
-   "id": "dependent-circumstances",
+   "id": "structural-tragedy",
    "metadata": {},
    "outputs": [
     {
@@ -1739,7 +1739,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "exciting-growth",
+   "id": "registered-siemens",
    "metadata": {},
    "source": [
     "You probably noticed that in addition to telling `ttest_ind` which means I wanted to compare, I also added the argument `equal_var = True` to the command. This wasn't strictly necessary in this case, because by default this argument is set to `True`. But I made it explicit anyway, because we will be using this argument again later. By saying `equal_var = True`, what we're really doing is telling Python to use the *Student* independent samples $t$-test. More on this later."
@@ -1747,7 +1747,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "square-scott",
+   "id": "south-supervision",
    "metadata": {},
    "source": [
     "In any case, the difference between the two groups is significant (just barely), so we might write up the result using text like this:\n",
@@ -1757,7 +1757,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "bright-maker",
+   "id": "appointed-shipping",
    "metadata": {},
    "source": [
     " \n",
@@ -1789,7 +1789,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "damaged-peter",
+   "id": "organizational-drill",
    "metadata": {},
    "source": [
     "(studentassumptions)= \n",
@@ -1805,7 +1805,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "binding-struggle",
+   "id": "distributed-lawrence",
    "metadata": {},
    "source": [
     "(welchttest)=\n",
@@ -1839,7 +1839,7 @@
   {
    "cell_type": "code",
    "execution_count": 26,
-   "id": "sublime-hospital",
+   "id": "functioning-andrews",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1849,7 +1849,7 @@
     {
      "data": {
       "text/plain": [
-       ""
+       ""
       ]
      },
      "execution_count": 26,
@@ -1913,7 +1913,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "standing-crawford",
+   "id": "serial-offset",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp2_fig\n",
@@ -1926,7 +1926,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "intelligent-profile",
+   "id": "inappropriate-knife",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1937,7 +1937,7 @@
   {
    "cell_type": "code",
    "execution_count": 27,
-   "id": "established-bryan",
+   "id": "simple-lloyd",
    "metadata": {},
    "outputs": [
     {
@@ -1958,7 +1958,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "fixed-elder",
+   "id": "moving-cabin",
    "metadata": {},
    "source": [
     "Not too difficult, right? Not surprisingly, the output has exactly the same format as it did last time too: a test statistic $t$, and a $p$-value. So that's all pretty easy. \n",
@@ -1973,7 +1973,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "lesbian-wright",
+   "id": "ceramic-favor",
    "metadata": {},
    "source": [
     "(pairedsamplesttest)=\n",
@@ -1984,7 +1984,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "central-interim",
+   "id": "transsexual-armor",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1995,7 +1995,7 @@
   {
    "cell_type": "code",
    "execution_count": 28,
-   "id": "graphic-wellington",
+   "id": "better-depth",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2005,7 +2005,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "failing-split",
+   "id": "prepared-sword",
    "metadata": {},
    "source": [
     "The data frame `chico` contains three variables: an `id` variable that identifies each student in the class, the `grade_test1` variable that records the student grade for the first test, and the `grade_test2` variable that has the grades for the second test. Here's the first five students:"
@@ -2014,7 +2014,7 @@
   {
    "cell_type": "code",
    "execution_count": 29,
-   "id": "higher-prefix",
+   "id": "chinese-missouri",
    "metadata": {},
    "outputs": [
     {
@@ -2098,7 +2098,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "aquatic-attention",
+   "id": "divided-deadline",
    "metadata": {},
    "source": [
     "At a glance, it does seem like the class is a hard one (most grades are between 50\\% and 60\\%), but it does look like there's an improvement from the first test to the second one. If we take a quick look at the descriptive statistics"
@@ -2107,7 +2107,7 @@
   {
    "cell_type": "code",
    "execution_count": 30,
-   "id": "apparent-partition",
+   "id": "endangered-theme",
    "metadata": {},
    "outputs": [
     {
@@ -2203,7 +2203,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "professional-mounting",
+   "id": "reserved-procedure",
    "metadata": {},
    "source": [
     "we see that this impression seems to be supported. Across all 20 students[^note12] the mean grade for the first test is 57\\%, but this rises to 58\\% for the second test. Although, given that the standard deviations are 6.6\\% and 6.4\\% respectively, it's starting to feel like maybe the improvement is just illusory; maybe just random variation. This impression is reinforced when you see the means and confidence intervals plotted in {numref}`pairedta` panel A. If we were to rely on this plot alone, we'd come to the same conclusion that we got from looking at the descriptive statistics that the `describe()` method produced. Looking at how wide those confidence intervals are, we'd be tempted to think that the apparent improvement in student performance is pure chance."
@@ -2212,7 +2212,7 @@
   {
    "cell_type": "code",
    "execution_count": 31,
-   "id": "empty-backup",
+   "id": "compact-employer",
    "metadata": {
     "tags": [
      "hide-input"
@@ -2221,7 +2221,7 @@
    "outputs": [
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "
"
       ]
@@ -2265,7 +2265,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "virtual-humor",
+   "id": "falling-language",
    "metadata": {},
    "source": [
     "```{glue:figure} pairedta_fig\n",
@@ -2279,7 +2279,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continued-temperature",
+   "id": "bearing-allowance",
    "metadata": {},
    "source": [
     "Nevertheless, this impression is wrong. To see why, take a look at the scatterplot of the grades for test 1 against the grades for test 2.  shown in {numref}`fig-pairedta` panel B. \n",
@@ -2292,7 +2292,7 @@
   {
    "cell_type": "code",
    "execution_count": 32,
-   "id": "median-pharmacy",
+   "id": "earlier-director",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2301,7 +2301,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "pharmaceutical-pickup",
+   "id": "white-sweden",
    "metadata": {},
    "source": [
     "Notice that I assigned the output to a variable called `df['improvement]`. That has the effect of creating a new column called `improvement` inside the `chico` data frame. Now that we've created and stored this `improvement` variable, we can draw a histogram showing the distribution of these improvement scores, shown in {numref}`fig-pairedta` panel C. \n",
@@ -2313,7 +2313,7 @@
   {
    "cell_type": "code",
    "execution_count": 33,
-   "id": "ceramic-spice",
+   "id": "plain-massage",
    "metadata": {},
    "outputs": [
     {
@@ -2338,7 +2338,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "reasonable-slide",
+   "id": "fixed-concentrate",
    "metadata": {},
    "source": [
     "we see that it is 95\\% certain that the true (population-wide) average improvement would lie between 0.95\\% and 1.86\\%. So you can see, qualitatively, what's going on: there is a real \"within student\" improvement (everyone improves by about 1\\%), but it is very small when set against the quite large \"between student\" differences (student grades vary by about 20\\% or so). "
@@ -2346,7 +2346,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "maritime-harvest",
+   "id": "dramatic-injury",
    "metadata": {},
    "source": [
     "### What is the paired samples $t$-test?\n",
@@ -2385,7 +2385,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "annual-inspector",
+   "id": "prompt-parade",
    "metadata": {},
    "source": [
     "### Doing the test in Python \n",
@@ -2396,7 +2396,7 @@
   {
    "cell_type": "code",
    "execution_count": 34,
-   "id": "about-restoration",
+   "id": "attached-catalyst",
    "metadata": {},
    "outputs": [
     {
@@ -2417,7 +2417,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "listed-bones",
+   "id": "incident-thailand",
    "metadata": {},
    "source": [
     "However, suppose you're lazy and you don't want to go to all the effort of creating a new variable. Or perhaps you just want to keep the difference between one-sample and paired-samples tests clear in your head. In that case, `scipy` also has a built-in method for conducting paired $t$-tests called `ttest_rel` (the `_rel` part is for \"related\"). Using this method, we get:"
@@ -2426,7 +2426,7 @@
   {
    "cell_type": "code",
    "execution_count": 35,
-   "id": "english-china",
+   "id": "handled-produce",
    "metadata": {},
    "outputs": [
     {
@@ -2448,7 +2448,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "eleven-governor",
+   "id": "joined-skiing",
    "metadata": {},
    "source": [
     "Either way, the result is exactly the same, which is strangely comforting, actually.  Not only that, but the result confirms our intuition. There’s an average improvement of 1.4% from test 1 to test 2, and this is significantly different from 0 ($t$(19) = 6.48, $p$ < .001). In fact, $p$ is quite a bit less than one, since the $p$-value has been given in scientific notation. The exact $p$-value is $3.32^{-06}$, that is, $p$ = 0.0000032."
@@ -2456,7 +2456,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "rational-carolina",
+   "id": "integral-intervention",
    "metadata": {},
    "source": [
     "## One sided tests\n",
@@ -2467,7 +2467,7 @@
   {
    "cell_type": "code",
    "execution_count": 36,
-   "id": "toxic-boating",
+   "id": "elect-oasis",
    "metadata": {},
    "outputs": [
     {
@@ -2492,7 +2492,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "possible-precipitation",
+   "id": "scientific-portfolio",
    "metadata": {},
    "source": [
     "The $t$-statistics are exactly the same, which makes sense, if you think about it, because the calculation of the $t$ is based on the mean and standard deviation, and these do not change. The $p$-value, on the other hand, is lower for the one-sided test. The only thing that changes between the two tests is the _expectation_ that we bring to data. The way that the $p$-value is calculated depends on those expectations, and they are the reason for choosing one test over the other. It should go without saying, but maybe is worth saying anyway, that our reasons for choosing one test over the other should be theoretical, and not based on which test is more likely to give us the $p$-value we want!\n",
@@ -2504,7 +2504,7 @@
   {
    "cell_type": "code",
    "execution_count": 37,
-   "id": "living-auditor",
+   "id": "covered-thanks",
    "metadata": {},
    "outputs": [
     {
@@ -2538,7 +2538,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "american-chrome",
+   "id": "frank-neutral",
    "metadata": {},
    "source": [
     "What about the paired samples $t$-test? Suppose we wanted to test the hypothesis that grades go *up* from test 1 to test 2 in Dr. Chico's class, and are not prepared to consider the idea that the grades go down. Again, we can use the `alternative` argument to specify the one-sided test, and it works the same way it does for the independent samples $t$-test. Since we are comparing test 1 to test 2 by substracting one from the other, it makes a difference whether we subract test 1 from test 2, or test 2 from test 1. So, to test the hypothesis that grades for test 2 are higher than test 2, we will need to enter the grades from test 2 first; otherwise we are testing the opposite hypothesis: "
@@ -2547,7 +2547,7 @@
   {
    "cell_type": "code",
    "execution_count": 38,
-   "id": "lovely-cylinder",
+   "id": "constitutional-newfoundland",
    "metadata": {},
    "outputs": [
     {
@@ -2573,7 +2573,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sporting-wednesday",
+   "id": "downtown-float",
    "metadata": {},
    "source": [
     "(cohensd)=\n",
@@ -2592,7 +2592,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "played-church",
+   "id": "awful-annual",
    "metadata": {},
    "source": [
     "(dinterpretation)=\n",
@@ -2602,7 +2602,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "treated-quebec",
+   "id": "corrected-reconstruction",
    "metadata": {},
    "source": [
     "| d-value   | rough interpretation |\n",
@@ -2614,7 +2614,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "advised-oasis",
+   "id": "precious-scholarship",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from one sample\n",
@@ -2629,7 +2629,7 @@
   {
    "cell_type": "code",
    "execution_count": 39,
-   "id": "featured-somerset",
+   "id": "rural-ratio",
    "metadata": {},
    "outputs": [
     {
@@ -2656,7 +2656,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "square-culture",
+   "id": "cooperative-final",
    "metadata": {},
    "source": [
     "What does this effect size mean? Overall, then, the psychology students in Dr Zeppo's class are achieving grades (mean = 72.3\\%) that are about .5 standard deviations higher than the level that you'd expect (67.5\\%) if they were performing at the same level as other students. Judged against Cohen's rough guide, this is a moderate effect size."
@@ -2664,7 +2664,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "administrative-chile",
+   "id": "noted-timeline",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Student $t$ test\n",
@@ -2686,7 +2686,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "foster-correction",
+   "id": "reliable-trigger",
    "metadata": {},
    "source": [
     "However, there are other possibilities, which I'll briefly describe. Firstly, you may have reason to want to use only one of the two groups as the basis for calculating the standard deviation. This approach (often called Glass' $\\Delta$) only makes most sense when you have good reason to treat one of the two groups as a purer reflection of \"natural variation\" than the other. This can happen if, for instance, one of the two groups is a control group. Secondly, recall that in the usual calculation of the pooled standard deviation we divide by $N-2$ to correct for the bias in the sample variance; in one version of Cohen's $d$ this correction is omitted. Instead, we divide by $N$. This version makes sense primarily when you're trying to calculate the effect size in the sample; rather than estimating an effect size in the population. Finally, there is a version based on @Hedges1985, who point out there is a small bias in the usual (pooled) estimation for Cohen's $d$. Thus they introduce a small correction, by multiplying the usual value of $d$ by $(N-3)/(N-2.25)$. \n",
@@ -2697,7 +2697,7 @@
   {
    "cell_type": "code",
    "execution_count": 40,
-   "id": "western-document",
+   "id": "suffering-technical",
    "metadata": {},
    "outputs": [
     {
@@ -2748,7 +2748,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "aboriginal-calcium",
+   "id": "present-kennedy",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Welch test\n",
@@ -2778,7 +2778,7 @@
   {
    "cell_type": "code",
    "execution_count": 657,
-   "id": "advance-cleanup",
+   "id": "romantic-reality",
    "metadata": {},
    "outputs": [
     {
@@ -2813,7 +2813,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "above-yeast",
+   "id": "diagnostic-processor",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a paired-samples test\n",
@@ -2831,7 +2831,7 @@
   {
    "cell_type": "code",
    "execution_count": 663,
-   "id": "legendary-crown",
+   "id": "classical-juice",
    "metadata": {},
    "outputs": [
     {
@@ -2860,7 +2860,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "altered-example",
+   "id": "animated-brunei",
    "metadata": {},
    "source": [
     "The only wrinkle is figuring out whether this is the measure you want or not. To the extent that you care about the practical consequences of your research, you often want to measure the effect size relative to the *original* variables, not the *difference* scores (e.g., the 1\\% improvement in Dr Chico's class is pretty small when measured against the amount of between-student variation in grades), in which case you use the same versions of Cohen's $d$ that you would use for a Student or Welch test. For instance, when we do that for Dr Chico's class, "
@@ -2869,7 +2869,7 @@
   {
    "cell_type": "code",
    "execution_count": 666,
-   "id": "noble-history",
+   "id": "structured-colon",
    "metadata": {},
    "outputs": [
     {
@@ -2903,7 +2903,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "northern-appeal",
+   "id": "functioning-circus",
    "metadata": {},
    "source": [
     "what we see is that the overall effect size is quite small, when assessed on the scale of the original variables."
@@ -2911,107 +2911,259 @@
   },
   {
    "cell_type": "markdown",
-   "id": "frequent-market",
+   "id": "economic-medicine",
    "metadata": {},
    "source": [
     "(shapiro)=\n",
     "## Checking the normality of a sample\n",
     "\n",
-    "All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section \\@ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is *actually* an average of lots of different things, there's a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using $t$-tests. However, life doesn't come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there's a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.^[This is a massive oversimplification.] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test.\n",
-    "\n",
+    "All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section \\@ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is *actually* an average of lots of different things, there's a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using $t$-tests. However, life doesn't come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there's a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.[^note13] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "northern-pressing",
+   "metadata": {},
+   "source": [
     "### QQ plots\n",
     "\n",
-    "```{r qq1a, fig.cap=\"Histogram of `normal.data`, a normally distributed sample with 100 observations.\", echo=FALSE}\n",
-    "  width <- 5\n",
-    "  height <- 5\n",
-    "  \n",
-    "  plotOne <- function( data, title ) {\n",
-    "    \n",
-    "\n",
-    "    hist( x = data ,\n",
-    "          xlab = \"Value\", # x-axis label\n",
-    "          main = title,                     \n",
-    "          border=\"white\", \n",
-    "          col=ifelse(colour,emphColLight,emphGrey),\n",
-    "          font.main = 1\n",
-    "    )         \n",
-    "    \n",
-    "    require(psych)\n",
-    "    cat(title,\"\\n\")\n",
-    "    cat( \"skew=\",skew(data),\"\\n\")\n",
-    "    cat( \"kurtosis=\",kurtosi(data),\"\\n\")\n",
-    "    print( shapiro.test(data) )\n",
-    "    \n",
-    "    \n",
-    "  }\n",
-    "  \n",
-    "  normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers\n",
-    "  plotOne( normal.data, \"Normally Distributed Data\" )\n",
-    "```\n",
+    "One way to check whether a sample violates the normality assumption is to draw a **_\"quantile-quantile\" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot, using `probplot` from `scipy`. We can compare this with a histogram of the data as well:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 706,
+   "id": "seeing-consideration",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       ""
+      ]
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qq_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "filenames": {
+       "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/05.02-ttest_126_1.png"
+      },
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "import numpy as np\n",
+    "import seaborn as sns\n",
+    "from scipy.stats import probplot\n",
     "\n",
-    "```{r qq1b, fig.cap=\"Normal QQ plot of `normal.data`, a normally distributed sample with 100 observations.\", echo=FALSE}\n",
-    "  plotTwo <- function( data ) {   \n",
-    "    qqnorm( y = data,\n",
-    "            pch=19,\n",
-    "            font.main = 1\n",
-    "    )        # draw the QQ plot\n",
-    "  \n",
-    "  }\n",
+    "np.random.seed(42)\n",
+    "normal_data = np.random.normal(size=100)\n",
     "\n",
-    "  plotTwo( normal.data )\n",
-    "```\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
     "\n",
-    "The Shapiro-Wilk statistic associated with the data in Figures \\@ref(fig:qq1a) and \\@ref(fig:qq1b) is $W = .99$, indicating that no significant departures from normality were detected ($p = .73$). As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the two data sets shown in Figures \\@ref(fig:qq2a), \\@ref(fig:qq2b), \\@ref(fig:qq2c), \\@ref(fig:qq2d). Figures \\@ref(fig:qq2a) and \\@ref(fig:qq2b) show the histogram and a QQ plot for a data set that is highly skewed: the QQ plot curves upwards. Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) show the same plots for a heavy tailed (i.e., high kurtosis) data set: in this case, the QQ plot flattens in the middle and curves sharply at either end.\n",
+    "qq = probplot(normal_data, dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(normal_data, axes=ax2)\n",
     "\n",
-    "```{r qq2a, fig.cap=\"A histogram of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  skewed.data <- rgamma( n = 100,1,2 ) \n",
-    "  plotOne( skewed.data, \"Skewed Data\" )\n",
-    "```\n",
+    "glue(\"qq_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "variable-cedar",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qq_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qq\n",
     "\n",
-    "```{r qq2b, fig.cap=\"A normal QQ plot of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  plotTwo( skewed.data )\n",
-    "```\n",
+    "QQ plot (left) and histogram (right) of `normal_data`, a normally distributed sample with 100 observations. The Shapiro-Wilk statistic associated with these data is $W = .99$, indicating that no significant departures from normality were detected ($p = .73$).\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "hired-classic",
+   "metadata": {},
+   "source": [
+    "And the results are shown in {numref}(`fig-qq`), above.\n",
+    "\n",
+    "As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the data shown in {numref}(qqskew) and {numref}(qqheavy), which show the histogram and a QQ plot for a data sets that are highly skewed and have a heavy tail (i.e., high kurtosis), respectively."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 704,
+   "id": "protected-sunday",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "filenames": {
+       "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/05.02-ttest_129_0.png"
+      },
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/skewed_data.csv\")\n",
     "\n",
-    "The skewness of the data in Figures \\@ref(fig:qq2a) and \\@ref(fig:qq2b) is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
     "\n",
-    "```{r qq2c, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  heavy.tailed.data <- c(rnorm( n = 80 ) , rnorm( 20, sd=5))\n",
-    "  plotOne( heavy.tailed.data, \"Heavy-Tailed Data\" )\n",
-    "```\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
     "\n",
-    "```{r qq2d, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  plotTwo( heavy.tailed.data )\n",
-    "```\n",
+    "glue(\"qqskew_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "improving-husband",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqskew_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqskew\n",
     "\n",
-    "Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) shows the same plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
+    "The skewness of these data of 100 observations is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
     "\n",
-    "One way to check whether a sample violates the normality assumption is to draw a **_\"quantile-quantile\" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot using the R function `qqnorm()`. The `qqnorm()` function has a few arguments, but the only one we really need to care about here is `y`, a vector specifying the data whose normality we're interested in checking. Here's the R commands:\n",
-    "```{r}\n",
-    "normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers\n",
-    "hist( x = normal.data )          # draw a histogram of these numbers\n",
-    "qqnorm( y = normal.data )        # draw the QQ plot\n",
     "```"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 707,
+   "id": "understood-michigan",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       ""
+      ]
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qqheavy_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "filenames": {
+       "image/png": "/Users/ethan/Documents/GitHub/pythonbook/Book/_build/jupyter_execute/05.02-ttest_131_1.png"
+      },
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/heavy_tailed_data.csv\")\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
+    "\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
+    "\n",
+    "glue(\"qqheavy_fig\", ax, display=False)"
+   ]
+  },
   {
    "cell_type": "markdown",
-   "id": "established-system",
+   "id": "precious-sally",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqheavy_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqheavy\n",
+    "\n",
+    "Plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
+    "\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "silent-desire",
    "metadata": {},
    "source": [
     "### Shapiro-Wilk tests\n",
     "\n",
-    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.^[Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality; although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests; in R it's implemented via the `ks.test()` function.] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.[^note14] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "\n",
     "$$\n",
     "W = \\frac{ \\left( \\sum_{i = 1}^N a_i X_i \\right)^2 }{ \\sum_{i = 1}^N (X_i - \\bar{X})^2}\n",
     "$$\n",
+    "\n",
     "where $\\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. \n",
     "\n",
-    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. \n",
+    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "stuck-prison",
+   "metadata": {},
+   "source": [
+    "```{figure} ../img/ttest2/shapirowilkdist.png\n",
+    ":name: fig-shapirowilk\n",
+    ":width: 600px\n",
+    ":align: center\n",
     "\n",
-    "```{r swdist, fig.cap=\"Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\", echo=FALSE}\n",
-    "knitr::include_graphics(file.path(projecthome, \"img/ttest2/shapirowilkdist.png\"))\n",
+    "Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\n",
+    "\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "lesbian-cooperation",
+   "metadata": {},
+   "source": [
     "\n",
-    "```\n",
     "\n",
     "To run the test in R, we use the `shapiro.test()` function. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal.data`, we get the following:\n",
     "```{r}\n",
@@ -3128,7 +3280,11 @@
     "\n",
     "[^note11]: At this point we have Drs Harpo, Chico and Zeppo. No prizes for guessing who Dr Groucho is.\n",
     "\n",
-    "[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students."
+    "[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students.\n",
+    "\n",
+    "[^note13]: This is a massive oversimplification.\n",
+    "\n",
+    "[^note14]: Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality, although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests."
    ]
   }
  ],
diff --git a/Book/_build/jupyter_execute/05.02-ttest.py b/Book/_build/jupyter_execute/05.02-ttest.py
index 7e4be09a..85952f3f 100644
--- a/Book/_build/jupyter_execute/05.02-ttest.py
+++ b/Book/_build/jupyter_execute/05.02-ttest.py
@@ -1195,96 +1195,94 @@
 (shapiro)=
 ## Checking the normality of a sample
 
-All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section \@ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is *actually* an average of lots of different things, there's a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using $t$-tests. However, life doesn't come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there's a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.^[This is a massive oversimplification.] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test.
+All of the tests that we have discussed so far in this chapter have assumed that the data are normally distributed. This assumption is often quite reasonable, because the central limit theorem (Section \@ref(clt)) does tend to ensure that many real world quantities are normally distributed: any time that you suspect that your variable is *actually* an average of lots of different things, there's a pretty good chance that it will be normally distributed; or at least close enough to normal that you can get away with using $t$-tests. However, life doesn't come with guarantees; and besides, there are lots of ways in which you can end up with variables that are highly non-normal. For example, any time you think that your variable is actually the minimum of lots of different things, there's a very good chance it will end up quite skewed. In psychology, response time (RT) data is a good example of this. If you suppose that there are lots of things that could trigger a response from a human participant, then the actual response will occur the first time one of these trigger events occurs.[^note13] This means that RT data are systematically non-normal.  Okay, so if normality is assumed by all the tests, and is mostly but not always satisfied (at least approximately) by real world data, how can we check the normality of a sample? In this section I discuss two methods: QQ plots, and the Shapiro-Wilk test.
 
 ### QQ plots
 
-```{r qq1a, fig.cap="Histogram of `normal.data`, a normally distributed sample with 100 observations.", echo=FALSE}
-  width <- 5
-  height <- 5
-  
-  plotOne <- function( data, title ) {
-    
-
-    hist( x = data ,
-          xlab = "Value", # x-axis label
-          main = title,                     
-          border="white", 
-          col=ifelse(colour,emphColLight,emphGrey),
-          font.main = 1
-    )         
-    
-    require(psych)
-    cat(title,"\n")
-    cat( "skew=",skew(data),"\n")
-    cat( "kurtosis=",kurtosi(data),"\n")
-    print( shapiro.test(data) )
-    
-    
-  }
-  
-  normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers
-  plotOne( normal.data, "Normally Distributed Data" )
-```
+One way to check whether a sample violates the normality assumption is to draw a **_"quantile-quantile" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot, using `probplot` from `scipy`. We can compare this with a histogram of the data as well:
 
-```{r qq1b, fig.cap="Normal QQ plot of `normal.data`, a normally distributed sample with 100 observations.", echo=FALSE}
-  plotTwo <- function( data ) {   
-    qqnorm( y = data,
-            pch=19,
-            font.main = 1
-    )        # draw the QQ plot
-  
-  }
+import numpy as np
+import seaborn as sns
+from scipy.stats import probplot
 
-  plotTwo( normal.data )
-```
+np.random.seed(42)
+normal_data = np.random.normal(size=100)
 
-The Shapiro-Wilk statistic associated with the data in Figures \@ref(fig:qq1a) and \@ref(fig:qq1b) is $W = .99$, indicating that no significant departures from normality were detected ($p = .73$). As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the two data sets shown in Figures \@ref(fig:qq2a), \@ref(fig:qq2b), \@ref(fig:qq2c), \@ref(fig:qq2d). Figures \@ref(fig:qq2a) and \@ref(fig:qq2b) show the histogram and a QQ plot for a data set that is highly skewed: the QQ plot curves upwards. Figures \@ref(fig:qq2c) and \@ref(fig:qq2d) show the same plots for a heavy tailed (i.e., high kurtosis) data set: in this case, the QQ plot flattens in the middle and curves sharply at either end.
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
 
-```{r qq2a, fig.cap="A histogram of the 100 observations in a `skewed.data` set", echo=FALSE}
-  skewed.data <- rgamma( n = 100,1,2 ) 
-  plotOne( skewed.data, "Skewed Data" )
-```
+qq = probplot(normal_data, dist="norm", plot = ax1)
+hist = sns.histplot(normal_data, axes=ax2)
 
-```{r qq2b, fig.cap="A normal QQ plot of the 100 observations in a `skewed.data` set", echo=FALSE}
-  plotTwo( skewed.data )
-```
+glue("qq_fig", ax, display=False)
 
-The skewness of the data in Figures \@ref(fig:qq2a) and \@ref(fig:qq2b) is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). 
+ ```{glue:figure} qq_fig
+:figwidth: 600px
+:name: fig-qq
 
-```{r qq2c, fig.cap="A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. ", echo=FALSE}
-  heavy.tailed.data <- c(rnorm( n = 80 ) , rnorm( 20, sd=5))
-  plotOne( heavy.tailed.data, "Heavy-Tailed Data" )
+QQ plot (left) and histogram (right) of `normal_data`, a normally distributed sample with 100 observations. The Shapiro-Wilk statistic associated with these data is $W = .99$, indicating that no significant departures from normality were detected ($p = .73$).
 ```
 
-```{r qq2d, fig.cap="A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. ", echo=FALSE}
-  plotTwo( heavy.tailed.data )
+And the results are shown in {numref}(`fig-qq`), above.
+
+As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the data shown in {numref}(qqskew) and {numref}(qqheavy), which show the histogram and a QQ plot for a data sets that are highly skewed and have a heavy tail (i.e., high kurtosis), respectively.
+
+df = pd.read_csv("https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/skewed_data.csv")
+
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
+
+qq = probplot(df['data'], dist="norm", plot = ax1)
+hist = sns.histplot(df['data'], axes=ax2)
+
+glue("qqskew_fig", ax, display=False)
+
+ ```{glue:figure} qqskew_fig
+:figwidth: 600px
+:name: fig-qqskew
+
+The skewness of these data of 100 observations is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). 
+
 ```
 
-Figures \@ref(fig:qq2c) and \@ref(fig:qq2d) shows the same plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).
+df = pd.read_csv("https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/heavy_tailed_data.csv")
+
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))
+
+qq = probplot(df['data'], dist="norm", plot = ax1)
+hist = sns.histplot(df['data'], axes=ax2)
+
+glue("qqheavy_fig", ax, display=False)
+
+ ```{glue:figure} qqheavy_fig
+:figwidth: 600px
+:name: fig-qqheavy
+
+Plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).
 
-One way to check whether a sample violates the normality assumption is to draw a **_"quantile-quantile" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot using the R function `qqnorm()`. The `qqnorm()` function has a few arguments, but the only one we really need to care about here is `y`, a vector specifying the data whose normality we're interested in checking. Here's the R commands:
-```{r}
-normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers
-hist( x = normal.data )          # draw a histogram of these numbers
-qqnorm( y = normal.data )        # draw the QQ plot
 ```
 
 ### Shapiro-Wilk tests
 
-Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.^[Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality; although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests; in R it's implemented via the `ks.test()` function.] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by
+Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.[^note14] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by
+
 $$
 W = \frac{ \left( \sum_{i = 1}^N a_i X_i \right)^2 }{ \sum_{i = 1}^N (X_i - \bar{X})^2}
 $$
+
 where $\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. 
 
 Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look "perfectly normal". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. 
 
-```{r swdist, fig.cap="Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.", echo=FALSE}
-knitr::include_graphics(file.path(projecthome, "img/ttest2/shapirowilkdist.png"))
+```{figure} ../img/ttest2/shapirowilkdist.png
+:name: fig-shapirowilk
+:width: 600px
+:align: center
+
+Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.
 
 ```
 
+
+
 To run the test in R, we use the `shapiro.test()` function. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal.data`, we get the following:
 ```{r}
 shapiro.test( x = normal.data )
@@ -1394,4 +1392,8 @@
 
 [^note11]: At this point we have Drs Harpo, Chico and Zeppo. No prizes for guessing who Dr Groucho is.
 
-[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students.
\ No newline at end of file
+[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students.
+
+[^note13]: This is a massive oversimplification.
+
+[^note14]: Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality, although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests.
\ No newline at end of file
diff --git a/Book/_build/jupyter_execute/05.02-ttest_126_0.png b/Book/_build/jupyter_execute/05.02-ttest_126_0.png
new file mode 100644
index 00000000..e52ee6ef
Binary files /dev/null and b/Book/_build/jupyter_execute/05.02-ttest_126_0.png differ
diff --git a/Book/_build/jupyter_execute/05.02-ttest_126_1.png b/Book/_build/jupyter_execute/05.02-ttest_126_1.png
new file mode 100644
index 00000000..e52ee6ef
Binary files /dev/null and b/Book/_build/jupyter_execute/05.02-ttest_126_1.png differ
diff --git a/Book/_build/jupyter_execute/05.02-ttest_129_0.png b/Book/_build/jupyter_execute/05.02-ttest_129_0.png
new file mode 100644
index 00000000..2e981f7a
Binary files /dev/null and b/Book/_build/jupyter_execute/05.02-ttest_129_0.png differ
diff --git a/Book/_build/jupyter_execute/05.02-ttest_131_0.png b/Book/_build/jupyter_execute/05.02-ttest_131_0.png
new file mode 100644
index 00000000..f79b40c9
Binary files /dev/null and b/Book/_build/jupyter_execute/05.02-ttest_131_0.png differ
diff --git a/Book/_build/jupyter_execute/05.02-ttest_131_1.png b/Book/_build/jupyter_execute/05.02-ttest_131_1.png
new file mode 100644
index 00000000..f79b40c9
Binary files /dev/null and b/Book/_build/jupyter_execute/05.02-ttest_131_1.png differ
diff --git a/Book/_build/jupyter_execute/05.02-ttest_58_0.png b/Book/_build/jupyter_execute/05.02-ttest_58_0.png
index f2ce2894..4bdb39da 100644
Binary files a/Book/_build/jupyter_execute/05.02-ttest_58_0.png and b/Book/_build/jupyter_execute/05.02-ttest_58_0.png differ
diff --git a/Book/_build/jupyter_execute/05.02-ttest_89_0.png b/Book/_build/jupyter_execute/05.02-ttest_89_0.png
index a09d07cc..5e2eae0a 100644
Binary files a/Book/_build/jupyter_execute/05.02-ttest_89_0.png and b/Book/_build/jupyter_execute/05.02-ttest_89_0.png differ
diff --git a/Chapters/.ipynb_checkpoints/05.02-ttest-checkpoint.ipynb b/Chapters/.ipynb_checkpoints/05.02-ttest-checkpoint.ipynb
index a69e2a8c..9aca68ba 100644
--- a/Chapters/.ipynb_checkpoints/05.02-ttest-checkpoint.ipynb
+++ b/Chapters/.ipynb_checkpoints/05.02-ttest-checkpoint.ipynb
@@ -230,7 +230,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "federal-filling",
+   "id": "finnish-original",
    "metadata": {},
    "source": [
     "### Constructing the hypothesis test\n",
@@ -255,7 +255,7 @@
   {
    "cell_type": "code",
    "execution_count": 253,
-   "id": "acknowledged-mongolia",
+   "id": "hungry-visiting",
    "metadata": {
     "tags": [
      "hide-input"
@@ -327,7 +327,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "ancient-norman",
+   "id": "challenging-milan",
    "metadata": {},
    "source": [
     "```{glue:figure} ztesthyp-fig\n",
@@ -344,7 +344,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "spread-tuner",
+   "id": "critical-practitioner",
    "metadata": {},
    "source": [
     "The next step is to figure out what we would be a good choice for a diagnostic test statistic; something that would help us discriminate between $H_0$ and $H_1$. Given that the hypotheses all refer to the population mean $\\mu$, you'd feel pretty confident that the sample mean $\\bar{X}$ would be a pretty useful place to start. What we could do, is look at the difference between the sample mean $\\bar{X}$ and the value that the null hypothesis predicts for the population mean. In our example, that would mean we calculate $\\bar{X} - 67.5$. More generally, if we let $\\mu_0$ refer to the value that the null hypothesis claims is our population mean, then we'd want to calculate\n",
@@ -399,7 +399,7 @@
   {
    "cell_type": "code",
    "execution_count": 254,
-   "id": "automated-karma",
+   "id": "decreased-girlfriend",
    "metadata": {
     "tags": [
      "hide-input"
@@ -478,7 +478,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "israeli-porcelain",
+   "id": "superb-lawyer",
    "metadata": {},
    "source": [
     "```{glue:figure} ztest-fig\n",
@@ -495,7 +495,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "dated-plain",
+   "id": "incorporated-professional",
    "metadata": {},
    "source": [
     "And what this meant, way back in the days where people did all their statistics by hand, is that someone could publish a table like this:"
@@ -503,7 +503,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "center-microphone",
+   "id": "fallen-newport",
    "metadata": {},
    "source": [
     "|                   || critical z value             |\n",
@@ -517,7 +517,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "editorial-avenue",
+   "id": "asian-scanning",
    "metadata": {},
    "source": [
     "which in turn meant that researchers could calculate their $z$-statistic by hand, and then look up the critical value in a text book.  That was an incredibly handy thing to be able to do back then, but it's kind of unnecessary these days, since it's trivially easy to do it with software like Python."
@@ -525,7 +525,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sized-infrastructure",
+   "id": "derived-travel",
    "metadata": {},
    "source": [
     "### A worked example using Python\n",
@@ -536,7 +536,7 @@
   {
    "cell_type": "code",
    "execution_count": 262,
-   "id": "photographic-insight",
+   "id": "forbidden-county",
    "metadata": {},
    "outputs": [
     {
@@ -558,7 +558,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "unable-release",
+   "id": "asian-allocation",
    "metadata": {},
    "source": [
     "Then, I create variables corresponding to known population standard deviation ($\\sigma = 9.5$), and the value of the population mean that the null hypothesis specifies ($\\mu_0 = 67.5$):"
@@ -567,7 +567,7 @@
   {
    "cell_type": "code",
    "execution_count": 261,
-   "id": "magnetic-bookmark",
+   "id": "amazing-ethnic",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -577,7 +577,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "operating-pilot",
+   "id": "generous-watts",
    "metadata": {},
    "source": [
     "Let's also create a variable for the sample size. We could count up the number of observations ourselves, and type `N = 20` at the command prompt, but counting is tedious and repetitive. Let's get Python to do the tedious repetitive bit by using the `len()` function, which tells us how many elements there are in a vector:"
@@ -586,7 +586,7 @@
   {
    "cell_type": "code",
    "execution_count": 263,
-   "id": "ambient-welcome",
+   "id": "otherwise-kitty",
    "metadata": {},
    "outputs": [
     {
@@ -607,7 +607,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "dangerous-hollywood",
+   "id": "boring-footage",
    "metadata": {},
    "source": [
     "Next, let's calculate the (true) standard error of the mean:"
@@ -616,7 +616,7 @@
   {
    "cell_type": "code",
    "execution_count": 266,
-   "id": "detailed-outreach",
+   "id": "lesbian-sussex",
    "metadata": {},
    "outputs": [
     {
@@ -638,7 +638,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "personal-manchester",
+   "id": "selected-thursday",
    "metadata": {},
    "source": [
     "And finally, we calculate our $z$-score:"
@@ -647,7 +647,7 @@
   {
    "cell_type": "code",
    "execution_count": 268,
-   "id": "anticipated-medication",
+   "id": "universal-assistant",
    "metadata": {},
    "outputs": [
     {
@@ -668,7 +668,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "protective-drawing",
+   "id": "hundred-property",
    "metadata": {},
    "source": [
     "At this point, we would traditionally look up the value 2.26 in our table of critical values. Our original hypothesis was two-sided (we didn't really have any theory about whether psych students would be better or worse at statistics than other students) so our hypothesis test is two-sided (or two-tailed) also. Looking at the little table that I showed earlier, we can see that 2.26 is bigger than the critical value of 1.96 that would be required to be significant at $\\alpha = .05$, but smaller than the value of 2.58 that would be required to be significant at a level of $\\alpha = .01$. Therefore, we can conclude that we have a significant effect, which we might write up by saying something like this:\n",
@@ -681,7 +681,7 @@
   {
    "cell_type": "code",
    "execution_count": 286,
-   "id": "social-diesel",
+   "id": "circular-broadway",
    "metadata": {},
    "outputs": [
     {
@@ -703,7 +703,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "emerging-world",
+   "id": "portuguese-hebrew",
    "metadata": {},
    "source": [
     "`NormalDist().cdf()` calculates the \"cumulative density function\" for a normal distribution. Translated to something slightly less opaque, this means that  `NormalDist().cdf()` gives us the probability that a random variable X will be less than or equal to a given value. In our case, the given value for the lower tail of the distribution was our z-score, $2.259$. So `NormalDist().cdf(-z_score)` gives us the probability that a random value draw from a normal distribution would be less than or equal to $-2.259$.\n",
@@ -716,7 +716,7 @@
   {
    "cell_type": "code",
    "execution_count": 288,
-   "id": "consecutive-stationery",
+   "id": "stone-accent",
    "metadata": {},
    "outputs": [
     {
@@ -739,7 +739,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "handled-chile",
+   "id": "latter-payday",
    "metadata": {},
    "source": [
     "(zassumptions)=\n",
@@ -756,7 +756,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "practical-meaning",
+   "id": "statutory-richardson",
    "metadata": {},
    "source": [
     "(onesamplettest)=\n",
@@ -768,7 +768,7 @@
   {
    "cell_type": "code",
    "execution_count": 289,
-   "id": "unexpected-minority",
+   "id": "primary-storage",
    "metadata": {},
    "outputs": [
     {
@@ -789,7 +789,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "attractive-dubai",
+   "id": "deluxe-willow",
    "metadata": {},
    "source": [
     "In other words, while I can't say that I know that $\\sigma = 9.5$, I *can* say that $\\hat\\sigma = 9.52$. \n",
@@ -800,7 +800,7 @@
   {
    "cell_type": "code",
    "execution_count": 291,
-   "id": "chicken-robin",
+   "id": "median-safety",
    "metadata": {
     "tags": [
      "hide-input"
@@ -873,7 +873,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "bacterial-somerset",
+   "id": "proud-courage",
    "metadata": {},
    "source": [
     "\n",
@@ -893,7 +893,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "helpful-module",
+   "id": "raised-merchant",
    "metadata": {},
    "source": [
     "### Introducing the $t$-test\n",
@@ -910,7 +910,7 @@
   {
    "cell_type": "code",
    "execution_count": 325,
-   "id": "behavioral-register",
+   "id": "visible-endorsement",
    "metadata": {
     "tags": [
      "hide-input"
@@ -966,7 +966,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "matched-graph",
+   "id": "finnish-headset",
    "metadata": {},
    "source": [
     "```{glue:figure} ttestdist-fig\n",
@@ -982,7 +982,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "coordinated-reduction",
+   "id": "scheduled-resolution",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -995,7 +995,7 @@
   {
    "cell_type": "code",
    "execution_count": 359,
-   "id": "going-springer",
+   "id": "wooden-table",
    "metadata": {},
    "outputs": [
     {
@@ -1016,7 +1016,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "initial-deployment",
+   "id": "accredited-bibliography",
    "metadata": {},
    "source": [
     "So that seems straightforward enough. Our calculation resulted in a $t$-statistic of 2.54, and a $p$-value of 0.36. Now what do we *do* with this output? Well, since we're pretending that we actually care about my toy example, we're overjoyed to discover that the result is statistically significant (i.e. $p$ value below .05), and we will probably want to report our result. We could report the result by saying something like this:\n",
@@ -1033,7 +1033,7 @@
   {
    "cell_type": "code",
    "execution_count": 352,
-   "id": "dramatic-closing",
+   "id": "maritime-patch",
    "metadata": {},
    "outputs": [
     {
@@ -1055,7 +1055,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "focal-dairy",
+   "id": "announced-coffee",
    "metadata": {},
    "source": [
     "Now at least we have the bare minimum of what is necessary to report our results. Still, it would be sweet if we could get those confidence intervals as well. `scipy` actually has all the tools we need, and why these are not just built into the `ttest_1samp()` method is beyond me. To find the confidence interval, we need to:\n",
@@ -1071,7 +1071,7 @@
   {
    "cell_type": "code",
    "execution_count": 356,
-   "id": "infrared-rouge",
+   "id": "inner-sierra",
    "metadata": {},
    "outputs": [
     {
@@ -1099,7 +1099,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "cloudy-collect",
+   "id": "ranking-display",
    "metadata": {},
    "source": [
     "Whew. Now at least we have everything we need for a full report of our results.\n",
@@ -1113,7 +1113,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "republican-corruption",
+   "id": "dying-correlation",
    "metadata": {},
    "source": [
     "(ttestoneassumptions)=\n",
@@ -1129,7 +1129,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "affiliated-cable",
+   "id": "finite-hygiene",
    "metadata": {},
    "source": [
     "(studentttest)=\n",
@@ -1140,7 +1140,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "greatest-picnic",
+   "id": "sexual-national",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1151,7 +1151,7 @@
   {
    "cell_type": "code",
    "execution_count": 360,
-   "id": "received-statement",
+   "id": "active-scott",
    "metadata": {},
    "outputs": [
     {
@@ -1232,7 +1232,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "funded-confidentiality",
+   "id": "superb-teach",
    "metadata": {},
    "source": [
     "As we can see, there's a single data frame with two variables, `grade` and `tutor`. The `grade` variable is a numeric vector, containing the grades for all $N = 33$ students taking Dr Harpo's class; the `tutor` variable is a factor that indicates who each student's tutor was. The first five observations in this data set are shown above, and below is a nice little table with some summary statistics:"
@@ -1241,7 +1241,7 @@
   {
    "cell_type": "code",
    "execution_count": 437,
-   "id": "musical-excerpt",
+   "id": "favorite-shore",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1320,7 +1320,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "illegal-memorabilia",
+   "id": "indie-vector",
    "metadata": {},
    "source": [
     "To give you a more detailed sense of what's going on here, I've plotted histograms showing the distribution of grades for both tutors {numref}`fig-harpohist`. Inspection of these histograms suggests that the students in Anastasia's class may be getting slightly better grades on average, though they also seem a little more variable."
@@ -1329,7 +1329,7 @@
   {
    "cell_type": "code",
    "execution_count": 408,
-   "id": "connected-rochester",
+   "id": "overhead-router",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1371,7 +1371,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "european-disposal",
+   "id": "distributed-greece",
    "metadata": {},
    "source": [
     " ```{glue:figure} harpohist_fig\n",
@@ -1384,7 +1384,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sophisticated-growth",
+   "id": "angry-signature",
    "metadata": {},
    "source": [
     "{numref}`fig-ttestci` is a simpler plot showing the means and corresponding confidence intervals for both groups of students."
@@ -1393,7 +1393,7 @@
   {
    "cell_type": "code",
    "execution_count": 413,
-   "id": "cognitive-paragraph",
+   "id": "impressed-lesson",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1430,7 +1430,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "elegant-steering",
+   "id": "appropriate-karaoke",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttestci-fig\n",
@@ -1444,7 +1444,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "middle-median",
+   "id": "quality-clock",
    "metadata": {},
    "source": [
     "### Introducing the test\n",
@@ -1464,7 +1464,7 @@
   {
    "cell_type": "code",
    "execution_count": 436,
-   "id": "hispanic-demonstration",
+   "id": "veterinary-willow",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1528,7 +1528,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "unsigned-patch",
+   "id": "surrounded-albany",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp_fig\n",
@@ -1541,7 +1541,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "inside-ranking",
+   "id": "worse-aluminum",
    "metadata": {},
    "source": [
     "To construct a hypothesis test that handles this scenario, we start by noting that if the null hypothesis is true, then the difference between the population means is *exactly* zero, \n",
@@ -1563,7 +1563,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "immediate-ceremony",
+   "id": "psychological-influence",
    "metadata": {},
    "source": [
     "### A \"pooled estimate\" of the standard deviation\n",
@@ -1594,7 +1594,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "modified-sussex",
+   "id": "suburban-irish",
    "metadata": {},
    "source": [
     "### The same pooled estimate, described differently\n",
@@ -1622,7 +1622,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "realistic-differential",
+   "id": "available-guess",
    "metadata": {},
    "source": [
     "(indsamplesttest_formula)=\n",
@@ -1645,7 +1645,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "usual-italian",
+   "id": "impaired-darwin",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1658,7 +1658,7 @@
   {
    "cell_type": "code",
    "execution_count": 454,
-   "id": "forced-honolulu",
+   "id": "lesser-cigarette",
    "metadata": {},
    "outputs": [
     {
@@ -1684,7 +1684,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "worldwide-prior",
+   "id": "amazing-editor",
    "metadata": {},
    "source": [
     "This is fairly straightforward, and just as it was for the one-sample $t$-test, `scipy` does very little to format its results or give you any information over the bare minimum. You get a $t$-statistic and a $p$-value and that's that. Luckily, as was the case with the one-sampel $t$-test, getting the other elements we need to report our results isn't too bad. We will need the $t$-statistic, the $p$-value, the mean of each group, and the degrees of freedom. The first two we already have, and the last two are easy to get. As we discussed [above](indsamplesttest_formula), the degrees of freedom for an independent samples $t$-test is $N-2$, so..."
@@ -1693,7 +1693,7 @@
   {
    "cell_type": "code",
    "execution_count": 465,
-   "id": "surrounded-setup",
+   "id": "improved-cambridge",
    "metadata": {},
    "outputs": [
     {
@@ -1717,7 +1717,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sacred-updating",
+   "id": "exclusive-storage",
    "metadata": {},
    "source": [
     "You probably noticed that in addition to telling `ttest_ind` which means I wanted to compare, I also added the argument `equal_var = True` to the command. This wasn't strictly necessary in this case, because by default this argument is set to `True`. But I made it explicit anyway, because we will be using this argument again later. By saying `equal_var = True`, what we're really doing is telling Python to use the *Student* independent samples $t$-test. More on this later."
@@ -1725,7 +1725,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sunrise-cargo",
+   "id": "cardiac-georgia",
    "metadata": {},
    "source": [
     "In any case, the difference between the two groups is significant (just barely), so we might write up the result using text like this:\n",
@@ -1735,7 +1735,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "surface-responsibility",
+   "id": "essential-pleasure",
    "metadata": {},
    "source": [
     " \n",
@@ -1767,7 +1767,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "spiritual-ceremony",
+   "id": "bigger-success",
    "metadata": {},
    "source": [
     "(studentassumptions)= \n",
@@ -1783,7 +1783,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "pending-thousand",
+   "id": "meaningful-polyester",
    "metadata": {},
    "source": [
     "(welchttest)=\n",
@@ -1817,7 +1817,7 @@
   {
    "cell_type": "code",
    "execution_count": 530,
-   "id": "broke-chest",
+   "id": "affiliated-overview",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1888,7 +1888,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "owned-harris",
+   "id": "speaking-nickname",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp2_fig\n",
@@ -1901,7 +1901,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "qualified-tender",
+   "id": "together-salon",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1912,7 +1912,7 @@
   {
    "cell_type": "code",
    "execution_count": 531,
-   "id": "rough-exemption",
+   "id": "above-float",
    "metadata": {},
    "outputs": [
     {
@@ -1933,7 +1933,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "forty-dragon",
+   "id": "polished-governor",
    "metadata": {},
    "source": [
     "Not too difficult, right? Not surprisingly, the output has exactly the same format as it did last time too: a test statistic $t$, and a $p$-value. So that's all pretty easy. \n",
@@ -1948,7 +1948,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "powerful-technique",
+   "id": "loose-shepherd",
    "metadata": {},
    "source": [
     "(pairedsamplesttest)=\n",
@@ -1959,7 +1959,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "convinced-entertainment",
+   "id": "artificial-miami",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1970,7 +1970,7 @@
   {
    "cell_type": "code",
    "execution_count": 563,
-   "id": "absolute-walnut",
+   "id": "later-interest",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1980,7 +1980,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "diverse-consortium",
+   "id": "honey-fishing",
    "metadata": {},
    "source": [
     "The data frame `chico` contains three variables: an `id` variable that identifies each student in the class, the `grade_test1` variable that records the student grade for the first test, and the `grade_test2` variable that has the grades for the second test. Here's the first five students:"
@@ -1989,7 +1989,7 @@
   {
    "cell_type": "code",
    "execution_count": 534,
-   "id": "innocent-elimination",
+   "id": "hollywood-speed",
    "metadata": {},
    "outputs": [
     {
@@ -2073,7 +2073,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "suspended-small",
+   "id": "radical-noise",
    "metadata": {},
    "source": [
     "At a glance, it does seem like the class is a hard one (most grades are between 50\\% and 60\\%), but it does look like there's an improvement from the first test to the second one. If we take a quick look at the descriptive statistics"
@@ -2082,7 +2082,7 @@
   {
    "cell_type": "code",
    "execution_count": 535,
-   "id": "cosmetic-partition",
+   "id": "governing-eating",
    "metadata": {},
    "outputs": [
     {
@@ -2178,7 +2178,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "confidential-windows",
+   "id": "precise-eleven",
    "metadata": {},
    "source": [
     "we see that this impression seems to be supported. Across all 20 students[^note12] the mean grade for the first test is 57\\%, but this rises to 58\\% for the second test. Although, given that the standard deviations are 6.6\\% and 6.4\\% respectively, it's starting to feel like maybe the improvement is just illusory; maybe just random variation. This impression is reinforced when you see the means and confidence intervals plotted in {numref}`pairedta` panel A. If we were to rely on this plot alone, we'd come to the same conclusion that we got from looking at the descriptive statistics that the `describe()` method produced. Looking at how wide those confidence intervals are, we'd be tempted to think that the apparent improvement in student performance is pure chance."
@@ -2187,7 +2187,7 @@
   {
    "cell_type": "code",
    "execution_count": 595,
-   "id": "exclusive-bread",
+   "id": "american-despite",
    "metadata": {
     "tags": [
      "hide-input"
@@ -2237,7 +2237,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "employed-valuation",
+   "id": "comparative-implement",
    "metadata": {},
    "source": [
     "```{glue:figure} pairedta_fig\n",
@@ -2251,7 +2251,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "noted-facial",
+   "id": "laughing-potter",
    "metadata": {},
    "source": [
     "Nevertheless, this impression is wrong. To see why, take a look at the scatterplot of the grades for test 1 against the grades for test 2.  shown in {numref}`fig-pairedta` panel B. \n",
@@ -2264,7 +2264,7 @@
   {
    "cell_type": "code",
    "execution_count": 596,
-   "id": "plain-climb",
+   "id": "composite-cylinder",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2273,7 +2273,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "aggregate-perspective",
+   "id": "impossible-possibility",
    "metadata": {},
    "source": [
     "Notice that I assigned the output to a variable called `df['improvement]`. That has the effect of creating a new column called `improvement` inside the `chico` data frame. Now that we've created and stored this `improvement` variable, we can draw a histogram showing the distribution of these improvement scores, shown in {numref}`fig-pairedta` panel C. \n",
@@ -2285,7 +2285,7 @@
   {
    "cell_type": "code",
    "execution_count": 598,
-   "id": "congressional-springfield",
+   "id": "failing-profession",
    "metadata": {},
    "outputs": [
     {
@@ -2310,7 +2310,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sporting-hampshire",
+   "id": "optimum-mission",
    "metadata": {},
    "source": [
     "we see that it is 95\\% certain that the true (population-wide) average improvement would lie between 0.95\\% and 1.86\\%. So you can see, qualitatively, what's going on: there is a real \"within student\" improvement (everyone improves by about 1\\%), but it is very small when set against the quite large \"between student\" differences (student grades vary by about 20\\% or so). "
@@ -2318,7 +2318,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "particular-peoples",
+   "id": "russian-activation",
    "metadata": {},
    "source": [
     "### What is the paired samples $t$-test?\n",
@@ -2357,7 +2357,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "smoking-funds",
+   "id": "amber-harbor",
    "metadata": {},
    "source": [
     "### Doing the test in Python \n",
@@ -2368,7 +2368,7 @@
   {
    "cell_type": "code",
    "execution_count": 600,
-   "id": "lyric-topic",
+   "id": "upper-protein",
    "metadata": {},
    "outputs": [
     {
@@ -2389,7 +2389,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "scheduled-assets",
+   "id": "orange-singer",
    "metadata": {},
    "source": [
     "However, suppose you're lazy and you don't want to go to all the effort of creating a new variable. Or perhaps you just want to keep the difference between one-sample and paired-samples tests clear in your head. In that case, `scipy` also has a built-in method for conducting paired $t$-tests called `ttest_rel` (the `_rel` part is for \"related\"). Using this method, we get:"
@@ -2398,7 +2398,7 @@
   {
    "cell_type": "code",
    "execution_count": 599,
-   "id": "metropolitan-founder",
+   "id": "architectural-geneva",
    "metadata": {},
    "outputs": [
     {
@@ -2420,7 +2420,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "documented-person",
+   "id": "interior-brighton",
    "metadata": {},
    "source": [
     "Either way, the result is exactly the same, which is strangely comforting, actually.  Not only that, but the result confirms our intuition. There’s an average improvement of 1.4% from test 1 to test 2, and this is significantly different from 0 ($t$(19) = 6.48, $p$ < .001). In fact, $p$ is quite a bit less than one, since the $p$-value has been given in scientific notation. The exact $p$-value is $3.32^{-06}$, that is, $p$ = 0.0000032."
@@ -2428,7 +2428,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "unknown-characteristic",
+   "id": "legislative-contents",
    "metadata": {},
    "source": [
     "## One sided tests\n",
@@ -2439,7 +2439,7 @@
   {
    "cell_type": "code",
    "execution_count": 606,
-   "id": "median-howard",
+   "id": "ecological-engineering",
    "metadata": {},
    "outputs": [
     {
@@ -2464,7 +2464,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "copyrighted-plenty",
+   "id": "finite-newsletter",
    "metadata": {},
    "source": [
     "The $t$-statistics are exactly the same, which makes sense, if you think about it, because the calculation of the $t$ is based on the mean and standard deviation, and these do not change. The $p$-value, on the other hand, is lower for the one-sided test. The only thing that changes between the two tests is the _expectation_ that we bring to data. The way that the $p$-value is calculated depends on those expectations, and they are the reason for choosing one test over the other. It should go without saying, but maybe is worth saying anyway, that our reasons for choosing one test over the other should be theoretical, and not based on which test is more likely to give us the $p$-value we want!\n",
@@ -2476,7 +2476,7 @@
   {
    "cell_type": "code",
    "execution_count": 613,
-   "id": "emerging-choir",
+   "id": "awful-investigation",
    "metadata": {},
    "outputs": [
     {
@@ -2510,7 +2510,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "informed-majority",
+   "id": "weekly-alpha",
    "metadata": {},
    "source": [
     "What about the paired samples $t$-test? Suppose we wanted to test the hypothesis that grades go *up* from test 1 to test 2 in Dr. Chico's class, and are not prepared to consider the idea that the grades go down. Again, we can use the `alternative` argument to specify the one-sided test, and it works the same way it does for the independent samples $t$-test. Since we are comparing test 1 to test 2 by substracting one from the other, it makes a difference whether we subract test 1 from test 2, or test 2 from test 1. So, to test the hypothesis that grades for test 2 are higher than test 2, we will need to enter the grades from test 2 first; otherwise we are testing the opposite hypothesis: "
@@ -2519,7 +2519,7 @@
   {
    "cell_type": "code",
    "execution_count": 616,
-   "id": "animated-republic",
+   "id": "loved-consumption",
    "metadata": {},
    "outputs": [
     {
@@ -2545,7 +2545,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "dangerous-cameroon",
+   "id": "diverse-gibson",
    "metadata": {},
    "source": [
     "(cohensd)=\n",
@@ -2564,7 +2564,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "distributed-remainder",
+   "id": "driving-salem",
    "metadata": {},
    "source": [
     "(dinterpretation)=\n",
@@ -2574,7 +2574,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "phantom-truck",
+   "id": "helpful-style",
    "metadata": {},
    "source": [
     "| d-value   | rough interpretation |\n",
@@ -2586,7 +2586,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "labeled-lottery",
+   "id": "moved-skirt",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from one sample\n",
@@ -2601,7 +2601,7 @@
   {
    "cell_type": "code",
    "execution_count": 624,
-   "id": "altered-movie",
+   "id": "spanish-experiment",
    "metadata": {},
    "outputs": [
     {
@@ -2628,7 +2628,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "responsible-track",
+   "id": "identical-cricket",
    "metadata": {},
    "source": [
     "What does this effect size mean? Overall, then, the psychology students in Dr Zeppo's class are achieving grades (mean = 72.3\\%) that are about .5 standard deviations higher than the level that you'd expect (67.5\\%) if they were performing at the same level as other students. Judged against Cohen's rough guide, this is a moderate effect size."
@@ -2636,7 +2636,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "electronic-healing",
+   "id": "miniature-fishing",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Student $t$ test\n",
@@ -2658,7 +2658,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "located-settle",
+   "id": "valued-house",
    "metadata": {},
    "source": [
     "However, there are other possibilities, which I'll briefly describe. Firstly, you may have reason to want to use only one of the two groups as the basis for calculating the standard deviation. This approach (often called Glass' $\\Delta$) only makes most sense when you have good reason to treat one of the two groups as a purer reflection of \"natural variation\" than the other. This can happen if, for instance, one of the two groups is a control group. Secondly, recall that in the usual calculation of the pooled standard deviation we divide by $N-2$ to correct for the bias in the sample variance; in one version of Cohen's $d$ this correction is omitted. Instead, we divide by $N$. This version makes sense primarily when you're trying to calculate the effect size in the sample; rather than estimating an effect size in the population. Finally, there is a version based on @Hedges1985, who point out there is a small bias in the usual (pooled) estimation for Cohen's $d$. Thus they introduce a small correction, by multiplying the usual value of $d$ by $(N-3)/(N-2.25)$. \n",
@@ -2669,7 +2669,7 @@
   {
    "cell_type": "code",
    "execution_count": 654,
-   "id": "necessary-plasma",
+   "id": "ignored-watershed",
    "metadata": {},
    "outputs": [
     {
@@ -2719,7 +2719,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "located-journey",
+   "id": "incoming-smith",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Welch test\n",
@@ -2749,7 +2749,7 @@
   {
    "cell_type": "code",
    "execution_count": 657,
-   "id": "useful-suicide",
+   "id": "tender-microphone",
    "metadata": {},
    "outputs": [
     {
@@ -2784,7 +2784,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "weighted-chain",
+   "id": "prepared-powell",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a paired-samples test\n",
@@ -2802,7 +2802,7 @@
   {
    "cell_type": "code",
    "execution_count": 663,
-   "id": "disciplinary-physics",
+   "id": "representative-bridges",
    "metadata": {},
    "outputs": [
     {
@@ -2831,7 +2831,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "acoustic-final",
+   "id": "little-vector",
    "metadata": {},
    "source": [
     "The only wrinkle is figuring out whether this is the measure you want or not. To the extent that you care about the practical consequences of your research, you often want to measure the effect size relative to the *original* variables, not the *difference* scores (e.g., the 1\\% improvement in Dr Chico's class is pretty small when measured against the amount of between-student variation in grades), in which case you use the same versions of Cohen's $d$ that you would use for a Student or Welch test. For instance, when we do that for Dr Chico's class, "
@@ -2840,7 +2840,7 @@
   {
    "cell_type": "code",
    "execution_count": 666,
-   "id": "divine-courage",
+   "id": "disturbed-genius",
    "metadata": {},
    "outputs": [
     {
@@ -2874,7 +2874,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "approximate-prisoner",
+   "id": "chemical-rolling",
    "metadata": {},
    "source": [
     "what we see is that the overall effect size is quite small, when assessed on the scale of the original variables."
@@ -2882,7 +2882,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "entitled-imaging",
+   "id": "ultimate-coordination",
    "metadata": {},
    "source": [
     "(shapiro)=\n",
@@ -2893,7 +2893,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "prostate-proxy",
+   "id": "medieval-tiffany",
    "metadata": {},
    "source": [
     "### QQ plots\n",
@@ -2903,14 +2903,26 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 687,
-   "id": "personal-explanation",
+   "execution_count": 706,
+   "id": "paperback-cancellation",
    "metadata": {
     "tags": [
      "hide-input"
     ]
    },
    "outputs": [
+    {
+     "data": {
+      "application/papermill.record/text/plain": ""
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qq_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
     {
      "data": {
       "image/png": "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\n",
@@ -2936,12 +2948,13 @@
     "\n",
     "qq = probplot(normal_data, dist=\"norm\", plot = ax1)\n",
     "hist = sns.histplot(normal_data, axes=ax2)\n",
-    "\n"
+    "\n",
+    "glue(\"qq_fig\", ax, display=False)"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "improved-consequence",
+   "id": "listed-barrier",
    "metadata": {},
    "source": [
     " ```{glue:figure} qq_fig\n",
@@ -2954,69 +2967,199 @@
   },
   {
    "cell_type": "markdown",
-   "id": "regulation-telling",
+   "id": "latest-spectrum",
    "metadata": {},
    "source": [
     "And the results are shown in {numref}(`fig-qq`), above.\n",
     "\n",
-    "As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the two data sets shown in {numref}(qqskew), which show the histogram and a QQ plot for a data set that is highly skewed: the QQ plot curves upwards. Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) show the same plots for a heavy tailed (i.e., high kurtosis) data set: in this case, the QQ plot flattens in the middle and curves sharply at either end.\n",
+    "As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the data shown in {numref}(qqskew) and {numref}(qqheavy), which show the histogram and a QQ plot for a data sets that are highly skewed and have a heavy tail (i.e., high kurtosis), respectively."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 704,
+   "id": "prepared-chaos",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/skewed_data.csv\")\n",
     "\n",
-    "```{r qq2a, fig.cap=\"A histogram of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  skewed.data <- rgamma( n = 100,1,2 ) \n",
-    "  plotOne( skewed.data, \"Skewed Data\" )\n",
-    "```\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
     "\n",
-    "```{r qq2b, fig.cap=\"A normal QQ plot of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  plotTwo( skewed.data )\n",
-    "```\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
     "\n",
-    "The skewness of the data in Figures \\@ref(fig:qq2a) and \\@ref(fig:qq2b) is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
+    "glue(\"qqskew_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "suspended-disorder",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqskew_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqskew\n",
     "\n",
-    "```{r qq2c, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  heavy.tailed.data <- c(rnorm( n = 80 ) , rnorm( 20, sd=5))\n",
-    "  plotOne( heavy.tailed.data, \"Heavy-Tailed Data\" )\n",
-    "```\n",
+    "The skewness of these data of 100 observations is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
     "\n",
-    "```{r qq2d, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  plotTwo( heavy.tailed.data )\n",
-    "```\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 707,
+   "id": "miniature-christmas",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "application/papermill.record/text/plain": ""
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qqheavy_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3cAAAFNCAYAAACjTZb0AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAABADUlEQVR4nO3deZhcdZm38fvJ4oQIIkJgNKQTdieiIgRfBEEFFHQQWUYBA4PAGGRAlhlASNgChEF2VLYWlSWtgIDKJoiIQUeMLEOCSEDEdECCQBwNGsYQ+nn/ONXQ6XR3qtPVfbqr7s919VV1Tp1z6qki5PQ3vy0yE0mSJEnS0Das7AIkSZIkSX1nuJMkSZKkOmC4kyRJkqQ6YLiTJEmSpDpguJMkSZKkOmC4kyRJkqQ6YLiT+iAiMiI2XsVz50fEzt28tn1EPNHVsRExNSKuXLWKe1XfhyPi2f5+H0mSJNWG4U4NpxKUXomIv0bEHyPiWxGxetl1dZSZP8vMzbp57azM/DeAiJhQCZgjVuV9IuJzEfFa5btYHBGPRMRuq3CdqyLizFWpQZIkSbVhuFOj+mRmrg5sCWwNnNT5gFUNTEPQ/ZXv4q3AN4AbIuJt5ZYkSZKk3jLcqaFl5h+AHwKbw+vdLA+PiN8Cv63s+3xEPBURf4qIWyLiHZ0u84mIeDoiXoqIcyNiWOW8jSLiJxGxqPJaS0S8tdO5W0fEbyLifystiKMq53bbJTIiTouImZXN+yqPf660vn2oUue7Oxy/bqWlcsxKvos24JvAasCGXbzvP0XETyPizxHxWETsXtk/BZgMHF+p4dae3keSJEn9w3CnhhYR44BPAP/TYfcewP8DJkbEjsB/AZ8B3g60Atd1usyewCSKVsBPAQe3X75y7juAfwLGAad1OncysAuwEbApXbQgrsQOlce3ZubqmTmrUt/+HY7ZD/hxZr7Y04UqLZX/BvyVSrDt8NpI4FbgR8C6wBeBlojYLDObgRbgnEoNn+zlZ5AkSVINGO7UqL4fEX8Gfg7MAs7q8Np/ZeafMvMVivD1zcx8ODP/DpwIfCAiJnQ4/suV4xcAF1GEKTLzqcy8OzP/XglWFwAf6lTH1zLzmcz8EzCj/dw+uhr4bHsLInAAcG0Px29T+S6er7z/npn5l87HAKsDZ2fm0sz8CXBbjeqVJElSDTTKmCKpsz0y88fdvPZMh+fvAB5u38jMv0bEImAsML+L41sr5xAR6wJfAbYH1qD4x5T/7eG9Xj+3LzJzdkT8DfhQRCwENgZu6eGUX2bmB1dy2XcAz1S6brZrpfgeJEmSNAjYcietKDs8fw4Y374REW8G1gb+0OGYcR2eN1XOgaJLZgLvycy3UHSVjE7v1d25q1JrR1dX3u8A4MbM/L9eXrez54BxHVoDoai3/Xvorg5JkiQNEMOd1LNvAwdFxBYR8Q8U3TdnZ+b8DsccFxFrVcbvHQVcX9m/BsX4tT9HxFjguC6uf3hErF+ZnXJqh3Or9SLQxooToFxLMRZwf+CaXl6zK7OBv1FMmjIyIj4MfJI3xh/+sYsaJEmSNIAMd1IPMvMe4GTgJmAhxcQn+3Y67AfAQ8AjwO0UywkATKeYZOUvlf03d/EW36aYpOTpyk+v1orLzCUUY/X+uzKL5TaV/c9SdCdN4Ge9uWY377MU2B34OPAScCnwr5k5r3LINygmoPlzRHy/r+8nSZKk3otMe1NJ9Sgivgk8l5m9nYFTkiRJQ5ATqkh1qDKb517A+0ouRZIkSQPEbplSnYmIM4BfA+dm5u/LrkeSJEkDw26ZkiRJklQHbLmTJEmSpDpguJMkSZKkOjCkJlRZZ511csKECWWXIUnqZw899NBLmTmm7DqGCu+PktQ4erpHDqlwN2HCBB588MGyy5Ak9bOIaC27hqHE+6MkNY6e7pF2y5QkSZKkOmC4kyRJkqQ6YLiTJEmSpDpguJMkSZKkOmC4kyRJkqQ6YLiTJEmSpDpguJMkSZKkOmC4kyRJkqQ6YLiTJNVESwtMmADDhhWPLS1lVyRJUmMx3ElSg+jP8NXSAlOmwJqtc9gnv0Nra7FtwJMkaeAY7iSpAbSHr9ZWyKTm4esbxz/BN5bsyxy24ByOZwSvsmQJTJtWm+tLkqSVM9xJUgOYNg2WLFl+X03CV2srHHwwdz83kd24jRlM5T3MZRkjAViwoI/Xl6RBYuy4JiKizz9jxzWV/VFUx0aUXYAkqf91F7JWOXwtXAgzZkBzMwwbxlVrHMmJL5/Ii6y73GFN/g4jqU489+wz7HPFL/p8nesP3bYG1Uhds+VOkhpAdyGr1+Fr0SI4/njYaCO44go46CB46ilGXXYhfxu9fLAbPbrIf5IkaWAY7iSpAcyYUYStjnoVvhYvhunTYYMN4LzzYO+9Yd68IuCtvz6TJxeNeOPHQ0Tx2NwMkyfX/KNIkqRu2C1TkhpAe8iaNq3oitnUVAS7lYavJUvgkkvgy18uWu322gtOPx3e9a4u38MwJ0lSeUoPdxExHHgQ+ENm7lZ2PZJUr3oVvpYuha9/Hc48E55/HnbZpXg+aVK/1ihJklbdYOiWeRTweNlFSJKAZcvgW9+CTTeFI46ATTaB++6DO+802EmSNMiVGu4iYn3gn4Ery6xDkoaami9I3tYG118Pm28OBx8M66xTBLpZs2D77WtQsSRJ6m9lt9xdBBwPtJVchyQNGTVdkDwTbrsNttwS9t0XRoyAm2+GBx4oumJG1Lx+SZLUP0oLdxGxG/BCZj60kuOmRMSDEfHgiy++OEDVSdLgVbMFyX/yE9h2W/jkJ+Gvf4WZM2HOHNhzT0OdJElDUJktd9sBu0fEfOA6YMeImNn5oMxszsxJmTlpzJgxA12jJA06fV6Q/Je/hJ12Kn6efbZYs+Dxx4vZVoYPr1mdkiRpYJUW7jLzxMxcPzMnAPsCP8nM/cuqR5KGilVekHzOnKKV7gMfgEcfhQsvhN/+Fj7/eRg5suZ1SpKkgVX2mDtJUi/1ekHyJ54oxtNtsQX8/OfFgU8/DUcfDaNG9XO1kiRpoAyKcJeZP3WNO0mqzuTJRU/K8eOLoXHjxxfbK6xh19pazHw5cWIxacq0afD738PUqbD66qXULkmS+s+gCHeSpEK1SxxMngzz5xcrGMyf3ynYLVz4xhp13/42HHVU0VJ35pnw1rf290eQJEklGVF2AZKkQvsSB+0zYbYvcQBdtMp1ZdEi+PKX4Wtfg1dfLVrtTj4Z1l+/32qWJEmDhy13kjRIrPISB4sXw/TpsMEGcN55sPfeMG8eXHGFwU6SpAZiy50kDRK9XuJgyRK45JKitW7RomJ9utNPh80377caJUnS4GXLnSQNElUvcbB0KVx6KWy8MRx/PEyaBA88ADffbLCTJKmBGe4kaZBY6RIHy5bBt74Fm24Khx9ehLtZs+DOO4uAJ0mSGprhTpIGiW6XONivDW64oWiVO/hgWGedItDNmgU77FB22VpFETEuIu6NiMcj4rGIOKqy/20RcXdE/LbyuFbZtUqShgbDnSQNIsstcfD7ZPKat8GWW8I++8CIEUXXywcegF12KRKghrJlwH9m5j8B2wCHR8RE4ATgnszcBLinsi1J0koZ7iRpMPrJT2C77eCTn4SXX4aZM2HOnGLSFENdXcjMhZn5cOX5y8DjwFjgU8DVlcOuBvYopUBJ0pBjuJOkweSXv4Sddip+FiwoljOYN69o0hs+vOzq1E8iYgLwPmA2sF5mLoQiAALrdnPOlIh4MCIefPHFFwesVknS4GW4k6TBYM4c2H13+MAH4NFH4cIL4amnilXMR44suzr1o4hYHbgJODozF1d7XmY2Z+akzJw0ZsyY/itQkjRkGO4kqUxPPAH77gtbbAE/+1kxNebTT8PRR8OoUWVXp34WESMpgl1LZt5c2f3HiHh75fW3Ay+UVZ8kaWgx3ElSGVpbi5kvJ06E226DqVOLUDd1Kqy+etnVaQBERADfAB7PzAs6vHQLcGDl+YHADwa6NknS0DSi7AIkqaEsXAhnnVWMpRs2DI48Ek48EdbtcliV6tt2wAHAoxHxSGXfVOBs4IaIOARYAHy6nPIkSUON4U6SBsKiRfDlL8PXvgavvlq02p10EowbV3ZlKklm/hzoburTnQayFklSfTDcSVJ/Wry4mBzl/PPhr38tZr087TTYaKOyK5MkSXXGMXeS1B+WLIFzz4UNNyzC3Ec/Co8+Ssuu1zJhp40YNgwmTICWlrILlSRJ9cJwJ0m1tHQpXHJJ0TJ3/PEwaRI88ADcdBMtj7yLKVOKuVQyi8cpUwx4kiSpNgx3klQLy5bBt74Fm24KRxwBm2wC990Hd95ZBDxg2rSiQa+jJUuK/ZIkSX1luJOkvmhrgxtugM03LyZJWWedItDNmgXbb7/coQsWdH2J7vZLkiT1huFOklZFJtx+O2y1FeyzDwwfDjfdBA88QMtLuzBhg1hhXF1TU9eX6m6/JElSbxjuJKm37r0XttsOdtutmA3z2mth7lzYay9avh3djqubMQNGj17+UqNHF/slSQ1i2Agios8/Y8f5L4NakUshSFK1Zs8uBsjdcw+MHVssRH7QQTBy5OuH9DSubv78N45ZsKBosZsxo1gdQZLUINqWsc8Vv+jzZa4/dNsaFKN6Y8udJK3M3Lmw++6wzTbF8wsvhKeegilTaLlhJBMm8HoXzNbWri/RPq5u8uQi5LW1FY8GO0mSVCuGO0nqzhNPwL77wnvfCz/7GY98egYTRz3NsP84mgnvHMW//zsrdMGM6PpSjquTJEn9zW6ZktRZaytMnw5XXw2rrQZTp/Ld8cfyuWPWer3LZWsrXH55Eeo6yiwCXsf9jquTJEkDwXAnSe0WLixSWHNz0c/yqKPghBNg3XU5bsKKY+k6B7uO+8ePd1ydJEkaWIY7SVq0CL78Zfja1+DVV4v16k46CcaNe/2Q3qxFN378G5OnSJIkDRTDnaTGtXhxMTnK+efDX/9aNK+ddhpstNEKhzY1dT1Zil0wJUnSYFHahCoRMS4i7o2IxyPisYg4qqxaJDWYJUvg3HNhww2LMLfzzsUsmNde22Wwg+7XqPvCF4qWuojisbnZLpiSJKkcZbbcLQP+MzMfjog1gIci4u7M/E2JNUmqZ0uXwte/XiS1hQthl13gjDNg661Xemp7YHONOkmSNFiV1nKXmQsz8+HK85eBx4GxZdUjqY4tWwZXXQWbbQZHHAEbbwyzZsGdd3Yb7FpaWG79upYW16iTJEmD26BY5y4iJgDvA2aXXIqketLWBjfcAJtvDgcdBGuvDT/8YRHsdtih29NaWlZcv27KlGK/JEnSYFV6uIuI1YGbgKMzc3EXr0+JiAcj4sEXX3xx4AuUNPRkwm23wZZbwj77wPDhcNNN8MADsOuu3a80XjFt2orLHixZUuyXJGlQGDaCiKjJz9hxTWV/GtVIqbNlRsRIimDXkpk3d3VMZjYDzQCTJk3qZlUpSaq4994ihd1/fzFhyrXXwn77FQGvSt0te9Cb5RAkSepXbcvY54pf1ORS1x+6bU2uo/KVOVtmAN8AHs/MC8qqQ1KdmD27mPVyxx2LFHbFFTBvHuy/f1XBruMYu2Hd/M3Y5D9sSpKkQazMbpnbAQcAO0bEI5WfT5RYj6ShaM4c2H132GabYjmDCy+Ep54qBsmNHFnVJTqPsXvttRWPcf06SZI02JXWLTMzfw70PPBFkrrz5JNwyilw/fXw1rcWyevII2H11as6vaXljWUNhg3rOtANH17MyeKyB5IkaSgodcydJPVaaytMnw5XXw2rrQZTp8Kxx8Jaa/V4Wscw97a3wcsvF8veQdfBDopg19ZW4/olSZL6ieFO0tCwcGHRfNbcXDS1HXkknHgirLvuSk9t73bZPgPmokXVvaVj7CRJ0lBiuJM0uC1aBOecA1/9atHUdsghcNJJMG5c1ZfoammDlXGMnSRJGmpKX+dOkrq0eDGcfnqxnMG558JeexWzX15xRdXBrn0GzNbW6t5y+PBiCbzx44sGQsfYSZKkocSWO0mDyyuvwCWXwNlnF612e+wBZ5wBm2/e5eGdx9IB/OlPK46rW5nRow10kiRpaDPcSRocli6FK6+EM88sxtftskvxfNKkbk/paSzdysbVjRwJb3lLEQSdDVOSJNUDw52kci1bBjNnFjNgzp8P229fLG+w/fYrPXVVxtJB0e3SMCdJkuqN4U5SOdra4MYb4dRTi7F0W20Fl18OH/tYMfCtCgsW9P5tx48vMqQkSVK9cUIVSQMrE267rQhz++xTLGtw003wwANFV8yI1ydCGTYM1lmn+ImAESOKx/Z9mb17a2fAlCRJ9cyWO0kD5957i76U999fzIJ57bWw337FNJUVPY2ja19svNp16hxXJ0mSGoktd5L63y9/CTvvDDvuCM88UyxnMG8e7L//68GuvbVu//1XbRwdwNprFz/tyxl861vw0ktFD9D58w12kiSpvtlyJ6n/zJ1bLDh+660wZgxceCF84QswatTrh7S0wFFHVd8a152IIshJkiQ1KsOdpNp74oliopTrr4c11yyWNDjqKFh99eUO69wFsy+amvp+DUmSpKHMcCepdlpbiyUNrr4aVlsNpk6FY4+Ftdbq8vBVXcqgMydKkSRJMtxJqoXnny/S1RVXFFNcHnkknHgirLtuj6dVs5TB2msXj4sWFcPzXnvtjX1OlCJJkvQGw52kVbdoEZxzDnz1q7B0KRx8MJx8MowbR0tL0TK3YAG87W1vHN4e0IYP73kpg9GjobnZ0CZJklQtw52k3lu8uJgc5YILyJdf5vujP8txr53G/G9uzGtfLyY36RjculrOoP2xK2uvDRdfbLCTJEnqDcOdpOq98gpccgmcfTYsWsRtb9qTE/J0Hvvb5sXrlcDW28XF240fbxdLSZKkVWW4k7RyS5fClVcWs14uXMhz796Fz/z1TP7775Nq9hYRxVp0kiRJWjUuYi6pe8uWwVVXwWabweGHw8Ybw333se3iO2sa7MClDCRJkvrKcCdpRW1tcMMN8O53w0EHMfe5tdmFOxnx37OIHbantbW2b+dSBpIkSX1nuJMEFAuKTxif/HPczv8M3wr22YdfzxvOXtzEe5c+wI/YhdfaYpWu/eY3v7F8wfDhyz+OH++smJIkSbXgmDupQbUvVdDaWox3+1Dey7eZxrbcz+/YkP25lu+wH20Mr/qaw4YVjX7tyx04QYokSdLAMdxJDaalBY466o3lCd7PbGbkNHbmHp5lLIdyOd/kYJYxslfXnTnTECdJklQmu2VKDaSlBaZMKYLdu5nLD9id2WzDe5jLMVzAxjxFM4f2OtiNH2+wkyRJKpvhTmoALS0wYQLsvz+sv+QJvs1+zOW97MB9TONMNuRpLuIY/s6oXl/byVAkSZIGB8OdVMdaWmCddYpQl62tXMkh/IaJfJJbOYsT2YDfcxbT+BurV3W9YZW/MZwMRZIkafAx3El1qGOoG7Hoeb7CF/ktmzCZFr7CkWzI00zjLP7MWl2e31WImzmzmCQls1j+LrNYdNxgJ626iPhmRLwQEb/usO+0iPhDRDxS+flEmTVKkoYOw51UR5ZrqVu0iLP5Ek+zIV/gcr7FQWzCb/kPLuRF1l3uvPYwZ4iTBtxVwK5d7L8wM7eo/NwxwDVJkoaoUmfLjIhdgYuB4cCVmXl2mfVIQ1HHJQ0A1mAxp3Ah/8EFrMHLfJvPcirTeZqNujx/7bXhpZcGsGBJr8vM+yJiQtl1SJLqQ2ktdxExHLgE+DgwEdgvIiaWVY80VLRPjhJRtLjtv38R7FZjCcdyLk+zIdM5jXvYifcwlwOY2W2wGz0aLr54YOuXVJUjImJupdtml/2nI2JKRDwYEQ+++OKLA12fJGkQKrNb5vuBpzLz6cxcClwHfKrEeqRBq2Ogaw9zUHSZHMlS/p1LeIqNOZfjeZBJTOIB9uZmHmPzFa7VsQumk6FIg9JlwEbAFsBC4PyuDsrM5syclJmTxowZM4DlSZIGqzLD3VjgmQ7bz1b2SQ2vu9a5joazjAO5iifYjEs4gqfYmB2Yxce5k4eYtMI11157+fF0jqOTBqfM/GNmvpaZbcDXKf4xVJKklepVuIuIYRHxlhq9d3SxL7t4T7udqCH01DrXUdDGv/BdHuXdXMVB/Im3sQt38iFm8TN2WOG67aHupZcMc9JQEBFv77C5J/Dr7o6VJKmjlYa7iPh2RLwlIt4M/AZ4IiKOq8F7PwuM67C9PvBc54PsdqJ60zHEjRjRc+vc8pJPcDsPsRXf5TO0MYy9uIlJPMiP2IXO/14SAYcdZqiTBrOI+A5wP7BZRDwbEYcA50TEoxExF/gIcEypRUqShoxqZsucmJmLI2IycAfwJeAh4Nw+vvcDwCYRsQHwB2Bf4LN9vKY0qLW0wJQpsGRJsf3aa8Vj59a5zj7MvcxgGttyP79jQ/bnWr7DfrQxfLnjhg2DtrZiPN2MGYY6abDLzP262P2NAS9EklQXqumWOTIiRgJ7AD/IzFfpovtkb2XmMuAI4C7gceCGzHysr9eVBrNp094IdtV4P7O5m525lx1pYgGHcjnvZB4t7L9csOu8Pp3j6SRJkhpPNS13VwDzgTnAfRExHlhcizevLMzq4qxqGAsWVHfce5jDGZzM7tzKC4zhGC7gMg7j74wqZru0dU6SJEmdrLTlLjO/kpljM/MTWWilGAMgaSU6j69bWffLTXiSb7Mfc9iCHbiPkziDDXmaiziGfxw/ytY5SZIkdauaCVXWi4hvRMQPK9sTgQP7vTJpiGsfX9c+SUr7+LquNNHKlRzCb5jIJ7mVs5jKBvyemeNP4oqZqxvmJEmStFLVjLm7imJc3Dsq208CR/dTPdKQ0dWsl51nv1zZ+Lr1eJ6LOZIn2ZT9mck1axzBXZf8jqk5g//NtQx0kiRJqlo1Y+7WycwbIuJEKCZCiYge2iCk+tTSUkyI0tpaBLiOXSzbW+Wqnf1yLf7E8ZzDkXyFN7GUEZ8/GE4+mYPHjev5REmSJKkb1YS7v0XE2lRmyIyIbYC/9GtV0iDS0gJHHQWLFr2xb2XhrTtrsJijuYj/5HzW4GW+zWf5+jtOY1bzxrUpVpIkSQ2rmnD3H8AtwEYR8d/AGOBf+rUqqUQ9tdCtqlG8wuFcwgmczTos4mb25BRO5/ejN6f5nL5fX5IkSVppuMvMhyPiQ8BmQABPVNa6k+pKLVvo2o1kKf/GlZzEmbyDhdzFxzhl2Ax+1TaJ8eOh2aUMJEmSVCMrDXcR8a+ddm0ZEWTmNf1UkzSgugp1fTWcZezPTE5lOhswn5/HB3ls2nXscsYO7FK7t5EkSZJeV81smVt3+NkeOA3YvR9rkmqqmlkt+xrshlX+TxoxrI1PcwOP8m6u4iAWsTafWeOHtF5zHx89Y4c+fxZJkiSpO9V0y/xix+2IWBO4tt8qkvqolrNadmfYMGhrg/HjYcYMmPzZhDvugJNOgkcegYkT4YybmLTnntwQ0afPI0mSJFWjmpa7zpYAm9S6EKkvOrbOHXDAGwuH12IylI7WXhtmzizC4esLi7/jXthuO9htN1i8GK69FubOhb32KgqSJEmSBkA1Y+5upbIMAkUYnAjc0J9FSdXqj0lQOlqhha7j5CezZxdNhPfcA2PHwuWXw8EHw8iRtStAkiRJqlI1SyGc1+H5MqA1M5/tp3qkqrW0wJQpsGRJ7a+99tpw8cXdzGQ5Zw6ccgrccguMGQMXXACHHQajRtW+EEmSJKlK1Yy5mzUQhUi9NW1a7YNdj6HuiSfg1FPh+uthzTXhjDPg6KNh9dVrW4QkSZK0CroNdxHxMm90x1zuJSAz8y39VpVUhQULen9OezfL4cOLcXPtj112u2zX2gqnnw5XX120zk2dCsceC2ut1efPIEmSJNVKt+EuM9cYyEKk3mhpKYJa+6yXnbXPkllVeOvO888XJzU3Fxc84ogi2K27bk0+gyRJklRL1Yy5AyAi1gVeH1SUmavQbiL1XselDdrDWuclDjrqsWtlNRYtgnPPha98BZYuLSZJOflkGDdulT+DJEmS1N+qmS1zd+B84B3AC8B44HHgXf1bmhpNe4hbsADe9rZi36JFywe5ntanGz686Dm5yqFu8WK46CI4/3x4+WXYbz+YPh023ngVLyhJkiQNnGpa7s4AtgF+nJnvi4iPAPv1b1lqFN0tOL4qSxu0ta1isHvlFbjkEjj77OKN99yzGGO3+earcDFJkiSpHNUsYv5qZi4ChkXEsMy8F9iif8tSI2hfyqBWC443NfXyhKVL4dJLYaON4LjjYKut4Fe/gptvNthJkiRpyKkm3P05IlYH7gNaIuJiivXupFXS0gITJsD++9duKYPRo4u5T6ry2mtF/83NNoPDDy/C3axZcNddsPXWtSlIUkOJiO2q2SdJUn/qNtxFxL9ExCjgU8AS4BjgTuB3wCcHpjzVi/ZAFwEHHPBGa11fRBSP48cXE1qutEtmWxvccEPRKve5zxUzr9x5J9x3H+ywQ98LktTIvlrlPkmS+k1PY+4mA5dSBLrvAD/KzKsHpCrVlfbul+2tdKvS/bJPSxtkwh13wEknwSOPwMSJcNNNxdi69oQoSasgIj4AbAuMiYj/6PDSW4Dh5VQlSWpU3bbcZeaewMbAPcCRwDMRcVlE2MShXpk2rfrul+1Za+21i5+IIshde22R0ZYtKx7nz68y2N17L2y3Hey2WzEb5rXXwty5sNdeBjtJtfAmYHWKfyxdo8PPYuBfSqxLktSAepwtMzMXA1cDV0fE2hQ3qq9GxNsy00W/1KOOM2FWY5UWGu/O7NnFm99zD4wdC1dcAQcdBCNH1uDiklTIzFnArIi4KjNr0OFckqRVV9Ui5hGxFrAXsA/wNuCm/ixKQ1/nrpg9GT26yjFz1Zg7t+h+eeutMGYMXHABHHYYjBpVg4tLUrf+ISKagQl0uLdm5o6lVSRJajjdhruIWAPYg2JNuy2BW4AzgXsz+zppverdyrpito+hq1lr3ZNPwqmnwvXXw1veAmeeCUcdBauv3scLS1JVvgtcDlwJvFZyLZKkBtVTy93vgbuAy4A7M/PVgSlJ9WDBgu5fq2n3y9bWYsHxq68uWudOPBGOPRbWWqsGF5ekqi3LzMvKLkKS1Nh6CndNmVmjVcjUaJqauh5rN358MRlKnz3/PJx1VjGWLgKOOAKmToV1163BxSWp126NiH8Hvgf8vX1nZv6pvJIkSY2m23DXn8EuIs6lWCtvKcW6eQdl5p/76/008GbMWHHMXa8WGu/On/4E55wDX/kKLF0KhxxSjLEb5/w+kkp1YOXxuA77EtiwhFokSQ2q26UQ+tndwOaZ+R7gSeDEkupQDbUvVD5sWDHm7sADi5a69uUM+jRpyuLFRffLDTYowt3ee8MTTxQtdwY7SSXLzA26+DHYSZIGVFWzZdZaZv6ow+YvcS2gIavjcgftk6RAsX311TWYBfOVV+CSS+Dss2HRomLh8dNPh803r0n9klQLEfGvXe3PzGsGuhZJUuPqabbMWym6lHQpM3evUQ0HA9fX6FoaQJ2XO+g8h+qSJUXwW6Vwt3QpXHllMevlwoWwyy7F80mT+ly3JPWDrTs8HwXsBDwMGO4kSQOmp5a78yqPewH/CMysbO8HzF/ZhSPix5XzOpuWmT+oHDMNWAa09HCdKcAUgKamppW9rQbQypY7gJ5nzezSsmUwcyZMn17MvPLBD8J118EOO6xqmZLU7zLzix23I2JN4NqSypEkNaieJlSZBRARZ2Rmx9+sb42I+1Z24czcuafXI+JAYDdgp57WzcvMZqAZYNKkSa6vN4hUE9yqzuNtbXDTTXDKKTBvHmy1FVx2WdFiF9GnOiWpBEuATcouQpLUWKoZczcmIjbMzKcBImIDYExf3jQidgW+BHzI5RaGru6WO2hX1eyYmXDHHcWMl488AhMnwo03wl57GeokDRmdhjIMB/4JuKG8iiRJjaiacHcM8NOIeLqyPQE4tI/v+zXgH4C7o/gF/peZ+YU+XlMDrKvlDtonValqofJ77y36dt5/P2y4IVxzDXz2szB8eL/XLkk1dl6H58uA1sx8tqxiJEmNaaXhLjPvjIhNgHdWds3LzL/3dE4V19y4L+drcGgPbtOmFV00m5qqCHQAs2cXLXU//jGMHVssZ3DQQTByZL/XLEn9ITNnRcR6vDGxym/LrEeS1JhWus5dRIymWJT1iMycAzRFxG79XpmGhMmTi3lP2tqKxx6D3dy58KlPwTbbwJw5cOGF8NRTRfOfwU7SEBYRnwF+BXwa+AwwOyJc5keSNKCqWcT8W8BS4AOV7WeBM/utIg1qHRcqnzCh2F6pJ5+E/faDLbaAWbPgjDPgd7+Do4+GUaP6tV5JGiDTgK0z88DM/Ffg/cDJJdckSWow1Yy52ygz94mI/QAy85UIZ7poRJ3XtWttLbahmxa71tZiwfGrr4Z/+Ac44QQ47jhYa60Bq1mSBsiwzHyhw/YiqvsHVEmSaqaacLc0IlajMgtYRGwE9GnMnYamrta163Kh8uefLwbfNTcX20ccASeeCOutN2C1StIAuzMi7gK+U9neB7ijxHokSQ2omnB3KnAnMC4iWoDtgM/1Z1EanLpb1+71/YsWwbnnwle+AkuXwsEHw8knw7hxA1ajJA2kiNgYWC8zj4uIvYAPAgHcD1TTcV2SpJrpMdxFxDBgLWAvYBuKG9ZRmfnSANSmQaa7de0mrr8Ypl8IF1wAL79cjK+bPh02dlJUSXXvImAqQGbeDNwMEBGTKq99sqzCJEmNp8fxAJnZRjFL5qLMvD0zbzPYNa4ZM4qFyduN4hVOHHkeD/7vhnDaabDTTsWMmC0tBjtJjWJCZs7tvDMzH6RYF1aSpAFTTbfMuyPiWOB64G/tOzPzT/1WlQal9nF1p01dyscWXMkpw89kvVcXwkc+BmeeCVtv3fMFJKn+9DTl72oDVoUkSVQX7g6uPB7eYV8CG9a+HA1qy5Yx+dWZTB42HZgPH/ggzLgOdtih7MokqSwPRMTnM/PrHXdGxCHAQyXVJElqUCsNd5m5wUAUosGnpaWYCfOZ1jYOXecm/utNp7Dmc/Ngyy3h0kth113BVTEkNbajge9FxGTeCHOTgDcBe67s5Ij4JrAb8EJmbl7Z9zaK3jITgPnAZzLzf2tduCSp/qx0DZ6IGB0RJ0VEc2V7k4jYrf9LU5laWmDK55N3td7OA0zi0pc+w3MLg/uOvBEefBA+/nGDnaSGl5l/zMxtgUqXBuYD0zPzA5n5fBWXuArYtdO+E4B7MnMT4J7KtiRJK1XNAqvfApYC21a2nwXO7LeKNCjc9p/3cvcr23E7u7Emf+EArmHzfJR//cHehjpJ6iQz783Mr1Z+ftKL8+4DOo9h/xRwdeX51cAetalSklTvqgl3G2XmOcCrAJn5CsWSCKozLS2wx9tn8+PYme/8cUeaWMChXM47mcdMDqCN4d2udSdJqpn1MnMhQOVx3ZLrkSQNEdVMqLI0IlajmESFiNgI+Hu/VqUBd/t/zeWtJ5/E91+7lRcYwzFcwGUcxt87TQTX1FRSgZKk5UTEFGAKQJN/OUuSqK7l7lTgTmBcRLRQ9P8/vl+rUr9raYEJE+Cd8QS3vnlf/nnqe9nutfuYxplsxO+4iGNWCHajRxdr3UmS+tUfI+LtAJXHF7o6KDObM3NSZk4aM2bMgBYoSRqcqpkt8+6IeBjYhqI75lEuZD60tbTAGf/Wysn/dzqf4ypeWbIaM5jKeRzLn1lrheMjiha7GTPeWOtOktRvbgEOBM6uPP6g3HIkSUNFt+EuIrbstGth5bEpIpoy8+H+K0u10r6cwYIFRUA7/7jnefXEGcz9vytIgq/yRf6LE3mB9bo8f/x4mD9/YGuWpEYREd8BPgysExHPUvSWORu4obJW3gLg0+VVKEkaSnpquTu/8jiKYs2eORQtd+8BZgMf7N/S1FctLTBlCixZAmvxJ77Qeg4fP+IrvImlfINDOJOTeJZx3Z5vN0xJ6l+ZuV83L+00oIVIkupCt2PuMvMjmfkRoBXYstKvfyvgfcBTA1WgVt20aTBsycuczOn8ng04nnO4mb1417B5fIErVgh2a69dtNRFFI/NzXbDlCRJkoaKambLfGdmPtq+kZm/jogt+q8k1cQrr/Dp1kv4EmezDou4mT05hdN5jM2hrWiVW7LkjcNHj4aLLzbMSZIkSUNVNbNlzouIKyPiwxHxoYj4OvB4fxemVbR0KVx6KWy0EedyHA+xFVvzK/bm5iLY8UarnK10kiRJUv2opuXuc8BhwFGV7fuAy/qrIK2iZctg5kyYPr2YAeWDH+TuQ65jrwt2WKGFrn3WS8OcJEmSVD96DHcRMRy4LTN3Bi4cmJLUK21tcOONcMop8MQTsOWWcNllsMsufDSC5ncuP1umyxlIkiRJ9anHcJeZr0XEkohYMzP/MlBFqQqZcMcdcNJJ8MgjMHFiEfL22qvoa1lhC50kSZLUGKrplvl/wKMRcTfwt/admXlkv1Wlnt17b9Ecd//9sOGGcM018NnPwvDhZVcmSZIkqSTVhLvbKz8q2+zZRai75x4YOxYuvxwOPhhGjiy7MkmSpEFp7Lgmnnv2mbLLGNyGjSA69PxaVe9Yfxx/eGZBDQrSqqom3F0PbAwk8LvM/L/+LUkrmDsXTj4ZbrkF1lkHLrgAvvAFWG21siuTJEka1J579hn2ueIXfb7O9YduW4NqBqm2ZX5HdaLbpRAiYkREnAM8C1wNzASeiYhzIsKmooHw5JOw336wxRYwaxaccQY8/TQcc4zBTpIkSdJyelrn7lzgbcAGmblVZr4P2Ah4K3DeANTWuFpb4ZBDiklSbrkFTjgBfv/7YvKUNdZY4fCWFpgwAYYNKx5bWga8YkmSJEkl6ync7QZ8PjNfbt+RmYsp1rz7RC3ePCKOjYiMiHVqcb0h7/nn4YtfhE03LdasO+KIoqXurLNgrbW6PKWlBaZMKfJgZvE4ZYoBT5IkSWo0PYW7zMzsYudrFOPv+iQixgEfBRx1uWgRfOlLxcyXl10GBx4ITz0FF10E663X46nTprHcIuVQbE+b1n/lSpIkSRp8egp3v4mIf+28MyL2B+bV4L0vBI6nBkFxyFq8GKZPL0LduecWa9TNmwfNzTBuXFWXWNBNNO5uvyRJkqT61NNsmYcDN0fEwcBDFCFsa2A1YM++vGlE7A78ITPnrGza1YiYAkwBaGpq6svbDh6vvAKXXAJnn1202u2xRzFZyuab9/pSTU1FV8yu9kuSJElqHN2Gu8z8A/D/ImJH4F1AAD/MzHuquXBE/Bj4xy5emgZMBT5WzXUysxloBpg0adLQbuVbuhSuvBLOPBMWLoSPfax4vvXWq3zJGTOKMXYdu2aOHl3slyRJktQ4VrrOXWb+BPhJby+cmTt3tT8i3g1sALS32q0PPBwR78/M53v7PkPCsmXFBCnTp8P8+fDBD8J118EOO/T50pMnF4/TphVdMZuaimDXvl+SJElSY6hmEfOaysxHgXXbtyNiPjApM18a6Fr6XVsb3HQTnHJKMZZuyy2LCVN22QVW0h21NyZPNsxJkiRJja6nCVW0qjLh9tthq63gM58pgtyNN8KDD8Kuu9Y02EmSJEkSDIJwl5kT6qrV7t57YbvtYLfditkwr7kGHn0U9t7bUCdJkiSp35Qe7urG7Nnw0Y/CjjsW01defnnRFfOAA2D48LKrkyRJklTnDHd9NXcufOpTsM028MgjcP75xQLkhx4KI0eWXZ0kSZKkBjHgE6rUjSefhFNPheuvh7e8pVin7qijYI01yq5MkiRJUgOy5a63WlvhkENg4kS45RY44QR4+mk46aR+C3YtLTBhAgwbVjy2tPTL20iSJEkawmy5q9bzzxcLyDU3F9tHHAEnngjrrdevb9vSsvwi5a2txTa4/IEkSZKkN9hytzKLFhWtcxtuWKxRd+CBxZi6iy7q92AHxeLk7cGu3ZIlxX5JkiRJamfLXXcWLy4C3Pnnw8svw377wWmnwSabDGgZCxb0br8kSZKkxmTLXWevvALnnVe01J16arG0wZw5Rf/IAQ52AE1NvdsvSZIkqTEZ7totXQqXXgobbQTHHQdbbQW/+hV873vw7neXVtaMGTB69PL7Ro8u9kuSJElSO8PdsmVw1VWw2WZw+OFFi91Pfwp33QVbb112dUyeXMzhMn48RBSPzc1OpiJJkiRpeY075q6tDW68seh6OW8ebLll0XK3665FihpEJk82zEmSJEnqWeO13GXC7bcX3S732acIcjfeCA8+CB//+KALdpIkSZJUjcYKdz/7GWy3Hey2WzEb5jXXwKOPwt57G+okSZIkDWmN1S1z9uxiDYHLL4eDD4aRI8uuSJIkSZJqorHC3RFHFJOmrLZa2ZVIkiRJUk01VrgbNarsCiRJkiSpXzTWmDtJkiRJqlOGO0mSJEmqA4Y7SZIkSaoDhjtJkiRJqgOGO0mSJEmqA4Y7SZIkSaoDhjtJkiRJqgOGO0mSJEmqA4Y7SZIkSaoDhjtJkiRJqgOGO0mSJEmqA4Y7SZIkSaoDpYW7iPhiRDwREY9FxDll1SFJkiRJ9aCUcBcRHwE+BbwnM98FnFdGHYNBSwtMmADDhhWPLS1lVyRJkiRpKBpR0vseBpydmX8HyMwXSqqjVC0tMGUKLFlSbLe2FtsAkyeXV5ckqXwRMR94GXgNWJaZk8qtSJI02JXVLXNTYPuImB0RsyJi65LqKNW0aW8Eu3ZLlhT7JUkCPpKZWxjsJEnV6LeWu4j4MfCPXbw0rfK+awHbAFsDN0TEhpmZXVxnCjAFoKmpqb/KLcWCBb3bL0mSJEnd6bdwl5k7d/daRBwG3FwJc7+KiDZgHeDFLq7TDDQDTJo0aYXwN5Q1NRVdMbvaL0lqeAn8KCISuKJyP5QkqVtldcv8PrAjQERsCrwJeKmkWkozYwaMHr38vtGji/2SpIa3XWZuCXwcODwiduj4YkRMiYgHI+LBF19c4d9GJUkNqKxw901gw4j4NXAdcGBXXTLr3eTJ0NwM48dDRPHY3OxkKpIkyMznKo8vAN8D3t/p9ebMnJSZk8aMGVNGiZKkQaaU2TIzcymwfxnvPdhMnmyYkyQtLyLeDAzLzJcrzz8GnF5yWZKkQa6spRAkSVL31gO+FxFQ3Ku/nZl3lluSJGmwM9xJkjTIZObTwHvLrkOSNLSUNeZOkiRJklRDhjtJkiRJqgOGO0mSJEmqA4Y7SZIkSaoDhjtJkiRJqgOGO0mSJEmqA4Y7SZIkSaoDhjtJkiRJfTdsBBHR55+x45rK/iRDlouYS5IkSeq7tmXsc8Uv+nyZ6w/dtgbFNCZb7iRJkiSpDhjuJEmSJKkOGO5qpKUFJkyAYcOKx5aWsiuSJEmS1Egcc1cDLS0wZQosWVJst7YW2wCTJ5dXlyRJkqTGYctdDUyb9kawa7dkSbFfkiRJkgaC4a4GFizo3X5JkiRJqjXDXQ00dbMUR3f7JUmSJKnWDHc1MGMGjB69/L7Ro4v9kiRJkjQQDHc1MHkyNDfD+PEQUTw2NzuZiiRJkqSB42yZNTJ5smFOkiRJUnlsuZMkSWpwY8c1ERF9/hk7rnYTDtSqJqmvavVnsdb/j3TFljtJkqQG99yzz7DPFb/o83WuP3TbGlRTGIw1qTHV6s8i9P+fR1vuJEmSJKkOGO4kSZIkqQ4Y7iRJkiSpDhjuJEmSJKkOGO4kSZIkqQ4Y7iRJkiSpDpQS7iJii4j4ZUQ8EhEPRsT7y6hDkiRJkupFWS135wDTM3ML4JTKtiRJkiRpFZUV7hJ4S+X5msBzJdUhSZIkSXVhREnvezRwV0ScRxEw+3epdkmSJPW/YSOIiLKr0FBXoz9H71h/HH94ZkENCho6+i3cRcSPgX/s4qVpwE7AMZl5U0R8BvgGsHM315kCTAFoamrqp2olSZLUZ23L2OeKX9TkUtcf6r/9N6wa/TlqxD9D/RbuMrPLsAYQEdcAR1U2vwtc2cN1moFmgEmTJmUta5QkSZKkelHWmLvngA9Vnu8I/LakOiRJkiSpLpQ15u7zwMURMQL4PyrdLiVJkiRJq6aUcJeZPwe2KuO9JUmSJKkeldUtU5IkSZJUQ4Y7SZIkSaoDDRPuWlpgwgQYNqx4bGkpuyJJkiRJqp2yJlQZUC0tMGUKLFlSbLe2FtsAkyeXV5ckSZIk1UpDtNxNm/ZGsGu3ZEmxX5IkSZLqQUOEuwULerdfkiRJkoaahgh3TU292y9JUiMaO66JiOjzz9hxtbnB1qqeEW8aVZPr1PNnk+rSsBEN9/9HQ4y5mzFj+TF3AKNHF/slSVLhuWefYZ8rftHn61x/6LY1qKa29dTiOu3XqoXB9tlq9bmkQaVtWcP9/9EQLXeTJ0NzM4wfDxHFY3Ozk6lIkiRJqh8N0XIHRZAzzEmSJEmqVw3RcidJkiRJ9c5wJ0mSJEl1wHAnSZIkSXXAcCdJkiRJdcBwJ0nSIBQRu0bEExHxVEScUHY9kqTBz3AnSdIgExHDgUuAjwMTgf0iYmK5VUmSBjvDnSRJg8/7gacy8+nMXApcB3yq5JokSYOc4U6SpMFnLPBMh+1nK/skSepWZGbZNVQtIl4EWsuuowbWAV4qu4hByO+la34vXfN76V49fDfjM3NM2UWUJSI+DeySmf9W2T4AeH9mfrHDMVOAKZXNzYAnBrzQN9TDn7nO/ExDRz1+Lj/T0FDWZ+r2HjlioCvpi3q50UfEg5k5qew6Bhu/l675vXTN76V7fjd14VlgXIft9YHnOh6Qmc1A80AW1Z16/DPnZxo66vFz+ZmGhsH4meyWKUnS4PMAsElEbBARbwL2BW4puSZJ0iA3pFruJElqBJm5LCKOAO4ChgPfzMzHSi5LkjTIGe7KMSi60QxCfi9d83vpmt9L9/xu6kBm3gHcUXYdVarHP3N+pqGjHj+Xn2loGHSfaUhNqCJJkiRJ6ppj7iRJkiSpDhjuShAR50bEvIiYGxHfi4i3ll3TYBERn46IxyKiLSIG1exDZYiIXSPiiYh4KiJOKLuewSAivhkRL0TEr8uuZTCJiHERcW9EPF75f+iosmtSY4mILSLilxHxSEQ8GBHvL7umWoiIL1b+Hn4sIs4pu55aiYhjIyIjYp2ya+mrevq9qt7u+/V8b4qI4RHxPxFxW9m1dGS4K8fdwOaZ+R7gSeDEkusZTH4N7AXcV3YhZYuI4cAlwMeBicB+ETGx3KoGhauAXcsuYhBaBvxnZv4TsA1wuH9eNMDOAaZn5hbAKZXtIS0iPgJ8CnhPZr4LOK/kkmoiIsYBHwUWlF1LjdTF71V1et+v53vTUcDjZRfRmeGuBJn5o8xcVtn8JcX6RQIy8/HMLHMh3sHk/cBTmfl0Zi4FrqP4JaOhZeZ9wJ/KrmOwycyFmflw5fnLFDecseVWpQaTwFsqz9ek07p8Q9RhwNmZ+XeAzHyh5Hpq5ULgeIr/ZkNeHf1eVXf3/Xq9N0XE+sA/A1eWXUtnhrvyHQz8sOwiNCiNBZ7psP0sdfAXovpfREwA3gfMLrkUNZajgXMj4hmKFq4h2XrSyabA9hExOyJmRcTWZRfUVxGxO/CHzJxTdi39ZCj/XlXX9/06uzddRPEPJG0l17ECl0LoJxHxY+Afu3hpWmb+oHLMNIrm6paBrK1s1Xw3AiC62FcX/8qq/hMRqwM3AUdn5uKy61F96envb2An4JjMvCkiPgN8A9h5IOtbFSv5TCOAtSi6k20N3BARG+Ygn2p8JZ9pKvCxga2o7xrk96q6ve/X070pInYDXsjMhyLiwyWXswLDXT/JzB5vaBFxILAbsNNgv0nU2sq+G73uWWBch+31qY9uTuonETGS4ubZkpk3l12P6k9Pf39HxDUUY1AAvssg7K7UlZV8psOAmyv36V9FRBuwDvDiQNW3Krr7TBHxbmADYE5EQHFfeTgi3p+Zzw9gib3WIL9X1eV9vw7vTdsBu0fEJ4BRwFsiYmZm7l9yXYDdMksREbsCXwJ2z8wlZdejQesBYJOI2CAi3gTsC9xSck0apKL4Te0bwOOZeUHZ9aghPQd8qPJ8R+C3JdZSK9+n+CxExKbAm4CXyiyoLzLz0cxcNzMnZOYEijCx5WAPditTR79X1d19vx7vTZl5YmauX/l/aF/gJ4Ml2IHhrixfA9YA7q5MGX152QUNFhGxZ0Q8C3wAuD0i7iq7prJUBocfAdxFMQD5hsx8rNyqyhcR3wHuBzaLiGcj4pCyaxoktgMOAHas/L3ySOVfFaWB8nng/IiYA5wFTCm5nlr4JrBhZemV64ADh3CrUD2ri9+r6vS+771pgIV/R0mSJEnS0GfLnSRJkiTVAcOdJEmSJNUBw50kSZIk1QHDnSRJkiTVAcOdJEmSJNUBw53qTkSs3WG63ecj4g+V53+OiN8McC17RMTEDtunR0SvF3GPiAmVqbhLERFTO23/ovJYal2SpPoXEadFxLE9vL7cvVZqZIY71Z3MXJSZW2TmFsDlwIWV51sAbbV+v4gY0cPLewCv33Ay85TM/HGtaxgAy4W7zNy2rEIkSepkDzrca6VGZrhToxkeEV+PiMci4kcRsRpARGwUEXdGxEMR8bOIeGdl//iIuCci5lYemyr7r4qICyLiXuDLXZ0fEdsCuwPnVloON6qc9y+Va2wdEb+IiDkR8auIWKPSEvaziHi48tNjiIrC1yLiNxFxe0Tc0eH68yNincrzSRHx08rz91fe938qj5tV9n8uIm6ufI7fRsQ5lf1nA6tVPkNLZd9fu6hleEScGxEPVL6vQyv73x4R91XO/3VEbN/H/4aSpDoXEdMi4omI+DHQfp/6fOUeMyciboqI0d3ca1c4rtQPIw0gw50azSbAJZn5LuDPwN6V/c3AFzNzK+BY4NLK/q8B12Tme4AW4CsdrrUpsHNm/mdX52fmL4BbgOMqLYm/az8xIt4EXA8clZnvBXYGXgFeAD6amVsC+3R6v67sSXHTezfweaCaFrV5wA6Z+T7gFOCsDq9tUXnfdwP7RMS4zDwBeKXyGSb3cN1DgL9k5tbA1sDnI2ID4LPAXZXW0/cCj1RRoySpQUXEVsC+wPuAvSjuKQA3Z+bWlfvm48Ah3dxrVzhu4D+FVI6eupNJ9ej3mflI5flDwISIWJ0iFH03ItqP+4fK4wcobiwA1wLndLjWdzPztZWc353NgIWZ+QBAZi4GiIg3A1+LiC2A1ygCZE92AL6Tma8Bz0XET1ZyPMCawNURsQmQwMgOr92TmX+p1PIbYDzwTBXXBPgY8J72lsPK+2wCPAB8MyJGAt/v8P1LktSV7YHvZeYSgIi4pbJ/84g4E3grsDpwVzfnV3ucVHcMd2o0f+/w/DVgNYoW7D9XWpZWJjs8/1vlsTfnt4tO12p3DPBHihauYcD/9bKmjpbxRuv8qA77zwDuzcw9I2IC8NMOr3X+fnrzd0RQtF6ucBONiB2AfwaujYhzM/OaXlxXktR4urq3XQXskZlzIuJzwIe7Obfa46S6Y7dMNbxKq9nvI+LT8Po4tvdWXv4FRdcQgMnAz3t5/svAGl287TzgHRGxdeWcNaKYmGVNiha9NuAAYPhKyr8P2Lcy3u3twEc6vDYf2KryfO8O+9cE/lB5/rmVXL/dq5WWt57cBRzWflxEbBoRb46I8cALmfl14BvAllW+pySpMd0H7BkRq0XEGsAnK/vXABZW7jMdhwl0vtd2d5xU9wx3UmEycEhEzAEeAz5V2X8kcFBEzKUIW0f18vzrgOMqk5ds1H5wZi6lGNv21co5d1O0rl0KHBgRv6Tokvk3evY94LfAo8BlwKwOr00HLo6In1G0wrU7B/iviPhvVh4e2zUDc9snVOnGlcBvgIejWB7hCoqWvw8Dj0TE/1CEzIurfE9JUgPKzIcpxqU/AtwE/Kzy0snAbIp75rwOp3S+13Z3nFT3IrO7Hl2ShpqIuAq4LTNvLLsWSZIkDSxb7iRJkiSpDthyJ0mSJEl1wJY7SZIkSaoDhjtJkiRJqgOGO0mSJEmqA4Y7SZIkSaoDhjtJkiRJqgOGO0mSJEmqA/8f1agE3Bo5C9wAAAAASUVORK5CYII=\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/heavy_tailed_data.csv\")\n",
     "\n",
-    "Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) shows the same plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
+    "\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
+    "\n",
+    "glue(\"qqheavy_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "distinguished-wheel",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqheavy_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqheavy\n",
+    "\n",
+    "Plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
     "\n",
-    "One way to check whether a sample violates the normality assumption is to draw a **_\"quantile-quantile\" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot using the R function `qqnorm()`. The `qqnorm()` function has a few arguments, but the only one we really need to care about here is `y`, a vector specifying the data whose normality we're interested in checking. Here's the R commands:\n",
-    "```{r}\n",
-    "normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers\n",
-    "hist( x = normal.data )          # draw a histogram of these numbers\n",
-    "qqnorm( y = normal.data )        # draw the QQ plot\n",
     "```"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "sapphire-advertiser",
+   "id": "russian-monaco",
    "metadata": {},
    "source": [
     "### Shapiro-Wilk tests\n",
     "\n",
-    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.^[Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality; although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests; in R it's implemented via the `ks.test()` function.] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.[^note14] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "\n",
     "$$\n",
     "W = \\frac{ \\left( \\sum_{i = 1}^N a_i X_i \\right)^2 }{ \\sum_{i = 1}^N (X_i - \\bar{X})^2}\n",
     "$$\n",
-    "where $\\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. \n",
     "\n",
-    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. \n",
+    "where $\\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. \n",
     "\n",
-    "```{r swdist, fig.cap=\"Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\", echo=FALSE}\n",
-    "knitr::include_graphics(file.path(projecthome, \"img/ttest2/shapirowilkdist.png\"))\n",
+    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "durable-delivery",
+   "metadata": {},
+   "source": [
+    "```{figure} ../img/ttest2/shapirowilkdist.png\n",
+    ":name: fig-shapirowilk\n",
+    ":width: 600px\n",
+    ":align: center\n",
     "\n",
-    "```\n",
+    "Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\n",
     "\n",
-    "To run the test in R, we use the `shapiro.test()` function. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal.data`, we get the following:\n",
-    "```{r}\n",
-    "shapiro.test( x = normal.data )\n",
-    "```\n",
-    "So, not surprisingly, we have no evidence that these data depart from normality. When reporting the results for a Shapiro-Wilk test, you should (as usual) make sure to include the test statistic $W$ and the $p$ value, though given that the sampling distribution depends so heavily on $N$ it would probably be a politeness to include $N$ as well.\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "clinical-stroke",
+   "metadata": {},
+   "source": [
+    "To run the test in Python, we use the `scipy.stats.shapiro` method. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal_data`, we get the following:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 708,
+   "id": "union-accordance",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "ShapiroResult(statistic=0.9898834228515625, pvalue=0.6551706790924072)"
+      ]
+     },
+     "execution_count": 708,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from scipy.stats import shapiro\n",
     "\n",
+    "shapiro(normal_data)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "bored-bristol",
+   "metadata": {},
+   "source": [
+    "So, not surprisingly, we have no evidence that these data depart from normality. When reporting the results for a Shapiro-Wilk test, you should (as usual) make sure to include the test statistic $W$ and the $p$ value, though given that the sampling distribution depends so heavily on $N$ it would probably be a politeness to include $N$ as well."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "renewable-calcium",
+   "metadata": {},
+   "source": [
     "## Testing non-normal data with Wilcoxon tests{#wilcox}\n",
     "\n",
     "Okay, suppose your data turn out to be pretty substantially non-normal, but you still want to run something like a $t$-test? This situation occurs a lot in real life: for the AFL winning margins data, for instance, the Shapiro-Wilk test made it very clear that the normality assumption is violated. This is the situation where you want to use Wilcoxon tests. \n",
@@ -3128,7 +3271,9 @@
     "\n",
     "[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students.\n",
     "\n",
-    "[^note13]: This is a massive oversimplification."
+    "[^note13]: This is a massive oversimplification.\n",
+    "\n",
+    "[^note14]: Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality, although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests."
    ]
   }
  ],
diff --git a/Chapters/05.02-ttest.ipynb b/Chapters/05.02-ttest.ipynb
index d8c5cb99..2a614d5a 100644
--- a/Chapters/05.02-ttest.ipynb
+++ b/Chapters/05.02-ttest.ipynb
@@ -230,7 +230,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "junior-petersburg",
+   "id": "mounted-eleven",
    "metadata": {},
    "source": [
     "### Constructing the hypothesis test\n",
@@ -255,7 +255,7 @@
   {
    "cell_type": "code",
    "execution_count": 253,
-   "id": "compliant-steps",
+   "id": "worldwide-hughes",
    "metadata": {
     "tags": [
      "hide-input"
@@ -327,7 +327,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "handled-reliance",
+   "id": "protective-seller",
    "metadata": {},
    "source": [
     "```{glue:figure} ztesthyp-fig\n",
@@ -344,7 +344,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "excellent-gothic",
+   "id": "wireless-share",
    "metadata": {},
    "source": [
     "The next step is to figure out what we would be a good choice for a diagnostic test statistic; something that would help us discriminate between $H_0$ and $H_1$. Given that the hypotheses all refer to the population mean $\\mu$, you'd feel pretty confident that the sample mean $\\bar{X}$ would be a pretty useful place to start. What we could do, is look at the difference between the sample mean $\\bar{X}$ and the value that the null hypothesis predicts for the population mean. In our example, that would mean we calculate $\\bar{X} - 67.5$. More generally, if we let $\\mu_0$ refer to the value that the null hypothesis claims is our population mean, then we'd want to calculate\n",
@@ -399,7 +399,7 @@
   {
    "cell_type": "code",
    "execution_count": 254,
-   "id": "occasional-medicine",
+   "id": "pressed-circular",
    "metadata": {
     "tags": [
      "hide-input"
@@ -478,7 +478,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "superb-norwegian",
+   "id": "lined-tooth",
    "metadata": {},
    "source": [
     "```{glue:figure} ztest-fig\n",
@@ -495,7 +495,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "hired-birmingham",
+   "id": "threaded-ranking",
    "metadata": {},
    "source": [
     "And what this meant, way back in the days where people did all their statistics by hand, is that someone could publish a table like this:"
@@ -503,7 +503,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "mental-textbook",
+   "id": "undefined-kennedy",
    "metadata": {},
    "source": [
     "|                   || critical z value             |\n",
@@ -517,7 +517,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "discrete-billy",
+   "id": "strange-demonstration",
    "metadata": {},
    "source": [
     "which in turn meant that researchers could calculate their $z$-statistic by hand, and then look up the critical value in a text book.  That was an incredibly handy thing to be able to do back then, but it's kind of unnecessary these days, since it's trivially easy to do it with software like Python."
@@ -525,7 +525,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "overhead-lexington",
+   "id": "pending-spiritual",
    "metadata": {},
    "source": [
     "### A worked example using Python\n",
@@ -536,7 +536,7 @@
   {
    "cell_type": "code",
    "execution_count": 262,
-   "id": "optical-attitude",
+   "id": "clinical-baseline",
    "metadata": {},
    "outputs": [
     {
@@ -558,7 +558,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "brave-throw",
+   "id": "settled-prince",
    "metadata": {},
    "source": [
     "Then, I create variables corresponding to known population standard deviation ($\\sigma = 9.5$), and the value of the population mean that the null hypothesis specifies ($\\mu_0 = 67.5$):"
@@ -567,7 +567,7 @@
   {
    "cell_type": "code",
    "execution_count": 261,
-   "id": "pointed-oliver",
+   "id": "thirty-error",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -577,7 +577,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "nuclear-trade",
+   "id": "living-treaty",
    "metadata": {},
    "source": [
     "Let's also create a variable for the sample size. We could count up the number of observations ourselves, and type `N = 20` at the command prompt, but counting is tedious and repetitive. Let's get Python to do the tedious repetitive bit by using the `len()` function, which tells us how many elements there are in a vector:"
@@ -586,7 +586,7 @@
   {
    "cell_type": "code",
    "execution_count": 263,
-   "id": "useful-costume",
+   "id": "acting-england",
    "metadata": {},
    "outputs": [
     {
@@ -607,7 +607,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "incorporated-acquisition",
+   "id": "intimate-volunteer",
    "metadata": {},
    "source": [
     "Next, let's calculate the (true) standard error of the mean:"
@@ -616,7 +616,7 @@
   {
    "cell_type": "code",
    "execution_count": 266,
-   "id": "opposite-advocacy",
+   "id": "opposed-williams",
    "metadata": {},
    "outputs": [
     {
@@ -638,7 +638,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "southern-lucas",
+   "id": "aboriginal-default",
    "metadata": {},
    "source": [
     "And finally, we calculate our $z$-score:"
@@ -647,7 +647,7 @@
   {
    "cell_type": "code",
    "execution_count": 268,
-   "id": "virgin-geometry",
+   "id": "discrete-missile",
    "metadata": {},
    "outputs": [
     {
@@ -668,7 +668,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "premier-contemporary",
+   "id": "enclosed-edwards",
    "metadata": {},
    "source": [
     "At this point, we would traditionally look up the value 2.26 in our table of critical values. Our original hypothesis was two-sided (we didn't really have any theory about whether psych students would be better or worse at statistics than other students) so our hypothesis test is two-sided (or two-tailed) also. Looking at the little table that I showed earlier, we can see that 2.26 is bigger than the critical value of 1.96 that would be required to be significant at $\\alpha = .05$, but smaller than the value of 2.58 that would be required to be significant at a level of $\\alpha = .01$. Therefore, we can conclude that we have a significant effect, which we might write up by saying something like this:\n",
@@ -681,7 +681,7 @@
   {
    "cell_type": "code",
    "execution_count": 286,
-   "id": "russian-victim",
+   "id": "subject-stretch",
    "metadata": {},
    "outputs": [
     {
@@ -703,7 +703,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "continuing-consequence",
+   "id": "associate-cornwall",
    "metadata": {},
    "source": [
     "`NormalDist().cdf()` calculates the \"cumulative density function\" for a normal distribution. Translated to something slightly less opaque, this means that  `NormalDist().cdf()` gives us the probability that a random variable X will be less than or equal to a given value. In our case, the given value for the lower tail of the distribution was our z-score, $2.259$. So `NormalDist().cdf(-z_score)` gives us the probability that a random value draw from a normal distribution would be less than or equal to $-2.259$.\n",
@@ -716,7 +716,7 @@
   {
    "cell_type": "code",
    "execution_count": 288,
-   "id": "arctic-chair",
+   "id": "disabled-proportion",
    "metadata": {},
    "outputs": [
     {
@@ -739,7 +739,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "posted-tongue",
+   "id": "adjustable-minimum",
    "metadata": {},
    "source": [
     "(zassumptions)=\n",
@@ -756,7 +756,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "bizarre-blocking",
+   "id": "regular-prior",
    "metadata": {},
    "source": [
     "(onesamplettest)=\n",
@@ -768,7 +768,7 @@
   {
    "cell_type": "code",
    "execution_count": 289,
-   "id": "sufficient-melissa",
+   "id": "literary-mainstream",
    "metadata": {},
    "outputs": [
     {
@@ -789,7 +789,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "domestic-eligibility",
+   "id": "rough-yeast",
    "metadata": {},
    "source": [
     "In other words, while I can't say that I know that $\\sigma = 9.5$, I *can* say that $\\hat\\sigma = 9.52$. \n",
@@ -800,7 +800,7 @@
   {
    "cell_type": "code",
    "execution_count": 291,
-   "id": "liquid-relief",
+   "id": "raising-forestry",
    "metadata": {
     "tags": [
      "hide-input"
@@ -873,7 +873,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "coastal-distributor",
+   "id": "alike-debut",
    "metadata": {},
    "source": [
     "\n",
@@ -893,7 +893,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "excess-philosophy",
+   "id": "systematic-milton",
    "metadata": {},
    "source": [
     "### Introducing the $t$-test\n",
@@ -910,7 +910,7 @@
   {
    "cell_type": "code",
    "execution_count": 325,
-   "id": "informative-karen",
+   "id": "ordinary-orleans",
    "metadata": {
     "tags": [
      "hide-input"
@@ -966,7 +966,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "looking-instrumentation",
+   "id": "aquatic-extraction",
    "metadata": {},
    "source": [
     "```{glue:figure} ttestdist-fig\n",
@@ -982,7 +982,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "about-commitment",
+   "id": "false-angola",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -995,7 +995,7 @@
   {
    "cell_type": "code",
    "execution_count": 359,
-   "id": "engaging-stadium",
+   "id": "modular-incident",
    "metadata": {},
    "outputs": [
     {
@@ -1016,7 +1016,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "parallel-collapse",
+   "id": "forward-playlist",
    "metadata": {},
    "source": [
     "So that seems straightforward enough. Our calculation resulted in a $t$-statistic of 2.54, and a $p$-value of 0.36. Now what do we *do* with this output? Well, since we're pretending that we actually care about my toy example, we're overjoyed to discover that the result is statistically significant (i.e. $p$ value below .05), and we will probably want to report our result. We could report the result by saying something like this:\n",
@@ -1033,7 +1033,7 @@
   {
    "cell_type": "code",
    "execution_count": 352,
-   "id": "acknowledged-japanese",
+   "id": "international-variable",
    "metadata": {},
    "outputs": [
     {
@@ -1055,7 +1055,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "serial-percentage",
+   "id": "improved-incident",
    "metadata": {},
    "source": [
     "Now at least we have the bare minimum of what is necessary to report our results. Still, it would be sweet if we could get those confidence intervals as well. `scipy` actually has all the tools we need, and why these are not just built into the `ttest_1samp()` method is beyond me. To find the confidence interval, we need to:\n",
@@ -1071,7 +1071,7 @@
   {
    "cell_type": "code",
    "execution_count": 356,
-   "id": "common-comedy",
+   "id": "enclosed-maple",
    "metadata": {},
    "outputs": [
     {
@@ -1099,7 +1099,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "japanese-highway",
+   "id": "collective-wallpaper",
    "metadata": {},
    "source": [
     "Whew. Now at least we have everything we need for a full report of our results.\n",
@@ -1113,7 +1113,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "dense-symbol",
+   "id": "previous-discretion",
    "metadata": {},
    "source": [
     "(ttestoneassumptions)=\n",
@@ -1129,7 +1129,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "southwest-peter",
+   "id": "female-health",
    "metadata": {},
    "source": [
     "(studentttest)=\n",
@@ -1140,7 +1140,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "engaging-peace",
+   "id": "critical-olympus",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1151,7 +1151,7 @@
   {
    "cell_type": "code",
    "execution_count": 360,
-   "id": "drawn-riding",
+   "id": "israeli-details",
    "metadata": {},
    "outputs": [
     {
@@ -1232,7 +1232,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "hawaiian-project",
+   "id": "lyric-copyright",
    "metadata": {},
    "source": [
     "As we can see, there's a single data frame with two variables, `grade` and `tutor`. The `grade` variable is a numeric vector, containing the grades for all $N = 33$ students taking Dr Harpo's class; the `tutor` variable is a factor that indicates who each student's tutor was. The first five observations in this data set are shown above, and below is a nice little table with some summary statistics:"
@@ -1241,7 +1241,7 @@
   {
    "cell_type": "code",
    "execution_count": 437,
-   "id": "honey-thomson",
+   "id": "forty-silly",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1320,7 +1320,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "romantic-refund",
+   "id": "empty-powell",
    "metadata": {},
    "source": [
     "To give you a more detailed sense of what's going on here, I've plotted histograms showing the distribution of grades for both tutors {numref}`fig-harpohist`. Inspection of these histograms suggests that the students in Anastasia's class may be getting slightly better grades on average, though they also seem a little more variable."
@@ -1329,7 +1329,7 @@
   {
    "cell_type": "code",
    "execution_count": 408,
-   "id": "cutting-effects",
+   "id": "promising-screw",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1371,7 +1371,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "demonstrated-night",
+   "id": "grateful-object",
    "metadata": {},
    "source": [
     " ```{glue:figure} harpohist_fig\n",
@@ -1384,7 +1384,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "variable-optimization",
+   "id": "attempted-knife",
    "metadata": {},
    "source": [
     "{numref}`fig-ttestci` is a simpler plot showing the means and corresponding confidence intervals for both groups of students."
@@ -1393,7 +1393,7 @@
   {
    "cell_type": "code",
    "execution_count": 413,
-   "id": "built-princess",
+   "id": "binary-relaxation",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1430,7 +1430,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "steady-compact",
+   "id": "hispanic-reynolds",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttestci-fig\n",
@@ -1444,7 +1444,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "informative-lawyer",
+   "id": "british-anime",
    "metadata": {},
    "source": [
     "### Introducing the test\n",
@@ -1464,7 +1464,7 @@
   {
    "cell_type": "code",
    "execution_count": 436,
-   "id": "stupid-check",
+   "id": "raising-scale",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1528,7 +1528,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "particular-finish",
+   "id": "joint-builder",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp_fig\n",
@@ -1541,7 +1541,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "laughing-approach",
+   "id": "friendly-bacon",
    "metadata": {},
    "source": [
     "To construct a hypothesis test that handles this scenario, we start by noting that if the null hypothesis is true, then the difference between the population means is *exactly* zero, \n",
@@ -1563,7 +1563,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "processed-candidate",
+   "id": "refined-width",
    "metadata": {},
    "source": [
     "### A \"pooled estimate\" of the standard deviation\n",
@@ -1594,7 +1594,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "psychological-trainer",
+   "id": "still-exhibition",
    "metadata": {},
    "source": [
     "### The same pooled estimate, described differently\n",
@@ -1622,7 +1622,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "swiss-uncle",
+   "id": "black-james",
    "metadata": {},
    "source": [
     "(indsamplesttest_formula)=\n",
@@ -1645,7 +1645,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "experimental-documentary",
+   "id": "indoor-reach",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1658,7 +1658,7 @@
   {
    "cell_type": "code",
    "execution_count": 454,
-   "id": "independent-alexandria",
+   "id": "concrete-revolution",
    "metadata": {},
    "outputs": [
     {
@@ -1684,7 +1684,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "settled-yield",
+   "id": "mysterious-minority",
    "metadata": {},
    "source": [
     "This is fairly straightforward, and just as it was for the one-sample $t$-test, `scipy` does very little to format its results or give you any information over the bare minimum. You get a $t$-statistic and a $p$-value and that's that. Luckily, as was the case with the one-sampel $t$-test, getting the other elements we need to report our results isn't too bad. We will need the $t$-statistic, the $p$-value, the mean of each group, and the degrees of freedom. The first two we already have, and the last two are easy to get. As we discussed [above](indsamplesttest_formula), the degrees of freedom for an independent samples $t$-test is $N-2$, so..."
@@ -1693,7 +1693,7 @@
   {
    "cell_type": "code",
    "execution_count": 465,
-   "id": "interracial-hobby",
+   "id": "strategic-alberta",
    "metadata": {},
    "outputs": [
     {
@@ -1717,7 +1717,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "wireless-worcester",
+   "id": "hindu-neighborhood",
    "metadata": {},
    "source": [
     "You probably noticed that in addition to telling `ttest_ind` which means I wanted to compare, I also added the argument `equal_var = True` to the command. This wasn't strictly necessary in this case, because by default this argument is set to `True`. But I made it explicit anyway, because we will be using this argument again later. By saying `equal_var = True`, what we're really doing is telling Python to use the *Student* independent samples $t$-test. More on this later."
@@ -1725,7 +1725,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "military-green",
+   "id": "graduate-keeping",
    "metadata": {},
    "source": [
     "In any case, the difference between the two groups is significant (just barely), so we might write up the result using text like this:\n",
@@ -1735,7 +1735,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "amino-sharing",
+   "id": "relevant-technology",
    "metadata": {},
    "source": [
     " \n",
@@ -1767,7 +1767,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "delayed-hygiene",
+   "id": "occupied-selection",
    "metadata": {},
    "source": [
     "(studentassumptions)= \n",
@@ -1783,7 +1783,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "color-participant",
+   "id": "after-assist",
    "metadata": {},
    "source": [
     "(welchttest)=\n",
@@ -1817,7 +1817,7 @@
   {
    "cell_type": "code",
    "execution_count": 530,
-   "id": "united-little",
+   "id": "later-horizontal",
    "metadata": {
     "tags": [
      "hide-input"
@@ -1888,7 +1888,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "ceramic-thesis",
+   "id": "enormous-istanbul",
    "metadata": {},
    "source": [
     " ```{glue:figure} ttesthyp2_fig\n",
@@ -1901,7 +1901,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "constant-beads",
+   "id": "golden-hawaiian",
    "metadata": {},
    "source": [
     "### Doing the test in Python\n",
@@ -1912,7 +1912,7 @@
   {
    "cell_type": "code",
    "execution_count": 531,
-   "id": "patient-needle",
+   "id": "correct-upper",
    "metadata": {},
    "outputs": [
     {
@@ -1933,7 +1933,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "cheap-simon",
+   "id": "excellent-fruit",
    "metadata": {},
    "source": [
     "Not too difficult, right? Not surprisingly, the output has exactly the same format as it did last time too: a test statistic $t$, and a $p$-value. So that's all pretty easy. \n",
@@ -1948,7 +1948,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "appointed-exploration",
+   "id": "becoming-right",
    "metadata": {},
    "source": [
     "(pairedsamplesttest)=\n",
@@ -1959,7 +1959,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "pharmaceutical-laundry",
+   "id": "hungarian-headline",
    "metadata": {},
    "source": [
     "### The data\n",
@@ -1970,7 +1970,7 @@
   {
    "cell_type": "code",
    "execution_count": 563,
-   "id": "proof-investment",
+   "id": "competitive-string",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1980,7 +1980,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "liberal-israel",
+   "id": "current-criminal",
    "metadata": {},
    "source": [
     "The data frame `chico` contains three variables: an `id` variable that identifies each student in the class, the `grade_test1` variable that records the student grade for the first test, and the `grade_test2` variable that has the grades for the second test. Here's the first five students:"
@@ -1989,7 +1989,7 @@
   {
    "cell_type": "code",
    "execution_count": 534,
-   "id": "electrical-harvard",
+   "id": "proof-shame",
    "metadata": {},
    "outputs": [
     {
@@ -2073,7 +2073,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "developing-baseball",
+   "id": "analyzed-wesley",
    "metadata": {},
    "source": [
     "At a glance, it does seem like the class is a hard one (most grades are between 50\\% and 60\\%), but it does look like there's an improvement from the first test to the second one. If we take a quick look at the descriptive statistics"
@@ -2082,7 +2082,7 @@
   {
    "cell_type": "code",
    "execution_count": 535,
-   "id": "solar-dodge",
+   "id": "smooth-budget",
    "metadata": {},
    "outputs": [
     {
@@ -2178,7 +2178,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "silent-orleans",
+   "id": "increasing-january",
    "metadata": {},
    "source": [
     "we see that this impression seems to be supported. Across all 20 students[^note12] the mean grade for the first test is 57\\%, but this rises to 58\\% for the second test. Although, given that the standard deviations are 6.6\\% and 6.4\\% respectively, it's starting to feel like maybe the improvement is just illusory; maybe just random variation. This impression is reinforced when you see the means and confidence intervals plotted in {numref}`pairedta` panel A. If we were to rely on this plot alone, we'd come to the same conclusion that we got from looking at the descriptive statistics that the `describe()` method produced. Looking at how wide those confidence intervals are, we'd be tempted to think that the apparent improvement in student performance is pure chance."
@@ -2187,7 +2187,7 @@
   {
    "cell_type": "code",
    "execution_count": 595,
-   "id": "greenhouse-filing",
+   "id": "identical-sympathy",
    "metadata": {
     "tags": [
      "hide-input"
@@ -2237,7 +2237,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "reasonable-adult",
+   "id": "emerging-opinion",
    "metadata": {},
    "source": [
     "```{glue:figure} pairedta_fig\n",
@@ -2251,7 +2251,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "understanding-situation",
+   "id": "signed-lawsuit",
    "metadata": {},
    "source": [
     "Nevertheless, this impression is wrong. To see why, take a look at the scatterplot of the grades for test 1 against the grades for test 2.  shown in {numref}`fig-pairedta` panel B. \n",
@@ -2264,7 +2264,7 @@
   {
    "cell_type": "code",
    "execution_count": 596,
-   "id": "martial-street",
+   "id": "discrete-observer",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -2273,7 +2273,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "abandoned-clearing",
+   "id": "pacific-smith",
    "metadata": {},
    "source": [
     "Notice that I assigned the output to a variable called `df['improvement]`. That has the effect of creating a new column called `improvement` inside the `chico` data frame. Now that we've created and stored this `improvement` variable, we can draw a histogram showing the distribution of these improvement scores, shown in {numref}`fig-pairedta` panel C. \n",
@@ -2285,7 +2285,7 @@
   {
    "cell_type": "code",
    "execution_count": 598,
-   "id": "dried-continuity",
+   "id": "analyzed-penguin",
    "metadata": {},
    "outputs": [
     {
@@ -2310,7 +2310,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "graduate-error",
+   "id": "accomplished-carol",
    "metadata": {},
    "source": [
     "we see that it is 95\\% certain that the true (population-wide) average improvement would lie between 0.95\\% and 1.86\\%. So you can see, qualitatively, what's going on: there is a real \"within student\" improvement (everyone improves by about 1\\%), but it is very small when set against the quite large \"between student\" differences (student grades vary by about 20\\% or so). "
@@ -2318,7 +2318,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "radical-shoulder",
+   "id": "proof-rubber",
    "metadata": {},
    "source": [
     "### What is the paired samples $t$-test?\n",
@@ -2357,7 +2357,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "organic-simpson",
+   "id": "behind-continent",
    "metadata": {},
    "source": [
     "### Doing the test in Python \n",
@@ -2368,7 +2368,7 @@
   {
    "cell_type": "code",
    "execution_count": 600,
-   "id": "painted-architecture",
+   "id": "dedicated-grade",
    "metadata": {},
    "outputs": [
     {
@@ -2389,7 +2389,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "forty-consortium",
+   "id": "sought-scott",
    "metadata": {},
    "source": [
     "However, suppose you're lazy and you don't want to go to all the effort of creating a new variable. Or perhaps you just want to keep the difference between one-sample and paired-samples tests clear in your head. In that case, `scipy` also has a built-in method for conducting paired $t$-tests called `ttest_rel` (the `_rel` part is for \"related\"). Using this method, we get:"
@@ -2398,7 +2398,7 @@
   {
    "cell_type": "code",
    "execution_count": 599,
-   "id": "annoying-interpretation",
+   "id": "roman-berkeley",
    "metadata": {},
    "outputs": [
     {
@@ -2420,7 +2420,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "accessory-forth",
+   "id": "younger-flour",
    "metadata": {},
    "source": [
     "Either way, the result is exactly the same, which is strangely comforting, actually.  Not only that, but the result confirms our intuition. There’s an average improvement of 1.4% from test 1 to test 2, and this is significantly different from 0 ($t$(19) = 6.48, $p$ < .001). In fact, $p$ is quite a bit less than one, since the $p$-value has been given in scientific notation. The exact $p$-value is $3.32^{-06}$, that is, $p$ = 0.0000032."
@@ -2428,7 +2428,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "small-crossing",
+   "id": "accompanied-contents",
    "metadata": {},
    "source": [
     "## One sided tests\n",
@@ -2439,7 +2439,7 @@
   {
    "cell_type": "code",
    "execution_count": 606,
-   "id": "capital-avenue",
+   "id": "careful-victor",
    "metadata": {},
    "outputs": [
     {
@@ -2464,7 +2464,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "abandoned-neighbor",
+   "id": "split-climate",
    "metadata": {},
    "source": [
     "The $t$-statistics are exactly the same, which makes sense, if you think about it, because the calculation of the $t$ is based on the mean and standard deviation, and these do not change. The $p$-value, on the other hand, is lower for the one-sided test. The only thing that changes between the two tests is the _expectation_ that we bring to data. The way that the $p$-value is calculated depends on those expectations, and they are the reason for choosing one test over the other. It should go without saying, but maybe is worth saying anyway, that our reasons for choosing one test over the other should be theoretical, and not based on which test is more likely to give us the $p$-value we want!\n",
@@ -2476,7 +2476,7 @@
   {
    "cell_type": "code",
    "execution_count": 613,
-   "id": "appreciated-witness",
+   "id": "discrete-playback",
    "metadata": {},
    "outputs": [
     {
@@ -2510,7 +2510,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "committed-interaction",
+   "id": "warming-insider",
    "metadata": {},
    "source": [
     "What about the paired samples $t$-test? Suppose we wanted to test the hypothesis that grades go *up* from test 1 to test 2 in Dr. Chico's class, and are not prepared to consider the idea that the grades go down. Again, we can use the `alternative` argument to specify the one-sided test, and it works the same way it does for the independent samples $t$-test. Since we are comparing test 1 to test 2 by substracting one from the other, it makes a difference whether we subract test 1 from test 2, or test 2 from test 1. So, to test the hypothesis that grades for test 2 are higher than test 2, we will need to enter the grades from test 2 first; otherwise we are testing the opposite hypothesis: "
@@ -2519,7 +2519,7 @@
   {
    "cell_type": "code",
    "execution_count": 616,
-   "id": "surface-senator",
+   "id": "upper-security",
    "metadata": {},
    "outputs": [
     {
@@ -2545,7 +2545,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "breeding-pizza",
+   "id": "muslim-population",
    "metadata": {},
    "source": [
     "(cohensd)=\n",
@@ -2564,7 +2564,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "separated-principle",
+   "id": "electrical-movie",
    "metadata": {},
    "source": [
     "(dinterpretation)=\n",
@@ -2574,7 +2574,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "favorite-dollar",
+   "id": "least-property",
    "metadata": {},
    "source": [
     "| d-value   | rough interpretation |\n",
@@ -2586,7 +2586,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "gorgeous-aspect",
+   "id": "fifteen-surgery",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from one sample\n",
@@ -2601,7 +2601,7 @@
   {
    "cell_type": "code",
    "execution_count": 624,
-   "id": "proof-gateway",
+   "id": "blind-confusion",
    "metadata": {},
    "outputs": [
     {
@@ -2628,7 +2628,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "simplified-extra",
+   "id": "empty-staff",
    "metadata": {},
    "source": [
     "What does this effect size mean? Overall, then, the psychology students in Dr Zeppo's class are achieving grades (mean = 72.3\\%) that are about .5 standard deviations higher than the level that you'd expect (67.5\\%) if they were performing at the same level as other students. Judged against Cohen's rough guide, this is a moderate effect size."
@@ -2636,7 +2636,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "foreign-tolerance",
+   "id": "original-minute",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Student $t$ test\n",
@@ -2658,7 +2658,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "sitting-highway",
+   "id": "scenic-positive",
    "metadata": {},
    "source": [
     "However, there are other possibilities, which I'll briefly describe. Firstly, you may have reason to want to use only one of the two groups as the basis for calculating the standard deviation. This approach (often called Glass' $\\Delta$) only makes most sense when you have good reason to treat one of the two groups as a purer reflection of \"natural variation\" than the other. This can happen if, for instance, one of the two groups is a control group. Secondly, recall that in the usual calculation of the pooled standard deviation we divide by $N-2$ to correct for the bias in the sample variance; in one version of Cohen's $d$ this correction is omitted. Instead, we divide by $N$. This version makes sense primarily when you're trying to calculate the effect size in the sample; rather than estimating an effect size in the population. Finally, there is a version based on @Hedges1985, who point out there is a small bias in the usual (pooled) estimation for Cohen's $d$. Thus they introduce a small correction, by multiplying the usual value of $d$ by $(N-3)/(N-2.25)$. \n",
@@ -2669,7 +2669,7 @@
   {
    "cell_type": "code",
    "execution_count": 654,
-   "id": "sensitive-midwest",
+   "id": "amended-agriculture",
    "metadata": {},
    "outputs": [
     {
@@ -2719,7 +2719,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "interim-madison",
+   "id": "extended-philippines",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a Welch test\n",
@@ -2749,7 +2749,7 @@
   {
    "cell_type": "code",
    "execution_count": 657,
-   "id": "canadian-fountain",
+   "id": "artificial-evidence",
    "metadata": {},
    "outputs": [
     {
@@ -2784,7 +2784,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "excessive-siemens",
+   "id": "constant-easter",
    "metadata": {},
    "source": [
     "### Cohen's $d$ from a paired-samples test\n",
@@ -2802,7 +2802,7 @@
   {
    "cell_type": "code",
    "execution_count": 663,
-   "id": "entire-chinese",
+   "id": "electric-astrology",
    "metadata": {},
    "outputs": [
     {
@@ -2831,7 +2831,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "valued-message",
+   "id": "attended-beginning",
    "metadata": {},
    "source": [
     "The only wrinkle is figuring out whether this is the measure you want or not. To the extent that you care about the practical consequences of your research, you often want to measure the effect size relative to the *original* variables, not the *difference* scores (e.g., the 1\\% improvement in Dr Chico's class is pretty small when measured against the amount of between-student variation in grades), in which case you use the same versions of Cohen's $d$ that you would use for a Student or Welch test. For instance, when we do that for Dr Chico's class, "
@@ -2840,7 +2840,7 @@
   {
    "cell_type": "code",
    "execution_count": 666,
-   "id": "civic-female",
+   "id": "parental-terrain",
    "metadata": {},
    "outputs": [
     {
@@ -2874,7 +2874,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "united-timber",
+   "id": "written-aquarium",
    "metadata": {},
    "source": [
     "what we see is that the overall effect size is quite small, when assessed on the scale of the original variables."
@@ -2882,7 +2882,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "necessary-coffee",
+   "id": "amended-genetics",
    "metadata": {},
    "source": [
     "(shapiro)=\n",
@@ -2893,7 +2893,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "greek-noise",
+   "id": "humanitarian-agenda",
    "metadata": {},
    "source": [
     "### QQ plots\n",
@@ -2903,14 +2903,26 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 687,
-   "id": "international-worst",
+   "execution_count": 706,
+   "id": "vulnerable-spanish",
    "metadata": {
     "tags": [
      "hide-input"
     ]
    },
    "outputs": [
+    {
+     "data": {
+      "application/papermill.record/text/plain": ""
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qq_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
     {
      "data": {
       "image/png": "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\n",
@@ -2936,12 +2948,13 @@
     "\n",
     "qq = probplot(normal_data, dist=\"norm\", plot = ax1)\n",
     "hist = sns.histplot(normal_data, axes=ax2)\n",
-    "\n"
+    "\n",
+    "glue(\"qq_fig\", ax, display=False)"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "cordless-dining",
+   "id": "welsh-bench",
    "metadata": {},
    "source": [
     " ```{glue:figure} qq_fig\n",
@@ -2954,118 +2967,235 @@
   },
   {
    "cell_type": "markdown",
-   "id": "municipal-mongolia",
+   "id": "primary-argument",
    "metadata": {},
    "source": [
     "And the results are shown in {numref}(`fig-qq`), above.\n",
     "\n",
-    "As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the two data sets shown in {numref}(qqskew), which show the histogram and a QQ plot for a data sets that are highly skewed the QQ plot curve upwards. Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) show the same plots for a heavy tailed (i.e., high kurtosis) data set: in this case, the QQ plot flattens in the middle and curves sharply at either end."
+    "As you can see, these data form a pretty straight line; which is no surprise given that we sampled them from a normal distribution! In contrast, have a look at the data shown in {numref}(qqskew) and {numref}(qqheavy), which show the histogram and a QQ plot for a data sets that are highly skewed and have a heavy tail (i.e., high kurtosis), respectively."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 699,
-   "id": "beneficial-perth",
-   "metadata": {},
+   "execution_count": 704,
+   "id": "macro-serum",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
    "outputs": [
     {
      "data": {
+      "image/png": "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\n",
       "text/plain": [
-       "7.714999391855671e-20"
+       "
"
       ]
      },
-     "execution_count": 699,
-     "metadata": {},
-     "output_type": "execute_result"
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
     }
    ],
    "source": [
-    "import numpy as np\n",
-    "import scipy.special as sps  \n",
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/skewed_data.csv\")\n",
+    "\n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
+    "\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
     "\n",
-    "shape = 2\n",
-    "scale = 2\n",
-    "s = np.random.gamma(shape, scale, 1000)\n",
-    "count = s\n",
-    "bins = 100\n",
-    "data = bins**(shape-1)*(np.exp(-bins/scale) / sps.gamma(shape)*scale**shape)\n"
+    "glue(\"qqskew_fig\", ax, display=False)"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "lucky-business",
+   "id": "gross-filling",
    "metadata": {},
    "source": [
-    "```{r qq2a, fig.cap=\"A histogram of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  skewed.data <- rgamma( n = 100,1,2 ) \n",
-    "  plotOne( skewed.data, \"Skewed Data\" )\n",
-    "```\n",
+    " ```{glue:figure} qqskew_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqskew\n",
     "\n",
-    "```{r qq2b, fig.cap=\"A normal QQ plot of the 100 observations in a `skewed.data` set\", echo=FALSE}\n",
-    "  plotTwo( skewed.data )\n",
-    "```\n",
+    "The skewness of these data of 100 observations is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
+    "\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 707,
+   "id": "emotional-franchise",
+   "metadata": {
+    "tags": [
+     "hide-input"
+    ]
+   },
+   "outputs": [
+    {
+     "data": {
+      "application/papermill.record/text/plain": ""
+     },
+     "metadata": {
+      "scrapbook": {
+       "mime_prefix": "application/papermill.record/",
+       "name": "qqheavy_fig"
+      }
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "
"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"https://raw.githubusercontent.com/ethanweed/pythonbook/main/Data/heavy_tailed_data.csv\")\n",
     "\n",
-    "The skewness of the data in Figures \\@ref(fig:qq2a) and \\@ref(fig:qq2b) is 1.94, and is reflected in a QQ plot that curves upwards. As a consequence, the Shapiro-Wilk statistic is $W=.80$, reflecting a significant departure from normality ($p<.001$). \n",
+    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 5))\n",
     "\n",
-    "```{r qq2c, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  heavy.tailed.data <- c(rnorm( n = 80 ) , rnorm( 20, sd=5))\n",
-    "  plotOne( heavy.tailed.data, \"Heavy-Tailed Data\" )\n",
-    "```\n",
+    "qq = probplot(df['data'], dist=\"norm\", plot = ax1)\n",
+    "hist = sns.histplot(df['data'], axes=ax2)\n",
     "\n",
-    "```{r qq2d, fig.cap=\"A histogram of the 100 observations in a *heavy tailed*` data set, again consisting of 100 observations. \", echo=FALSE}\n",
-    "  plotTwo( heavy.tailed.data )\n",
-    "```\n",
+    "glue(\"qqheavy_fig\", ax, display=False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "behavioral-chester",
+   "metadata": {},
+   "source": [
+    " ```{glue:figure} qqheavy_fig\n",
+    ":figwidth: 600px\n",
+    ":name: fig-qqheavy\n",
     "\n",
-    "Figures \\@ref(fig:qq2c) and \\@ref(fig:qq2d) shows the same plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
+    "Plots for a heavy tailed data set, again consisting of 100 observations. In this case, the heavy tails in the data produce a high kurtosis (2.80), and cause the QQ plot to flatten in the middle, and curve away sharply on either side. The resulting Shapiro-Wilk statistic is $W = .93$, again reflecting significant non-normality ($p < .001$).\n",
     "\n",
-    "One way to check whether a sample violates the normality assumption is to draw a **_\"quantile-quantile\" plot_** (QQ plot). This allows you to visually check whether you're seeing any systematic violations. In a QQ plot, each observation is plotted as a single dot. The x co-ordinate is the theoretical quantile that the observation should fall in, if the data were normally distributed (with mean and variance estimated from the sample) and on the y co-ordinate is the actual quantile of the data within the sample. If the data are normal, the dots should form a straight line. For instance, lets see what happens if we generate data by sampling from a normal distribution, and then drawing a QQ plot using the R function `qqnorm()`. The `qqnorm()` function has a few arguments, but the only one we really need to care about here is `y`, a vector specifying the data whose normality we're interested in checking. Here's the R commands:\n",
-    "```{r}\n",
-    "normal.data <- rnorm( n = 100 )  # generate N = 100 normally distributed numbers\n",
-    "hist( x = normal.data )          # draw a histogram of these numbers\n",
-    "qqnorm( y = normal.data )        # draw the QQ plot\n",
     "```"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "bronze-shore",
+   "id": "loving-regard",
    "metadata": {},
    "source": [
     "### Shapiro-Wilk tests\n",
     "\n",
-    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.^[Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality; although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests; in R it's implemented via the `ks.test()` function.] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "Although QQ plots provide a nice way to informally check the normality of your data, sometimes you'll want to do something a bit more formal. And when that moment comes, the **_Shapiro-Wilk test_** [@Shapiro1965] is probably what you're looking for.[^note14] As you'd expect, the null hypothesis being tested is that a set of $N$ observations is normally distributed. The test statistic that it calculates is conventionally denoted as $W$, and it's calculated as follows. First, we sort the observations in order of increasing size, and let $X_1$ be the smallest value in the sample, $X_2$ be the second smallest and so on. Then the value of $W$ is given by\n",
+    "\n",
     "$$\n",
     "W = \\frac{ \\left( \\sum_{i = 1}^N a_i X_i \\right)^2 }{ \\sum_{i = 1}^N (X_i - \\bar{X})^2}\n",
     "$$\n",
-    "where $\\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. \n",
     "\n",
-    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. \n",
+    "where $\\bar{X}$ is the mean of the observations, and the $a_i$ values are ... mumble, mumble ... something complicated that is a bit beyond the scope of an introductory text. \n",
     "\n",
-    "```{r swdist, fig.cap=\"Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\", echo=FALSE}\n",
-    "knitr::include_graphics(file.path(projecthome, \"img/ttest2/shapirowilkdist.png\"))\n",
+    "Because it's a little hard to explain the maths behind the $W$ statistic, a better idea is to give a broad brush description of how it behaves. Unlike most of the test statistics that we'll encounter in this book, it's actually *small* values of $W$ that indicated departure from normality. The $W$ statistic has a maximum value of 1, which arises when the data look \"perfectly normal\". The smaller the value of $W$, the less normal the data are. However, the sampling distribution for $W$  -- which is not one of the standard ones that I discussed in Chapter \\@ref(probability) and is in fact a complete pain in the arse to work with -- does depend on the sample size $N$. To give you a feel for what these sampling distributions look like, I've plotted three of them in Figure \\@ref(fig:swdist). Notice that, as the sample size starts to get large, the sampling distribution becomes very tightly clumped up near $W=1$, and as a consequence, for larger samples $W$ doesn't have to be very much smaller than 1 in order for the test to be significant. "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "overhead-python",
+   "metadata": {},
+   "source": [
+    "```{figure} ../img/ttest2/shapirowilkdist.png\n",
+    ":name: fig-shapirowilk\n",
+    ":width: 600px\n",
+    ":align: center\n",
     "\n",
-    "```\n",
+    "Sampling distribution of the Shapiro-Wilk $W$ statistic, under the null hypothesis that the data are normally distributed, for samples of size 10, 20 and 50. Note that *small* values of $W$ indicate departure from normality.\n",
     "\n",
-    "To run the test in R, we use the `shapiro.test()` function. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal.data`, we get the following:\n",
-    "```{r}\n",
-    "shapiro.test( x = normal.data )\n",
-    "```\n",
-    "So, not surprisingly, we have no evidence that these data depart from normality. When reporting the results for a Shapiro-Wilk test, you should (as usual) make sure to include the test statistic $W$ and the $p$ value, though given that the sampling distribution depends so heavily on $N$ it would probably be a politeness to include $N$ as well.\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "federal-jenny",
+   "metadata": {},
+   "source": [
+    "To run the test in Python, we use the `scipy.stats.shapiro` method. It has only a single argument `x`, which is a numeric vector containing the data whose normality needs to be tested. For example, when we apply this function to  our `normal_data`, we get the following:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 708,
+   "id": "restricted-baseball",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "ShapiroResult(statistic=0.9898834228515625, pvalue=0.6551706790924072)"
+      ]
+     },
+     "execution_count": 708,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from scipy.stats import shapiro\n",
     "\n",
-    "## Testing non-normal data with Wilcoxon tests{#wilcox}\n",
+    "shapiro(normal_data)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "touched-mailing",
+   "metadata": {},
+   "source": [
+    "So, not surprisingly, we have no evidence that these data depart from normality. When reporting the results for a Shapiro-Wilk test, you should (as usual) make sure to include the test statistic $W$ and the $p$ value, though given that the sampling distribution depends so heavily on $N$ it would probably be a politeness to include $N$ as well."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "executed-sierra",
+   "metadata": {},
+   "source": [
+    "(wilcox)=\n",
+    "## Testing non-normal data with Wilcoxon tests\n",
     "\n",
     "Okay, suppose your data turn out to be pretty substantially non-normal, but you still want to run something like a $t$-test? This situation occurs a lot in real life: for the AFL winning margins data, for instance, the Shapiro-Wilk test made it very clear that the normality assumption is violated. This is the situation where you want to use Wilcoxon tests. \n",
     "\n",
-    "Like the $t$-test, the Wilcoxon test comes in two forms, one-sample and two-sample, and they're used in more or less the exact same situations as the corresponding $t$-tests. Unlike the $t$-test, the Wilcoxon test doesn't assume normality, which is nice. In fact, they don't make any assumptions about what kind of distribution is involved: in statistical jargon, this makes them **_nonparametric tests_**. While avoiding the normality assumption is nice, there's a drawback: the Wilcoxon test is usually less powerful than the $t$-test (i.e., higher Type II error rate). I won't discuss the Wilcoxon tests in as much detail as the $t$-tests, but I'll give you a brief overview.\n",
-    "\n",
-    "\n",
+    "Like the $t$-test, the Wilcoxon test comes in two forms, one-sample and two-sample, and they're used in more or less the exact same situations as the corresponding $t$-tests. Unlike the $t$-test, the Wilcoxon test doesn't assume normality, which is nice. In fact, they don't make any assumptions about what kind of distribution is involved: in statistical jargon, this makes them **_nonparametric tests_**. While avoiding the normality assumption is nice, there's a drawback: the Wilcoxon test is usually less powerful than the $t$-test (i.e., higher Type II error rate). I won't discuss the Wilcoxon tests in as much detail as the $t$-tests, but I'll give you a brief overview."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "driven-plain",
+   "metadata": {},
+   "source": [
     "### Two sample Wilcoxon test\n",
     "\n",
-    "I'll start by describing the **_two sample Wilcoxon test_** (also known as the Mann-Whitney test), since it's actually simpler than the one sample version. Suppose we're looking at the scores of 10 people on some test. Since my imagination has now failed me completely, let's pretend it's a \"test of awesomeness\", and there are two groups of people, \"A\" and \"B\". I'm curious to know which group is more awesome. The data are included in the file `awesome.Rdata`, and like many of the data sets I've been using, it contains only a single data frame, in this case called `awesome`. Here's the data:\n",
+    "I'll start by describing the **_two sample Wilcoxon test_** (also known as the Mann-Whitney test), since it's actually simpler than the one sample version. Suppose we're looking at the scores of 10 people on some test. Since my imagination has now failed me completely, let's pretend it's a \"test of awesomeness\", and there are two groups of people, \"A\" and \"B\". I'm curious to know which group is more awesome. The data are included in the file `awesome.Rdata`, and like many of the data sets I've been using, it contains only a single data frame, in this case called `awesome`. Here's the data:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "unsigned-finger",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "markdown",
+   "id": "smart-vacation",
+   "metadata": {},
+   "source": [
     "```{r}\n",
     "load(file.path(projecthome, \"data/awesome.Rdata\"))\n",
     "print( awesome )\n",
     "```\n",
+    "\n",
     "As long as there are no ties (i.e., people with the exact same awesomeness score), then the test that we want to do is surprisingly simple. All we have to do is construct a table that compares every observation in group $A$ against every observation in group $B$. Whenever the group $A$ datum is larger, we place a check mark in the table:\n",
     "\n",
     "```{r echo=FALSE}\n",
@@ -3163,7 +3293,9 @@
     "\n",
     "[^note12]: This is obviously a class being taught at a very small or very expensive university, or else is a postgraduate class. *I've* never taught an intro stats class with less than 350 students.\n",
     "\n",
-    "[^note13]: This is a massive oversimplification."
+    "[^note13]: This is a massive oversimplification.\n",
+    "\n",
+    "[^note14]: Either that, or the Kolmogorov-Smirnov test, which is probably more traditional than the Shapiro-Wilk, though most things I've read seem to suggest Shapiro-Wilk is the better test of normality, although Kolomogorov-Smirnov is a general purpose test of distributional equivalence, so it can be adapted to handle other kinds of distribution tests."
    ]
   }
  ],
diff --git a/Data/awesome.csv b/Data/awesome.csv
new file mode 100644
index 00000000..cb26bb82
--- /dev/null
+++ b/Data/awesome.csv
@@ -0,0 +1,11 @@
+scores,group
+6.4,A
+10.7,A
+11.9,A
+7.3,A
+10,A
+14.5,B
+10.4,B
+12.9,B
+11.7,B
+13,B
diff --git a/Data/awesome2.csv b/Data/awesome2.csv
new file mode 100644
index 00000000..8990a224
--- /dev/null
+++ b/Data/awesome2.csv
@@ -0,0 +1,6 @@
+score.A,score_B
+6.4,14.5
+10.7,10.4
+11.9,12.9
+7.3,11.7
+10,13
diff --git a/img/.DS_Store b/img/.DS_Store
index a08e268f..d5d7def2 100644
Binary files a/img/.DS_Store and b/img/.DS_Store differ
diff --git a/img/ttest2/shapirowilkdist.png b/img/ttest2/shapirowilkdist.png
new file mode 100644
index 00000000..c47849c1
Binary files /dev/null and b/img/ttest2/shapirowilkdist.png differ