-
Notifications
You must be signed in to change notification settings - Fork 92
/
Animation of DIDID.R
141 lines (129 loc) · 9.41 KB
/
Animation of DIDID.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
library(tidyverse)
library(gganimate)
library(ggthemes)
set.seed(32322)
df <- data.frame(Control = c(rep("Control",300),rep("Treatment",300)),
Time=rep(c(rep("Before",150),rep("After",150)),2),
Group = rep(c(rep("Group A",75),rep("Group B",75)),4),
GroupTime=rep(c(rep("Group A Before",75),rep("Group B Before",75),
rep("Group A After",75),rep("Group B After",75)),2)) %>%
mutate(Y = 2+.5*(Control=="Treatment")+.5*(Time=="After") +
.2*(Control=="Treatment")*(Time=="After")+
.4*(GroupTime=="Group B After" & Control == "Treatment")+
2.6*(Group=="Group B")+
rnorm(600)/4,state="1",
xaxisTime = (Time == "Before") + 2*(Time == "After") + (runif(600)-.5)*.95) %>%
group_by(Control,GroupTime) %>%
mutate(mean_Y=mean(Y)) %>%
ungroup()
df$Time <- factor(df$Time,levels=c("Before","After"))
#Create segments
dfseg <- df %>%
group_by(Control,GroupTime) %>%
summarize(mean_Y = mean(mean_Y)) %>%
ungroup()
diffA <- filter(dfseg,GroupTime=='Group A After',Control=='Control')$mean_Y[1] - filter(dfseg,GroupTime=='Group A Before',Control=='Control')$mean_Y[1]
diffB <- filter(dfseg,GroupTime=='Group B After',Control=='Control')$mean_Y[1] - filter(dfseg,GroupTime=='Group B Before',Control=='Control')$mean_Y[1]
DIDA <- filter(dfseg,GroupTime=='Group A After',Control=='Treatment')$mean_Y[1] - filter(dfseg,GroupTime=='Group A Before',Control=='Treatment')$mean_Y[1] - diffA
dffull <- rbind(
#Step 1: Raw data only
df %>% mutate(state='1. Start with raw data.'),
#Step 2: Add Y-lines
df %>% mutate(state='2. See how Y is explained by Treatment, Time and/or Group.'),
#Step 3: Collapse to means
df %>% mutate(Y = mean_Y,state="3. Keep only what's explained by Treatment, Time, and/or Group."),
#Step 4: Display time effect
df %>% mutate(Y = mean_Y,state="4. See how Control changed over Time in each Group."),
#Step 5: Shift to remove time effect
df %>% mutate(Y = mean_Y
- (Time=='After')*ifelse(Group=="Group A",diffA,diffB),
state="5. Remove the Before/After difference in each groups."),
#Step 6: Raw demeaned data only
df %>% mutate(Y = mean_Y
- (Time=='After')*ifelse(Group=="Group A",diffA,diffB),
state='6. The remaining Before/After difference is the diff-in-diff in each Group.'),
#Step 7: And get rid of the group A difference
df %>% mutate(Y = mean_Y
- (Time=='After')*ifelse(Group=="Group A",diffA,diffB) - (Time=='After' & Control == "Treatment")*DIDA,
state='7. Remove the difference-in-difference effect from Group A.'),
#Step 8: And rest
df %>% mutate(Y = mean_Y
- (Time=='After')*ifelse(Group=="Group A",diffA,diffB) - (Time=='After' & Control == "Treatment")*DIDA,
state='8. The remaining Before/After in Group B is the diff-in-diff-in-diff.')
)
p <- ggplot(dffull,aes(y=Y,x=xaxisTime,color=as.factor(Control)))+geom_point()+
guides(color=guide_legend(title="Group"))+
geom_vline(aes(xintercept=1.5),linetype='dashed')+
geom_hline(aes(yintercept=4),col='black',size=1)+
scale_color_colorblind()+
scale_x_continuous(
breaks = c(1, 2),
label = c("Before Treatment", "After Treatment")
)+xlab("Time")+
scale_y_continuous(
breaks = c(2.5, 5.5),
label = c("Y for Group A", "Y for Group B")
)+ylab("")+
theme(axis.text.y=element_text(angle=90,hjust=.5,size=14),
axis.text.x=element_text(size=14),
legend.text=element_text(size=14))+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
.5,NA),
xend=1.5,y=filter(dfseg,GroupTime=='Group A Before',Control=='Control')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group A Before',Control=='Control')$mean_Y[1]),size=1,color='black')+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
.5,NA),
xend=1.5,y=filter(dfseg,GroupTime=='Group A Before',Control=='Treatment')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group A Before',Control=='Treatment')$mean_Y[1]),size=1,color="#E69F00")+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
1.5,NA),
xend=2.5,y=filter(dfseg,GroupTime=='Group A After',Control=='Control')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group A After',Control=='Control')$mean_Y[1]),size=1,color='black')+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
1.5,NA),
xend=2.5,y=filter(dfseg,GroupTime=='Group A After',Control=='Treatment')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group A After',Control=='Treatment')$mean_Y[1]),size=1,color="#E69F00")+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
.5,NA),
xend=1.5,y=filter(dfseg,GroupTime=='Group B Before',Control=='Control')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group B Before',Control=='Control')$mean_Y[1]),size=1,color='black')+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
.5,NA),
xend=1.5,y=filter(dfseg,GroupTime=='Group B Before',Control=='Treatment')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group B Before',Control=='Treatment')$mean_Y[1]),size=1,color="#E69F00")+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
1.5,NA),
xend=2.5,y=filter(dfseg,GroupTime=='Group B After',Control=='Control')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group B After',Control=='Control')$mean_Y[1]),size=1,color='black')+
geom_segment(aes(x=ifelse(state %in% c('2. See how Y is explained by Treatment, Time and/or Group.',"3. Keep only what's explained by Treatment, Time, and/or Group."),
1.5,NA),
xend=2.5,y=filter(dfseg,GroupTime=='Group B After',Control=='Treatment')$mean_Y[1],
yend=filter(dfseg,GroupTime=='Group B After',Control=='Treatment')$mean_Y[1]),size=1,color="#E69F00")+
#And the vertical lines for effects
geom_segment(aes(x=1.5,xend=1.5,
y=ifelse(state=="4. See how Control changed over Time in each Group.",
filter(dfseg,GroupTime=='Group B Before',Control=='Control')$mean_Y[1]+diffB,
ifelse(state=="5. Remove the Before/After difference in each groups.",
filter(dfseg,GroupTime=='Group B Before',Control=='Control')$mean_Y[1],NA)),
yend=filter(dfseg,GroupTime=='Group B Before',Control=='Control')$mean_Y[1]),size=1.5,color='blue')+
geom_segment(aes(x=1.5,xend=1.5,
y=ifelse(state=="4. See how Control changed over Time in each Group.",
filter(dfseg,GroupTime=='Group A Before',Control=='Control')$mean_Y[1]+diffA,
ifelse(state=="5. Remove the Before/After difference in each groups.",
filter(dfseg,GroupTime=='Group A Before',Control=='Control')$mean_Y[1],NA)),
yend=filter(dfseg,GroupTime=='Group A Before',Control=='Control')$mean_Y[1]),size=1.5,color='blue')+
geom_segment(aes(x=1.5,xend=1.5,
y=ifelse(state=='6. The remaining Before/After difference is the diff-in-diff in each Group.',
filter(dfseg,GroupTime=='Group A After',Control=='Treatment')$mean_Y[1]-diffA,
ifelse(state=='7. Remove the difference-in-difference effect from Group A.',
filter(dfseg,GroupTime=='Group A Before',Control=='Treatment')$mean_Y[1],NA)),
yend=filter(dfseg,GroupTime=='Group A Before',Control=='Treatment')$mean_Y[1]),size=1.5,color='blue')+
geom_segment(aes(x=1.5,xend=1.5,
y=ifelse(state=='8. The remaining Before/After in Group B is the diff-in-diff-in-diff.',
filter(dfseg,GroupTime=='Group B After',Control=='Treatment')$mean_Y[1]-diffA-DIDA,NA),
yend=filter(dfseg,GroupTime=='Group B Before',Control=='Treatment')$mean_Y[1]),size=1.5,color='blue')+
labs(title = 'The Difference-in-Difference-in-Difference Effect of Treatment \n{next_state}')+
transition_states(state,transition_length=c(6,16,6,16,6,16,16,6),state_length=c(50,22,12,22,30,50,50,50),wrap=FALSE)+
ease_aes('sine-in-out')+
exit_fade()+enter_fade()
animate(p,nframes=250)