Improve missingness and measurement chapter

[malcolmbarrett]

#252 introduces a foundation for discussing missingness and measurement error. I'm happy with this pedagogically, but I think there are a few areas to improve after @LucyMcGowan works on the imputation section.

• I reference the canonical m-DAGs. I show a greatly simplified version of the point of that paper, but I'm leaning towards including an appendix that has a full version that matches the paper. This way we can point people to something as a starting point.
• I mention using simulations. I think we could do that to investigate what we can recover in the touringplans example
• I want to make sure we link these ideas to when and why to use multiple imputation
• I don't want to cover bayesian methods here but I think we should shout out that you can use brms/Stan for both measurement error (using SDs) and imputation
• Show measurement error correction with multiple imputation

Some links:
https://stefvanbuuren.name/fimd/sec-true.html
https://stefvanbuuren.name/fimd/sec-prevalence.html
https://stefvanbuuren.name/fimd/sec-when.html
https://onlinelibrary.wiley.com/doi/10.1002/bimj.202200326
https://pubmed.ncbi.nlm.nih.gov/21389091/
https://github.com/moreno-betancur/missingness_DAG
https://cameronpatrick.com/post/2023/06/untangling-mar-mcar-mnar/

Also see #44