search

egap / methods-guides

EGAP methods guides
https://egap.org/methods-guides/
21 stars 12 forks source link

Hypothesis testing #57

Closed amwilk closed 2 years ago

amwilk commented 2 years ago

@jwbowers @malisi I made some small updates here, but for some reason I cannot compile the Rmd (even though I pulled the newest version of the Rmd file today before making changes). So, I only made changes in the Rmd file but did not compile it. Based on your previous troubleshooting, do you have ideas on how I can get this to compile? Three other thoughts: 1) In the version that is currently online, the equation label {#eq:identity} seems broken. Because I couldn't compile, I didn't fix that but am happy to take another look at this after that problem is solved. 2) I made some updates to the sections on one- and two-tailed p-values. I wasn't quite sure why the example calculates the one-tailed p-values as the upper p-value (for the diff-in-means test statistic) but as the minimum between the lower and upper p-values (for the diff-in-ranks test statistic). It may be worth clarifying this in the text. 3) Also it may be worth mentioning that one-tailed p-values correspond to a hypothesis of the form "the treatment effect is no larger than X" against an alternative of "the treatment effect is larger than X."

I am happy to make any additional changes along these lines if you would like me to @jwbowers - these were just my thoughts. Thanks!

amwilk commented 2 years ago

@jwbowers @malisi I manage to compile this and fix the equation. So, this is ready for review. Thanks!