1. Look closely at the Table 1 in a paper or all descriptive stats: Are the distributions what you expected? If not, see if you can figure out why. Maybe theres a methods gap or limitation the authors & reviewers didn’t note. #RobinsonTwitterTakeover 2/n
Or maybe the final conclusions don’t even fit with the descriptive data (see J. Ward, @danydoodledo, et al. @AnnalsOfEpi) maybe you could do better. #RobinsonTwitterTakeover 3/n
Or maybe you’ll see something interesting - a sex difference, a strangely high prevalence of exposure, could be anything - that makes you think, I wonder if anyone ever looked at the data *this* way? #RobinsonTwitterTakeover 4/n
The moment I saw a descriptive table re: Add Health data in a paper by @P_Gordon_Larsen, I knew that my dissertation would work. #RobinsonTwitterTakeover 5/n
• • •
Missing some Tweet in this thread? You can try to
force a refresh
