I wonder if relaxing the "same data" condition (so that researchers would have to pick it) would introduce more or less variation compared to the question formulation
doesn't this just add salt to the wound in that social scientists who talk to each other and think they are in the same research subfields aren't actually even studying the same estimands
results from the Many-Economists project coming out soon; one of our steps was inspired by this paper, but we do not find that a unified research question (and design!) meaningfully improves agreement.
Comments
I wonder if relaxing the "same data" condition (so that researchers would have to pick it) would introduce more or less variation compared to the question formulation
https://pubmed.ncbi.nlm.nih.gov/29262287/
https://pmc.ncbi.nlm.nih.gov/articles/PMC5888052/
https://journals.sagepub.com/doi/10.1177/00031224211004187