Simonsohn, Uri (in press) "Just post it: The lesson from two cases of fabricated data detected by statistics alone", Psychological Science


first posted:2013 02 27
last update: 2013 02 27

SAS code behind simulations and bootstraps

Raw data (includes my replication of the t-shirts study)

Get the paper in SSRN

Terms of use:
Use the data for whatever you'd like.
Inform me first of any errors you encounter. Paper is still in press, any errors can still be corrected.

Author’s note (in paper):
I thank Leif Nelson and Joe Simmons for invaluable support and contributions to this project from beginning to end. Nick Epley, Eric Johnson (Columbia), Christopher Van den Bulte, Hal Pashler and Henry L. Roediger, III helped dramatically improve the writing and positioning of the paper. Conversations with Rolf Zwaan, who headed the committee investigating Smeesters, proved instrumental for some of the analyses here presented. The co-authors of both Sanna and Smeesters were responsive and collaborative throughout, despite the obvious difficulty of confronting the possibility that a trusted friend, colleague and advisor had betrayed their trust and endangered their reputations for years.  This project did not benefit from the University of Michigan nor the University of Northern Carolina refusal to share the outcome of their investigations into possible misconduct by Lawrence Sanna.  I am exclusively responsible for all errors that remain.