‘Data detective’ Uri Simonsohn has posted a paper on the Social Science Research Network describing the statistical method he used to expose suspicious data in the work of social psychologists Dirk Smeesters and Lawrence Sanna.
Smeesters resigned from Erasmus University in Rotterdam, the Netherlands, last month after an investigation found that he had massaged data to produce positive outcomes in his research, such as the effect of colour on consumer behaviour (see ‘The data detective‘). In May, Sanna resigned from the University of Michigan in Ann Arbor for unknown reasons. But Simonsohn had also found odd statistical patterns in his data, and Sanna has asked the Journal of Experimental Social Psychology to retract three of his papers (see ‘Uncertainty shrouds psychologist’s resignation‘).
While the string of high-profile fraud cases in psychology, including that of Diederik Stapel, has some worried that psychology will earn a reputation for dishonesty, Simonsohn has previously told Nature that he thinks that is unwarranted, pointing out that at least in psychology, they are trying to do something about the problem (see ‘Replication studies: Bad copy’).
Report this comment
I’ve posted a few thoughts about Simonsohn’s paper.
The first highlights the series of steps that he took to avoid making a false accusation (https://www.davenussbaum.com/simonsohns-fraud-detection-technique-revealed/)
The second is a non-technical explanation of why Simonsohn’s analysis suggests that Smeesters’ data was not generated by real subjects (https://www.davenussbaum.com/just-post-it-update/)