Last year a well-known Dutch psychologist, Diederik Stapel, was caught manufacturing his own data. Today a similar scandal was announced (Dutch, briefer English version), although smaller in scope, concerning the Belgian marketing professor Dirk Smeesters at Rotterdam’s Erasmus University.
According to the report (in Dutch) a researcher raised suspicions because some results were “too good to be true.” A commission was formed to replicate Smeesters’ findings. However, channeling John Lott, Smeesters had lost the raw computer data and the paper copies for all data for all of the suspicious articles due to respectively a computer crash and a residential move. Smeesters finally admitted to “massaging” the data (i.e. removing some respondents to get the results he wanted) and resigned. One of the articles that has to be withdrawn from publication was published in the prestigious Journal of Personality and Social Psychology, a favored outlet for Stapel as well.
The Dutch report highlights that Smeesters does not feel individually responsible for his actions as there is a culture in his department and the field of marketing to massage data in order to get statistically significant results. I have no way of evaluating this claim but it is a remarkable and sad story nonetheless.