Social psychology has recently gained considerable attention in the mass media - not only because of its findings but also because of research fraud. In the Netherlands the highly respected Social Psychologist Diederik Stapel admitted to invent data that were subsequently used for publication in the leading Journals in the discipline as well as in “Science”. Likewise in the Netherlands the work of Jens Förster (formerly Jacobs University Bremen) has been subject to investigation.

On behalf of the 'Committee Levelt' I examined 46 papers of Stapel and six dissertations to detect evidence of fraud and questionable research practices. I will present the framework of the investigation and some procedures that were particularly useful in the Stapel case, but were also helpful in other cases. The framework and procedures boil down to finding inconsistencies in the pairs 'research materials-data', 'research materials-article', 'data-article', or detecting extremely unlikely patterns in the data. The framework and procedure is illustrated with some examples of the Stapel case; the Förster case is discussed as well.

I will focus on the implications of my research misconduct investigation for (i) investigations of research misconduct in general, (ii) incentives and behavior of scientists, (iii) researchers' (mis)understanding of statistics.