At least twelve articles and three dissertations with which former professor Diederik Stapel was involved, turn out to have been based on fake data. This is one of the initial results published by the Levelt committee, working for Tilburg University.
“Highly unlikely design of experiments”, “highly implausible results”, “too clean data”, “fabricated data”; these are just a selection of qualifications used in the initial conclusions on the articles and dissertations investigated by the Committee. The Committee was assisted by statisticians who checked the datasets for irregularities and implausibilities.
The batch concerns dissertations and articles that were published between 2007 and 2011. Six of the twelve articles came directly from the dissertations investigated. It was already known in October that Stapel supplied complete datasets to his PhD candidates. He allegedly conducted the experiments at schools, but his accounts and the data turned out to be completely fictitious. At least two of the three dissertations were based on such data. At the end of October, the board of Tilburg University stated that the candidates had acted in good faith and did not need to hand in their degrees.
The Levelt Committee, however, is deferring the answer to the question whether the Stapel's co-authors and PhD candidates are also implicated in the fraud, to its final report. Last October, the Committee stated in its primary findings that the ready-to-use datasets supplied by the former professor could have given rise to doubts at an earlier stage. There was praise for the young researchers who eventually reported their suspicions of fraud and it was concluded that they “showed more courage, alertness and investigative capabilities than incumbent professors”.
There is good news too. Eight publications with Stapel as co-author that were investigated, turned out to be fine. It was already clear in October that seven of Stapel's former PhD candidates could heave a sigh of relief. They had collected their data themselves and therefore had ‘clean’ dissertations.