In November 2019, the Committee already found that Colzato had acted in violation of academic integrity. The list of misconducts was long: tampering with data and grant applications, fabricated experiments, missing test subjects, co-author names that were omitted and blood samples that were illegally taken.

The fraud was brought to light thanks to three whistle-blowers who raised the alarm with the scientific director, who then filed a complaint with the Executive Board. The case was then forwarded to the Committee for Academic Integrity. In the meantime, Colzato had already left Leiden.

Before this news came out at the end of 2019, the Committee had already given the advice to withdraw two of the psychologist’s scientific articles. The Executive Board then instructed the Committee to investigate all of her work for fraud.

OMITTING DATA

For this investigation, the Faculty of Social and Behavioural Sciences provided a list of 174 publications, out of which 53 were actually examined. These 53 were all published after 2015 and the data collection had taken place in Leiden. It would involve a ‘disproportionate effort’ to obtain the required data for the older articles.

The Committee concludes that at least 15 of these 53 publications involved fraud. This includes changes in the research design, the subsequent addition of a control group or the omission of data without justification or notice.

In seven of the fifteen articles, the manipulations are of ‘such a serious nature that it is recommended that the journals concerned are asked to retract them’ because, due to the fraud, the conclusions are incorrect or likely to be so. As the report has been anonymised, it remains unclear which articles are concerned and who are involved as co-authors.

NON-ADHERENCE TO RESEARCH DESIGN

The Committee does, however, explain for each article how it came to its conclusion. For example, in one of the publications, 21 test subjects were excluded. The researcher thought they were unsuitable test subjects because they ‘did not show a positive association as a group’. In another article, the reported research design was ‘substantially different from the one in the protocol’, a control group was added and important data and test subjects were omitted.

In a third article, data from two of the studied groups were omitted from the publication because they ‘showed a different result than presumed’. Thus, 49 test subjects were excluded. In two other publications, 16 and 12 test subjects were excluded, respectively. In the last two articles, the research design was not adhered to and seven and sixteen test subjects disappeared, respectively.

In the other eight articles, malpractices undermined the conclusions, the Committee states. However, the manipulations ‘are of such a nature that the editorial board of the journals should be notified, so that it can decide how the readers are to be informed, or whether the articles should be retracted after all’.

In most of these publications, the fraud is mainly due to the omission of test subjects (varying from two to twelve). For one of the articles, it is not possible to determine how many test subjects were excluded, but it is clear that the original research design was not adhered to, half of the measurements were removed and ‘more genetic variations were measured than reported’.

For 27 articles, no further investigation was possible because of a lack of additional information and data sets or because there was no direct reason for an investigation based on the information provided by the Faculty of Social and Behavioural Sciences. There were no indications of fraud in the other articles.

FALSE RESULTS

The Committee also examined a third grant application, after an earlier investigation revealed that Colzato had presented the results of research that had never been carried out as preliminary results in two research proposals submitted to the Dutch Research Council, a science funding agency. The Committee did not find any evidence of malpractice in the third application, but did advise the Executive Board to inform the Research Council of its findings.

The Committee attributes two causes to the fact that the report took so long to complete. For one thing, the required sources had not been systematically archived, which meant that data files could only be obtained via individual employees. Furthermore, access to the Institute of Psychology was limited during the pandemic.

The Executive Board will adopt the recommendations. This became clear earlier this week when a list of decisions was shared with the University Council, in which the issue was briefly touched upon. The Board requests the publishers of the fifteen fraudulent articles to retract the publications and informs the Research Council about the questionable grant applications.

NO PUBLICITY

The co-authors who were involved have also been informed. One of them is assistant professor Laura Steenbergen, who is also one of the three whistle-blowers. On 12 November, she received the non-anonymised advice with an accompanying letter from the Executive Board’s lawyer. ‘It was totally unexpected,’ she says. ‘During the summer, I had asked the Committee if there was a timeline available so I would have some idea of when to expect something, but I never received a reply.’

A week later, the document was quietly and anonymously published on the university’s website, as Steenbergen knows. ‘I checked that myself. I was refreshing the page every hour. Apparently, the university thought it was being transparent by putting an anonymous report on the internet and that was that. Even internally, at the Institute of Psychology, there was no publicity. Many of my colleagues don’t even know about it.’

She is not allowed to reveal which articles have been identified as fraudulent. ‘That is strictly confidential, I’ve been told.’ She does say, however, that fourteen of the fifteen articles in which Colzato committed fraud involved one of the three whistle-blowers - Roberta Sellaro, Bryant Jongkees and herself.

NO RETRACTIONS YET

She is a co-author on six of the fifteen publications. ‘Only one of these is currently being investigated,’ says Steenbergen. ‘None of the articles have been retracted yet. And those other two articles, which were already recommended for retraction in 2019, were not retracted until November 2020 and January 2021. So, I don’t think things will go much faster this time.’

Chairman of the Committee for Academic Integrity Frits Rosendaal explains that retracting publications often takes a long time. ‘Firstly, because the journals responsible also want to examine them carefully. And secondly, because some journals unfortunately consider this to be a disgrace and delay the whole thing.’

Lorenza Colzato did not respond to a request for comment. After the first Committee’s first investigation in 2019, she informed Mare that she considers ‘good research practices and academic integrity to be very important’ and always ‘acts accordingly’. According to her, there had been ‘no proper investigation or fair trial in any way’.

In the present report, the Committee writes that the findings have been presented to Colzato. She ‘did not present any substantive arguments’ in her response. Therefore, the advice remained unchanged.

(Written with cooperation of Vincent Bongers)