I. Issues with evaluation forms

It is early December 2018. The Programme Committee of the Master's Statistical Science for the Life and Behavioural Sciences is discussing course evaluations when one of the student members of the committee notices something odd.

According to the summary of the evaluations, no student has given Marta Fiocco, one of the teachers of the Statistics and Probability course, a grade lower than seven out of ten. But the student is certain that she only gave the teacher a six.

She knows that some classmates have given lower grades as well. Comments made by students on their evaluation forms are not mentioned either. In fact, there isn't any comment present in the evaluation summary of the Statistics and Probability course, even thoug every other course discussed by the commission has at least some comments.

Odd Man Out

This is enough for the committee to organise an investigation. The committee members look the original forms in the archive. The originals they find correspond to the summary. All negative reviews are missing.

Statisticians would not be statisticians if they didn’t want to analyse the data. Given that the number of comments for Statistics and Probability is remarkably lower than in other subjects, two teachers on the committee decide to compare the number of comments with previous years and other courses.

Their conclusion is clear. On average, one-third to two-thirds of students leave a comment on an average evaluation form, either positive or negative. With a total of zero comments on 35 forms, Statistics and Probability is a notable outlier. The number of comments in evaluations from previous years is also very low.

With this observation, committee members decide to contact the course teachers and ask for an explanation. Two of them declare that they never had anything to do with the evaluations and refer to a third teacher who handled the forms. That third teacher is Marta Fiocco, who happens to be the person whose grades are being looked into.

Missing 

Fiocco says she has nothing to do with missing forms but offers an explanation. The exam during which evaluations filled out took place in two different rooms at the same time. According to Fiocco, the forms from one of the two rooms might have gone missing.

Soon enough, this turns out to be incorrect. Two students whose evaluation forms are absent, were sitting in different rooms. Both their forms are missing.

On December 18, the programme committee, together with the programme director Jacqueline Meulman, share their findings with Aad van der Vaart, the then scientific director of the Mathematical Institute. He further the issue further and finds that on some forms, grades have been changed.

Suspicion

A month later, the programme committee organizes a new evaluation of Statistics and Probability. To avoid raising suspicion among students, they are told there was a technological error with the previous forms, and the evaluation must be repeated. Although not all students fill out the form again, the commission considers it a valid evaluation.

For the most part, this new evaluation is identical to the first, except for one point: the grade that the students gave to Marta Fiocco has changed. In the first evaluation, Fiocco did not receive a grade lower than 7 but now, only very few students have given her even a passing grade. On top of that, 15 comments have been made, some of which explicitly criticize Fiocco’s didactic skills.

The Programme Committee concldudes its investigation and defers further actions to the scientific director Van der Vaart, who lays down his findings and forwards them to dean Geert De Snoo. And while Van der Vaart's report addresses the irregularities surrounding the evaluation forms in detail, the question as to how the negative evaluation forms disappeared remains open.

II. A “cowboy-investigation” becomes a divisive issue

In the months following the report, the investigation itno the evaluation forms grows into a divisive issue within the Mathematical Institute. Although the case remains small at the beginning of 2019, matters escalate when Aad van der Vaart completes his term as scientific director and is succeeded by professor of pure mathematics Peter Stevenhagen in September.

Stevenhagen was already involved in the investigation of the evaluation forms before taking office. In the wake of the investigation, Stevenhagen presents himself as a confidential counsellor in the matter.

Although several committee members trust him initially, they later discover Stevenhagen has assumed this role of his own accord. He does not act on behalf of the Institute or the Faculty Board, nor is he impartial. Any information he gathers, he passes on to Fiocco.

Lies

The fact that Stevenhagen had chosen a side in the issue becomes clear when in October 2019, shortly after becoming the scientific director, he sends a letter to the programme committee and the oversight committee. In the letter, which Mare has been able to obtain, he accuses members of the programme committee of being biased against Fiocco. Stevenhagen calls for the replacement of the student members in the committee and for the appointment of 'a new chairperson who has our confidence.

A remarkable request, because the composition of the programme committee does not fall within the purview of one scientific director but rather, is decided jointly with the Faculty Board and all the institutes that are involved in the programme. One of the committee members says that the letter is ‘full of lies.’

Three weeks later, Stevenhagen withdraws the letter, but in another letter he had written to dean Geert de Snoo before becoming scientific director, he expands on why he thinks the committee is investigating Fiocco. In that letter, also in possession of Mare, he disparages the investigation and calls it a ‘cowboy investigation’.

Disinformation

Stevenhagen writes that students are egging each other on in WhatsApp groups to criticise Fiocco, creating a ‘kill Marta’ atmosphere. This is partly due to their background, he tells the dean.

Quote: ‘The Statistical Science students in 2019 are no longer the Dutch boys and girls I taught in 1990, who filled out evaluation forms and sometimes came up with very reasonable comments that I could use. We now have Chinese, who are not allowed to criticize a teacher at home, and think here they can blame a teacher for not being in her office often enough. Or Russians, who come from a society where spreading disinformation is a way of life and believe it is not reprehensible.’

One of the students who pointed out the irregularities in the programme committee is Russian.

Not Public

In the autumn of 2019, the faculty completes its own investigation over the evaluation forms. An HR assistant and an external lawyer review the documents and interview several students. The faculty does not disclose the result of the investigation, nor does it share it with the programme committee, to the great frustration of the members who criticize the lack of transparency.

In a joint mail to the Faculty Board, the programme committee requests to see the results: ‘the lack of a firm conclusion on [who the culprit is] does not diminish the importance [of whether fraud has been committed] for us. By analogy, if the identity of the burglar cannot be established, it does not mean that a burglary did not occur.’

Whistle-blower committee

Ultimately, the programme committee has to contact a special whistle-blower committee before the faculty board is willing to share any information. Vice-dean Bart de Smit sends a letter containing the conclusion of the investigation to the programme committee, but any details about persons involved in the affair remain confidential.

He writes that the Faculty Board draws three conclusions. Firstly, that irregularities could be observed in the first evaluation, due to which it was considered invalid. However, because there also are ‘justified doubts’ about the second evaluation organised by the programme committee, that is considered invalid as well. Secondly, the Faculty Board states that it is necessary to improve the evaluation procedure and that digital evaluations will be used in future. Finally, he informs that ‘the investigation has not been able to determine the origin of the irregularities in the forms’ with certainty. No further steps will be taken.

Under the leadership of Stevenhagen meanwhile, the relationship between the management and some professors at the Mathematical Institute, and the statistics team have worsened. As a scientific director, Stevenhagen openly criticizes decisions concerning promotions and appointments made by his predecessor and opposes an already made job offer.

Previously, Aad van der Vaart had made agreements with various statisticians about appointments to new positions. This kind of informal agreement was common within the institute. Details of an appointment could be formally discussed later, but the agreement was always honoured.

Disproportionately severe blows

Several staff members had recently received Veni- and Vidi-grants. It was expected that with these grants, they would receive a appointment as assistant professor or promotion to associate professor. Other research groups from outside Leiden had offered them similar positions.

Under Stevenhagen however, the standing agreements were not upheld. Contracts for new appointments were left unsigned. Promised promotions postponed.

A scientific director must sometimes make unpopular decisions, but the statisticians think their research group is bearing an unfair portion of the brunt. Some call it revenge for the investigation on evaluation forms. Doubt starts to form whether Leiden is still the right place for them.

III. Profound discord and inevitable departures

What began as an investigation into shady evaluations forms has grown into a major affair by the spring of 2020, with a divide running right through the Mathematical Institute as a result. The statistics team has lost all trust in the management team and the relationships within the institute have been so strained that the faculty feels obliged to appoint a mediator to bridge the gap between different employees.

In total, nine employees are invited to a voluntary and confidential interview with an external mediator. On March 21, Dean Michiel Kreutzer goes to the Snellius building to speak to staff members of the institute and assure them that the faculty is taking the issue seriously.

Failed Mediation

The medation does not go as planned. When dean Kreutzer puts a halt on all new contracts, which results in a young researcher losing her position, several statisticians withdraw from the mediation.

One month after his speech at the Snellius, the Dean sends a mail saying that the ‘discussions have led to the conclusion that it has not been possible to find a solution that is satisfactory for all parties.'

On May 26 2020, Kreutzer holds a last meeting with Stevenhagen and Van der Vaart, to whom he makes a proposal: both must step down from executive positions. On June 1st, Stevenhagen announces that he relinquishes his position as scientific director. Both Van der Vaart and Stevenhagen are prohibited from entering the Mathematical Institute and from contacting colleagues.

The faculty board appoints Frank van der Duijn Schouten as temporary scientific director. The former rector of the Vrije Universiteit is regarded as an experienced executive who has been acting ad interim in various roles since his retirement.

An exodus can no longer be prevented, however. More and more statisticians start to leave. Due to the uncertainty around postponed appointments and the toxic work environment, their Mathematical Institute is no longer a safe haven. Finding a new position is not difficult, skilled statisticians are highly sought after.

At least six employees leave because of the affair, which amounts to roughly three-quarters of the permanent staff (excluding emeriti, postdocs and PhDs). Some transfer to another faculty, but most go to other universities. PhDs move with them.

This month, Aad van der Vaart left the university. Six years ago, he won the Spinoza Prize, the biggest Dutch scientific award. At the time he said in Mare about the award of €2.5 million: ‘What will I do with the money? That's going towards hiring staff. We’re developing mathematics here, and you need people to do that.’ He now holds a new position as a professor at TU Delft.

The statisticians suspect that the unexpected departure of Kreutzer is also a result of the affair. At the end of March, it was announced that he would step down as Dean as of May 1st, without knowing yet who would succeed him. The university refutes that affair at the Mathematical Institute has anything to do with Kreutzer’s departure.

Eternal Shame

What remains is a shadow of what the Leiden statistics group once was. According to emeritus professor Richard Gill, it is an ‘eternal shame that such a good statistics team, one of the best in Europe, has now completely fallen apart. I don’t think people realise the damage this has caused to the reputation of the university both at home and abroad.’

Professor Johannes Schmidt-Hieber also laments the course of events. ‘It was by far the best group of theoretical statisticians in the Netherlands. Many of the researchers working there had received grants and awards, it is a disaster for the institute that they have left.’

Schmidt-Hieber has been professor at the Mathematical Institute since November 2020, but has already decided to return to the University of Twente. With the arrival of interim scientific director Van der Duijn Schouten, he hoped that the toxic work environment would disappear. It turned out he was mistaken.

Undesirable

‘When he was informed many statisticians wanted to leave, he barely did anything', says Schmidt-Hieber. ‘Apparently, it fit within his plan to re-structure the institute. When I informed him that I was in negotiations to return to the University of Twente, he announced my departure in the newsletter before it was official. He later called the statisticians’ departure “undesirable, but inevitable"’

This month, the Science Faculty contacted the staff who had quit. The objective is to organise an ‘exit’ conversation. The invitation does not explain why these conversations have only been scheduled now, despite many researchers leaving over a year ago.

Only two researchers from the statistics team of early 2019 are still working at the institute. One of them is Marta Fiocco. She has been appointed professor of applied mathematics on April 1st.