Why some PhD students may decide to leave academia

November 10, 2022

TU/e researcher Andrea Kis discusses her research into the correlation between bad workplace experiences and the decisions PhD researchers make to abandon their PhD research.

A day-to-day image of international students on the TU/e campus. Photo: Bart van Overbeeke

The number of young, talented doctoral candidates who decide to leave academia, often before graduating as a PhD, is worrying. The fact that approximately 60 percent of PhD’s worldwide consider quitting their research, or at least do not graduate within 6 years of study, is a risk for the future of knowledge and research institutions such as universities. In her thesis, psychologist Andrea Kis is researching what factors significantly influence these career decisions in order to better understand why some young researchers leave or stay.

In her research, Kis found a correlation between experiencing problems in supervision or other socially unsafe situations as a PhD, and the decision to leave academia. Her recommendations of improving supervisor training and skills are very practical. However, the stress stemming from the current pervasive publish-or-perish culture, in all levels of academia, may not be something that any one institution is able change on its own. She published the first part of her research recently in PLOS ONE.

Andrea Kis. Photo: Angeline Swinkels

Safe working environment

There is a lot of focus on social safety in academia these days, and universities are slowly coming to realize that the current working climate in academia is perhaps not as perfect as everyone wanted to believe for a long time. “Working in a (socially) unsafe working environment is very impactful, in many ways. For this part of my research, I wanted to know if the safety of your working environment was a deciding, significant factor in the decision of PhD students to leave academia,” says Andrea Kis, psychologist and part of the Technology, Innovation & Society group at the department of Industrial Engineering and Innovation Sciences.

“We had around 1600 PhD students at TU/e according to the website in 2020. Almost 400 of them sent back valid survey results. That is an enormous commitment regarding the subject at hand and makes for a good, and higher than typical, response rate.

“When processing the results we had to account for the fact that we are a Dutch STEM university. STEM researchers constitute an understudied population that might differ substantially from non-STEM populations, as natural and laboratory sciences appear to have (much) lower attrition rates than, for example, the social sciences and humanities.

Commitment and response

“PhD candidates in the Netherlands are generally employees and earn a salary, which means that ‘funding’ is much less of a concern for them. Studies show that PhD candidates who have more stable financial resources are more likely to complete their education. We found that most, around 90 percent, of our respondents were very happy in their work, and about their supervision.

“We found that only 21 percent of our respondents considered leaving academia, which is significantly less than the worldwide number of 60 percent or the 34 percent attrition rates in the literature. However, the group that did report irregularities, ranging from bad or absent supervision and a high workload to unethical research practices, also turned out to have an above-average chance of leaving academia.”

The correlation between bad experiences at work and the decision to leave may not come across as very surprising, it is however extra poignant when realizing that this group not only ‘changes jobs’, but leaves their scientific career behind to go do something else entirely. “Students who decide to do a PhD are people who dream about becoming a scientist, or at least enjoy doing research for a living. That is their ambition, although they do not necessarily want to stay on in academia after graduation. It is crushing to see these driven, talented people leave behind the PhD research they set out to do, because of bad workplace experiences,” adds Kis.

“Being a PhD can be a lonely existence, especially when you leave your family and friends behind and move to a different city or country to follow your dream. When you feel completely on your own, it can be much harder to deal with a bad vibe in the team, a workplace incident, or a supervisor with little time for you.”

Working to create a socialy safe environment

Even though the percentage of PhD candidates considering leaving academia at Eindhoven University of Technology (TU/e) is well below average (21 percent at TU/e vs. a reported average of 34 percent in Europe and 60 percent worldwide), there is still a drive to understand and prevent the reasons why young and talented researchers leave their chosen career path behind.

"TU/e’s HR department and the Graduate School are jointly responsible for the education and working environments of the PhD’s. They value the insights and recommendations offered by Andrea Kis," explains Ineke van der Vegt, TU/e HR policy advisor. "Her policy related reflections have been carefully considered by a committee and have influenced the position paper that is being written, as we speak, for Paul Koenraad’s second term as dean for the Graduate School. Although the contents are not final yet, it is safe to assume that there will be a lot of focus on social safety."

"But, also in the present day, there are a lot of measures in place to ensure that PhD students can thrive during their research. The PhD supervisor training course has been around for years and has been recently updated; it is freely available for every new supervisor. There is also focus on developing healthy leadership skills for all researchers at assistant professor level and beyond," says van der Vegt. "Additionally, every PhD is required to have a second supervisor (co-promotor), to limit how much any PhD student is dependent on just the first supervisor. And, last but certainly not least, there is the PhD counselor, who is there for every PhD researcher who wants to discuss their situation in a confidential setting with someone who can help and offer advice."

Researching your own

“It is very interesting to be a psychologist at a STEM university,” Kis adds enthusiastically. “For me, in my experience in social sciences and humanities studies, it is very commonplace to research your own environment and population. Every psychology department anywhere will research the students of its own institution, the employees, etcetera. I found that researching our own students and employees like we did is in many ways a first for TU/e. Nevertheless, the results have been welcomed to help improve our institution.

“And to that end I’d love to see more psychologists join me in researching our TU/e community. It will help to normalize this type of research at our university too. And the results can only help policymakers to understand their own institution better and help them to make the right decisions. It also explains why I have four different supervisors, because there is no single professor who covers the same domain that I’m studying.”

Students on the TU/e campus. Photo: Bart van Overbeeke

Using the results for the better

Nevertheless, TU/e is not alone in having to address the situation. When asked about the root causes, Kis clarifies: “It is my belief that the publish-or-perish culture in science is at the root of many of the workplace issues we see in academia today, such as normalizing overtime work. It pushes people to unconsciously have unhealthy work-life balances, to favor PhD students who publish more, to take on more work and more students, and to put your name on as many papers as you can. This can result in workplace environments no one wants.”

“Added to that, we do not yet have a mandatory training or course in place to train PhD supervisors in their new role. They may be the best researchers in their particular field, but effectively being the manager of a team of brand new scientists and coaching them requires a different set of skills. We should allow people to acquire those too, before we give them PhD students to supervise.”

“What my research can do is to grow awareness of the root causes and the effects. That way we can help people become better in their role as a supervisor, instead of blaming them for things that go wrong and leaving them to sort it out themselves. It is very, very rare that people create a bad working environment on purpose. Educating and supporting supervisors is a far more constructive way forward, according to all literature I found on the subject,” adds Kis.

“I really am an optimistic person and I believe we can help to change academia become a better place to work in. Better for excellent science, and kinder to the people who work there. I hope my research helps to bring that about,” concludes Kis.

More information

The articleLeaving academia: PhD attrition and unhealthy research environments’ was published in PLOS ONE on October 5, 2022. Andrea Kis is continuing her PhD research under supervision of Elena M. Tur, Daniël Lakens, Krist Vaesen and Wybo Houkes.

HR and the Graduate School are gathering everything a PhD may need in one place, a portal. This is a work-in-progress, but already offers a shortcut to help, or someone to talk to. More info:

PhD Programs (tue.nl) – Graduate School - PhDs
PROOF Training Program (tue.nl) – courses for PhD’s
Scientific Staff Management Training Program (tue.nl) – o.a. academic leadership courses and the recently updated Supervising PhD’s course.
PhD-EngD Counselor (tue.nl) – link to the PhD-EngD counselor

Nicole van Overveld
(Science Information Officer)

More on our strategy

Latest news

Keep following us