Academics underestimate willingness of PhDs to use fake data

About one in 12 PhD students would publish fraudulent results if it helped them to get ahead in academia, a study suggests.

In an international study that surveyed almost 800 doctoral candidates, researchers presented PhD researchers with a scenario in which data had been fabricated and asked whether they would be happy to proceed to publication.

In the first part of the study, involving 440 PhD candidates recruited from social science or psychology departments in Dutch universities, almost all spotted the use of fraudulent data but 8 per cent said they would publish if they felt under pressure to do so, explains the study, published in Frontiers in Psychology.

A replication study involving 198 PhD candidates from the medical and psychology faculties at a Dutch university found similar results, while a third study that polled 127 social science PhD students in Belgium found that 13.4 per cent would publish the dodgy data.

“Many of those we interviewed came up with good arguments for publishing what they knew was fabricated data, such as ‘if this is what it takes to finish my PhD’,” said the study’s lead author, Rens van de Schoot, professor of statistics at Utrecht University.

While the proportion of those willing to use fake data “was not high, it is also not zero”, he added, stating that most of the 36 Dutch academic leaders they also interviewed predicted that cheating would be unthinkable for PhD students.

While the study was confined to the Netherlands and Belgium, Professor van de Schoot said he believed the trend “could be even worse in other parts of the world as PhDs in the Netherlands are university employees and protected by certain rights under Dutch law”. “They don’t have that same status in many countries,” he said.

The study sought to explore whether ethical leadership made a difference to dishonesty levels, describing various checks on research transparency and ethics to respondents. “You might have expected training or education in ethics to make a difference, but it didn’t,” Professor van de Schoot said.

Instead, he said, universities should seek to create a “safe space for PhD researchers to share their uncertainties” where “someone with power can be made available to ask if what we’re doing here is right”.

jack.grove@timeshighereducation.com