Research integrity training is a “box-ticking” exercise that overlooks key practicalities of experimental design and data management, many Australian academics believe.
University staff have conflicting perceptions of who provides the training, what form it takes, whether it is compulsory and even if it exists, according to research by Springer Nature and the Australian Academy of Science.
The survey of almost 1,000 staff from dozens of universities and other research organisations found that internal training programmes were focused mainly on high-level principles such as the importance and definition of research integrity. Fewer than one-third of respondents said their training covered practicalities such as curating data or finding time to manage it.
Little was offered on topics like “determining statistical power”, “metadata descriptions” or “determining the scale of the experimental cohort”. Yet these are the areas where researchers say they need help.
“[The training] was brief, general and more focused on the definition of integrity than how to make decisions,” one respondent said. “[It] seems to come from a risk management, cover-your-arse place, especially as it relates to research ethics,” said another. “It tends to leave unexplored many of the actual questions and issues of research integrity I struggle with in my own work.”
Springer’s director of research environment alliances, Ed Gerstner, said that the survey had revealed a desire for “the very under-the-hood, nitty-gritty things that researchers in particular fields need”.
He said things had improved since his 1990s physics studies, when “I was never taught anything about research integrity” and postdoctoral colleagues were unaware of principles like “don’t cherry-pick” data. “Any training is better than no training,” Dr Gerstner said. “Even if it is a box-ticking exercise, these are boxes that need to be ticked. But it could be better.”
The survey revealed divergent views on whether training existed and who was obliged to take it. Of almost 100 responses from one university, 65 per cent said that their institution provided research integrity training while 14 per cent disagreed and 21 per cent could not say.
Overall, of the 650-odd respondents aware of training at their institutions, 46 per cent said that it was online while the same proportion said it was blended.
The academy’s secretary for science policy, former Australian National University vice-chancellor and Australian chief scientist Ian Chubb, said these results suggested that institutions left issues of research integrity “to the discretion of the individuals down the chain”.
“If a PhD student isn’t motivated to ask, because they’re busy and have a part-time job to make ends meet, and their supervisor doesn’t tell them that there is an institutional programme or policy, then they’re not going to find out,” he said.
“Organisations…have a responsibility for ensuring that the researchers they are training have a complete, comprehensive understanding of the need to conduct research with high integrity. We work with a social licence and a social licence isn’t something that’s given to you – it’s something that’s earned.”
Dr Gerstner said it was difficult to know how research integrity training compared elsewhere, because very little benchmarking had been done anywhere in the world. “This, we hope, is the first of many surveys we’ll be doing, comparing contexts and helping identify where universities could be doing better.”
