More than two-thirds of researchers now view preprints as a “valued source of communication”, a share that was below half at the start of the pandemic, according to a global survey of scientists.

More than 2,000 researchers took part in for the academic publisher Elsevier in 2020 and 2021 to help gauge how the Covid-19 crisis was affecting changes in science.

The exercise found that 67 per cent of respondents in 2021 agreed that preprints were a valued source, with 17 per cent strongly agreeing, up from 43 per cent the year before, including 7 per cent strongly agreeing. The share disagreeing with the statement almost halved, to 12 per cent.

Preprints quickly became an important way to disseminate research findings in the early days of the pandemic, further boosting use that had already been growing strongly before Covid struck.

̽����Ƶ

The increasing acceptance of preprints has led to a growing debate in the academic community about whether they could become the main way to share findings, given their speed and free access, or whether peer-reviewed publications should remain central to the scientific process.

Of those responding to the survey who agreed that preprints were valuable, three-quarters referred to their free access and speed of dissemination as being key factors, while the major reason for people disagreeing was the lack of peer review, with 55 per cent citing this.

̽����Ƶ

Elsewhere, the survey results, published in a report called Research Futures 2.0 – a follow-up to a similar Elsevier study in 2019 – appear to find a growing acceptance towards the use of artificial intelligence in peer review, albeit this still representing a significant minority of researchers.

In all, 21 per cent of researchers agreed that they would read papers peer reviewed by AI, a 5 percentage point increase on 2020. However, most of those surveyed continued to say that they would object to AI peer review, with 58 per cent unwilling to read to such articles.

Meanwhile, the proportion of those reporting that they were collaborating more than in the past grew over the period of the survey, from 48 per cent in 2020 to 63 per cent in 2021.

This included increases across geographies and disciplines, with researchers in computer science in particular seeing a big rise, from 41 per cent agreeing that there was more collaboration in 2020 to 76 per cent in 2021.

̽����Ƶ

However, the report also points to continued major concerns about funding in the research community, with about half of those questioned in 2021 and 2020 disagreeing that there was “sufficient funding available in my field”. The percentage agreeing also fell in the period, from 30 per cent to 24 per cent.

Researchers pointed to fewer grants, increased competition, changing disciplinary priorities and the reallocation of funds to Covid-19 research as reasons for a dearth of funding.

Adrian Mulligan, research director at Elsevier and co-author of the report, said it was clear from the findings that the pandemic had created “added pressure” on the research community, but there were also “new levels of collaboration and openness” apparent in the wake of the crisis.

simon.baker@timeshighereducation.com