Survey shows 86% of researchers view at least some research outputs as untrustworthy, and they are taking extra measures to verify them

A lack of confidence in scientific findings seems to be increasing the workload of researchers, according to new analysis by Dutch publishing giant Elsevier and the independent UK-based charity Sense about Science. More than 60% of 3000 researchers surveyed in May and 1500 polled in March describe all or most of the research outputs they come across as trustworthy, while more than a third say they only view ‘about half’ or ‘some of them’ as trustworthy, and 1% had no confidence at all. Overall, 86% of the researchers rated at least some research outputs as untrustworthy, and the primary reasons cited were poor interpretation, lack of clarity of the peer review process and methodological flaws.

The survey results indicate that these doubts have caused researchers to work overtime to ensure that the findings they use are reliable. For example, nearly 60% of those polled reported checking supplementary data carefully, just over half said they seek corroboration from other trusted sources, and over a third admitted to only reading and accessing the work of researchers they know.

The data also show that it is taking researchers almost as long to search for scientific articles as to actually read them. Each week, they are spending just over four hours finding these papers and more than five hours reading them. Indeed, the situation appears to be getting worse. The survey responses reveal that researchers read 10% fewer articles between 2011 and 2019, it took them 11% more time to locate this literature.