An analysis of hundreds of scientific papers over seven years has found a large number might need to be retracted due to duplicated images. About 35,000, to be exact.

Published on the preprint server bioRxiv, the researchers looked at 960 biomedical papers published from 2009 to 2016 in the American Society for Microbiology’s journal Molecular and Cellular Biology. Of these they found about 6 percent had been “inappropriately duplicated”, using software to spot the images, and extrapolated for their final figure.

In other words, a number of papers were using images from other research and passing it off as their own. This, of course, leads to questions about the veracity of research in the process – and has led to a number of retractions already.

The researchers said they reported the 59 instances of image duplication to the journal, of which 42 were corrected, 12 had no action taken (due to things such as the laboratory in question closing down), and five of which were retracted.

In an interview with the website Retraction Watch, the researchers noted this low rate of retraction was not surprising, because “most image problems were the result of error in assembling figures.”

Elisabeth Bik of uBiome, lead author on the paper, added: “We trusted the authors when they said that the duplication was the result of an error. Our goal is to make sure that the science is correct, not to punish.

“The cases that were retracted were the papers where we felt that there were too many errors to be corrected, or where misconduct was suspected.”

The team noted that in an earlier study, they had used software to analyze 20,000 papers, and found that 3.8 percent of those contained duplicated images. Taking that figure across the almost 9 million biomedical papers published from 2009 to 2016, and suggesting that up to 11 percent contain errors worthy of retraction, they come to their final figure of 35,000 papers being suspect.

While this is still a small fraction of the overall papers, it’s still obviously a cause for concern. They note that screening for issues before publication, having one researcher assemble the images, or training peer reviewers to spot duplications could help prevent issues like this in future.

“At the very least, our findings suggest the need for both authors and journals to redouble their efforts [to] prevent inappropriate image duplications,” the team noted in their paper.

[H/T: Retraction Watch / Science Alert]