At least one in 25 papers includes an image that has been improperly duplicated, a new analysis has found.
A trio of researchers led by Stanford University's Elisabeth Bik manually screened the images in more than 20,600 papers from 40 scientific journals published between 1995 and 2014. As the researchers note in their paper, which is posted at bioRxiv, they found that nearly 4 percent of papers include a problematic image and about 2 percent of papers have an image that appears to have been deliberately manipulated.
"My initial screen is without software, just quickly flipping through images," Bik tells Retraction Watch. "When I see something that might be a duplication, I will download the figure(s) and inspect them using Preview (Mac software)."
For an image to be included on their list, Bik and her co-authors, Johns Hopkins University's Arturo Casadevall and Ferric Fang from the University of Washington School of Medicine all had to agree that the image was a duplicate. They focused in particular on western blots.
They further report that if an author was found to have one paper with a suspicious image, that author was more likely to have additional papers with improperly duplicated pictures. The rate of inappropriate image duplications ranged by journal from 0.3 percent at Journal of Cell Biology to 12.4 percent at the International Journal of Oncology.
Retraction Watch notes that at the time Bik and her colleagues screened these articles none had been retracted. She since wrote 700 reports to journal editors describing the duplications and 10 reports to institutions pointing out labs that had multiple problematic images in their papers. So far, six papers — four from Fang's journal Infection & Immunity — have been retracted and 60 images have been corrected.