Harvard researchers have condemned a 2015 study that claimed psychologists are having an alarmingly difficult time replicating their results.
In 2015, we reported on a study that reviewed over 100 psychology research publications and found that more than half of them were impossible to replicate. According to a report from the Washington Post, a team of scientists from Harvard University and the University of Virginia recently issued a scathing letter condemning the results of the study, asserting that the studies weren’t impossible to replicate – just poorly designed.
Scientists wrote that the research methods used to reproduce the published studies led to a gross overestimation of the failure rate of reproducing psychological research.
The 2015 study was carried out by the nonprofit organization, the Center for Open Science. It was published in the journal Science, and convinced millions that maybe some conclusions of modern psychological research weren’t all that reliable. According to the editor of Science, “We should be less confident about many of the experimental results that were provided as empirical evidence in support of those theories.”
The recent letter, however, says that the 2015 study jumped the gun on saying that psychological studies could not be replicated. According to Harvard psychologist Daniel Gilbert, a conclusion so serious could do a significant amount of damage to current work in psychological research.
“This paper has had an extraordinary impact. It led to changes in policy at many scientific journals, changes in priorities at funding agencies, and it seriously undermined public perceptions of psychology,” he wrote.
Gilbert wrote that the 2015 study failed to take a random sample of publications to assess, which led to an incomplete picture of the field as a whole. Even worse, the study didn’t even really try to replicate the research methods it sought to discredit. The title of the study was misleading as a result, and the authors did nothing to clarify the impression that headlines gave.