Posted at 11:00 a.m.
This concept has been overgeneralized over the years, but to find its origin, you have to go back 40 years, to 1979. Two young American psychologists, Lauren B. Alloy and Lyn Y. Abramson, published a study whose title made an impression: “Sadder but Wiser? [Plus tristes, mais plus sages ?] “.
The study reported the results of four experiments conducted with college students, some reporting depressive symptoms, others not. Participants were given a button and had to rate the degree of control they had over a small green light that came on. The goal was to test learned helplessness, a well-known theory in psychology that depressed people underestimate the control they have over their environment.
Surprisingly, the group of depressed students achieved accurate results in all four experiments, while the non-depressed students in some cases overestimated the degree of control they had. The hypothesis has been named “depressive realism”: under certain conditions, depressed people would judge the control they have relatively more realistically than those who are not depressed.
“The study had a huge influence, not only in popular culture, but also in the scientific literature,” explains The Press Don Moore, psychology researcher at the University of California. She is frequently quoted, and her conclusions are repeated as if it were established scientific fact that depressed people are more realistic about all sorts of things. »
Difficult to reproduce
University of California researchers — including Professor Moore and student Amelia S. Dev — have made it their mission to replicate the 1979 study, taking into account current “methodological innovations.” The results of the Dev. et al. were published this month in the newspaper Collaboration: Psychology. They were also the subject of a report by the New York Times.
To our surprise, we found that the original result could not be reproduced under any condition. There is no evidence that participants who reported experiencing depression were more accurate in their control assessment.
Don Moore, psychology researcher at the University of California
The University of California team is not alone in trying to replicate the study in recent decades. “And we’re not the only ones who have had trouble doing it,” says Don Moore.
Does this mean that the initial experiment does not meet the criterion of reproducibility—and is therefore not scientifically valid? Joined by The Press, Professor Lauren B. Alloy, co-author of the 1979 study, disagrees. First, she says, since 1979, her team and other independent researchers have “reproduced the results of depressive realism time and time again.” Their work, she says, has clarified certain “conditions” under which depressive realism does or does not emerge.
According to Lauren B. Alloy, changes the University of California team made to the original studies likely account for the difference in results. The team asked participants to rate the likelihood of the light turning on throughout the task, not at the end, she says. She also changed the instructions given to participants and did not measure depressive symptoms just before the task. “Of the four Alloy & Abramson experiments of 1979, Dev et al. chose to model their study with the experiment whose results are the least robust”, adds Lauren B. Alloy by email.
No reproduction is perfect, argues researcher Don Moore. “If science doesn’t correct itself when scientists correct it, and if journals don’t publish breeding failures, then we’re stuck with a literature full of false positives that never get corrected,” he says.
Psychological phenomena — invisible and complex — are difficult to pin down, points out Anna Weinberg, associate professor in the Department of Psychology at McGill University. It is not alarming, she says, to see such back and forth in the scientific literature. “Usually when you see something like this happening with certain replications, it means there may be an effect, but it may be weaker than what the original study showed,” says Anna. Weinberg.
The original experiments were carried out with undergraduate students who presented with depressive symptoms. That these people are more aware of the “abyssal reality” is something intuitive, underlines Anna Weinberg. “But I don’t think it applies to people who have severe clinical depression,” she concludes.