HSE: Less competent people tend to overestimate themselves, but there is an explanation for this

People who are poorly versed in the subject are more likely to sin by overestimating their knowledge, the HSE noted. Scientists refer to the work of Justin Krueger and David Dunning, who described this effect back in 1999.

More than 20 years ago, an experiment was conducted in which researchers asked students to take tests that assessed their sense of humor, logical thinking, and knowledge of English grammar and guess how well they did compared to others. Participants, in general, significantly overestimated their abilities. On average, they believed that they completed the task better than 66% of their classmates. Moreover, the lower the test result, the more the estimated score differed from the real one. Only the strongest participants underestimated themselves. Then the authors of the article explained this by the fact that people with low abilities not only cope worse with tasks, but also inadequately assess their knowledge.

Anatoly Peresetsky, a research professor at the HSE Faculty of Economics, and his co-author Jan Magnus from the University of Amsterdam, offered another explanation for this effect: in terms of statistics, not psychology. In their report, made at a seminar of the HSE Department of Applied Economics, the scientists noted that in experiments to study this effect, it was possible to get scores in a strictly specified range for the test: from 0 to 10 or from 0 to 100.

“Suppose that people can guess their result with some error, but in general it is correct, that is, they are equally likely to underestimate and overestimate themselves. In this case, due to the fact that it is impossible to get less than 0 or more than 100 points, two groups of participants are limited in their assumptions: those who wrote very badly cannot say that they completed the tasks by -5 or -10, and those who solved everything correctly – what they expect by 110. It turns out that the estimated estimates of the first will be on average higher, and the second – lower than they could be if the scale were not limited, ”the authors of the report noted.

They tested this assumption on the data of 665 HSE students who took an exam in statistics in 2016-2019. In the middle of the exam, students were asked to write down what grade they expected to get in the end. For a correct guess, they were promised a small bonus. The results showed the presence of the Dunning-Kruger effect.

“The Dunning-Kruger effect does exist, but it is a statistical artifact and does not require concepts from the field of psychology and cognitive science to explain it. The question of why “psychological” explanations of this effect are still popular in the psychological literature, perhaps, can be the subject of research by psychologists, ”summed up Anatoly Peresetsky, co-author of the work.