You may not be as smart as you think : an alternative account of the Dunning-Kruger effect

Abstract

Perception of abilities plays an important role in informing one’s decisions at times and often in forming one’s self perception. The Dunning-Kruger effect is a fascinating and empirically observable bias in which top performers tend to make more accurate estimations of their ability than bottom performers. The current theory states that the effect is caused by top performers possessing greater metacognitive ability than bottom performers. There have been many alternative theories and explanations proposed to explain the observed Dunning-Kruger effect. The current study is the first to test whether top and bottom performers base their predictions on inflated preconceived notions of ability, rather than their metacognitive ability. This theory proposes that if top and bottom performers both based their predictions of performance on their preconceived notions of ability it would create a Dunning-Kruger effect. This presupposes that that both top and bottom performers make above average estimates of performance as they hold preconceived notions of above average ability. Thus, top performers’ predictions of performance would be most accurate as their performance would be above average, whereas bottom performers would most overestimate their performance as their performance would be below average. The intention of this study was, thus, to assess whether either top or bottom performers based their predictions of performance on preconceived notions of ability or using metacognitive ability. A total of 97 university students were divided into two groups and given an identical test, one group containing 49 participants were told the test measured Logical Thinking and the remaining 48 participants were told the test measured Computational Mathematics. After completing the test, which was a 23 item preparatory test for the LSAT, participants were asked to estimate their ability in the domain being assessed, their performance relative to their peers and their score out of 23. A t-test was used to compare the two groups and it was found that the Logical Thinking and Computational Mathematical group made significantly different predictions of ability and therefore held significantly different preconceived notions of ability. Further t-tests were used to compare the estimates of ability and predictions of performance of the two groups of top and bottom performers. A significant difference was found between the two groups of top performers’ prediction of ability. However, there was no significant difference between any of the other scores of the two groups of top and bottom performers. Therefore, the alternative theory that top and bottom performers base their predictions of performance on preconceived notions of ability was found to be invalid. Therefore, the current theory which states that top performers’ superior metacognitive ability allow them to make more accurate estimates of performance than bottom performers is still the best account for the Dunning-Kruger effect.