The humble estimate: Humility predicts higher self-assessment accuracy.

Abstract

Humility is thought to be associated with greater accuracy in self-assessment. However, clear evidence is lacking. Two studies tested this central proposition. In Study 1 (N=258), participants completed a task on logical reasoning before estimating both their raw and relative performance. Study 2 (N=214) was aimed at replicating Study 1 with a task on English fluency. Results from both studies were consistent. There was evidence of overestimation bias across each sample in which participants' estimated performance was higher than their actual performance. More importantly, humility was associated with less overestimation bias, such that the difference in estimated and actual performance was smaller or non-existent among those higher in humility. We also replicated the Dunning-Kruger effect in which participants of lower ability in these skills (i.e., lower actual scores) were most likely to overestimate their performance. Further analyses found that the negative relationship between humility and overestimation bias was not moderated by actual performances. However, the same analyses revealed that the Dunning-Kruger effect was also not moderated by humility. Hence, there is strong replicable evidence that humility is associated with less overestimation bias, supporting the view that greater accuracy in self-assessment is an attribute of humility, and that this relationship is independent of actual ability, but humility does not affect the robust Dunning-Kruger effect.