Estimating the relationship between skill and overconfidence

TLDR: In this article, the authors restate the Dunning-kruger effect in terms of skill and overconfidence and show that the unskilled are more overconfident than the skilled.

About: This article is published in Journal of Behavioral and Experimental Finance.The article was published on 2017-06-01 and is currently open access. It has received 28 citations till now. The article focuses on the topics: Overconfidence effect & Dunning–Kruger effect.

Figures

Table 2: Estimates of the Dunning–Kruger Effect

Table 1: Predictions, Grades, and Overestimation

Figure 1: Distribution of Exam Grades

Figure 2: Actual versus Expected Exam Grades

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "Estimating the relationship between skill and overconfidence" ?

The authors show that they can correct for bias caused by measurement error with an instrumental variable approach that uses a second performance as instrument. However, as the authors predict in their methodological discussion, this relationship is significantly weaker than ordinary least squares estimates suggest. 

Q2. What is the importance of a second performance?

All the authors need is a second performance as an instrument, as long as the measurement error of this performance is uncorrelated with skill, overconfidence, and the known and unknown luck on the test. 

Q3. What is the effect of feedback on the test?

Having taken a test provides feedback about one’s skill and Ryvkin, Krajč, and Ortmann (2012) have shown that feedback improves calibration, particularly among the low-skilled. 

Q4. What is the effect of skill on the GPA?

An increase in skill of one grade point is related to a decrease in overconfidence by 0.60 grade points, a large effect that is substantially smaller (i.e., less negative) than OLS estimates would suggest. 

Q5. What is the main finding of the study?

The main finding is that bottom quartile performers vastly overestimate their performance while top quartile performers more accurately assess their performance. 

Q6. What is the argument of Low2 Krueger and Mueller?

Low2 Krueger and Mueller (2002) argue that the Dunning–Kruger effect may be a statistical artifact caused by regression effects and the better-than-average effect. 

Q7. How many students were not present on the day the questionnaire was distributed?

The remaining 14 students were not present on the day the questionnaire was distributed in the classroom, either because they missed the particular session or had already dropped out of the course.