Except that when you ask someone "how well do you think you did on that test?", they hear "how well would you like me to think you think you did on that test?". The results are better explained, IMO, by noting that people tend to declare themselves as "about average" when asked about their intelligence, in an attempt to not appear snobby or lacking in confidence. For both extremes, evaluating oneself in the edge quartile is socially inappropriate.
Also, consider the correlation between schooling, ability to take grammar/logic tests, and modesty. In my experience, modesty is a social norm significantly correlated with schooling: have you ever heard someone brag about having studied at Princeton?
The experiment lacked the proper controls - for example, asking Princeton math majors how well they thought they knew political science, physics, or astronomy. I would hypothesize that they overestimate their competence (saying "about average") as much as anyone else. Another way to test this hypothesis would be to ask people to rate their own performance, then declare publicly how close their guesses were, and publicly award those with the best self-awareness.
Comments
Except that when you ask someone "how well do you think you did on that test?", they hear "how well would you like me to think you think you did on that test?". The results are better explained, IMO, by noting that people tend to declare themselves as "about average" when asked about their intelligence, in an attempt to not appear snobby or lacking in confidence. For both extremes, evaluating oneself in the edge quartile is socially inappropriate.
Also, consider the correlation between schooling, ability to take grammar/logic tests, and modesty. In my experience, modesty is a social norm significantly correlated with schooling: have you ever heard someone brag about having studied at Princeton?
The experiment lacked the proper controls - for example, asking Princeton math majors how well they thought they knew political science, physics, or astronomy. I would hypothesize that they overestimate their competence (saying "about average") as much as anyone else. Another way to test this hypothesis would be to ask people to rate their own performance, then declare publicly how close their guesses were, and publicly award those with the best self-awareness.