Academic journal article Economic Inquiry

The Role of Noncognitive Skills in Explaining Cognitive Test Scores

Academic journal article Economic Inquiry

The Role of Noncognitive Skills in Explaining Cognitive Test Scores

Article excerpt


This article reports findings of an experiment to examine whether measured cognitive test scores are influenced by noncognitive skills. The basic idea of our analysis is that the performance on a cognitive test not only depends on the actual cognitive abilities but also on the willingness to put mental effort in answering difficult questions in the absence of extrinsic rewards. A relationship between noncognitive skills and cognitive test scores can exist for two reasons. First, people who are motivated to perform well and who have a positive attitude toward work might be more inclined to do their best at tests, irrespective of the rewards offered. If so, high IQ scores might partly reflect favorable noncognitive skills, and correlations between cognitive skills and outcomes are upward biased. Second, people with favorable behavioral or labor-market outcomes might have an attitude to put in effort only when there are sufficient rewards. This could serve as an explanation for a successful career, despite lower cognitive test scores at school.

To investigate the relationship between noncognitive skills and cognitive test scores, we performed an experiment in which we first measured psychological traits and economic preference parameters of 128 students. Next, these students carried out a cognitive test. Initially, there were no rewards for right answers, but later on, we introduced payments for right answers. To disentangle the effect of increased mental effort from increased time investments, we also varied the time available for each question. To investigate whether our results are affected by heterogeneity in the marginal value of time, we ended the experiment by measuring the marginal price for the willingness to spend time doing nothing.

We find that students put substantially more time in answering IQ questions when rewards are higher. The effect of extra time investments on test scores is less obvious and depends on the type and difficulty of the IQ question. When time constraints are very binding, the effects disappear, suggesting that people cannot increase mental effort as a substitute for investing more time. We find several personality traits for which the effect of rewards on the time spent to answer a question is significantly smaller than average: performance motivation, internal locus of control, and curiosity. Also, components from the 5-factor model of personality structure, such as emotional stability and conscientiousness, are associated with a low effect of rewards on extra time investments. Students with a high preference for leisure (measured by psychological tests) and a negative fear of failure increase their efforts to answer questions more than average when rewards go up. This also holds for the component measuring openness to new experiences from the 5-factor model of personality structure.

For the economic preference parameters, we find the opposite result. Students with high discount rates, high risk aversion, and a high preference for leisure (measured by questions in which the respondent has to trade-off time and money) tend to decrease time spent on an IQ question more than others, when rewards are increased. Again, the incentives do not always increase performance. Since in general, low discount rates and low levels of risk aversion are also associated with favorable behavioral or labor-market outcomes, this is surprising. A potential explanation is that economic preference parameters are measured by questions about economic trade-offs. Possibly, people differ in their ability to deal with such trade-offs, explaining why the psychological measures might pick up other aspects of noncognitive skills than the economics preference parameters.


A. Design

We conducted an experiment with different time constraints and financial incentives on IQ questions to examine the influence of different preferences and types on the performance in this cognitive test. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.