Revelle (1993) stated that intelligence researchers assume that all participants who complete their IQ tests are sufficiently motivated. However, Duckworth et al. (2011) provided evidence to suggest that individual differences in test-taking motivation correlate positively with IQ scores. Furthermore, Duckworth et al. suggested that the opportunity to win incentives can increase IQ scores in a material way (d=.64). A limitation associated with existing research on test-taking motivation is that it is based principally upon samples of children. Additionally, the measures of test-taking motivation are usually administered after the IQ tests have been administered (i.e., confounded). Finally, the positive correlation between test-taking motivation and IQ scores has not investigated to determine whether it is more consistent with the interference (causal) or the deficit (non-causal) hypotheses. The purpose of this series of investigations was to examine the above issues through a combination of individual differences and experimental studies.
The individual differences (N=180) and experimental study (N=100) samples were based on first-year undergraduate psychology students. Individual differences in test-taking motivation were measured with the Student Opinion Scale (SOS; trait and state), which is comprised of two subtests – effort and importance. Four subtests of intelligence were administered: Advanced Vocabulary Test (Gc), Advanced Progressive Matrices (Gf), Letter-Number Sequencing (Gsm), and Connections (Gs). In the experimental study, parallel versions of the subtests were used to facilitate pre- and post-testing.
In the individual differences study, a trait version of the SOS was administered prior to the IQ testing. Then, the more conventional state version of the SOS was administered post-IQ testing. The experimental study consisted of a mixed-design: Half the participants (N=50) received the financial incentive at time 1 and the other (N=50) half at time 2.
Correlations of approximately .25 to .40 were observed between the test-taking motivation subscales and general intelligence. Furthermore, the state SOS subscales tended to correlate more substantially with general intelligence than the trait SOS subscales, which suggested that the participants’ experience of IQ testing impacted their testing-taking motivation. Based on the mixed-design ANOVAs, none of the four subtests yielded the hypothesized interaction effect, which suggested that enhanced motivation through financial incentive failed to improve IQ test performance. Finally, the comparison of the ‘no-bias’ versus ‘general-bias’ latent-variable models failed to yield evidence to support the notion that individual differences in test-taking motivation impact IQ test performance in a manner to suggest bias.
A substantial, positive association between self-reported test-taking motivation and performance on cognitive ability tests likely exists in adults. However, the association is likely not best interpreted as a causal effect of test-taking motivation onto IQ test performance, at least not in adults. Instead, intelligence may be suggested to impact test-taking motivation. Thus, the validity of IQ test scores may not be compromised by the fact that individual differences in test-taking motivation exist.
Cognition and Attention , Measurement and Psychometrics
