Abstract
The reorientation of experimental psychology from studying performance to studying cognitive processes has created a new potential for under standing ability tests in terms of the nature of the cognitive events which contribute to individual dif ferences in solving the test items. The results from the present study suggest the feasibility of explain ing individual differences in performance on a pro totypic intelligence test item—verbal analogies— from the success and efficiency of processing infor mation on hypothesized component events. The data revealed that at least three types of processing events are needed to describe individual differences in the component task durations, but that probably only one factor is needed to describe accuracy in completing the components. More critically, both the accuracy and duration of the component tasks were significantly related to solving psychometric analogies. The results are discussed with respect to the nature of successful performance on analogy test items and the need for more complex models to fully account for individual differences in test per formance.