A comparison of error based and time based learning measures as predictors of general intelligence

Two methodological issues involved in determining the relationship between learning and general intelligence (g) are examined. They are firstly, the use of subject learning strategies in a paired-associated learning task and secondly the importance of time in the measurement of learning efficiency. Concerning the first issue, paired-associates learning scores of subjects who were instructed in the use of an appropriate learning strategy correlated significantly with general intelligence scores while those not so instructed did not. Concerning the second issue, a time based learning measure was developed to predice the time required for complete learning in the paired-associates learning task. This measure was compared learning with a number of error-based learning measures in their ability to predict general intelligence. The time based learning measure was clearly superior to the error based measures yielding a correlation of .59 with general intelligence.