Summary. Reviews of recent aptitude‐treatment interaction research indicate a low level of consistency of results even when considered within a particular age level. It is suggested that an aspect of the problem may be the lack of comparability among the variables commonly employed. Two experiments are described where comparable teaching strategies were employed to teach students of similar ages material from two different subject areas. Comparable disordinal interactions between measured extroversion and teaching strategy were demonstrated across the two experiments where the criterion measure was a test of problem‐solving but not where it was a test of recall.