Stability of intelligence: A 40-year follow-up.

The Revised Examination M, a measure of intelligence, was re- administered to 260 men approximately 40 years after they had received this test as World War II army recruits. Their average age and educational level was 64.7 years and Grade 8.7, respectively. Three sets of scores were compared: those obtained at the time of army enlistment; current scores obtained within the regular time limits of the test; and current scores obtained within double the test's regular time limits. The results indicated a reliable but minor decline in total score in the regular speeded condition and a reliable improvement over wartime total scores in the double-time condition. Slower performance did not account for the decline observed on the nonverbal cognitive tasks, particularly spatial problem-solving, but was associated with the decrements recorded for mathematical and verbal reasoning skills. Significant gains in score were observed on the measures of vocabulary and mechanical knowledge in the speeded condition, with further gains noted in the double-time extension. Test-retest correlation coefficients, in general, attested to the long-term stability of individual differences among members of the sample, irrespective of the direction of cognitive performance changes. Gains in the double- time extension were associated with educational status but not with current age. Both age and educational level, in addition to early adult intellectual status, predicted current intellectual functioning. Finally, self-reported current activity level, as an indicator of personal lifestyle, was related both to early and to late adult intellectual ability.