Constancy of adult personality structure in males: longitudinal, cross-sectional and times-of-measurement analyses.

Constancy or change in adult personality organization can be assessed by comparing the factor structure of personality instruments at different ages, and some studies have reported cross-sectional differences in structure. The present study compares the factor structure of the Guilford-Zimmerman Temperament Survey scales longitudinally in three administrations 6 years apart and cross-sectionally in three age cohorts. Additional analyses compare first administration data collected in two successive decades to test for variation in structure resulting from cultural change over that period. Subjects were 769 male volunteers in the Baltimore Longitudinal Study of Aging, aged 17 to 97 at the time of the first administration. Three varimax-rotated principal components were extracted in each of eight analyses. Results show no systematic evidence of variation in structure in any of these groups, with coefficients of factor congruence ranging from .83 to .99. This longitudinally demonstrated invariance of personality structure is discussed in terms of the implications for the stability of personality organization throughout the adult years and for the use of personality tests in elderly groups.