of Work: Although developmental theories and popular accounts suggest that midlife is a time of turmoil and change, longitudinal studies of personality traits have generally found stability of rank order and little or no change in mean levels. A pair of studies on 2,274 men and women initially in their 40s enrolled in the University of North Carolina Alumni Heart Study (UNCAHS) examined the stability of personality traits over the mid-life period. The first study (Costa PT Jr, Herbst JH, McCrae RR, Siegler IC. Personality at midlife: Stability, intrinsic maturation, and response to life events. Assessment 2000;7:367-380) also explored the effects of recalled life events on subsequent personality scores. Results found a high degree of rank order stability; the uncorrected average retest correlation for the 30 traits measured by the NEO-PI-R was .72, and .84 for the domain scores or Big 5 factors. Life events in general showed very little influence on the levels of personality traits, although some effects were seen for changes in job and marital status that warrant further research. A second study (Herbst JH, McCrae RR, Costa PT Jr., Feaganes JR, and Siegler IC: Self-perceptions of stability and change in personality at midlife: The UNC Alumni Heart Study. Assessment 2000;7:381-390) examined whether self-reported changes in personality are associated with actual changes. UNCAHS respondents were asked to reflect back over a 6-year period and assess any changes in their personality. A majority (52.5%, N=1,177) reported they had """"""""stayed the same"""""""" while 38.5% (N=863) reported they had """"""""changed a little"""""""" and 202 or 9% reported they had """"""""changed a good deal"""""""". Coefficients of profile agreement, which combine information on all five factor scores and are sensitive to differences between corresponding profile elements and mean extremeness, were computed. Individuals who claimed they changed a good deal should have shown lower mean coefficients of profile agreement than others. In fact, however, the coefficients of profile agreement for the NEO PI-R personality profiles of the changed a good deal were virtually identical to the other two change groups. The lack of agreement or concordance between self-perceptions and measured changes in personality traits indicate that self-perceptions of change are not an adequate substitute for repeated objective assessments in studies of personality change or in medical outcome research.
Showing the most recent 10 out of 21 publications
