Abstract
The present study used longitudinal data on 182 adults between the ages of 20 and 54 (104 men, 78 women) from the Rochester Adult Longitudinal Study (RALS), assessed on four occasions, to test the hypothesis that identity and intimacy during the course of early and middle adulthood predict well-being at midlife. A cross-lagged panel model was estimated yielding the following findings: (a) Scores on both scales during the college years predicted midlife satisfaction-intimacy directly, and identity through the course of development from ages 20 to 54; moreover, identity in midlife, but not intimacy, was significantly linked with well-being at this same point in time; and (b) identity and intimacy unexpectedly did not predict one another over time, having been controlled for factor stability in identity and intimacy over time. The findings are discussed in terms of Erikson's psychosocial theory of development and the developmental moments and historical cohorts that characterize the present sample.
