Ego development in adolescence: Longitudinal studies.

Adolescents from four schools (one inner-city, two suburban, and one private) were tested two or three times using the Washington University Sentence Completion Test of Ego Development. Testing intervals ranged from 1.5 to 6 years. About half (N=193) the original pool was retested at twelfth grade. Every sample showed a mean rise in ego level; for six of eight samples that rise was statistically significant. Every pair of testings showed a postive correlation between scores; 10 of 14 correlations were significantly different from zero. Thus both test-retest correlations (about 0.5) and change scores support the hypothesized sequence of ego development. Significant correlation between ego level and intelligence occurred in two schools (0.6 at elementary and junior high grades and 0.4 at twelfth grade), but correlation was about zero in the private school.