Adolescent males are at risk for a number of academic, social, and emotional problems. Existing research provides evidence that a number of these problems are related to conflicts experienced by adolescent males through the gender socialization process, called gender role conflict. Whereas there is ample empirical literature on gender role conflict in adult males, few researchers have investigated this hypothesized conflict in adolescent males. This study, involving small interview groups of adolescent males, was designed to examine the validity of the gender role construct for adolescent males.

Idiographic personal construct (PC) measures of self-image disparity were hypothesized to be stronger than nonidiographic conventional construct (CC) measures in predicting neuroticism. Ninety-six college students completed PC and CC measures of real self, ideal self, social self, and ideal social self; the NEO Five-Factor Inventory; and the Marlowe-Crowne Social Desirability Scale. Content analysis showed that individuals' PC measures were unique. The disparity between real self and ideal self and the disparity between social self and ideal social self were computed for PC and CC; test-retest reliabilities were .76 to .81. Results of simultaneous multiple regression analyses supported the hypothesis. Also, the two-variable PC self-image disparity scores were stronger than one-variable PC esteem scores in predicting neuroticism. Implications were discussed for research method, self-concept theory, personal construct theory, neuroticism, and psychotherapy.