The developmental course of gender differentiation: conceptualizing, measuring, and evaluating constructs and pathways.

Gender differentiation is pervasive, and understanding how and why it develops is important for both theoretical and practical reasons. The work described here is rooted in constructivist accounts of gender differentiation. Past research provides considerable support for constructivist predictions concerning (a) developmental changes in gender attitudes and (b) the relation between gender attitudes and information processing. Little work, however, has addressed the more fundamental question of how children's developing gender attitudes about others are related to developing gender characterizations of self. The focus of the current Monograph is on this other-self relation during middle childhood. A brief review of past theory and empirical work on gender differentiation is provided. It is argued that a major explanation of the limitations and inconsistencies evident in earlier work may be traced to restrictions in the measures available to assess key constructs. A conceptual analysis of the specific limitations of past measures is presented. The Monograph then offers alternative models of the developmental relation between attitudes toward others and characterization of self (the attitudinal and the personal pathway models), and identifies conditions expected to influence the strength of the observed other-self relation. Four studies establish the reliability and validity of a suite of measures that provides comparable methods for assessing attitudes toward others (attitude measures, or AM) and sex typing of self (personal measures, or PM) in three domains: occupations, activities, and traits (or OAT). Parallel forms are provided for adults (the OAT-AM and OAT-PM) and for children of middle-school age, roughly 11-13 years old (the COAT-AM and COAT-PM). A fifth study provides longitudinal data from children tested at four times, beginning at the start of grade 6 (approximately age 11 years) and ending at the close of grade 7 (approximately age 13 years). These data are used to examine the developmental relation between children's sex typing of others and sex typing of the self, and to test the predictions concerning the factors hypothesized to affect the strength of the relation between the two types of sex typing. Overall, the data supported the conceptual distinctions among individuals' (a) gender attitudes toward others, (b) feminine self, and (c) masculine self, and, additionally, revealed some intriguing differences across domains. Interestingly, the data concerning the other-self relation differed by sex of participant. Among girls, analyses of concurrent relations showed that those girls who held fewer stereotypes of masculine activities for others showed greater endorsement of masculine items for self, a finding compatible with both the other-to-self attitudinal pathway model and the self-to-other personal pathway model. The prospective regression analyses, however, showed no effects. That is, preadolescent girls' gender attitudes about others did not predict their later self-endorsements, nor did self-endorsements predict later attitudes. Data from boys showed a strikingly different pattern, one consistent with the self-to-other personal pathway model: There was no evidence of concurrent other-self relations, but prospective analyses indicated that preadolescent boys who endorsed greater numbers of feminine traits as self-descriptive early in grade 6 developed increasingly egalitarian gender attitudes by the end of grade 7. The Monograph closes with discussions of additional implications of the empirical data, of preliminary work on developing parallel measures for younger children, and of the need to design research that illuminates the cognitive-developmental mechanisms underlying age-related changes in sex typing.