The Measurement of Role Identity

Abstract

This paper outlines a procedure for measuring those components of the self-concept called role/identities. Conceptualized from a symbolic interaction framework and using the semantic differential and discriminant analysis, this procedure overcomes many of the problems usually asociated with other procedures such as profile analysis, the d statistic, or the TST: the procedure provides substantive anchor points, incorporates the multidimensional nature of role/identity, and provides a numerical score for quantitative analysis. The method is illustrated by measuring the gender role/identity of a sample of sixth, seventh, and eighth grade school children. The validity of the resulting measures are checked by investigating certain of the children's characteristics. Although the constructs of self, self-concept, and identity are now widely used in social psychology and sociology, empirical research with these concepts has been quite limited. Undoubtedly the largest obstacle to research on the self is the measurement problem. Because the self is so multifaceted and rich in content, abstract theoretical discussion has far outstripped concrete empirical research. The purpose of the present paper is to overcome this problem by presenting a procedure for measuring those components of the self-concept called role/identities. Con-theoretical discussion has far outstripped concrete empirical research. The purpose of the present paper is to overcome this problem by presenting a procedure for measuring those components of the self-concept called role/identities. Conceptualized from a symbolic interaction framework and using the semantic differential and discriminant analysis, this procedure overcomes many of the problems usually associated with other procedures such as profile analysis, the d statistic or the Twenty Statements Test (TST). In the second part of the paper the method is illustratively used to measure gender role/identity of a sample of sixth, seventh, and eighth grade school children. PREVIOUS MEASUREMENT TECHNIQUES If empirical research is to catch up with theoretical developments, any technique for measuring identities must have four important properties: it must (1) produce a quantitative measure that can be used in multivariate data analysis, (2) produce measures that incorporate the multidimensional character of role/identities, (3) adequately define the underlying anchor points or dimensions that give meaning to the quantitative scores, and (4) it should integrate the concepts of self and role. Unfortunately most empirical research on self-concepts or identities has used measurement procedures that, though making research possible, tend to limit that research by failing on one or another of these four criteria. *We wish to thank Carolyn Mullins for her editorial assistance in the preparation of this manuscript.