Abstract

Coding schemes developed in the literature for content analysis of spontaneous self-descriptions have provided research results that are hard to interpret, because they have not considered some basic statistical assumptions. Using data from a sample of 426 young adult participants, we illustrate a new coding scheme for spontaneous selfdescriptions and develop a set of indices for statistical analyses (completion, salience, density, and contrast indices). The present coding scheme uses a more theoretically-based definition of interdependence and provides a more balanced frequency distribution than two traditional coding schemes in the literature (i.e., Kuhn and McPartland; Triandis). Most participants reported a «mixture» of self-descriptions (independent and interdependent). Overall, no significant gender differences were found. However, females used more emotional-affective content than did males in their independent self-concept. In conclusion we discuss the constructs of consensual and interdependent self-conceptions, as well as the empirical evidence in relation to gender differences in self-description.