Theoretical and methodological considerations of self-concept measurement.

Partially replicating a study by Savin-Williams and Jaquish (1981), assessment of self-concept was explored by investigating the relationships of "presented" and "experienced" selves among seven adolescent girls participating in a team sport at a high school in Southern California. Behavior observations and self- and peer ratings were used to assess three dimensions of self (self-confidence, popularity, and athletic skill) and examine relationships between these multimethods of self-concept measurement. In general, significant correlations between behavior observations and peer ratings were found, but not between behavior observations and self-ratings, or peer and self-ratings. A behavioral approach to measuring self-concept across situations appeared to be more indicative of the multidimensionality of the self than sole reliance on self-report. The authors concluded that self-concept measurement requires increased sensitivity to definition of, saliency of, and vacillation within a domain, the reference group used for social comparison, and the impact of previous experience on current views of self.