Gender Bias in the Rosenberg Self-Esteem Scale

Indianapolis, IN

There is a consistent finding in the research literature that, overall, females have lower global self-esteem than males, and the Rosenberg Self-Esteem Scale (RSES) is by far the most widely used measure of global self-esteem in psychological research. My goal is to determine whether or not this gender difference in self-esteem scores can be partially explained by the wording of the items of the RSES. In particular, self-esteem scores appear to vary with the use of the verbs "think" versus "feel" in the items of the RSES. Four versions of the RSES were created: the original RSES, a version using only the verb "feel" in appropriate items, a version using only the verb "think" in such items, and a version that eliminates the verbs "think" and "feel." Self-esteem scores will be examined using a 4 (RSES version) x 2 (gender) ANOVA. It is expected that rewording the items of the RSES to eliminate these verbs will decrease the gender difference that emerges in self-esteem scores. If my hypothesis is supported, then it is possible that the reported gender difference in self-esteem is not as large as previously believed and that the RSES is biased against women. In this case, a critical reexamination of the validity of the RSES is required.