Many social phenomena – reactions to positive and negative feedback, avoidance of evaluative circumstances – are presumed to reflect individual differences in self-esteem. Discussions of these phenomena often assume that an individual's esteem-relevant thoughts and behaviors can be predicted based on a single number, namely, the individual's level of self-esteem. Some approaches have gone beyond this “point estimate” to a multifaceted conceptualization, measuring self-esteem separately in various domains (e.g., Rosenberg, Schooler, Schoenbach, & Rosenberg, 1995; reviewed by Harter, 1996). Still, the level of self-esteem in each specific area is a single number. Research reviewed in this chapter suggests that understanding people's psychological experiences and predicting their responses requires knowledge not only of their level of self-esteem, but also of the certainty, clarity, consistency, and stability of the self-image.

From his extensive study of adolescents, Rosenberg (1989) concluded that difficulties experienced by low self-esteem individuals often resulted not from low self-esteem per se, but from concomitant self-image disturbances. He viewed self-esteem as one part of a multi-aspect self-image. Self-esteem is reflected in the direction – “whether he [sic] has a favorable or unfavorable opinion of himself” – and intensity – “how strongly favorable or unfavorable these feelings are” – aspects of the self-image (Rosenberg, 1979, p. 24).