Over the last decade an increased interest in the psychology of self-concept has led to numerous studies which examined self-esteem as an evaluative component of the self-concept. A review of recent studies which examined self-esteem during adolescence leads to three conclusions. First, the results of these studies are not systematized. Several variables found to be related to self-esteem differ with regard to their conceptual level: locus of control, physical appearance, relationship with parents and peers, and change of residence or school (Hirsch & Rapkin, 1987; Houston, 1984; Kulas, 1982; Marron & Kayson, 1984). Second, very few studies examined the relative contribution of these variables to the prediction of self-esteem (Buri, Kirchner, & Walsh, 1987; Wade, Thompson, Tashakkori, & Valente, 1989). Given that the correlates of self-esteem tend to be interrelated, examination of their independent effects seems to be the next necessary step. Third, the studies that examined stability of self-esteem during adolescence produced inconsistent results. This could be due to differences in design and in the measures used to assess self-esteem. Cross-sectional studies showed changes in self-esteem and decrease of self-respect during middle adolescence (Rosenberg, 1979). On the other hand, findings from the studies that used a longitudinal design indicated relative stability or even a slight increase in self-esteem during adolescence (Nottelman, 1987; Savin-Williams & Demo, 1984).
Mostly, changes of self-esteem during adolescence was seen as the consequences of physical maturation or radical changes in the way of living. The interactionist theory of self proposes, however, that the interactions with significant others play a crucial role in the development of self-esteem. This assumption has been tested and confirmed in our research (Lackovic-Grgin, 1987; 1988; Lackovic-Grgin & Dekovic, 1990). It seems therefore reasonable to expect that changes in self-esteem could be related to changes in adolescents' interactions with significant others.
From the psychological studies come evidence that the quality of interaction between children and their parents is related to the descriptive and evaluative aspects of self-concept (Burns, 1982; Papini & Sebby, 1987). During adolescence the quality of interactions with peers also becomes important (Lackovic-Grgin, 1988). Positive aspects of interactions such as intimacy, acceptance, and nurturance are related to higher self-esteem.
The biosocial studies pointed out that physical maturation affects the quality of interactions between adolescents and significant others, though these changes appear to be more subtle than is suggested by the traditional conception of adolescence. Contrary to the traditional view of adolescence as a period of stress and turmoil, it appears that the increase in conflicts, larger distance, and less intimacy within the parent-adolescent relationship characterized only the period at the apex of pubertal maturation. These changes are especially evident in girls (Steinberg, 1988; Steinberg & Hill, 1978).
With regard to the relationship between physical maturation and self-esteem, no clear findings have been reported. Ulman (cited in McGrory, 1990) found that girls who have been menstruating for two years were more depressed than those who have not started to menstruate. Early pubertal development has also been associated with lower self-esteem (Wade et al., 1989). On the other hand, Simmons, Blyth, and McKinney (1983) found no effects of onset and presence of menses on girls' global self-image in early adolescence. Similarly, Brack, Donald, & Ingersoll (1988) found no variation in self-esteem across Tanner's five stages of physical development.
Based on these findings it seems unlikely that physical maturation would have a direct effect on self-esteem of adolescent girls. It is more logical to assume that the effects are indirect, mediated by the changes in the adolescents' interactions with significant others. …