Key points are not available for this paper at this time.
This study used path analysis to test predictive models of depressive symptoms and loneliness in college students. A model of depression, which included the variables of instrumentality, social self–efficacy, expressiveness, and self–esteem, accounted for 41% of the variance in depressive symptoms. Instrumentality was strongly related to social self–efficacy, which in turn was related to both global self–esteem and expressiveness. In a model that included loneliness as well as depression, self–esteem was significantly related to both dependent variables. Social self–efficacy was directly related to loneliness and was indirectly related to depression though the mediating variable of self-esteem. This model accounted for 56% of the variance in loneliness and 41% of the variance in depressive symptoms. Implications for understanding and treatment are discussed.
Hermann et al. (Fri,) studied this question.