Modeling the relationship between mentoring and depression symptoms among adolescent North Korean refugees*

This study examined the conditions for, and mechanisms of, mentoring efficacy against depression symptoms in a random sample of 202 adolescent North Korean refugees living in South Korea. Rhodes (2005) argued that mentoring may benefit adolescents with high parental attachment because social engagement comes easily to them, and that it also may benefit adolescents with weak attachment because of greater need. The present study drew on Rhodes' (2005) theory to postulate curvilinear moderation of the mentoring-depression relationship by parental attachment. Teacher and peer attachment were posited as mediators of the relationship between mentoring and depression. Using fixed effects regression models, Sobel tests, and nonparametric bootstraps to test the hypotheses, a negative association was identified between mentoring and depression symptoms. Results suggested mediation by social withdrawal, and models also found a significant curvilinear moderating effect of parental attachment on the mentoring-depression relationship. These findings suggest that adolescent North Korean refugees with strong parental attachment and adolescent North Korean refugees with weak parental attachment may disproportionately benefit from mentoring as compared with adolescent North Korean refugees who scored in the middle in parental attachment. These results are also consistent with the idea that mentoring may reduce social withdrawal which, in turn, reduces depression.