Abstract
BACKGROUND: Loneliness can cause severe mental and physical health problems and is of particular concern among vulnerable groups such as left-behind children. Research has suggested important person-level characteristics and attributes, such as self-esteem and self-efficacy, to be protective factors of loneliness in children. However, existing research is limited in the use of a cross-sectional design and the lack of consideration of the self-esteem × self-efficacy interaction on loneliness, as well as the potential divergent effects of domain-specific (e.g., general vs. academic) self-efficacy in loneliness. METHODS: We used a longitudinal design to examine the moderation role of general vs. academic self-efficacy in the influence of self-esteem on left-behind children's loneliness. In a sample of 405 left-behind children (M(age) = 10.51, SD = 1.36; 49.6% girls), we assessed their self-esteem, general self-efficacy, academic self-efficacy and loneliness at baseline, with follow-up measure implemented to assess changes in loneliness in one year time. Longitudinal path analysis was conducted for hypothesis testing. RESULTS: Academic self-efficacy was a more proximal predictor of loneliness at baseline and its change over time. Importantly, self-esteem predicted lowered loneliness and a more significant reduction over one year only when academic (not general) self-efficacy was high. CONCLUSIONS: A profile of high self-esteem and low academic self-efficacy appeared to be the most at-risk for loneliness in left-behind children. The findings should inform future research and policy/intervention efforts regarding strategies for addressing loneliness in left-behind children by enhancing their academic self-efficacy and closing the esteem-efficacy discrepancy.