Effects of core self-evaluation on tobacco and alcohol use among left-behind children: mediating role of fear of missing out and moderating role of parental migration category and gender.

Abstract

Background: Tobacco and alcohol use are quite prevalent among left-behind children, and core self-evaluation is an important contributing factor. Current research has rarely focused on the underlying mechanisms between these variables. This study aims to explore the relationships between core self-evaluation and tobacco and alcohol use among left-behind children, and the mediating effect of fear of missing out as well as the moderating effect of parental migration category and gender.

Methods: For this study, data were collected from January 2024 to February 2025 at 25 rural primary and junior high schools in 5 cities in Jiangsu Province via a random cluster sampling method. A total of 1795 left-behind children were surveyed with the Core Self-Evaluation Scale, the Tobacco and Alcohol Use Questionnaire and the Multidimensional Fear of Missing Out Scale. SPSS 26.0 software was used for descriptive statistics and correlation analysis, and Mplus 8.4 software was used to analyse the mediation effect and moderated mediation effect.

Results: Core self-evaluation was significantly associated with tobacco and alcohol use; the fear of missing out partly mediated the relationships between core self-evaluation and tobacco and alcohol use; the parental migration category served as a moderating factor in the mediation model, specifically, in households with dual-parent migration, the associations between fear of missing out and tobacco and alcohol use were more significant; no significant gender differences were observed in the relationships between fear of missing out and tobacco and alcohol use.

Conclusions: This finding indicated that core self-evaluation contributed to tobacco and alcohol use among left-behind children through fear of missing out, and the parental migration category moderated the latter half of the mediating model.