잠재성장모형(LGM)을 활용한 게임이용 청소년의 건강, 학업, 가족 및 교우관계에 대한 종단연구
This study longitudinally examined how adolescents’ game use over time affects self-control, health status and health management, academic achievement, and family and peer relationships. To this end, we analyzed five-wave panel data from a national game user panel survey and sequentially applied latent growth models (LGMs), parallel-process LGMs, and time-varying covariate models. First, unconditional LGMs were estimated for eight variables—game use, self-control, health status, health management, academic achievement, parent–child communication, academic stress, and peer stress. The results showed that game use increased slightly but significantly over time, whereas self-control, health management behaviors, and academic achievement exhibited overall declining trajectories. Parent–child communication started at a relatively high level and showed no significant mean change, while academic and peer stress remained at relatively low levels with minimal mean-level fluctuation across waves. Because the mean of the linear slope factor was not significant for parent relationship, academic stress, or peer relationship, these variables were treated as comparatively stable contexts and excluded from subsequent models focusing on change dynamics. Next, we estimated conditional LGMs in which game use was included as a time-varying covariate in the growth models for self-control, health management, and academic achievement. Across all waves, higher game use was consistently associated with lower concurrent self-control, and with poorer health management and lower school grades at most time points. For health management, the negative association with game use was pronounced at the first three waves but its magnitude attenuated over time and became statistically nonsignificant at later waves. By contrast, academic achievement displayed a robust pattern of “increased game use → lower concurrent grades” at all five waves, indicating a stable negative association between frequent game use and short-term academic performance in adolescence. Finally, to investigate the long-term interrelations among game use, self-control, health management, and academic achievement, we estimated parallel-process LGMs. The results indicated that higher initial game use was associated with lower initial levels of self-control, health management, and academic achievement. Moreover, adolescents whose game use increased more rapidly over time showed steeper declines in self-control, greater deterioration in health management, and sharper drops in academic performance. A specific indirect pathway from initial game use to the slope of health management via the linear slope of self-control was statistically significant, suggesting that heavy game use can undermine health management behaviors partly by eroding self-regulation. In contrast, a mediation model specifying the slope of academic achievement as the final outcome did not converge, limiting interpretation of the game use–self-control–achievement pathway. Taken together, the findings demonstrate that the quantitative aspect of game use—how long adolescents play—has meaningful longitudinal implications for the developmental trajectories of self-control, health management, and academic achievement. At the same time, parent–child communication and academic and peer stress displayed relatively stable mean levels, implying that in this sample family and peer relationships functioned more as background contexts than as domains that deteriorate rapidly in tandem with game use. Rather than framing game use solely as a pathological disorder, the results underscore the importance of early monitoring of initial game-use levels and growth rates, as well as preventive interventions aimed at strengthening self-control and digital literacy, to protect adolescents’ health, learning, and family and peer relationships.