ABSTRACT
The primary concern about violent media is their potential long-term harmful effects on children, which are investigated in longitudinal studies. In this article, we describe best practice procedures for longitudinal studies of violent media effects using the Lacko et al. (2024) study as a case study. Lacko et al. concluded that exposure to violent video games had no significant effect on aggression. However, a critical examination of their study reveals six major concerns: (1) inferring cause from correlational data, (2) inaccurate claims about the General Aggression Model, (3) an unexpected positive association between empathy and aggression, (4) questionable measurement validity, (5) excessive missing data, and (6) overparameterization in their longitudinal model. Using their publicly available raw data (https://osf.io/q5wv6/), we reanalyzed their study using two Random Intercept Cross-Lagged Panel Models (RI-CLPMs). The reanalysis revealed a strong positive association between violent video game exposure and aggression at the between-person level, but a negative association at the within-person level. No significant relationship was found between violent game exposure and empathy at the between-person level. However, a negative within-person effect of violent game exposure on subsequent cognitive empathy was observed, whereas no such effect was found for affective empathy. We discuss our results in comparison with the original results reported by Lacko et al. More generally, we offer recommendations for analyzing longitudinal data and for interpreting the potential relationships—or causal implications—of violent video game exposure on aggression.