Predicting Cyberbullying Behavior From Attitudes: A 3-Year Longitudinal Cross-Lagged Analysis of Singaporean Youth

There is a paucity of research testing (a) the longitudinal stability in positive cyberbullying attitudes, (b) whether any change in positive cyberbullying attitudes over time predict subsequent cyberbullying perpetration, and (c) the cross-lagged relations between positive attitudes toward cyberbullying attitudes and behavior over time. The current study focused on empirically testing these theoretical gaps and sampled over 3,000 Singaporean youth participants (at Wave 1) who completed measures of cyberbullying behavior and positive attitudes consecutively for 3 years. Correlations and path analyses showed modest stability in positive cyberbullying attitudes and perpetration over time. Also, latent class analysis classified participants into either stable high attitudes, stable low attitudes, increasing attitudes, or decreasing attitudes. Results using this classification showed that changes in positive cyberbullying attitudes across Waves 1 and 2 predicted Wave 3 cyberbullying, such that those who endorsed cyberbullying attitudes were more likely to cyberbully than those who did not advocate such attitudes. Finally, path analysis results showed significant longitudinal cross-lags between positive attitudes toward cyberbullying and behaviors.