Educational big data mining: Mediation of academic performance in crime among digital age young adults

Abstract

The connection between academic performance and crime has been proven in many studies, but there is a lack of recent literature documenting academic performance as a mediator between predictors and crime, especially in a large population that has grown up in the digital age of which the high frequency of communication exchange differentiates it from previous times. The present study aims to discover the mediating role of academic performance in the connections between each predictor (the Internet usage, biological sex, duration of sleep, fast food intake, medical history, alternate drug use, and illegal drug use) and crime using a large data set from the national longitudinal study of adolescent to adult health (>90,000 sample size). The bivariate correlation between each predictor and crime is presented, as well as the direct and indirect effects for each predictor on crime through academic performance as the mediator. The results highlighted that there is a significant relationship between the Internet usage with crime, a negative correlation of academic performance with crime and positive correlations of fast food intake, alternative medicine use, and illegal drug use with crime. In terms of mediation analyses, academic performance mediated the relationship between all the predictors (except for duration of sleep) and crime. Results may aid education policy makers in deterring crime in the digital age by improving academic performance while targeting its predictors.