Predictors of injurious assault committed during or after drinking alcohol: a case-control study of young offenders

Studies of causal links between alcohol and aggression are often handicapped by threats to internal and external validity. Case-control methods employ an event-level analysis that can reduce some of these validity threats by the use of within-subject controls. This study used a case-control approach, asking 39 male inmates in a Young Offenders' Institution to compare drinking behaviour before incidents where they reported commission of an injurious assault and a matched incident where they did not. After controlling contextual differences, participants reported personally drinking more heavily and heavier drinking within their group, but not being more impaired when an assault was committed. The assault incidents were more likely to involve spontaneous, rather than planned, drinking and a higher proportion of males in the group. They were less likely to involve drinking in a pub. Our confirmation of previous findings using a case-control methodology strengthens those findings. Limitations of this methodology are also discussed.