The belief among personality researchers and mental health practitioners that behavior is a function of both personality and situation is generally accepted as an obvious truth. Yet, because aggression and violence are relatively rare in human interaction, the prediction of dangerousness has not been possible using clinical methods or personality measurement alone. As suggested recently by Mulvey and Cauffman (2001), school violence is difficult to predict because it typically occurs in a social context rather than in isolation, and changes rather dynamically with daily occurrences in the lives of adolescents, thus suggesting the importance of careful attention to situational variation. Predicting violence might be comparable to that of charting the path of a leaf (i.e., personality) in the wind (situation), where under some circumstances one or the other may give you the most useful information, sometimes one or the other may be best, and more often than not both will be most helpful.

Consider the following example of a 17-year-old frequent alcohol drinker and fighter who was recently arrested for allegedly attacking and robbing a fast food deliveryman. Due to his recent arrest and 5 weeks of sobriety, a threatening encounter with a rather large acquaintance led to a different encounter than usual. Cornered outside a restaurant, despite accusations that he had snitched on the antagonist, direct invitations to fight, and threats to “bitch slap” his girl friends in front of male companions, the 17-year-old did not fight.