Study Finds Predictors For Youth Violence And Drugs

The research, done long before the Columbine shootings by a group at the USC medical school's Institute for Health Promotion and Disease Prevention Research (IPR), followed groups of at-risk high school students who had completed detailed questionnaires asking questions about a wide spectrum of behavior and attitudes, looking for variables that predicted violent behavior or drug taking in the subsequent year. The results appear in forthcoming papers in two journals.

The racially mixed group of students studied, boys (55 percent) and girls between the ages of 15 and 19, were all from Southern California "continuation" high schools - special schools for youth with discipline or academic problems. Unlike ordinary high school populations in which violence is a rare exception, for these students it was a rule: fully 70 percent reported having been involved in violence against either persons or property in their first interview.

In a set of initial questionnaires, researchers asked questions about a wide range of behavior, attitudes and activities. In the follow-up interviews a year later, the same students were asked about their experience with violent episodes since the initial interview. The student's original responses were then tabulated to see which were associated, within the sample, with renewed violence.

Since the group was one in which violence had previously been common, statistical tests were applied to cancel out what Steve Sussman, associate professor at the IPR and lead author on both papers refers as the high "baseline" violence of the sample, to see which variables could be viewed as predictors of continued violence.

A number of variables stood out. One in particular had never been systematically studied before: self-identification with on-campus subgroups or social sets.

The USC researchers found that students who identified themselves as part of certain "high risk" campus groups showed a high tendency over the course of the subsequent year to become involved with violence, drugs, or both.

The "high risk" groups included "gang members" and "stoners."

Other categories listed on the questionnaire included identification terms such as "brains," "jocks," "popular," and "loners."

In the follow-up reports, the high risk groups reported significantly more alcohol, marijuana and other drug use than others (though not more cigarette smoking); and a much higher number of reports of violence, both offensive and defensive. The high risk group was particularly like to be found carrying weapons onto campus.

"While it may not seem surprising that students who say they identify with high risk group like gang members should be found to be involved in drugs and violence," commented Sussman, "the degree to which this is true, and particularly, the degree to which violence can be predicted from this identification, has never been scientifically studied."

This element of the research will appear in a forthcoming issue of the journal Psychology of Addictive Behavior.

In addition to violence the identification with high risk groups was a predictor of later drug use.

It has been speculated, Sussman added, that students who self-identify as loners or outsiders were violence risks. "Our data doesn't show that," he said. Sussman's co-authors on the study were Clyde W. Dent, associate professor at IPR; and project data analyst William J. McCullar, a graduate student at California State University Los Angeles.

Other variables found in initial interviews to be associated during the study period with violence included a range of items, discussed along with group identification in a separate article in a forthcoming issue of the American Journal of Health Behavior.

Most obviously, members of the group who had initially reported to have engaged in violent behavior in the past were likely to continue to engage in such behavior.

The study asked a number of questions about drug use, including frequency of use of a range of other drugs, from cigarettes and alcohol to methamphetamine and heroin. Surprisingly, the only drug whose use was found to significantly predict later violence, when contrasted with other drug use predictors was marijuana.

In the study the authors comment that "the relevance of marijuana use to the perpetuation of violence has been debated for at least 60 years in the United States. Some work has suggested that marijuana use might inhibit expression of aggression, but most studies do indicate a positive association between marijuana use and violence perpetuation. Another predictor was having been a victim of either of violent attack or having had property stolen or damaged.

Also measured at the same time was a high feeling of vulnerability to physical attack. All these were predictive of violent behavior, as was the expression of the belief by students that the police could not protect them. A logical concomitant of such a belief, carrying weapons, was also a violence predictor.

The students were also asked in earlier interviews whether they went to places - parties, streets, parks - which they knew were unsafe. An affirmative answer on this question was also associated with violence reported in the follow-up interviews.

The large role played by victimization as a predictor of future violence suggested to the researchers that these youth may become embedded in a lifestyle infected with violence, or that they simply learn to do as has been done to them.

The authors of the study append no less than eight limitations of the study, ranging from the racial heterogeneity of the study sample to the problems involved in self-reporting of violence, to the study not checking its findings against school records of actual events.

"While prospective studies such as the present one are sorely needed," the authors say, "theoretically rich studies are imperative to better understand the roots of adolescent violence."

Working with Sussman on this study were Dent; IPR research associate professor Alan W. Stacy; the recently graduated Thomas R. Simon Ph.D; and graduate student Jill M. Steinberg.

Both studies were supported by the National Institute on Drug Abuse.