An Effective Approach to Violence Prevention: Traditional Martial Arts in Middle School

Adolescence, Fall, 2001 by Gail Zivin, Nimr R. Hassan, Geraldine F. DePaula, Daniel A. Monti, Carmen Harlan, Kashfia D. Hossain, Ksai Patterson

The boys who took the course in the first semester improved over baseline on 12 variables. The waiting boys improved over baseline on only 5 variables, and to a comparatively lower degree. The waiting boys deteriorated from baseline on teacher-rated violence and inappropriate social behavior, on self-reported positive behavior, happiness, and schoolwork, and on three of four attention measures. At minimum, it might be that the martial arts course prevented, for the boys who took the course first, a common trend toward deterioration shown by the waiting boys at the end of the first semester.

Deterioration of the boys on the waiting list appears to have been halted by their taking the martial arts course in the second semester. In terms of teacher-rated conduct, Group B improved greatly over their baselines: they showed postcourse ratings that were remarkably similar to those of Group A. When both groups were pooled after having taken the course, statistically significant improvements were seen on teacher-rated resistance to rules, inappropriate social behavior, and impulsivity--although not on violence (apparently, the primary effect was the reversal of Group B's increased violence from baseline to end of the first semester). Thus, according to teachers' ratings, the two groups appear comparably improved from baseline after the same treatment. Group B's improvement might be seen as a small internal replication of Group A's success.

Design

Because of the low power of the t tests and the low representativeness of our small ns, it would be premature to emphasize those variables showing statistically significant results over those that did not. It is worth noting, however, that on the six psychological risk factors often found correlated with violence, there were modest (p < .012) to strong (p < .002) significant results (but only four had ps below their Bonferroni alpha). Most confidence can be placed in the more general finding: many risk factors improved among these high-risk middle school boys after as few as 30 classes in martial arts training that emphasizes nonaggression, respect, peaceful philosophy, and meditation.

Some limitations of the study dictate that even the findings in regard to consistent improvement patterns be considered tentative. The control group (Group B during Semester 1) was imperfect in three ways. First, it was not given placebo attention during the first semester (when waiting to take the martial arts course). Second, being placed in the control group may have been perceived as a negative experience. Although they had been told that it was due to random drawing, they may have resented the wait, which detracted from their second measurement scores. Third, while the groups were matched for problem types and intensities, the same was not true for age and grade. The 10 eighth graders and their matched controls were not randomly assigned--all eighth graders were placed in Group A to ensure that they took the course before graduation. Recall that the significant differences that were found depended on the poor performance of the controls as well as on the course takers' improvement. Perhaps if an equal nu mber of eighth graders had been wait-listed (i.e., placed in Group B), the two groups might have performed more similarly.