Paetsch J J, Bertrand L D
Canadian Research Institute for Law, University of Calgary, Alberta, Canada.
J Sch Health. 1997 Jan;67(1):27-32. doi: 10.1111/j.1746-1561.1997.tb06291.x.
This article examines the relationships between self-reported delinquency and peer, social, and school factors among junior and senior high school students in Calgary, Alberta, Canada, using both cross-tabulation and multiple-regression analyses. A greater number of ninth grade students were delinquent compared to students in the other grades studied. Predictably, peers' level of delinquency correlated strongly and positively with students' level of delinquency. Strong, positive correlations with delinquency also were found for students reporting high levels of involvement in sports, and high levels of involvement in commercial/entertainment activities. Multivariate analyses indicated the strongest predictor of delinquent behavior within the group of peer and social variables was peer delinquency. Within the group of school variables, the strongest predictors of delinquent behavior were how often classes were skipped and the number of times students were suspended from school. However, students' grades in school did not significantly increase the accuracy of the regression equation.
本文运用交叉列表分析和多元回归分析方法，研究了加拿大艾伯塔省卡尔加里市初中和高中学生自我报告的犯罪行为与同伴、社会及学校因素之间的关系。与其他被研究年级的学生相比，九年级学生中有更多人存在犯罪行为。不出所料，同伴的犯罪水平与学生的犯罪水平呈强正相关。对于那些报告称大量参与体育活动以及大量参与商业/娱乐活动的学生，其犯罪行为也呈现出强正相关。多变量分析表明，在同伴和社会变量组中，犯罪行为的最强预测因素是同伴犯罪。在学校变量组中，犯罪行为的最强预测因素是逃课频率以及学生被学校停学的次数。然而，学生的在校成绩并未显著提高回归方程的准确性。
