Psychology Department, Suffolk University, Boston, MA 02114, United States.
J Sch Psychol. 2010 Apr;48(2):135-61. doi: 10.1016/j.jsp.2009.12.001. Epub 2010 Jan 13.
Peer nominations and demographic information were collected from a diverse sample of 1493 elementary school participants to examine behavior (overt and relational aggression, impulsivity, and prosociality), context (peer status), and demographic characteristics (race and gender) as predictors of teacher and administrator decisions about discipline. Exploratory results using classification tree analyses indicated students nominated as average or highly overtly aggressive were more likely to be disciplined than others. Among these students, race was the most significant predictor, with African American students more likely to be disciplined than Caucasians, Hispanics, or Others. Among the students nominated as low in overt aggression, a lack of prosocial behavior was the most significant predictor. Confirmatory analysis using hierarchical logistic regression supported the exploratory results. Similarities with other biased referral patterns, proactive classroom management strategies, and culturally sensitive recommendations are discussed.
从一个由 1493 名小学生组成的多样化样本中收集了同伴提名和人口统计信息，以考察行为（公然和关系攻击、冲动和亲社会行为）、背景（同伴地位）和人口统计特征（种族和性别）对教师和管理人员关于纪律处分决定的预测作用。使用分类树分析的探索性结果表明，被提名表现为中等或高度公然攻击性的学生比其他学生更有可能受到纪律处分。在这些学生中，种族是最显著的预测因素，非洲裔美国学生比白种人、西班牙裔或其他种族的学生更有可能受到纪律处分。在被提名表现出低公然攻击性的学生中，缺乏亲社会行为是最显著的预测因素。使用分层逻辑回归进行的验证性分析支持了探索性结果。讨论了与其他有偏见的转介模式、积极主动的课堂管理策略和文化敏感的建议的相似之处。
