Psychological Science ›› 2011, Vol. 34 ›› Issue (3): 625-630.

Previous Articles     Next Articles

A Forecasting model of Children’s Peer Relationship Constructed by Latent Variables

  

  • Received:2010-06-11 Revised:2010-10-28 Online:2011-05-20 Published:2011-06-19

儿童同伴关系的隐变量预测模型

张奇勇   

  1. 扬州大学教育科学学院
  • 通讯作者: 张奇勇

Abstract:

Abstract: The preceding researches on the factors impacting on children’s peer relationship show a general tendency of elementalism. In order to study the factors systematically, we selected 4 questionnaires and 6 measurable variables on the basis of prevenient researches, and adopted 478 available questionnaires from elementary school children from the forth grade to the sixth. The results indicate that, the first factor impacting on peer relationship is school-label, the sequent variable is competence trait and anxiety, the above three factors’ cumulative percentage of variance to the model is 74.887%. Confirmatory factor analysis shows the model has a good model fit and all regression weights are significant. The three latent variables can explain the variance of ZLM (Z score of Like Most) by 62.2%, and explain the variance of ZLL (Z score of Like Least) by 47.4%. The model has a good forecasting effect. School-label takes on a most powerful forecasting effect among three latent variables.

摘要:

摘要:以往对影响儿童同伴关系的因素的研究具有明显的元素主义倾向,为系统研究影响儿童同伴关系的因素,在总结以往研究成果的基础上,选用了4大量表和6个可测量指标,收集了478名小学四至六年级学生的有效问卷和可测量指标。统计结果表明,影响儿童同伴关系的首要因素是学校标签,然后依次是能力特质和焦虑,三个因素对模型的方差累积贡献率为74.887%;验证性因素分析模型的拟合度高,路径系数均极其显著;三个隐变量能够解释ZLM(同伴正向提名Z分数)62.2%的变异,能够解释ZLL(同伴反向提名Z分数)47.4%的变异,其中学校标签的预测力最强。