中文版Grasmick自我控制量表在罪犯群体的跨样本信效度检验  

Reliability and validity test of Chinese Grasmick’s Self-Control Scale across multiple samples of offenders

作　　者：杨学锋[1] 杨茗美 YANG Xuefeng;YANG Mingmei(Department of Fundamentals of Public Security,Criminal Investigation Police University of China,Shenyang 110035;Department of Sociology,University of Macao,Macao 999078)

机构地区：[1]中国刑事警察学院公安基础部,沈阳110035 [2]澳门大学社会学系,中国澳门999078

出　　处：《心理研究》2023年第1期73-82,共10页Psychological Research

摘　　要：为促进本土犯罪学对于自我控制理论的实证研究,基于总容量为1219的四个罪犯样本对中文版Grasmick自我控制量表进行了跨样本的信效度检验。研究发现,尽管内部一致性指标和探索性因子分析等经典检验方法证实了该量表具有良好的信度和一阶单维度结构,然而,验证性因子分析表明,单维度假设只能在二阶模型下才能得到经验支持;此外,多组验证性因子分析表明,从原本的24道题目中挑选出来的18道题目所构成的修订中文版Grasmick量表的二阶单因子模型同时满足了因子结构恒等假设以及因子载荷恒等假设。因此,在本土语境下使用中文版Grasmick量表时,不应再延续简单加总的计分方式,而应该优先选用由18道题目构成的修订版量表的二阶单因子测量模型。In order to ameliorate indigenous researches of the self-control theory,the measurement qualities such as reliability and construct validity of the Chinese Grasmick’s scale were examined across four samples of offenders with overall size of 1219.It was found that classical tests such as internal consistency index and exploratory factor analysis seemed to confirm the scale with good reliability and first-order unidimensional structure,however,the confirmatory factor analysis suggested that only under the second-order measurement model could the unidimensional hypothesis be empirically supported;Moreover,multiple-group confirmatory factor analysis indicated that the second-order single factorial model of the revised Chinese Grasmick’s scale,which only consists of 18 items instead of the original 24 items,satisfied both factor structure and factor loading invariance hypotheses.In conclusion,upon using the Chinese Grasmick’s scale in indigenous context,the simple adding-up method should not be continued,and instead the second-order single-factor measurement model should be constructed from the revised 18-item Chinese Grasmick’s scale.

关 键 词：自我控制 中文版Grasmick量表 信度 结构效度 测量恒等性 

分 类 号：B841.7[哲学宗教—基础心理学]