正负性表述的方法效应:以核心自我评价量表的结构为例  

The Method Effect of Positively and Negatively Worded Items: The Case of Core Self-Evaluations Scale

作　　者：张春雨[1] 韦嘉[1] 赵清清[1] 张进辅[1] 

机构地区：[1]西南大学心理学部,重庆400715

出　　处：《心理学探新》2015年第1期78-83,共6页Psychological Exploration

基　　金：国家重点学科基础心理学科研基金重点项目(西国重04003)

摘　　要：为了平衡反应偏向,负性表述题目被引入量表,但这可能增加新的方法变异。本研究以核心自我评价量表(CSES)为例,检验正负性表述方法效应对其结构的影响。研究表明,CSES在结构效度和效标关联效度上均受到了表述方法的潜在影响。本研究支持了控制方法效应后的CSES单因子结构。本研究不仅为一直无法统一的CSES中文版提供了新的解释,也为自陈式心理测量工具在设置负性题目方面提供了新的思考。Negatively worded items were used to balance the response bias in psychological assessment. However, it may result in new method variances. In the present study, we examined the importance of this method effect by applying evolving methodological approaches to psychological assessment to resolve competing claims about the factor structure of a widely used instrument： the Core Self - Evaluations Scale（CSES）. Further,we examined the extent to which the method effect affects the relationship between the CSES and life satisfaction. We investigated 1223 college students from five universities. We first conducted exploratory factor analysis with a half of the sample（ N1 = 626 ）and obtained two substantively trait factors（ positive and negative core self- evaluation）. Next ,we posited five alternative models using another half of the sample（ N2 = 597 ）. The results provided no support for the unidimensional model with no method effect and the 2 -factor model ,but support for models positing one substantive（ core self- evaluation）factor and response -style method factors. Additionally,the results supported that the method effect may affect the relationship between core self - evaluation and life satisfaction. Our findings resolve the unstable factor solution of the Chinese version CSES and support the wording effect in the self - reported questionnaires with a mix of positively and negatively worded items.

关 键 词：核心自我评价 正负性表述 方法效应 共同方法变异 效标关联效度 

分 类 号：B841.2[哲学宗教—基础心理学]