多维测评工具聚敛和区分效度的SEM分析——以领悟社会支持量表为例  

On Analysis of Convergent and Discriminant Validity of Multidimensional Measurement Tools with the Perceived Social Support Scale as an Example

作　　者：陈维[1,2] 黄程琰[2] 毛天欣 罗杰[1] 张进辅[2] 

机构地区：[1]贵州师范大学教育科学学院/贵州省普通高校基础心理与认知神经科学重点实验室,贵阳550001 [2]西南大学心理学部,重庆400715

出　　处：《西南师范大学学报（自然科学版）》2016年第2期136-140,共5页Journal of Southwest China Normal University(Natural Science Edition)

摘　　要：目的:以领悟社会支持量表实例,探讨了量表聚敛效度和区分效度的评估方法,为测评工具效度的评估提供一种新的思路.方法:用领悟社会支持量表对1 500名在校高中生进行问卷调查,并计算用于评价聚敛和区分效度的各项指标,如因素负荷量、组合信度、平均方差抽取量、卡方值和相关系数的Bootstrap置信区间.结果:各项目的因素负荷量均大于0.5,各分量表的组合信度均大于0.7,其他支持和朋友支持的相关系数的Bootstrap区间包含1.结论:朋友支持的聚敛效度较好,其他支持和家庭支持的聚敛效度相对较差,但达到了可接受水平;朋友支持与家庭支持的区分效度较好,但其他支持与朋友支持和家庭支持的区分效度均较差,量表的改善可针对项目1和11进行修订.This article deals with the exploration of the evaluation methods of convergent and discriminant validity with an example,in order to provide a new idea for assessing the validity.1500 students have been surveyed with the perceived social support scale.Then,the indicators of convergent and discriminant validity,such as factor loading,average variance extracted,composite reliability,chi-square value and the bootstrap confidence interval of correlation coefficient have been calculated.To all items,the factor loadings are big than 0.5.The composite reliability of all sub-scales are greater than 0.7.The 1is among the bootstrap confidence interval of correlation coefficients of others and friends support.The generalizability coefficient and reliability index were about 0.90.The convergent validity of friends support is good,but the validity of others and family supports are relatively bad.The discriminant validity is good among the friends and family support,but it is relatively poor among others and friends support.We can revise the item 1and 11 for improving the scale.

关 键 词：聚敛效度 区分效度 因素负荷 平均抽取量 组合信度 

分 类 号：B841[哲学宗教—基础心理学]