摘要
针对目前利用多源数据的测量方法,指出现有研究方法存在的问题和可能导致的结果偏差,提出应对测量同一事物(构念、变量)的多源数据进行合成,并基于量表信度和结构效度检验的思想,给出其合成的合理性和可行性,探讨来自不同测评方所含信息的权重设计,并以变革型领导和个体创造力的研究为例进行分析.
In order to integrate the data information from multiple sources, this paper suggests an approach to produce a multiple-source measurement variable based on the principles of reliability and validity. Firstly, we discuss the rationality and feasibility of the synthetic method; then we explore the approaches of determining weights; finally, we apply the synthetic method of multiple-source data to the study of transformational leadership and individual creativity.
出处
《数学的实践与认识》
CSCD
北大核心
2011年第9期76-83,共8页
Mathematics in Practice and Theory
基金
中国科学院研究生院院长基金(085101UM03)
关键词
多源数据
合成方法
信度分析
结构效度
因子分析
个体创造力
multiple-source data
synthetic method
reliability
construct validity
factor analysis
individual creativity
