Carter Stephen R
Faculty of Pharmacy, University of Sydney, Sydney, Australia.
Int J Clin Pharm. 2016 Jun;38(3):731-7. doi: 10.1007/s11096-016-0302-9. Epub 2016 May 4.
Background Confirmatory factory analysis (CFA) and structural equation modelling (SEM) are increasingly used in social pharmacy research. One of the key benefits of CFA is that it allows researchers to provide evidence for the validity of internal factor structure of measurement scales. In particular, CFA can be used to provide evidence for the validity of the assertion that a hypothesized multi-dimensional scale discriminates between sub-scales. Aim This manuscript aims to provide guidance for researchers who wish to use CFA to provide evidence for the internal factor structure of measurement scales. Methods The manuscript places discriminant validity in the context of providing overall validity evidence for measurement scales. Four examples from the recent social pharmacy literature are used to critically examine the various methods which are used to establish discriminant validity. Using a hypothetical scenario, the manuscript demonstrates how commonly used output from CFA computer programs can be used to provide evidence for separateness of sub-scales within a multi-dimensional scale. Conclusion The manuscript concludes with recommendations for the conduct and reporting of studies which use CFA to provide evidence of internal factor structure of measurement scales.
背景
验证性因子分析（CFA）和结构方程建模（SEM）在社会药学研究中的应用越来越广泛。CFA的一个关键优势在于，它能使研究人员为测量量表内部因子结构的有效性提供证据。特别是，CFA可用于为如下断言的有效性提供证据：一个假设的多维量表能够区分各子量表。
目的
本文旨在为希望使用CFA为测量量表的内部因子结构提供证据的研究人员提供指导。
方法
本文将区分效度置于为测量量表提供整体效度证据的背景下。通过近期社会药学文献中的四个例子，对用于建立区分效度的各种方法进行批判性审视。利用一个假设场景，本文展示了如何使用CFA计算机程序的常用输出结果，为多维量表内各子量表的独立性提供证据。
结论
本文最后针对使用CFA为测量量表的内部因子结构提供证据的研究的开展和报告提出了建议。
