結構方程模式的應用－驗證性因素分析

Application of Structural Equation Modeling: Confirmatory Factor Analysis

高美玲(Meei-Ling Gau) ； 葉美玲(Mei-Ling Yeh)

《護理研究》 7卷6期 (1999/12) Pp. 594-605

https://doi.org/10.7081/NR.199912.0594

摘要

驗證性因素分析(CFA)乃爲建立測量工具的建構效度的常用方法之一。然而，於國內的護理研究卻很少使用這個統計方法，國內研究者們在建立工具的建構效度上多半使用探索性因素分析(EFA)(96.7%)。而實際上，就統計學的角度而方，EFA的目的乃是在一組變項中找出他們之間的共同因素；因此，EFA較適用於一個測量工具剛建立時的結構探討及初步理論的形成。而CFA，其名如其法，乃適用於測試一個無假設好的因素架構，亦爲驗證理論的方法。本文的主要目的乃是對驗證性因素分析做一簡介並讀者論壇通信班下性因素分析應用時應攷慮的三個原則：模式的規劃、測試假設的羅輯性及測量模式的統計假定。最後，並以LISREL8統計軟體說明執行CFA的基本步驟。

關鍵字

驗證性因素分析 ； 探索性因素分析 ； 建構效度 ； 結構方程模式 ； LISREL

並列摘要

Confirmatory factor analysis (CFA) is one of the statistical methods available to establish the construct validity of a measurement tool. However, it is seldom used in the field of nursing research in Taiwan. Most researchers (96.7%) conduct exploratory factor analysis (EFA) to establish construct validity instead of CFA. In fact, from the statistical viewpoint, EFA is concerned with the question of how many factors are necessary to explain the relations among a set of indicators. Thus, the purpose of EFA is to explore the factor structure of an instrument and it is appropriate as an initial process of theory generation. The purpose of CFA, as the name implies, is to test hypotheses regarding the number of factors underlying the relations among a set of indicators; it is a theory-testing method. The purpose of this paper is to introduce CFA and also to discuss three major principles for conducting CFA: model specification, the logic of hypothesis testing, and the statistical assumptions of CFA. Lastly, it illustrates the basic steps of using the LISREL 8 computer program to perform CFA.

並列關鍵字

confirmatory factor analysis ； exploratory factor analysis ； construct validity ； structural equation modeling ； LISREL