方法对比

	探索性因子分析（EFA）	验证性因子分析（CFA）	主成分分析
领域	统计学	心理测量学	机器学习
方法族	Latent structure	Latent structure	Machine learning
起源年份	—	1969	2002
提出者	—	Karl Gustav Jöreskog	Jolliffe, I.T. (textbook); Pearson & Hotelling (origins)
类型	Latent variable / dimension reduction	Hypothesis-testing latent variable model	Unsupervised dimensionality reduction
开创性文献	Fabrigar, L. R., Wegener, D. T., MacCallum, R. C. & Strahan, E. J. (1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4(3), 272–299.	Jöreskog, K. G. (1969). A general approach to confirmatory maximum likelihood factor analysis. Psychometrika, 34(2), 183–202.	Jolliffe, I.T. (2002). Principal Component Analysis (2nd ed.). Springer.
别名	common factor analysis, açımlayıcı faktör analizi, factor analysis	CFA, confirmatory FA, measurement model, restricted factor analysis	Temel Bileşenler Analizi (PCA), PCA, principal components analysis, Karhunen-Loève transform
相关	4	4	3
摘要	Exploratory factor analysis reduces a large set of observed variables into a smaller number of latent common factors. It is widely used in scale development and psychometrics to uncover the dimensional structure that underlies a set of correlated items, without specifying that structure in advance.	Confirmatory factor analysis tests a researcher-specified factor structure against observed data. Unlike exploratory approaches, the researcher decides in advance which indicators load on which latent factor, and the model is evaluated by how closely the implied covariance matrix reproduces the sample covariance matrix. CFA is central to scale validation, construct validity assessment, and measurement invariance testing.	Principal Component Analysis (PCA) is an unsupervised dimensionality-reduction method — given its modern textbook treatment by Ian Jolliffe (2002) — that compresses high-dimensional data into fewer dimensions while preserving the maximum possible variance. It re-expresses correlated variables as a small set of uncorrelated principal components ordered by how much of the data's variation each one captures.