方法对比

	多项式项目分析	验证性因子分析（CFA）
领域	心理测量学	心理测量学
方法族	Latent structure	Latent structure
起源年份	1969–1982	1969
提出者	Fumiko Samejima (graded response model, 1969); David Andrich (rating scale model, 1978); Geoffrey Masters (partial credit model, 1982)	Karl Gustav Jöreskog
类型	Item-level psychometric analysis	Hypothesis-testing latent variable model
开创性文献	Samejima, F. (1969). Estimation of latent ability using a response pattern of graded scores. Psychometrika Monograph Supplement, 34(4, Pt. 2), 1–97.	Jöreskog, K. G. (1969). A general approach to confirmatory maximum likelihood factor analysis. Psychometrika, 34(2), 183–202.
别名	ordered-category item analysis, graded response analysis, polytomous IRT, rated-scale item analysis	CFA, confirmatory FA, measurement model, restricted factor analysis
相关	4	4
摘要	Polytomous item analysis examines the psychometric behavior of items that have more than two ordered response categories — such as Likert-type scales or partial-credit tasks. It evaluates each item's difficulty thresholds, discriminating power, and category functioning to determine whether the full response scale is being used as intended and whether each item contributes reliably to measuring the underlying construct.	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.