手法を比較

	順序ロジスティック回帰（比例オッズモデル）	潜在クラス分析 (LCA)
分野	統計学	統計学
系統	Regression model	Latent structure
提唱年	2010	1950s–1968
提唱者	Agresti (textbook treatment); proportional odds model	Paul F. Lazarsfeld
種類	Ordinal logistic regression	Latent variable / person-centered classification
原典	Agresti, A. (2010). Analysis of Ordinal Categorical Data (2nd ed.). Wiley.	Goodman, L. A. (1974). Exploratory latent structure analysis using both identifiable and unidentifiable models. Biometrika, 61(2), 215–231.
別名	proportional odds model, ordered logit, ordinal logistic regression, Ordinal Regresyon (Proportional Odds)	LCA, latent class model, latent categorical analysis, finite mixture of multinomials
関連	5	6
概要	Ordinal logistic regression models an ordered categorical outcome — such as a Likert rating, a satisfaction level, or an education tier — as a function of predictors. It is the ordinal extension of logistic regression, developed in standard treatments such as Agresti's Analysis of Ordinal Categorical Data (2010), and in its most common form it is the proportional odds model.	Latent class analysis identifies unobserved subgroups — latent classes — within a population by finding patterns of responses across a set of categorical observed indicators. It is the categorical-variable counterpart of cluster analysis, but grounded in an explicit probabilistic model, and is widely used in social, health, and behavioral sciences to discover typologies in survey or diagnostic data.