Sammenlign metoder
Gennemgå dine valgte metoder side om side; rækker, der afviger, er fremhævet.
| Latent Class Analysis (LCA)× | Konstruktiv faktoranalyse (CFA)× | |
|---|---|---|
| Fagområde≠ | Statistik | Psykometri |
| Familie | Latent structure | Latent structure |
| Oprindelsesår≠ | 1950s–1968 | 1969 |
| Ophavsperson≠ | Paul F. Lazarsfeld | Karl Gustav Jöreskog |
| Type≠ | Latent variable / person-centered classification | Hypothesis-testing latent variable model |
| Oprindelig kilde≠ | Goodman, L. A. (1974). Exploratory latent structure analysis using both identifiable and unidentifiable models. Biometrika, 61(2), 215–231. DOI ↗ | Jöreskog, K. G. (1969). A general approach to confirmatory maximum likelihood factor analysis. Psychometrika, 34(2), 183–202. DOI ↗ |
| Aliasser | LCA, latent class model, latent categorical analysis, finite mixture of multinomials | CFA, confirmatory FA, measurement model, restricted factor analysis |
| Relaterede≠ | 6 | 4 |
| Resumé≠ | 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. | 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. |
| ScholarGateDatasæt ↗ |
|
|