Comparer des méthodes

	Recherche causale-comparative multivariée	Recherche corrélationnelle multivariée
Domaine	Conception de la recherche	Conception de la recherche
Famille	Process / pipeline	Process / pipeline
Année d'origine	Mid-20th century onward; multivariate extension systematized 1970s–1990s	1920s–1930s (multivariate extensions); consolidated in applied social science by 1970s
Auteur d'origine	Extension of causal-comparative tradition (cf. Chapin, 1947; Gay, Mills & Airasian)	Developed from Galton and Pearson's bivariate correlation work, extended to multivariate contexts by R.A. Fisher, Harold Hotelling, and others
Type	Quantitative non-experimental comparative design	Non-experimental quantitative research design
Source fondatrice	Fraenkel, J. R., Wallen, N. E., & Hyun, H. H. (2019). How to Design and Evaluate Research in Education (10th ed.). McGraw-Hill. ISBN: 978-1260085594	Tabachnick, B. G., & Fidell, L. S. (2019). Using Multivariate Statistics (7th ed.). Pearson. ISBN: 978-0134790541
Alias	multivariate causal-comparative design, MANOVA causal-comparative study, multi-outcome ex post facto research, multivariate ex post facto design	multivariate correlational design, multivariate relational research, multiple-variable correlational study, multivariate associational research
Apparentées	6	2
Résumé	Multivariate causal-comparative research is a quantitative, non-experimental design that investigates whether pre-existing group differences (defined by a naturally occurring categorical variable) are associated with differences across multiple outcome variables considered simultaneously. By extending the classic causal-comparative framework to several dependent variables at once, it reduces Type I error inflation and captures the correlated structure of outcomes that univariate comparisons would miss.	Multivariate correlational research is a non-experimental quantitative design that examines the simultaneous associations among three or more variables. Rather than manipulating conditions, the researcher measures naturally occurring variables and uses techniques such as multiple regression, canonical correlation, or structural equation modeling to map the pattern and strength of their interrelationships. It is the dominant design when the goal is to understand how a set of predictors jointly relates to one or more outcome variables.