Comparer des méthodes

	Plan de recherche ex post facto transversal	Conception Longitudinale Ex Post Facto
Domaine	Conception de la recherche	Conception de la recherche
Famille	Process / pipeline	Process / pipeline
Année d'origine	1964–1973	1964–1986 (Kerlinger 1964 first edition; Campbell & Stanley 1966)
Auteur d'origine	Fred N. Kerlinger (formalized ex post facto methodology)	Fred N. Kerlinger (systematized); Donald T. Campbell & Julian C. Stanley (quasi-experimental framework)
Type	Non-experimental quantitative research design	Non-experimental quantitative research design
Source fondatrice	Kerlinger, F. N. (1973). Foundations of Behavioral Research (2nd ed.). Holt, Rinehart and Winston. ISBN: 978-0030862731	Kerlinger, F. N. (1986). Foundations of Behavioral Research (3rd ed.). Holt, Rinehart and Winston. ISBN: 978-0030417498
Alias	cross-sectional causal-comparative design, retrospective cross-sectional design, after-the-fact cross-sectional study, cross-sectional EPF design	longitudinal causal-comparative design, longitudinal after-the-fact design, longitudinal retrospective design, LEPF design
Apparentées	4	5
Résumé	A cross-sectional ex post facto design investigates presumed causal relationships by comparing groups that already differ on a key characteristic — all measured at a single point in time. Because the independent variable (e.g., smoking history, prior educational attainment) has already occurred and cannot be manipulated, the researcher works backward from observed outcomes to infer probable antecedents. It is widely used in education, public health, and the social sciences when experimental control is ethically or practically impossible.	A longitudinal ex post facto design combines the time-depth of longitudinal research with the retrospective logic of ex post facto inquiry. Participants are grouped by a naturally occurring characteristic or past event — not randomly assigned — and then observed or measured at multiple points over time. The goal is to trace how pre-existing differences between groups unfold or predict outcomes across an extended period, without the researcher ever manipulating the independent variable.