So sánh phương pháp

	Thiết kế thử nghiệm Tiền kiểm-Hậu kiểm Thích ứng	Thiết kế thực nghiệm Crossover Pretest-Posttest
Lĩnh vực	Thiết kế thí nghiệm	Thiết kế thí nghiệm
Họ	Process / pipeline	Process / pipeline
Năm ra đời	2000s (integration of adaptive principles with classic pre-post structure)	1963 (Campbell & Stanley framework); crossover methodology formalized 1980s–2000s
Người khởi xướng	Synthesizes Campbell & Stanley (1963) pretest-posttest framework with adaptive design methodology formalized by Chow & Chang (2000s)	Donald T. Campbell & Julian C. Stanley (pretest-posttest framework); Stephen Senn (crossover trial methodology)
Loại	Experimental design	Within-subjects experimental design
Công trình gốc	Campbell, D. T., & Stanley, J. C. (1963). Experimental and Quasi-Experimental Designs for Research. Rand McNally.	Senn, S. (2002). Cross-over Trials in Clinical Research (2nd ed.). Wiley. ISBN: 978-0471496533
Tên gọi khác	adaptive pre-post design, adaptive pretest-posttest trial, adaptive two-period design, pre-post adaptive experiment	within-subjects pretest-posttest design, repeated-measures crossover design, AB/BA pretest-posttest design, crossover repeated-measures design
Liên quan	6	5
Tóm tắt	An adaptive pretest-posttest experimental design measures all participants before and after an intervention while allowing pre-specified modifications to the trial — such as sample size re-estimation, treatment arm dropping, or randomization ratio adjustment — based on accumulated interim data. It combines the interpretive power of change-score analysis with the efficiency gains and ethical safeguards of adaptive methodology, making it particularly valuable in clinical, educational, and behavioral research where early data can inform better resource allocation.	A crossover pretest-posttest experimental design is a within-subjects experiment in which each participant receives two or more treatments in a randomized sequence, with outcome measurements taken both before and after each treatment period. By serving as their own control across conditions, participants allow direct intra-individual comparison, dramatically increasing statistical power while reducing the sample size required relative to a parallel-group design.