Methoden vergleichen

	Single-Blind A/B-Test	Adaptiver A/B-Test
Fachgebiet	Versuchsplanung	Versuchsplanung
Familie	Process / pipeline	Process / pipeline
Entstehungsjahr	mid-20th century (blinded RCT framework); A/B test nomenclature ~1990s–2000s	1952 (Robbins); applied to A/B testing from ~2010s onward
Urheber	Fisher, R. A. (randomisation basis); blinding practice formalised in clinical trials mid-20th century	Herbert Robbins (bandit framework); Thompson Sampling formalized by William R. Thompson
Typ	Controlled experiment with partial blinding	Adaptive experimental design
Wegweisende Quelle	Kohavi, R., Tang, D., & Xu, Y. (2020). Trustworthy Online Controlled Experiments: A Practical Guide to A/B Testing. Cambridge University Press. ISBN: 978-1108724265	Russo, D., Van Roy, B., Kazerouni, A., Osband, I., & Wen, Z. (2018). A Tutorial on Thompson Sampling. Foundations and Trends in Machine Learning, 11(1), 1–96.
Aliasnamen	single-masked A/B test, single-blind split test, blinded two-condition experiment, participant-blind A/B test	adaptive AB test, bandit A/B test, multi-armed bandit testing, online adaptive experiment
Verwandt	6	6
Zusammenfassung	A single-blind A/B test is a controlled two-condition experiment in which participants are randomised to condition A (control) or condition B (treatment) but are kept unaware of which condition they have received, while researchers and analysts remain aware. The blind prevents participants from changing their behaviour in response to knowledge of their assignment, reducing demand characteristics and response bias while still allowing the investigator to monitor the trial.	An Adaptive A/B test is an experimental design that dynamically reallocates traffic or participants toward better-performing variants during the experiment itself, rather than holding allocations fixed until the end. Drawing on multi-armed bandit algorithms such as Thompson Sampling or Upper Confidence Bound (UCB), it balances the exploration of uncertain variants with the exploitation of those already showing superior performance, typically yielding higher aggregate outcomes while still producing valid inferential conclusions.