方法对比

	Adaptive Conjoint Analysis	MaxDiff / Best-Worst Scaling
领域	Marketing Research	Marketing Research
方法族	Process / pipeline	Regression model
起源年份	1987	2005
提出者	Richard M. Johnson (Sawtooth Software)	Jordan J. Louviere; A. A. J. Marley & Jordan Louviere
类型	Computer-adaptive conjoint combining self-explication and paired comparisons	Best-worst choice task for scaling relative importance of items
开创性文献	Green, P. E., Krieger, A. M., & Agarwal, M. K. (1991). Adaptive Conjoint Analysis: Some Caveats and Suggestions. Journal of Marketing Research, 28(2), 215-222.	Louviere, J. J., Flynn, T. N., & Marley, A. A. J. (2015). Best-Worst Scaling: Theory, Methods and Applications. Cambridge: Cambridge University Press.
别名	ACA, Adaptive Conjoint, Computer-Adaptive Conjoint, Self-Explicated and Paired-Comparison Conjoint	MaxDiff, Best-Worst Scaling, BWS, Maximum Difference Scaling
相关	4	4
摘要	Adaptive Conjoint Analysis (ACA) is a hybrid, computer-administered conjoint method that builds each respondent's part-worth utilities by combining a self-explicated priors stage with a sequence of adaptively chosen paired-comparison trade-offs. Developed by Richard Johnson at Sawtooth Software in the mid-1980s, ACA was designed to handle many more attributes than a respondent could realistically evaluate in full-profile or choice tasks. The interview first asks people to rate the desirability of attribute levels and the importance of attributes, then uses those answers to generate paired product comparisons that are roughly balanced in utility, which are the most informative trade-offs. Respondents indicate graded preference between each pair, and the program updates the utilities in real time, focusing later questions where uncertainty is greatest. Green, Krieger, and Agarwal's 1991 evaluation in the Journal of Marketing Research documented both ACA's strengths and important caveats about its self-explicated component and attribute-importance estimates. ACA produces individual-level utilities that can drive purchase-likelihood calibration and market simulation.	MaxDiff, also known as best-worst scaling (BWS), measures the relative importance or preference of a set of items by repeatedly asking respondents to identify the best (most important or most preferred) and worst (least) item within small subsets. Introduced by Jordan Louviere and formalized by Marley and Louviere's 2005 probabilistic models, the method exploits the fact that people are far better at picking extremes than at rating many items on a scale. Each best-worst judgment reveals the maximum-difference pair in a set, and across many balanced subsets the choices pin down a single interval scale of item utilities. Because every respondent is forced to make trade-offs, MaxDiff sidesteps the scale-use bias and lack of discrimination that plague rating grids, where respondents often call everything important. Item scores can be computed by simple best-minus-worst counts or, more rigorously, by fitting a multinomial logit choice model, with hierarchical Bayes yielding individual-level, probability-scaled importances. The result is a clear, discriminating ranking of items that supports prioritization, segmentation, and feature selection.