Incentive alignment in anchored MaxDiff yields superior predictive validity

• Published in: Marketing letters Vol. 36; no. 1; pp. 1 - 16
• Main Authors: Schramm, Joshua Benjamin, Lichters, Marcel
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.03.2025; Springer Nature B.V
• Subjects: Anchored maximum difference scaling (MaxDiff); Anchoring; Best-worst scaling (BWS); Business and Management; Conjoint analysis; Consumers; Incentive alignment; Incentives; Limitations; Market research methods; Marketing; Predictive validity; Preference measurement; Purchase intention; Usefulness; Variants; Best-worst scaling (BWS); Market research methods; Anchored maximum difference scaling (MaxDiff); Predictive validity; Preference measurement; Incentive alignment
• ISSN: 1573-059X; 0923-0645; 1573-059X
• DOI: 10.1007/s11002-023-09714-2

Abstract Maximum Difference Scaling (MaxDiff) is an essential method in marketing concerning forecasting consumer purchase decisions and general product demand. However, the usefulness of traditional MaxDiff studies suffers from two limitations. First, it measures relative preferences, which prevents predicting how many consumers would actually buy a product and impedes comparing results across respondents. Second, market researchers apply MaxDiff in hypothetical settings that might not reveal valid preferences due to hypothetical bias. The first limitation has been addressed by implementing anchored MaxDiff variants. In contrast, the latter limitation has only been targeted in other preference measurement procedures such as conjoint analysis by applying incentive alignment. By integrating anchored MaxDiff (i.e., direct vs. indirect anchoring) with incentive alignment (present vs. absent) in a 2 × 2 between-subjects preregistered online experiment ( n  = 448), the current study is the first to address both threats. The results show that incentive-aligning MaxDiff increases the predictive validity regarding consequential product choices—importantly—independently of the anchoring method. In contrast, hypothetical MaxDiff variants overestimate general product demand. The article concludes by showcasing how the managerial implications drawn from anchored MaxDiff differ depending on the four tested variants. In addition, we provide the first incentive-aligned MaxDiff benchmark dataset in the field.