Market Research: Foundations to Practice · Advanced Quantitative Techniques
MaxDiff (Best-Worst Scaling)
MaxDiff has become the workhorse method for prioritizing a long list of items precisely because it's easier on respondents than conjoint while still forcing genuine trade-offs. This chapter covers how it actually works.
MaxDiff, also called best-worst scaling and developed by researcher J.J. Louviere, presents respondents with a subset of items from a longer list and asks them to identify which item is most preferred (best) and which is least preferred (worst) within that specific subset, repeating this task across several different subsets to build a complete picture of preference across the entire original list [22].
Key Takeaways
- MaxDiff (best-worst scaling), developed by J.J. Louviere, asks respondents to identify the best and worst item within repeated subsets of a longer list.
- This repeated best-worst choice across subsets produces a genuine rank-ordering based on revealed choices, directly solving the 'everything is important' problem.
- MaxDiff is generally less mentally taxing per choice than conjoint analysis, since it evaluates simple items rather than full multi-attribute profiles.
- Anchored MaxDiff adds a follow-up question to distinguish relatively preferred items from those also considered absolutely important, which standard MaxDiff alone cannot show.