Have you ever been asked to measure respondents' preferences for things such as brands, product features, community concerns, job-related benefits, or product packaging? What about prioritizing a list of items, policies, or attributes based on importance? If so, MaxDiff (Maximum Difference Scaling, also known as best-worst scaling) achieves superior item discrimination and precision compared to typical ratings scales, while at the same time avoiding scale use bias. And, MaxDiff is simple to do!
With MaxDiff, respondents are shown groups (subsets) of items and are asked to indicate the best and worst items (or the most and least important, etc.). Below is an example, involving a set of four items (e.g. four out of potentially a dozen or more in your study).
Respondents typically see six to fifteen MaxDiff questions like the example above. The items are rotated across the questions such that each item typically appears a total of 2 to 4 times for each respondent.
MaxDiff questionnaires are relatively easy for respondents to understand and take advantage of the fact that humans are better at judging extremes rather than items of middling importance or preference. Respondents make choices in MaxDiff questionnaires rather than expressing strength of preference on a rating scale, so there is no opportunity for scale use bias. This is an extremely valuable property for cross-cultural research studies!