Problem of Many – Max Diff
Wonder how consumers react to different features or offerings? How to decide what best to offer, is there a better way to determine the optimized offering beyond the traditional paired comparison or rating and the ranking scales.
Old Man at Rescue, Max Diff.
Max Diff is an approach for obtaining derived preference/importance scores for multiple items (brand preferences, brand images, product features, advertising claims, etc.). Compared to rating and ranking scenarios, it gives robust analysis and applicable to a wider variety. It is also known as "best-worst” scaling.
The Max Diff is a mathematical theory with very specific assumptions about how people make choices. It assumes that respondents make their decision after comparing all possible pairs of items within the displayed set and choose the pair that reflects the maximum difference in preference or importance. It is an evolved variation of the method of Paired Comparisons.
The respondents are asked to share their preference of the best and worst among 3 to 5 items on one screen and multiple such screens are shown to an individual. The study assimilates all these responses which are analyzed to determine the utility estimations for the items tested in the survey.
Consider a set in which a respondent evaluates four items: A, B, C and D. If the respondent says that A is best and D is worst, this response informs us the five out of six possible implied paired comparisons:
A > B, A > C, A > D, B > D, C > D
The only paired comparison that cannot be inferred here is B>C. In a choice among five items, Max Diff questioning informs us the seven out of ten implied paired comparisons. The component of method involving the most different pair may be properly called “Max Diff” in contrast to a “most-least” or “best-worst” method where both the most different pair and the direction of difference are obtained.
It is a powerful technique which with its advanced versions can test out more than 150 different features or offerings.
How many screens?
A quick calculation for the number of screens on Max diff, 2 M/N where M is the total number of items and N is the number of items displayed per screen.
Why Max Diff
- Easy for respondents. Respondents only need to judge the best and worst options, and no more have to spend time ranking things they may not care about
- No scale bias
- Induces trade off mechanism
- Robust mechanism
- Gives respondent level choice utility for clustering and prediction
Max Diff question are simple to understand, so respondents from children to adults with a variety of educational and cultural backgrounds can provide reliable data. Since respondents make choices rather than expressing strength of preference using some numeric scale, there is no opportunity for scale use bias. This is an extremely valuable property for cross-cultural research studies.
The resulting item scores/utilities are also easy to interpret, as they can be placed on a 0 to 100 point common scale and sum to 100.
When to use?
The problem of many, has a one-word solution Max Diff. Its extensively used by researchers to carry out
- Concept test
- Ideas test
- Optimize offerings
- Flavors test
- Prioritize features
- Extensions - Cluster analysis, Bundle Optimization, Reach Determination, TURF etc.