Sawtooth Software Conference 2019 | San Diego, CA, USA | September 23–27
 

How MaxDiff Is a Better Measurement Technique than Rating Scales

At the April 2019 Quirk’s Event in Chicago, David Hengehold (P&G) and Megan Peitz (Numerious) showed how MaxDiff (best-worst scaling) often is a better survey measurement technique vs. ratings scales. David shared how P&G uses MaxDiff with great results. In Chicago, David & Megan showed an interactive MaxDiff survey demo for Ice Cream preferences, very similar to this:

https://www.sawtoothsoftware.com/icecream

This demo shows your preference scores along with a respondent consistency grade.

Rating scales suffer from straight-lining (lots of 4s and 5s, leading to low discrimination) and scale use bias (people use ratings differently, especially cross-culture). MaxDiff is much better for prioritizing a long list of items (or product concepts), market segmentation, and cross-cultural research.

For further reading, I recently published an article in Quirk’s showing that MaxDiff gets you much more than ranking data:

https://www.quirks.com/articles/how-good-is-best-worst-scaling