by Tim
31. May 2010 11:05
Max-Diff, sometimes known as a best-worst scaling, presents people with short lists of attributes and asks which they like best and which they like worst. This is done multiple times, with different sets of attributes each time. The sets of attributes are determined from an experimental design.
You can analyse Max-Diff questions in Q just like any other type of question, such as by crosstabbing with other questions, trees, latent class analysis, smart tables, mixture models, individual-level parameters and mapping.
Max-diff and anchored max-diff experiments are setup in Q as Ranking questions, unless they are best-worst conjoint, in which case they are setup as an Experiment.
This feature is only available in Q4; contact support@q-researchsoftware.com for a free upgrade.
