We would like to program a partial conjoint exercise with 15-20 attributes in total. Most attributes will have 2 levels each, although some may have 3 levels. The attributes will be descriptions of features, and we want to identify the most important features.
In addition, we would like to use a master attribute to represent brand, whereby different brands would have a different description of the features. We may show the same brand brands multiple times within each task, using a different combination of features.
This lends itself to an alternative specific design. However, given the number of features we also want to use a partial profile design so that we only show 5-8 at once. Is it possible to combine these two design types?
We have considered MaxDiff and TURF but we want to know whether certain combinations of features have a higher appeal to specific individuals, rather than across the sample.
All help and advice would be greatly appreciated