Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Max-Diff or MPC With 6 Items

I have a request for a max-diff question with 6 items. I realize, with less than 7 items it would be preferable to use an allocation or ranking but a max-diff is required. Are there advantages to using a best/worst max-diff with 6-10 sets (3-5 exposures per respondent) or removing the worst question like a paired-comparison and asking 9-15 sets?

Also with so few items, in either case is it overkill to increase the number of sets beyond 3 exposures?
asked Aug 31, 2017 by liskd (125 points)

1 Answer

+1 vote
For a 6-item study, I think MaxDiff showing 3 items at a time could work.  Also, paired comparisons could work.  Our MaxDiff software supports either approach.

Warning: do not try to show 4 or 5 items per set if only 6 items are in the study!  Bad results.  The software will warn you if you try to do this.

If respondents have the patience to answer enough questions such that each item appears 4 or 5 times, that would be good from a statistical standpoint.  True, most of your gains would be had by showing each item 3x, but there would be some additional resolution by asking each item 4x or 5x.
answered Aug 31, 2017 by Bryan Orme Platinum Sawtooth Software, Inc. (154,305 points)
Thank you Bryan.