Q: How many participants do we need for our cardsort?
Tullis and Woods conducted a real data simulation study to determine the practically optimal number of participants need for cardsorting.
First they collected data from a large set of users–168 people to be exact. Then they ran cardsort “studies” of different sizes by randomly drawing data for individual participants from the larger data set. Since it was all mathematical–based on similarity scores within data pairs and sets– it was easy (sort of) to “run the studies” enough times to understand the average variation that occurrs in the groups of different sizes.
Based on their simulations, Tullis & Woods concluded cardsorting studies with just 20-30 participants are robust and predictive.
Tullis, T., and Wood, L. (2004), “How Many Users Are Enough for a Card-Sorting Study?” Proceedings UPA’2004 (Minneapolis, MN, June 7-11, 2004