I'm looking to run a card sort to help gain understanding of how people would group & categorise courses. I currently have a list of 142 courses. I'm worried that the list is too long and participants will not complete the task. I could break it down and randomly surface say 50 different titles to sort for each participant, but would this skew the results?
You could use the Delphi method:
In this approach, the first participant creates an initial sort, and then the remaining participants review and modify that sort. This imposes a much lower mental load on the remaining participants, rendering a sort of 200+ items a much more tractable proposition (in theory).
It's an alternative supported by Donna Spencer who wrote the Card Sorting Book.
Breaking the card sort down into random groups of ~50 course titles shouldn't skew the results.
Bearing in mind the average times taken to complete a card sort exercise based on number of items;
you could be looking at 1-2 hours for the participants to complete the exercise. The potential drop-off in participant engagement over this time would have more of a negative effect on your data then you would get by breaking the task down.
Hope this helps.