We are currently working on an internal test of looking at a normal max diff, an express max diff and a sparse max diff with a large number of items. What I'm hoping to learn is when a client comes to us with a large number of items (80+), which method should we move forward with.
We are planning on asking some follow-up questions on the back-end and we will of course be comparing ranks, but I'm struggling on what to look for in terms of how each one actually compares to one another, since we really can't compare rescaled scores (or can we?). I was thinking about maybe an anchor would help set a level playing field, but was wondering if there are any other tests that I could from the output to compare these three exercises. I'm very familiar with the Max Diff program, but definitely not an expert on how to compare these numbers. Is there anything you have done before or anywhere you could direct me?
Thanks so much