1. Too small a group to be significant. 2. Selection process can tilt results. 3. Cause and effect relationship must be addressed. 4. Interesting, but would the same group test differently with a good book?