a standard set of questions they would like answered: can you really divide people up into a particular set of categories? Are the tests for these categories reliable; if you take the test twice will you come out the same both times? Are the categories you are trying to use related to how people learn? If you use a theory of learning styles, do people learn better? Can you use learning styles to predict who will benefit most from particular styles of instruction? Does using a learning styles system - any system - for teaching have other effects on learners or teachings, such as making them more confident or making them expend more effort?
These questions stem from the way academic psychologists systematically approach topics: we like to establish the truth of psychological claims. If someone comes to us with a theory about learning styles we want to know (a) if learning styles really exist, (b) if they really are associated with better learning and also (c) if, when learning styles are taken into account, learning is better because of something about the specific learing style theory rather than just being a side effect of an increase in teacher confidence, effort or somesuch.
