The idea is to look at "change intent" of young people and their preferences towards certain careers. So with this in mind what limitations would such data have? Should inferences be made only to this population (currently registered students); or can one go beyond this?