Some of the things that would make the experiment invalid would be, if there was a variation in the behavior of the class with respect to food - for example, if some people ate their breakfast earlier than the others. This then would mean that some students would have already digested their breakfast during the morning classes and the others would still be digesting.