a survey basing on plain data cant be "wrong". you could say that the samples/size/requirements to make a list are wrong, but not the data itself.
honestly, do you know a graduated person listening to lil john/wayne or whatever?? and the other way round, do you know a construction worker (nothing against them, really, i worship people that can work with their hands) that listens to classical stuff??
so for the main stuff, i think this survey gives a good overlook over the correlation between quality of your school/music taste.
some bands/genres might fall out of the context, but that just a characteristic of such a survey