<p>^^well, you’ve made two arguments, one in favor of collecting data on matriculants in general, and another in favor of doing so by graduating class. It would seem to me, that including the data by matriculants as whole would mitigate against wild swings from year to year in favor of any one “big school”.</p>