<p>
</p>
<p>No, I used actual baccalaureate degrees awarded over a ten year period. I started out trying it with enrollment, but it produces some major errors for schools that have rapidly grown (like University of Chicago) in size. It also produces some anomolies with schools that have low graduation rates (although arguably, it might be more informative to use “entering freshmen”) as the demoninator.</p>
<p>I first realized that current enrollment had problems as a denominator when my lists kept coming up with University of Chicago as a weak PhD producer, a result that made no sense. I looked into why that might be happening and realized that their enrollment had grown dramatically in the last few years, so they were getting hammered with denominator much much larger than would have been representative of the actual number of graduates “back in the day”. Most schools, with fairly stable enrollment, didn’t change much. A few changed dramatically when I went to actual graduates as the denominator.</p>