<p>the usnews-reported s/f ratios for research universities are mostly incorrect, a result of a SIGNIFICANT number of schools NOT including graduate students in their s/f calculations despite clear directions on the common data set telling them to do otherwise. this difference in counting ‘method’ significantly outweighs any issues in the faculty calculations, which were, when i looked into this a while ago, pretty much spot on. (the question, of course, is which top schools are misreporting intentionally and which have lazy institutional research officers.)</p>
<p>of course, the other significant piece in the class size puzzle is the number of undergraduate classes a professor teaches per year. this can vary quite significantly.</p>