<p>I would hesitate to draw the conclusions cellardwellar draws about LACs from the study he cites (in particular, from the table labeled “Proportion from Named Institution Obtaining PhDs at Universities with 10 or More Top Programs”).</p>
<p>The trouble is, many highly ranked graduate programs are distributed among universities that do not have 10 or more top departments (if by that we mean 10 or more in the NRC-95 top 10). For example, Hopkins has only 8 (including 4 in the Biological Sciences).</p>
<p>On the other hand, Yale has 19 top 10 programs, but none of them are in the Physical Sciences (unless we count Math). Princeton has 16 top 10 programs, but only one of them is in the Biological Sciences (and its other Bio programs have very low ranks, or no rank). The university whose departments have the highest average NRC-95 ranking is Stanford; however, even Stanford’s average NRC department ranking is only #9.</p>
<p>I count only 13 universities that meet the criterion of having 10 or more departments in the NRC top 10. At all but 6 of these, more than half of the NRC-ranked departments are not in the top 10. Now, it’s entirely possible that LAC graduates tend to be as discriminating in their choices of graduate schools as they are in their choice of colleges. They may tend to favor highly ranked programs, but not necessarily ones at universities with 10 or more NRC-95 top 10 departments. </p>
<p>I think the HEDS findings remain operative in this discussion. The schools with the best per capita track records for “Ph.D. production” tend to be LACs. The ones in the top 10 for Ph.D. production that are not LACs are technical schools (MIT, CalTech) and one relatively small, LAC-like university (Chicago). The cited study does not present enough evidence to conclude that LAC graduates gravitate toward second-rate Ph.D. programs.</p>