<p>Both R and S are weighted measures.</p>
<p>The S measure is more of a scholarly measure per capita.
The R measure is more related to the size of the program and S more to the quality of the program although there is a lot of overlap. </p>
<p>The S measure has the the following top weights:
- Grants per faculty member.
- Publications per faculty member.
- Citations per publication.
- Awards per faculty member.
- percent of graduates in academic position.
- Average GRE scores</p>
<p>The R measure has the following top weights:
- Average number of PhDs granted.
- Publications per faculty member.
- Grants per faculty member.
- Number of student support services offered.
- Percent of first year students with full support.</p>
<p>Large publics tend to do better in the R measure than S measure. Large privates will be about even and small privates may have better S scores than R scores.</p>
<p>Berkeley for instance had 15 top R scores versus 6 top S scores. (excluding agricultural sciences).
Harvard had 19 top R scores and 18 Top S scores.
Caltech has 6 top R scores and 9 top S scores.</p>
<p>There seems to be a lot of ways to rig the results, especially as Sakky has identified by creating multiple programs in a particular field. This is particularly the case in the sciences where the boundaries are often ill defined unlike say engineering. </p>
<p>Berkeley seems to have mastered that art especially well.</p>
<p>As an example comparing MIT and Berkeley in the broad field of Biological and Health Sciences.
MIT has five total programs in five fields:
S scores:4 #1 and 1#5
R scores: 4 # 1 and 1 #2 </p>
<p>Berkeley , in the same broad field has 15 programs spread across 8 fields:
S scores: 1 #1, 2 #2, 3#3
R scores: 3#1, 5#2, 1#3</p>
<p>In the broad field of Biological & Health Sciences and Sub-field of Biochemistry, Biophysics and Structural Biology, Berkeley has 3 entries as compared to MIT’s single program. MIT’s program is ranked (1,1) while Berkeley’s programs are ranked (1,2), (2,2) and (11,7). Some of Berkeley’s programs exist nowhere else such as Comparative Biochemistry! </p>
<p>By splicing and dicing, Berkeley has created programs which may no longer be statistically significant. More than half of Berkeley’s programs in biology/health sciences graduate less than 5 PhDs annually (8 out of 15), some less than 2! This compares to zero programs with less than 5 PhDs/year at MIT. </p>
<p>Number of ranked programs is therefore a pretty meaningless statistic. You really need to dig deep into the data to get any useful comparative data among institutions. Who cares if an institution has X number of highly ranked programs if they are in obscure or phantom fields with no comps!</p>