<p>Okay sure, but a) those numbers aren't normalized to student population (which I would argue is generally more relavent), and b) they don't include LACs, several of which have very high PhD rates in the sciences.</p>
<p>To cghen; Huh?</p>
<p>^ I was just surprised to read that RPI students get the third most PhDs (in engineering and chemistry) of any school. I know, for example, that Harvey Mudd College has a significantly higher PhD rate, so I wanted to see what methodology gellino's source was using.</p>
<p>I was clarifying that I think PhDs/student population is a better indication for academic quality (and student body strength) than total PhDs, and so I was suggesting that being ranked number 3 in total PhDs (excluding LACs as I mentioned) isn't that useful of a metric for identifying good technical schools (not to say RPI isn't one by any means, but I'm sure there are better ways of demonstrating that strength).</p>
<p>cghen,
1. class size of RPI is smaller than MIT or Cornell, so if you want to calculate a per capita figure, RPI is even closer to the top two.
2. not many LACs offer engineering, so not too many students from them are going to pursue PhDs in it. However, there is a section earlier in that study ~ page 35 that examines LACs and how many students pursue PhDs in a variety of disciplines including many different science subjects.</p>
<p>gellino,
1. Right, but that might not be the case for say the schools ranked below RPI in the study, which might have higher rates.
2. Fair enough, I was speaking more to chemistry with that comment.</p>
<p>Also, in general, when comparing tech schools to non-tech schools, PhD rate (or total PhDs) is kind of a silly thing to use anyway. It penalizes schools like Harvard (which has great science and good engineering) which offer a plethora of other majors - catering to people who often go to business, law, or med schools - which artificially dilute its PhD rate.</p>
<p>Harvey Mudd, Cooper Union and other specialized schools are also in that study further back around page 90.</p>
<p>Except for Case Western and Rice, no school is smaller than RPI that is below them on the list and RPI is higher than them per capita.</p>
<p>Okay, but let's look for example at Princeton.</p>
<p>RPI and Princeton both have roughly 4,700 undergrads.</p>
<p>RPI produced 328 engineering PhD students.
Princeton produced 178 engineering PhD students.</p>
<p>Now I believe you're claiming (implicitly) that RPI is better in engineering because of this statistic. I'm not convinced because I'm guessing that a higher percentage of students at RPI want a PhD in engineering than in Princeton. I'm pretty confident of this on the basis of comparing the offerings of the two schools, and I'm sure if anyone looked through commencement data, RPI would have a higher percentage of engineering degrees awarded than Princeton. Thus, for any applicant who is interested in a PhD in engineering, knowing that RPI pumps out more PhD earning students than Princeton says nothing of that student's relative chance of getting a PhD with an education from one of these two schools. Consequently, you haven't provided any real measure of the quality of the education at RPI in terms of producing qualified graduates - your numbers could be entirely an artifact of incoming student preferences and scope of course offerings at RPI. </p>
<p>Now you could argue that it's better to study engineering around lots of engineering students (and I wouldn't disagree), but that's a separate, less quantifiable claim than suggesting that RPI is one of the best engineering and chemistry schools simply because it cranks out a lot of PhD students. Again, I'm not saying RPI isn't one of the best schools for that; I'm just saying that your data doesn't necessarily support that claim.</p>
<p>
[quote]
I don't think Harvard is in the top 20 in any speciality.
[/quote]
</p>
<p>Harvard is ranked #15 in grad computer science (higher than Penn). True, CS is not strictly an "engineering specialty", but CS and engineering (especially EE) are spoken in the same breadth often enough that CS is treated as a 'de-facto' engineering specialty.</p>
<p>
[quote]
MIT
Berkeley
Stanford
Caltech
Georgia Tech
CMU
Purdue
Michigan
UIUC
Wisconsin
UCSD/UCLA
[/quote]
</p>
<p>Harvard is higher ranked in grad CS than Purdue (#18) and is tied with Michigan and UCLA (#15). It is also not that far off from Wisconsin (#9), Caltech and Georgia Tech (#11) or UCSD (#13).</p>
<p>The point is not to say that Harvard CS is the greatest thing. Nobody is arguing that it is as good as CS at places like MIT or Stanford.</p>
<p>But my central point is that Harvard CS is pretty decent, and we ought to have more respect for it. Honestly, given the choice between studying CS at Michigan/UCLA vs. studying it at Harvard, and I was quite certain that my computer interests wouldn't shift over to hardware (i.e. over to the EE side), then I would almost certainly go to Harvard. This is true whether we're talking about undergrad or grad. Why not? The CS rankings are the same, and, let's face it, the Harvard brand name is more powerful than that of Michigan or UCLA. The same analysis would hold for places like UCSD, Georgia Tech, or Wisconsin. You would only be losing a few spots in the rankings, and you would be gaining that Harvard brand-name in exchange. I think that's a quite fair trade. Harvard has that killer brand, and the CS program is actually quite decent.</p>
<p>
[quote]
. class size of RPI is smaller than MIT
[/quote]
</p>
<p>False. According to wikipedia, RPI has 4888 undergrads. MIT has 4136. </p>
<p><a href="http://en.wikipedia.org/wiki/MIT%5B/url%5D">http://en.wikipedia.org/wiki/MIT</a>
<a href="http://en.wikipedia.org/wiki/Rensselaer_Polytechnic_Institute%5B/url%5D">http://en.wikipedia.org/wiki/Rensselaer_Polytechnic_Institute</a></p>
<p>True, MIT has more grad students than RPI does. But we're not talking about that. We're only talking about where people got their UNDERgrad degrees, which is why only undergrad population is relevant.</p>
<p>
[quote]
Except for Case Western and Rice, no school is smaller than RPI that is below them on the list and RPI is higher than them per capita.
[/quote]
</p>
<p>Also false. Caltech, having only 896 undergrads, is clearly much much smaller than RPI, and if you normalize the numbers, you will see that Caltech is much more represented than is RPI on a per-capita basis.</p>
<p>If you look at the appendix of that study which lists the schools considered, you will see that many many large research universities are not on the list. The study actually only includes a small subset of research universities.</p>
<p>
[quote]
Now you could argue that it's better to study engineering around lots of engineering students (and I wouldn't disagree), but that's a separate, less quantifiable claim than suggesting that RPI is one of the best engineering and chemistry schools simply because it cranks out a lot of PhD students. Again, I'm not saying RPI isn't one of the best schools for that; I'm just saying that your data doesn't necessarily support that claim
[/quote]
</p>
<p>I think perhaps a better way to analyze the data is to look at the very back of the study where it discusses where school grads get their PhD's in ALL disciplines (not just engineering or chemistry). </p>
<p>For example, according to p.115 of the study, between 1920-1995, RPI had 3127 total undergrads go off to get PhD's in ANY disciplines (whether it be engineering, science, humanities, whatever). </p>
<p>Contrast that to the number from, say, MIT, which had 8916, or Caltech which had 3171, despite the fact that both MIT and Caltech have smaller undergrad populations than does RPI (and in the case of Caltech, a SUBSTANTIALLY smaller population). </p>
<p>Figures for other schools:</p>
<p>Princeton - 5237
Harvard - 9773
Case - 3654
Rice - 2600
Cornell - 10326 (but admittedly with a student body that is 3 times the size of RPI's)
Yale - 6435
Stanford - 6645</p>
<p>Now, whither Harvey Mudd? The basic problem with Mudd is that Mudd wasn't even founded until 1955. Obviously any study that looks at where grads from the years 1920-1995 get Phd's is not going to properly reflect the strength of Mudd when Mudd didn't even exist for many of those years. It would be far more fair to look at the years from 1986-1995 when Mudd actually existed in all of the years covered. </p>
<p>In those years, Mudd is credited with 241 undergrads who receive PhD's in any discipline (p.117 of the study). RPI had 711. But of course Mudd has only 720 undergrads, so RPI has literally almost 7 times the number of undergrads that Mudd has. </p>
<p>Hence, the point is, the only thing the evidence seems to show is that RPI grads tend to be strongly weighted towards technical disciplines, and especially towards engineering. Yet even looking at just engineering, on a per-capita basis, I think it's quite clear that RPI is not as good as Caltech or Mudd, although I agree that on a per-capita engineering basis, RPI is doing better than Cornell. However, on an overall per-capita PhD basis from all PhD disciplines, RPI is decent but by no means spectacular.</p>
<p>
[quote]
If you look at the appendix of that study which lists the schools considered, you will see that many many large research universities are not on the list. The study actually only includes a small subset of research universities.
[/quote]
</p>
<p>Also true. I was wondering where the heck schools like Berkeley, UCLA, and Michigan were.</p>
<p>Sakky, come back to me on chosing Harvard over Michigan for CS in a couple of years! hehe!!!</p>
<p>Why's that, Alexandre? Is something dramatic going to happen in the near future?</p>
<p>Yes, there are some significant developments taking place. Here are two that I can think of:</p>
<p>1) New, state of the art, multi-million dollar facility purely devoted to the advancement of Computer Science.</p>
<p>2) Google opening a new, $35 million center, in downtown Ann Arbor that will employ over 1,000 computer scientists.</p>
<p>Add to this the fact that Larry Page (Google co-founder) is a rabid University alum who is worth close to $5 billion at the age 35, and you can be sure that sometime in the next couple of years, Michigan is going to get him to spearhead the next fund raiser, his portion in all probability going entirely to the College of Engineering.</p>
<p>So yes, I'd say that Michigan's CS (and Engineering programs) is probably on its way up. However, I understand that Harvard is also currently trying to improve its Engineering programs, and if that's the case, given its formidable resources, I'd say that Harvard will probably improve dramatically over the coming years too.</p>
<p>Yeah, you are right that Michigan and Harvard will probably dramatically improve rather quickly. Let's not forget that Bill Gates has announced he will retire to run his charity, and he has already announced that he plans to give away most of his fortune, and I get the strong feeling that quite a big chunk of it will be earmarked to Harvard (although, strangely, Gates has in the past also given quite a bit of money to MIT to Stanford, despite having no real connection to either).</p>
<p>But anyway, the point is this. Just like I think the Harvard brand-name is sometimes overrated, I think that Harvard engineering/CS is equally underrated. What I see is that those people who don't know much will think that everything that Harvard has is top-notch. However, those people who a bit more about colleges have heard that Harvard's relatively weak subjects are engineering and CS, and so they translate that to mean that Harvard isn't good at all in those subjects. This is false. It is true that, relative to its other powerhouse subjects, engineering and CS at Harvard are not as good. But they are still quite good. It's just that they stand in the shadows of the giant Harvard programs in business, medicine, law, humanities, natural sciences, etc. It's like being Scottie Pippen standing in the shadow of Michael Jordan. Yeah, Pippen was not as good as MJ, but he was still pretty darn good. He was named to the 50th anniversary all-time team, after all.</p>
<p>I don't mind when people say that Harvard is not as good as engineering as the veritable superstars like MIT or Stanford. But when people start saying that Harvard is not as good as places like Duke or RPI or Virginia Tech, I mean, come on now.</p>
<p>I wasn't trying to 'prove' that RPI was better at engineering than Princeton, CalTech or Harvey Mudd (which I had never even heard of before this board), but merely trying to point out to the OP that RPI would be a solid choice to consider and learn more about.</p>