<p>
[quote]
Its also interesting that Berkeley, the first public on that list, doesn't show up until #49. Food for thought.
[/quote]
</p>
<p>Actually, very large universities are at a real disadvantage in per capita PhD production statistics. Think about it. The very nature of a large state university is that you have a huge student population with a wide range of interests: academic fields, nursing, engineering, agriculture, and on and on and on. That's the opposite of a liberal arts college where you have a subset of students -- usually a subset that is highly represented in academic and science PhD fields, med school, law school, etc.</p>
<p>Keeping that in mind, I would say that Berkeley's position as a top PhD producer suggests that it is unusually focused on "academic" or "geeky" directions compared to most large state universities.</p>
<p>Obviously, a school where 20% of the students go on to get PhDs is going to have a different overall campus feel than a school where 5% go on to get PhDs. Those are the "campus culture" issues that I think potential applicants should be thinking about....not whether a school is ranked #5 or #6 in USNEWS.</p>
<p>Exactly! That's why I'm so intrigued by your list. </p>
<p>I referenced in the thread about honors colleges. Maybe there are top honors colleges with 10% or more of the students going on for PhD's that would have comparable culture.</p>
<p>
[quote]
InterestedDad: "Keeping that in mind, I would say that Berkeley's position as a top PhD producer suggests that it is unusually focused on "academic" or "geeky" directions compared to most large state universities."
[/quote]
Actually what I found amazing in that list was the spread between the top sixty Phd producers. The spread between the first 25 schools is over 25 percentage points, while in the bottom 35 it's only about 2.5 points. That means to me you've really only got about twenty schools where the environment is that much different than all the rest on the list---if you believe in using this statistic as a criterion. Or, to put it another way, take the yield from the bottom school on the list and you have to go all the way up to position thirteen just to double that figure. When you look at the populations of some of these schools, you may be talking about a difference from the top quarter to the absolute bottom school of something around twenty students per graduating class. </p>
<p>I'll also take issue with the suggestion that Berkeley is the top "public" Phd producer. Look at the list and note that The College of William and Mary (a Virginia state school) is tied with Berkeley, and it's a LAC . . . and it's also on my original list near the beginning of this thread. See, even I can twist these statistics to prove your point, as well as mine! ;)</p>
<p>
[quote]
... Phd producers ... if you believe in using this statistic as a criterion ...
[/quote]
It's not whether you believe in it or not, it's whether it applies or not. For most students it is not such an interesting stat, for others it can be very important.</p>
<p>Well, actually the way you state it, it is a belief:
[quote]
I would say that Berkeley's position as a top PhD producer suggests that it is unusually focused on "academic" or "geeky" directions compared to most large state universities.
[/quote]
</p>
<p>And I might suggest, since you mentioned it, if you were to add law and medical degrees to that data set, you'd be looking at different results. That's my belief.</p>
<p>well obviously if you added a number of criteria it would change the list. Are you saying it would be a better list? I think it probably would be.</p>
<p>
[quote]
Is data on law and med degrees available anywhere?
[/quote]
</p>
<p>Nope. You have to go to each college's website and see if they provide alumni data. Some do, some don't. You have to be careful, because many times these stats come from surveys graduating seniors fill out in April. These numbers don't accurately reflect grad school rates, especially with the tendency for more and more students to work for a while before going to med school, law school, or biz school.</p>
<p>The PhD data is available because the National Science Foundation has collected data from every new PhD in the United States since the 1920s. Their database includes a response rate of over 90%.</p>
<p>My guess is that going on to earn a graduate degree, whether it be a Ph. D, masters, MBA, JD, MD, or something else, is a fairly good indicator of success. Of course it's not perfect, but I would guess that a list of schools by % of students earning grad degrees would look pretty good, if that data could be found of course.</p>
<p>One other alteration I would suggest is to use freshman class size instead of overall enrollment. If you use total enrollment, a college which loses 90% of its freshman after one year, but sends most of the survivors to grad school would rank higher than a school which retains all of its freshman through graduation, but only sends 30% of those to grad school. Using freshman enrollment, perhaps multiplying by 4 to get an 'expected total enrollment' number, would account for attrition.</p>
<p>Unless those schools have something that you want that the LACs don't provide, no. Eliminate away. However if you want some aspects of a larger school atmosphere than be cautious making sweeping cuts.</p>
<p>
[quote]
Using freshman enrollment, perhaps multiplying by 4 to get an 'expected total enrollment' number, would account for attrition.
[/quote]
</p>
<p>The PhD data I posted earlier in this thread uses actual graduate numbers. It's the total PhDs over a ten year period divided by the total number of graduates over a ten year period offset five years earlier.</p>
<p>Thus, low graduation rate schools and schools that have recently grown dramatically, are not penalized. It's as close to a real number as you can get.</p>
<p>Interesteddad, I see I misread. I thought you used total enrollment. Nevertheless, I think we *should * penalize schools with low graduation rates in the data. Allow me to give an unrealistic but illustrative example:</p>
<p>Let's say we have two schools: College A and College B, both of which enroll 1000 freshman per year. Let's say A loses 50% of those freshman before they graduate, while B keeps 100% through graduation. Let's also say that 50 of those A grads and 100 of those B grads go on to get Ph. D's, and assume this data is consistent for all periods of the study.</p>
<p>Using the method you posted, both schools would get a 10% Ph. D productivity number (50/500, 100/1000). But we would not consider A and B equivalently good schools for a Ph. D seeking undergrad. With the same freshman enrollment, B is producing twice as many Ph. D's. If we use freshman enrollment, we see that B ranks much higher than A - 10% to 5%. Obviously, this method is limited because it doesn't account for the strength of the incoming student body. However, the method you posted doesn't either, except perhaps indirectly through attrition.</p>
<p>If we offset the freshman enrollment data by a further five years - for a total of nine - we continue to avoid penalizing schools with recent dramatic growth.</p>
<p>I had done an earlier version of the spreadsheet using current enrollment numbers. Frankly, not much changed. </p>
<p>UChicago moved up using actual grads...because they have been massively increasing their enrollment.</p>
<p>BTW, I don't necessarily look at the PhD rates as being useful only to students planning to get a PhD. Rather, the list seems to correlate with schools that are notable for a high degree of academic engagement and challenging programs. I find it interesting that the signficant PhD producer schools cut across a fairly wide swath of admissions difficulty, etc and, conversely, schools with similar admissions standards often have quite different rates. I think the lists tell us something about campus cultures. A great example might be the crosstown Chicago schools: UChicago and Northwestern.</p>
<p>There are some who insist the reason behind the high-percentage of higher degrees among LAC grads is that they finish four years with no marketable skills. I'm not saying that's true (I'll have two kids in LACs next year), but it does give another perspective on why the dichotomy exists and why those stats may not indicate the level of education excellence everyone here assumes they imply.</p>
<p>Shimer College is one of fifty schools profiled in the book All American Colleges:Top Schools For Conservatives,Old Fashioned Liberals,and People of Faith.</p>
<p>
[quote]
Shimer College is one of fifty schools profiled in the book All American Colleges:Top Schools For Conservatives,Old Fashioned Liberals,and People of Faith.
<p>
[quote]
There are some who insist the reason behind the high-percentage of higher degrees among LAC grads is that they finish four years with no marketable skills.
[/quote]
Certain colleges have mostly pre-professional students that prefer to pursue advanced degrees in medicine, law, or business, which hurts PhD productivity. A better way to judge a department would be to see where their students go to graduate school rather than how many, but unfortunately most don't offer this information.</p>