@warblersrule Thank you for combining those two rankings. I looked at both independently, but for some reason, I never got the idea to combine them.
And yes, warblers are cool birds. :))
Why not Oxford? I hear the food has improved in England! :bz
@warblersrule, unless I am mistaken the USNWR math rankings are only for graduate programs, so LAC’s would never be included.
@tk21769, how can IPEDS be accurate as Amherst has graduated 10+% of the class as math majors for years, so that’s 40-50 per year. There are many other highly selective LAC’s that consistently produce the same percentages.
@retiredfarmer, I agree with your thoughts in posting #2 - there are many tried and true paths to take, with the top LAC’s consistently bearing the same fruit.
OK, so look instead at the Common Data set files (section J, “Degrees Conferred”) on Amherst’s own site.
- 2002 (Mathematics: 1% of degrees conferred)
- 2003 (degrees data not found)
- 2004 (Mathematics: 2% of degrees conferred)
https://www.amherst.edu/media/view/26537/original/DEGREES+CONFERRED.pdf - 2005 (Mathematics: 2% of degrees conferred)
https://www.amherst.edu/media/view/7868/original/J_Degrees_Conferred.pdf - 2006 (Mathematics: 3.1% of degrees conferred)
https://www.amherst.edu/media/view/7722/original/CDS2006_J_Degrees_Conferred.pdf - 2007 (Mathematics: 3% of degrees conferred)
https://www.amherst.edu/media/view/37007/original/Book12.pdf
etc.
(n.b.: I’m trying to establish a PhD production rate for the whole decade 2006-2015, not just for the most recent year. So if we’re trying to normalize by the number of math majors, we’d be going back at least to students who graduated in the early 2000s.)
Maybe someone can cite a better data source than IPEDS or the CDS to shed light on the PhD production rates? As far as I can tell from the data I’ve seen, top LACs seem to compare well with top research universities in math PhD production rates, whether we normalize by undergraduate student body size or by the number of math majors.
@tk21769, like many of the data sources provided the CDS doesn’t square with other information provided, here the profile of the Class of 2016 https://www.amherst.edu/media/view/669797. Of the 262 single major graduates, 20 are Math majors, so ~10%. Then there are another 36 that are either double major, including Math and 1 that is a Math triple major. Overall, that’s 57 of 432 which is well over 10%.
@retiredfarmer If you haven’t gone beyond Calc BC in high school, then you’d be starting off with a potentially significant disadvantage compared to UK students in Oxbridge undergrad math. For math and philosophy “the standard in mathematics for admission to the joint course is the same as for admission to the single-subject Mathematics course” (in the UK this will usually mean A* in both math and further math A-levels: “Further Mathematics is highly recommended”). PPE on the other hand doesn’t expect the same depth in math (few applicants would have done math and further math A-levels, and further math is not required).
tk, it is mathematically implausible for just 2% or 3% of LAC students to major in Mathematics. LACs simply do not offer that many options. According to Amherst’s more recent CDS reports, 8-10% of its graduates majored in Mathematics. When I visited their campus three years ago, the admissions officer who gave the information session said that Econ and Mathematics were to the most common majors, each accounting for over 10% of the students.
Generally speaking, Amherst or Williams are not going to produce PhDs at a significantly higher rate than large public elites like Cal or Michigan. The small gap in the rate of PhD production can be attributed to the type of student that is drawn to LACs vs the type of student that is drawn to a large research university.
https://nces.ed.gov/collegenavigator/?q=amherst&s=all&id=164465 says that 33 (7.6%) out of 432 bachelor’s degrees conferred by Amherst in 2015-2016 were in math (32) or statistics (1).
The only larger majors were economics (63 or 14.6%) and English (39 or 9.0%).
Philosophy is very small at Amherst, with only 6 (1.4%) graduates for 2015-2016.
Chembiodad, according to CDS data and counting methods, Amherst math/stat majors represented 7.6% of 2016 degrees conferred (https://www.amherst.edu/system/files/J%2520Degrees%2520Conferred.pdf). Presumably their denominator is the total number of graduating seniors, not just graduating single majors. But again, I’m using an earlier reference year, which I think is appropriate if we’re counting alumni-earned doctorates for the decade 2006-2015. I’m not persuaded that the data sources I used were wrong, only that more recent data or different counting methods might produce different denominators for at least one college. If we’re going to use those numbers/methods, then we should apply them consistently to all schools we’re comparing. Can anyone show that this would lead to very different conclusions about LAC vs. RU outcomes?
@ucbalumnus, thanks for providing the link as the comparisons are very interesting. As an example, U Penn had 32 in Math and 39 in Statistics for a total of 71 out of 2927 or 2.4% - does that make it a more robust program than at Amherst?
@tk21769, the link I included provides the detail for everyone of the 432 graduates in the Class of 2016 https://www.amherst.edu/media/view/669797 - the CDS only includes certain categories and doesn’t distinguish double majors. Yes, it would be great if all schools provided the same methodology as its more accurate.
As another example, there were 31, or 2.6%, Math majors at Dartmouth during the same year. I think 7.6% of the all majors at Amherst is significant.
It does look larger in statistics than Amherst – perhaps not surprising given the pre-professional nature of the school (i.e. fast track to Wall Street).
How “robust” it is for any given student’s interests and goals may better be checked by comparing what the department offers, rather than just its size. A very small department may have substantial limitations, but a larger department may focus on subareas of the subject other than what the student is most interested in.
@ucbalumnus, with regards to course offerings, Amherst lists 50 math/statistics courses including twenty-two 300 and 400 level course , and U Penn list 39 math/statistics courses including twenty-two 300 and 400 level courses - I will leave it others to parse the courses.
What may also be relevant is how frequently each upper level course is offered.
Once every two years is probably the bare minimum for an upper level course to be accessible as a student progresses through the college. But more often allows the student more scheduling flexibility, since a student may miss out if the once every two years course is at the same time slot as another once every two years course, or is offered in the semester the student is away (study abroad, co-op job, etc.).
Re: Amherst/Penn/Michigan/Berkeley PhD production
(skip if you’re finding this tedious!)
According to NSF/WebCaspar data for the 5 years 2011-2015,
UPenn alumni earned 22 math/stat doctorates for that period;
Amherst alumni earned 12.
If only 2.4 % of Penn’s ~10K undergrads major in math/stat, and
7.6% of Amherst’s ~1800 undergrads major in math/stat (see 2016-17 CDS),
then it would appear that UPenn generates about the same number of math PhDs per 1000 math/stat majors as Amherst does (~88 per thousand for Penn v. 85 per thousand for Amherst.)
According to NSF/WebCaspar data for the same 5 years,
Michigan-AA alumni earned 37 math/stat doctorates;
Berkeley alumni earned 94.
If 3.3% of UMichigan’s ~29K undergrads major in math/stat (see 2016-17 CDS),
and 5% of Berkeley’s ~31K undergrads major in math/stat (see 2016-17 CDS),
then it would appear that Michigan and Berkeley generate fewer math PhDs per 1000 math/stat majors than either Penn or Amherst do (~39 per thousand for Michigan and 61 per thousand for Berkeley). Perhaps selection effects do account for at least some of this difference.
If the CDS is only counting single majors, it may be more appropriate to also include double majors (including math minors) in the count … as long as we apply the same sources and methods consistently in comparing colleges.
@tk21769, much appreciated and nope don’t feel that its tedious to confirm that a university and a LAC can get a Math student to the same place - so its all about fit.
I am skeptical of any study or article that implies or insists that you need to go to an elite undergrad school to get a PhD at a top school.
That Psychology Today article, for example
-
Only included any analysis of two PhD programs out of the top 10. Not only does that leave out 80% of the PhD programs in the top 10, it also leaves out a whole lot of other PhD programs. Even if we say that a student needs to attend a top 25 or so school to have a decent shot at a tenure-track job later, 2 out of 25 is still less than 10% of the total.
-
It looks like a significant number of students had graduate work, and we’re ignoring that effect for now.
The biggest, issue, though, is that articles like this tend to assume that the effect is causal - aka students who went to universities with top-ranked philosophy departments (at the grad level) are more likely to get into top PhD programs. But my argument is that students who go to the kinds of universities with top-ranked philosophy departments - generally speaking expensive private universities like Princeton, Stanford, Columbia, MIT, NYU, Duke, etc., as well as a few very competitive public flagships like Michigan and Berkeley - are more likely to want to get a PhD in philosophy and apply in the first place, before they even set foot on a college campus, because they come from wealthier families and are more likely to have role models with advanced degrees who are encouraging that sort of achievement. I’m not saying that they all dreamed of getting a PhD in philosophy at age 15, but they’re certainly more likely than a Cal-State-bound kid or a SUNY-bound kid to have a PhD or other advanced degree included in the universe of things they aspire to. For the average Cal State kid, just GOING to college may be enough of a fairy tale.
I can’t find the statistics with a cursory search, but I think it’s safe to say that Cal State and SUNY have higher percentages of students who come from low-income families, speak English as a second (or third) language, and/or are the first in their family to attend college at all than students who went to Columbia, Princeton, Michigan, or Berkeley. (And frankly, if it was really about the quality of the philosophy department at the undergrad, you’d see more students from Rutgers - which has top 5 philosophy department. There are also two UC campuses (San Diego and Berkeley) that have top 15 philosophy departments. Yet the students from both schools are not present in large numbers, which to me indicates it’s not about that.)
I’m not saying there’s zero effect, but I am saying that I don’t think there’s a strong causal effect. I think a sufficiently bright and motivated student can get into a great philosophy PhD program from anywhere. There’s probably an advantage to attending a high-quality philosophy department, but there are plenty of those and it doesn’t have to be at an elite top 50 university.
A MUCH better way to do this would be to gather data on what % of students from Cal State and SUNY (and every school) who apply to philosophy PhD programs are admitted, and where do they go.
@juillet, I’m sure you’ve scoured the NYT article “Some Colleges Have More Students From the Top 1 Percent Than the Bottom 60. Find Yours” as it does offer some insight as to how access to the most selective colleges does lift all boats.
I think this is one of those situations where people tend to be biased by their own discipline. Some fields are quite picky about undergraduate origins, whereas others are much more egalitarian. You’ll notice a sharp difference between math students and engineering students about the importance of the strength of your undergrad institution, to cite some specific examples. For example, see this post by b@r!um with regard to math.
The many differences between disciplines are a major reason my typical answer to grad school questions is “it depends.”
To address your point more directly, the UCR prof’s blog touches upon the issue.
@warblersrule, I don’t think the majority of math undergraduates at highly selective LAc’s intend on pursuing Math PHD’s. A BS in math is the most coveted major on Wall Street and at the pedigree consulting firms, and the second most in Silicon Valley. It’s also one of the top majors for those pursuing MBA’s.