Monday, January 17, 2011

Demographic Shift?

** warning: This blog post contains anecdotal information involving the statistics-of-small-numbers **

In my field, it is typical for graduate applicants to indicate which subfield they are most interested in pursuing for graduate research. Most applicants also typically indicate which professor(s) they are most interested in having as advisors. Although an admissions committee (in consultation with the department chair) makes the final decisions and offers, individual faculty (or, at least, representative faculty from research groups) also have a say in the admissions process.

Not long ago, I was perusing applications, and there were several surprises in them, all pleasant.

Surprise 1: Although my particular research subfield comprises less than 15% of the faculty in my department, we got about 30% of the graduate applications. This % of applications is higher than in previous years, but there has been an upward trajectory. I was therefore not surprised that the % was high, but I was surprised by how high it was. Possible interpretations other than that these data have no meaning: The subfield is hot, we are hot, or both. I am OK with any of those possibilities.

Surprise 2: Most of the applicants were female. And, when I made a first pass through the applications and considered only those that excelled according to the standard on-paper criteria (GPA, GRE scores, letters etc.), the resulting group of apparently outstanding applicants was 100% female. In my nearly 20 years of experience with graduate applications, this has never happened. [For those who are wondering/cynical: A male colleague who independently looked through the applications came up with the exact same list that I did.]

Surprise 3: Of the first-pass group of outstanding applicants, 50% of the US applicants identify themselves as minorities or biracial. In my nearly 20 years of experience with graduate applications, this has never happened.

Trend or blip? Significant of insignificant? I don't know, but I am intrigued.

Has anyone else in the physical sciences seen an increase -- dramatic or not -- in the number of female/minority applicants (i.e., members of underrepresented groups) to doctoral programs this year?


James Annan said...

I'm genuinely surprised that you ask for, and see, information regarding the race of the applicants. In the UK, this does not generally happen, and using such info is in principle illegal.

(Some quasi-governmental institutes do ask for this sort of info, which goes on a separate form, and is collected by HR for statistical and monitoring purposes. It does not form part of the assessment of applicants and is not seen by the panel.)

Anonymous said...

Interesting observations, I do hope it is a trend.

I'm not familiar with the US graduate application process and I am curious as to why applicants are asked to identify their ethnicity at all? Is it just to track stats such as these? I think I would have been quite shocked to see that question on an application.

Anonymous said...

Interesting -- a couple of comments and questions on the above.

1) Since all the best applicants are female, will you apply "reverse-affirmative-action" to make sure a good proportion of men are admitted? For the sake of argument, let's assume that these best applicants all accept an offer and there are enough of them to fill up a class.

2) Some hypotheses come to mind that could explain this surprising upsurge in the percent of outstanding female and minority applicants:
(a) there simply are relatively more outstanding female and minority college grads than ever before (I know that women have now surpassed men in the total number of college and post-graduate degrees obtained, although some subjects continue to be either male- or female- dominated).
(b) a larger proportion of women/minorities are choosing to go to grad school in your subject for some reason,
(c) The (white) men think they have better opportunities elsewhere this year (maybe in the private sector) or
(d) (white) men have stopped studying your subject in favor of some other (perhaps more lucrative?) subject,
(e) your department has gained a reputation for being a good (fair, supportive, open?) place for female/minority students, so it is attracting more than its fair share of such students (maybe information leaked out about where FSP works :) ). Alternatively, maybe men have decided to avoid your department for some reason.
(f) It's a statistical blip.

I'm sure other hypotheses exist, too. Who knows?

Anonymous said...


Perhaps in your non-FSP life (ie in real world), you also promote the same values as you do here in the blog. Apparently, it has influenced your subfield in your department, so much so that it is getting reflected in admission process. That—in addition to field and faculty getting hot—explains 1, 2 and 3.

Anonymous said...

I began my graduate training in 2001 and was there for *ahem* several years. Throughout the time, our biochemistry doctoral program was always about 90% female. I believe that trend continues today.

Anonymous said...

I am in biomedical, not physical sciences but thought I'd comment about our apps this year, which also were a pleasant surprise. Although admissions is, as far as we can make it, gender blind and driven by research experience and numbers, we have had an entering class that was more than 50% female for several years now--we have even joked about having an affirmative action program for male applicants. This reflects our undergraduate Biology majors--well over half are female. In a particularly stark example, I taught a first year seminar class this fall with 21 students, 20 of whom were women.

I would also note that our under-represented minority applicants paralleled our overall applicants in being significantly better on paper than in years past. We had a significantly higher cut-off for an interview, and had no drop off in URMs who made it. In our case, I think the hard work of two people who created and maintained our minority outreach and support programs, now about 12 years old, has increased both our profile and the cohort of existing students, making this place a more attractive place for a top URM student to apply.

Mark P

quasihumanist said...

Very speculative interpretation:

The economy has cracked up so much that graduate school is now actually a reasonable option from the practical financial viewpoint?

(It has been speculated that white male upper middle class students are overrepresented in academia because they can most afford to make $20K a year for 5-7 years for a small chance at a $60K a year job rather than jumping into a $70K job right away. If that $70K job is no longer there, this changes.)

Anonymous said...

I've read only the applications in my particular subfield so far this year. Of those 7, 5 were female, and all 5 of those were foreign students. The 2 male applicants were also the only 2 US applicants, and neither self-identified as a minority.

Alex said...

Colleges and universities have been noting for some time that females are out-performing males in the competition for undergraduate admission, with some colleges reportedly pondering "affirmative action" for male applicants. This trend obviously hasn't been very visible in STEM (except perhaps in life sciences), but perhaps you're seeing its first effects in grad applicants in physical sciences? Even if there aren't as many female undergrads in physical science, given the stats on entering freshmen I'd be less than shocked if the high achievers in physical science are or soon will be disproportionately female.

On a cynical note, I can explain the differences in academic performance between male and female undergrads in 2 words: Video games.

Anonymous said...

Very interesting. I wonder if a lot of female undergrads who read this blog figured out who you are and want to work for you! :)

Anonymous said...

In my field, applied mathematics, the highest proportion of females in any cohort is in grad school (typically > 50%). This is not true at the undergrad level nor in pdfs (~20%) and definitely not at the faculty level. I have no idea why this is the case but I see it at most schools I am familiar with and for at least 10 years now.

When I first took a grad class (as an undergrad) I was one of two white males (in a class of about 20) and just assumed that there was about to be a huge shift in faculty ...

Alex said...

In my field, applied mathematics, the highest proportion of females in any cohort is in grad school (typically > 50%). This is not true at the undergrad level nor in pdfs (~20%) and definitely not at the faculty level.

Are you saying that only 20% of women in your field are available in an Adobe format? Or that the integral of the probability density function over the female domain gives 0.2?

Methinks you're using an unusual abbreviation for "postdoctoral fellow", but the jokes are too good to pass up.

Anonymous said...

Hard to tell here, as gender is not included in the summary information in the applicant data base, and I don't feel like opening 94 applications to see what the gender balance in the pool is.

Of the top 10 domestic applicants, we have a 5:5 split. Of the top 20, it is more like 13:7 (more male than female).

Of course, our department is in a computational field, and computer science still has a large gender imbalance (and is more subject to market-driven fluctuations than most fields).

We have ethnicity info, but are not legally allowed to use it for admissions (so I'd prefer that it didn't appear in applications, as I'm SURE some faculty and staff use it to color decisions). I didn't check, but I believe that few of our applicants met NIH's definition for underrepresented minorities.

Anonymous said...

My program is about 50% female, and I'd say it represents the field as a whole pretty well (except at the faculty level, of course), so I can't really comment from that point of view.

Then again, my personal experience has been that the male overachievers among my undergrad friends wanted to go on and make big bucks (typically got a first job in management consulting, investment banking, etc.) while the female overachievers either joined start-ups or went to grad school. This would seem to reflect the fact that women are either less into money or somehow less risk averse, but somehow this strikes me as counter-intuitive.

Anonymous said...

Hey James,

I needed to include a photograph for my MIT application. I was told it was in case I was accepted and needed to get a student ID, keys, etc ... BUT, my supervisor met me at the train station when I arrived and recognized me from my application photo ...

FrauTech said...

I have a depressing theory that this is because women traditionally populate underpaid, "service" jobs and where going to grad school in the past and getting your PhD might have meant a well respected job in your profession with tenure some day it now leads to uncertainty and adjunctification. Numbers are up overall due to the economy, but it's possible top dudes are flocking to other fields where they might get more respect and higher pay.

EliRabett said...

FWIW, if you go through the the NSF science and engineering statistics, you find that the number of male US citizen graduate students in the physical sciences has been dropping for decades. There has been a bit of a bounce recently, the low poin was in 2000. OTOH, the number of females (and minorities) has grown thanks, in no small part to the funding agencies efforts in encouraging them. for some information

Anonymous said...

I'm curious about the number of multiracial students. As the parent of a biracial young child, I was shocked to take my kid to her parent child class and find that the majority of the other kids were also some form of biracial (white/east asian, black/white, black/east asian,south asian/black, black/latino). The coming student demographics will surely be a real mix that I don't think anybody has predicted yet.

Amanda said...

I'm a graduate student in a physical science. At my department's reception for visiting accepted students this year, I realized that the group was probably about 80% women! I think the group of first-years that actually ended up here was not as skewed, but it is still more than 50% women, I think. My year is probably right around 50% women.

However, our department is extremely low on minorities. Of that group of prospective first-years that was 80%+ women, I think that maybe there was one Asian guy? My year is, I think, about 90% white.

This has pretty much been my observation of my field in undergrad, grad school, and internships (all in the U.S.): Quite a lot of women in grad school, but very few minorities.

Anonymous said...

yes, very much so

Eve said...

Fascinating. I wonder if they're having trouble finding work due to recession-embiggened discrimination and have decided to go to grad school?