New APPS

Art, Politics, Philosophy, Science. Mostly Philosophy.

about

Some figures on prestige bias in academia

February 14, 2015

I've put together some numbers on pedigree bias from various websites and sources, and it seems the problem is pervasive in academia:

In computer science, business and history, 25% of doctoral granting institutions provide 71-86% of all tenure track jobs (Clauset et al.)
In computer science, business and history, only 9-14% of candidates get placed at places that are higher-ranked than their institutions
In philosophy, 88% of initially reported tenure track hires in 2013-2014 were from Leiter-ranked programs
In philosophy, 37% of initially reported tenure track hires in 2013-2014 had their PhDs from the top-5 schools
[UPDATE: Based on Carolyn Dicey Jenning's more complete dataset, this turns out to be a skewed number. She finds 31% of tenure track hires come from top 10-departments
In English, the top-6 programs get 60% of their hires from other top-6 programs, and 90% are from the top-28 – nobody from the 65+ ranked programs ever gets hired in a top-6 faculty.
Older data in sociology suggest that the prestige of PhD granting department is one of the main factors in hiring decisions (the other is the selectivity of the undergraduate institution. The authors conclude (rather dryly) "job placement in sociology values academic origins over performance."

Posted in Helen De Cruz, Improving the philosophy profession

12 responses to “Some figures on prestige bias in academia”

Soon to be unemployed Leiter Top 5 PhD

February 14, 2015

I must have some truly awful letters.

LikeLike

Reply
anon

February 14, 2015

It would be interesting to see how the ranking of philosophy departments would be according to the criteria used by Clauset et al.!

LikeLike

Reply
Joe

February 14, 2015

Thanks, Helen, for collecting this data. What follows is something I’ve been puzzling over with respect to the current discussion of prestige bias.
So, a few months back, we were (rightly) worrying about Carolyn Dicey Jennings’ placement vs. ranking data. Prestige, it seems, was not as highly correlated with placement as we thought. Certain schools in the top-20 were clearly failing their graduate students in certain respects, absurd apologetics for the status quo notwithstanding.
Now, it appears as though the very thing that we were hoping for a few months ago (a decent correlation between prestige and placement) is now called ‘prestige bias’. There is a certain tension between these two lines of thought.
Put it this way: I don’t want there to be prestige bias, I want a meritocracy, as I believe we all do. But we need to get clearer on what kind of distribution-pattern we think will reflect a meritocracy. What if 25% of TT hires came from top-5 schools? Would this be OK? When does a desirable prestige-correlation become a prestige-bias?

LikeLike

Reply
Anonymous

February 14, 2015

The philosophy numbers are off the mark. Those numbers were not for “all” reported TT hires. They were for the first 100 reported TT hires.

LikeLike

Reply
Helen De Cruz

February 14, 2015

Do you know the figures for the rest? 100 seems a good sample to me (except if the first 100 hires are skewed in some respect).

LikeLike

Reply
Helen De Cruz

February 14, 2015

Hi Joe: Thanks – that’s good food for thought. Personally, I found Carolyn’s numbers comforting (I thought, well at least prestige isn’t everything and some lower-ranked programs still manage to place their candidates well). But if Marcus’ figures for the first 100 hires extrapolate to the total job season, to have nearly 40% of new TT hires from only 5 schools seems to me pretty skewed. To assess whether 25% would be fair, we’d have to know how much % of new candidates the top 5 schools jointly produce, and to assess how much of a skew towards those programs we would deem fair.

LikeLike

Reply
Sara L. Uckelman

February 14, 2015

to have nearly 40% of new TT hires from only 5 schools
I confess to being dubious that the numbers involved could make this possible. How many new Ph.D.s do the top 5 schools mint each year? How many TT hires are there each year? Are there really enough of the former and so few of the latter that a 40% figure could even be mathematically possible?

LikeLike

Reply
Helen De Cruz

February 15, 2015

Hi Sara: One way those numbers might be distorted is if people from top-5 schools got earlier offers (since Marcus’ figures only report the first 100 hires).

LikeLike

Reply
Anonymous

February 16, 2015

I wonder if the sample of the first 100 also might be skewed because the placements from a particular school are all reported around the same time? Regardless, it is not difficult to look at current faculties and see how many tenure track assistant professors received their degrees from the same top-ranked schools – the same is true for each generation, with slightly different Ph.D. granting schools considered part of the elite few.
Members of search committees doubtless find it daunting to scrutinize the details of candidate dossiers, much less to read all the writing samples thoroughly. I worry that in the initial stages of screening applications, preference given to an elite Ph.D. granting institution is perhaps the single most important factor. I know this is true for many of my colleagues because they say so flatly when explaining how they chose a short list of candidates; they pay closest attention to the files of candidates from top-ranked schools. I believe we owe each other something better. Across the discipline we should adopt a different set of “best practices”. A quick review of each candidate’s dissertation abstract and at least the first few pages of the writing sample is a much more accurate and merit-based approach, and it doesn’t take very long if the files are pre-screened and the work is divided among teams of reviewers. I pre-screen applications according to the “easy” factors: things like appropriate AOS/AOC, a complete application, and a personal letter with a meaningful expression of interest in working at the school. I’m surprised by how many applications can be screened out this way, leaving a more reasonable number for review of writing samples. I have found in my experience with several searches that most but not all candidates who attended high-ranking institutions will pass through that screening process, along with a number of other excellent candidates who would have been disqualified by a narrow focus on pedigree. I have also noticed that the remaining applicant pool is more diverse in a number of ways (including intellectually) than it would have been had I only chosen candidates from top-ranked schools.

LikeLike

Reply
Anonymous

February 16, 2015

In writing the above comment (13:20), I thought of a question I wish someone could answer. Occasionally I have encountered polished dossiers I assume are prepared by an applicant’s institution but are not accompanied by a personal letter from the applicant. Is it possible that some schools send out files to all the jobs listed in an eligible candidates AOS/AOC without coordinating this effort with the applicant so that a personal letter is sent? Or were these anomalies? I do not take those applications as seriously because I expect the candidate to examine the department before applying and personally explain reasons for interest in this specific position. I would like to know if I am unaware of a new practice and I would like candidates to know that a personal letter is important to at least one reviewer.

LikeLike

Reply
Anonymous Coward

February 16, 2015

My PhD program does not help us send anything out, in any way. Are you not counting cover letters as “personal letters”? Also, what qualifies as a meaningful expression of interest?
Keep in mind that many of us have to teach as many as 4 or even 5 courses per term and continue to research while applying to 60-100 jobs, where we’ll be evaluated by search committees so idiosyncratic that it’s impossible for anyone on the outside (and some people on the inside) to decipher how the decisions get made.

LikeLike

Reply
anonymous grad student

February 16, 2015

Comment number 9 is important for interpreting this data. The “first 100” thing is going to be particularly skewed because many placement directors (especially at higher ranked places) reported all of their placements for the entire year all at once.

LikeLike

Reply