The following is a guest post* by Carolyn Dicey Jennings (UC Merced), who has led a team of academics in producing and organizing a trove of data related to the graduation and placement records of English-language philosophy Ph.D. programs (previously). The team just published an update to its 2015 report, “Academic Placement Data and Analysis” (APDA). Among other things, the update contains the graduation and placement data for specific programs, some of which I’ve included below the post. Thanks to Professor Jennings and the rest of the team—Patrice Cobb, Chelsea Gordon, Bryan Kerster, Angelo Kyrilov, Evette Montes, Sam Spevack, David W. Vinson, and Justin Vlasits—for their work on this project.
An Update to Academic Placement Data and Analysis
by Carolyn Dicey Jennings
I am happy to report that the Academic Placement Data and Analysis (APDA) project has been able to release an update of its 2015 report, which includes program-specific placement rates. We decided not to include those in the original 2015 report because of errors we found just before its release. Those errors were in fact for records entered by placement officers and department chairs over the summer (e.g. missed placements), so we decided to spend the next six months or so checking the placement records in our database against other available sources, such as the PhilJobs Appointments page and program-specific placement pages. Having integrated that data, while also checking for duplicates and other errors, we feel confident enough in the data we have to release program-specific placements. This is a first for our discipline, and I am very grateful to all who have helped make it happen, including the American Philosophical Association (APA), all the APDA project personnel, and numerous people in philosophy who have helped the project in different ways (and me, when I needed it).
A first note: this does not mean that our database is error free. Only a week ago, long after removing duplicates in the database, Angelo Kyrilov noted that I was in the database twice—once for Boston College and once for Boston University. This was a coding error that was not caught in our system of checking for duplicates because one entry had my last name as “Jennings” and the other had my last name as “Dicey Jennings.” While we did check for misspellings of names, these were far enough apart in the alphabet to miss the duplication. I think it is a sign of the scientific nature of our project that the project head (me!) could have an incorrect entry. Yet, it is also a reminder that errors are inevitable in a project of this nature. We have made significant efforts to reduce and remove them, but when we add the option of individual editing in May we hope that still more will be noticed and removed.
So what did we find? One notable difference between this report and the one released in August is that we did not find a significant effect of gender on placement this time. There were some differences in how we made our comparisons—in the last report we grouped all but permanent in one category, comparing this to permanent placements. In this report we separated out temporary academic positions, nonacademic positions, and no reported positions (we have a small number of the latter two categories—251 total nonacademic positions and 204 graduates with no reported positions). We did this because we hope to expand our project to take better account of nonacademic positions. If we had used the previous model we would likely have found significance. In our database, 54% of the women have permanent academic positions, whereas only 49% of the men have permanent academic positions, and this difference is significant using a chi-squared test (p<.01). Yet, I was surprised that a number of programs (7, or around 7%) had no reported women graduates in the time period we looked at: 2012-2015. It is important to remember that disparity in philosophy is still high. (See this recent draft by Eric Schwitzgebel and myself for an in-depth accounting of that disparity.)
When reading our tables and graphs, keep in mind that 2015 is less-well represented in terms of number of graduates in our database: the 2015 graduates make up only 15% of the total graduates between 2012 and 2015 (the other three years each make up around the same proportion—28%). Since we did not add a time dimension to our analyses this year (but plan to do this in the future, as with the 2015 report), and since graduates of 2012 have had more time to secure permanent academic positions, those programs with fewer graduates from 2012 and more graduates from 2015 will likely appear to be worse off on placement relative to their peers than they actually are. (For this reason, I ordered the graduation year graphs by proportion of graduates from 2012.)
Also notable is that Value Theory remains the most popular of our first-listed area of specialization categories for this time period (2012-2015), with Language, Epistemology, Metaphysics, and Mind in second place, History and Traditions in third, and Science, Logic, and Math as the least populated of the categories. (This is excluding the “Unknown” category, which in fact beats Value Theory.) Nearly every program had some graduates with first-listed AOS in this category, with only 12 with no such graduates (excluding two programs for which we have only unknown AOS). Yet, Science, Logic, and Math was the category closest to significance, with more likelihood of those in our database with this AOS being reported as having a placement than not having a placement (in comparison with LEMM). (Value Theory and History and Traditions each had lower likelihoods for the academic positions and higher likelihoods for the nonacademic positions, but these values were not significant.)
Finally, we are still looking for the best way of capturing graduation data. Our current method is to take the mean for three external sources and then compare that to the numbers of graduates in our own database, taking the higher of these numbers. All three sources are good ones: the APA Guide to Graduate Programs, PhilJobs Appointments, and the Survey of Earned Doctorates (SED). Yet, there were lots of inconsistencies in the data. For example, in the same year a single program might have reported twice or three times the graduates in one source as in another (comparing PhilJobs and APA, since the data for both of these sources come from the departments themselves). This fact makes the search for reliable information about graduation numbers particularly tricky. If we were to use only SED data we might avoid some problems, but potentially fall prey to others, such as the fact that SED groups graduates by “field of doctorate” rather than department (and so might have higher or lower numbers for philosophy departments than they should, depending on how graduates see themselves). I am open to suggestions on this front. Needless to say, placement rates are sensitive to graduation numbers, so this is an important part of the project.
So what is the overall placement rate? Including graduates from all four years for 108 doctoral programs, and using external graduation data, 32% of graduates from 2012-2015 have by now found permanent academic positions and 68% have found academic positions of some kind. My supposition is that a large chunk of the remaining 32% are now in nonacademic positions. Yet our database only has record of a small proportion of these. For this reason, I am starting to think of the project somewhat differently, and will aim to do a better job collecting non-academic placements in 2016 by enabling individual reporting and editing of placement records. (I am supposing that placement officers and department chairs know less about these graduates.) We will begin thinking of how to categorize nonacademic positions so that we can, in future, have fewer reported unknowns and a better understanding of this large proportion of graduates.
Exciting developments are to come for APDA: we are going to send out a qualitative survey to all graduates in our database in May, reporting results in June. We are also going to add applications to the website that will allow users to interact more directly with the data, along with other website upgrades. If you have any feedback for the project, please let me know! We have turned off editing for now as we prepare for individual editing, but if you find any errors feel free to email me directly, preferably at the project email account.
See the rest of the data in the official APDA update.