Graph of AOS by Gender

Academic Placement Data and Analysis: An Update with a Focus on Gender

by Carolyn Dicey Jennings, Patrice Cobb, and David W. Vinson

Since 2015, we have taken part in a project that collects and analyzes placement data for doctoral graduates in philosophy, with funding provided by the small grants fund of the American Philosophical Association (APA). Philosophy has several qualities that make it a good discipline for such a project:

  1. With dramatic funding cuts across the academy, all doctoral graduates likely find it more difficult to achieve academic placement. Yet, philosophy, as one of the oldest and most theoretical of academic disciplines, is often perceived as less essential to the modern university and so in more danger than many other disciplines of losing academic positions. Stephen Hawking recently declared that “Philosophy is dead”. In discussing higher education, presidential candidate Marco Rubio said that the country needs more welders and fewer philosophers. Treating philosophy as a proverbial canary in the coal mine, tracking the placement of philosophy graduates can help us understand and act on shifts in the academy as a whole.
  2. Philosophy is a relatively small discipline, with perhaps 250 English-language doctoral programs in philosophy worldwide, and around 500 yearly graduates in the United States. This makes the project of tracking these graduates more manageable than for larger disciplines.
  3. Philosophy programs tend to make their placement information public. For this reason, one can verify placement records for most programs, which is something that is not possible for most other disciplines.
  4. Philosophy has one of the lowest proportions of women graduate students, with an even smaller proportion of faculty. In an assessment of 23 disciplines in 2009, Kieran Healy found that the proportion of women graduates in philosophy (29%) was the fourth lowest following engineering (22%), computer science (19%), and physics (18%) (i.e. there were higher proportions of women in astronomy/astrophysics and mathematics). Yet, women make up an even smaller proportion of faculty in philosophy: 17% of full-time instructional faculty and 21% of all instructional faculty according to NCES data from 2009. Thus, philosophy is a good test case for examining the pipeline between graduate school and academic placement.

This last feature is the focus of this post. In 2015, the project developed a website and database to help keep track of this data, as well as launching a successful data-gathering campaign. We found that women who graduated between 2012 and 2014 were more likely than men to have a permanent academic placement within two years of graduation, a finding that we intend to explore further. So far in 2016, we have released an update to the 2015 report, including data on 2015 graduates and a more finely-grained set of analyses. Some of the effects that were significant in our original analysis were not significant in these new analyses, including the gender effect. For that reason, we decided to release a new set of analyses and results, with approximately the same structure of those released in August 2015. These updated results include the new data from 2015 as well as improved data for previous years (see below). Before discussing these results, here are two other ongoing Academic Placement Data and Analysis (APDA) projects, along with a data gathering form:

  • We are adding individual editing to the website in May 2016. Up to March 2016, placement data were edited by project personnel, placement officers, or department chairs. In the future, individual graduates will have the option to claim their entry. To do this, we require a contact email for the graduates in our database. We currently have email addresses for roughly one quarter of the database. For graduates: to ensure that you are included among those who have access to individual editing, please provide your email address here.
  • Along with individual editing, in May 2016 we will add a brief qualitative survey for graduates. We will use linguistic analysis to compare these responses across graduates, connecting them to metadata on graduating institution, gender, graduation year, area of specialization, and placement type. Participants will be compensated for their time. Again, to do this, we require the contact email for the graduates in our database. For graduates: To ensure that you are sent the qualitative survey, please provide your email address here.

Please feel free to send the form to past philosophy graduates you know who may want to be included. And without further ado, here are the new results (with methods posted here):

Sample Demographics

For this analysis, only cases receiving doctorates between 2012 and 2015 were included. There were 1,187 men and 466 women across 127 programs. The number of men and women and the number in each first-reported  area of specialization (AOS) category by placement type are provided in Table 1 below. (See also Tables 2 and 3 here.)

Men Women LEMM Value Theory History & Traditions SLM
None 100 27 23 32 24 5
Permanent Academic 457 225 159 171 109 85
Temporary Academic 549 177 148 188 125 80
Nonacademic 69 31 14 21 19 6
Temp & Nonacademic 12 6 4 5 5 1
Total 1187 466 348 417 282 177

Table 1. Demographic make-up of the study sample.

Notes: LEMM – Language, Epistemology, Metaphysics, and Mind; SLM – Science, Logic, and Math. Columns do not total to the entire sample size as they do not incorporate missing data.

An Overview of the Results

While in the 2016 report analyses drew comparisons between and across all placement types, in the 2015 and present analyses, permanent academic placements were contrasted with the set of all other types of placements. Out of a total of 1,185 individuals, most were not placed in permanent academic positions within two years of graduation, and more men than women were placed in such positions. (See Figure 1 below).

Graph of AOS by Gender

Figure 1: The proportion of individuals who were not placed within 2 years into a permanent academic job as well as the proportion of those placed who were women and men. 472 were removed for NA values in either Gender, AOS or Permanent Placement (i.e. year of permanent placement is unknown).

Although more men than women were placed in permanent academic positions within two years of graduation, it is estimated that women have a 0.50 unit increase in the expected log odds of finding such placement. Many prefer to view this in terms of odds ratios. In terms of odd ratios, our findings (in Table 4 here) show the following:

  • The odds of women obtaining a permanent academic placement within two years is 65% greater than men when all else is held constant.
  • The odds of obtaining a permanent academic placement for the 2014 cohort compared to the 2012 cohort decreased by 33%.
  • Similarly, the 2015 cohort has a 54% decrease in odds of obtaining a permanent academic placement. It should be noted, however, that at the time of this analysis the 2015 cohort has not yet had two years post-graduation to secure placement.
  • And lastly, History and Traditions has 35% decreased odds of obtaining a permanent academic placement compared to Language, Epistemology, Metaphysics, and Mind.

Hypotheses and Future Directions

As stated above, the gender effect found in 2015 was replicated in 2016. (See Figure 2 below).

Graph of proportion of placements by gender

Figure 2: The proportion of placements by Gender for graduates 2012-2015. It shows that a higher proportion of women are placed in permanent positions within two years of graduation than men. Of 1,657 individuals, 46 were removed for NA values in either Placement or Gender (NA value in placement means that permanent placement was obtained but the year of that placement is unknown).

We have two hypotheses regarding this result that we hope to explore:

  1. Women philosophy graduates are more likely to find permanent academic placements because those women who would have been less likely to find permanent academic placement are more likely to leave the discipline before graduating than men in the same position. Here is one example of how this might occur: women are less likely to receive positive feedback or more likely to face a hostile environment than men such that less confident women are more likely to leave the field than less confident men. Women graduates are thus more confident, on average, than men graduates and confidence boosts likelihood of placement. (Thanks to “another commenter” at Daily Nous for this hypothesis.) To test this we will need attrition data that include gender. We intend to ask philosophy programs for this information in our next round of data gathering.
  2. Women philosophy graduates are more likely to find permanent academic placements because women are more likely to specialize in areas sought by hiring programs. Although our analyses accounted for first-reported area of specialization, they did not account for the area of specialization sought by the hiring program. To test whether hiring AOS helps to explain the gender effect, we intend to match our placement data to the job ads from the same time period.

Another possibility is that women philosophy graduates are more likely to find permanent academic placements because hiring programs have a preference for hiring women, all else being equal. This hypothesis has found some support in STEM fields. We do not now have plans to test this hypothesis, but could attempt in future to gather information from graduates relevant to hiring, such as publication and teaching records.

Supplementary materials and methods for this post can be found here.

Current project personnel include Carolyn Dicey Jennings, who leads the project; Angelo Kyrilov of Electrical Engineering and Computer Science and Bryan Kerster of Cognitive and Information Sciences, who work on database development and interface; Patrice Cobb of Psychological Sciences, who works on statistical analyses; Chelsea Gordon, Samuel Spevack, and David Vinson of Cognitive and Information Sciences, who work on the qualitative survey and linguistic analyses; Justin Vlasits of Philosophy at UC Berkeley, who serves primarily as an advisor; and Evette Montes, who is an undergraduate research assistant. Your feedback is welcome! You can contact us at the project email account: apda@ucmerced.edu or in the comments below. 

5 thoughts on “Academic Placement Data and Analysis: An Update with a Focus on Gender

  1. Another possibility is that women philosophy graduates are more likely to find permanent academic placements because hiring programs have a preference for hiring women, all else being equal. This hypothesis has found some support in STEM fields. We do not now have plans to test this hypothesis, but could attempt in future to gather information from graduates relevant to hiring, such as publication and teaching records.

    I’m surprised to read that the authors have no plans to test this hypothesis, which they admit is supported by evidence in STEM fields, while they do have plans to test the previous two hypotheses, for which they cite no preliminary evidence.

    Also, it would be great if the authors could pre-register the studies they intend to conduct.

  2. (To clarify: the authors do cite a study when mentioning the first of the two hypotheses they intend to test, but this study is only indirectly relevant to that hypothesis. What I find surprising is that the hypothesis they show less interest in testing is precisely the hypothesis that appears to have the strongest evidence behind it.)

  3. So the one hypothesis put to one side for the 65% difference is explicit or implicit bias in favor of women. Progress!

  4. I think the second hypothesis is clearly falsified by the Supplementary data from Tables 2 and 3 (following your link).
    Women make up 28.2 % of the population.
    They constitute 27. 3 % of those identifying as LEMM
    They constitute 28.4 % of those identifying as VALUES
    They constitute 28.9 % of those identifying as HISTORY
    and they constitute 28.2 % of those identifying as SCIENCE
    There is little variability in the proportion of women in each area. Further, women seemed to secure a disproportionately high percentage of the jobs in LEMM and SCIENCE. In SCIENCE the percentage of women is identical with the percentage in the population as a whole. In LEMM, on the other hand, they constitute a smaller proportion than they do in the population as a whole. This is contrary to what you would predict from Hypothesis 2.

Leave a Comment

WordPress Anti-Spam by WP-SpamShield