New APPS

Art, Politics, Philosophy, Science. Mostly Philosophy.

recent posts

about

    Women and Minorities in Philosophy: Which Programs Do Best? (Updated 1/21/16)

    Most of us know about efforts to sort philosophy programs according to placement rate or prestige, but what of the percentage of PhD graduates from each program who are women or other underrepresented minorities? Thanks to Eric Schwitzgebel's efforts in contacting the National Center for Science Engineering Statistics (see here and here), we have access to some numbers on this issue. Specifically, the NCSES supplied the number of women and minority graduates from doctoral philosophy programs in the United States between the years 1973 and 2014 (but not broken down by year). Below, I provide the top programs in the United States from this list of 96 programs in terms of % of women graduates in this period, as well as the top programs in terms of % of non-white graduates, where for "non-white" I am aggregating the NCSES categories of "Hispanic," "Asian," "Asian or Pacific Islander," "Black" and "two or more races." (I omitted institutions from the NCSES data that no longer offer doctoral degrees in philosophy.) One striking feature of these lists is how many of the programs show up on Brian Leiter's list of PhD programs "whose existence is not easy to explain."  A provocative rhetorical question follows: Should we be closing PhD programs that better serve women and minorities in philosophy? I welcome discussion below.

    Three notes of caution (added 1/21/16):

    1) These data are from 1973 to 2014, and so give only a longterm view of diversity. All programs are likely to have increased in diversity and many programs likely look much better with a more shortterm view (thanks to Chris Stephens for this point).

    2) Programs may be above average without this deviation from average being statistically significant (thanks to David Wallace for this point). 

    3) Aggregating the NCSES categories of "Hispanic," "Asian," "Asian or Pacific Islander," "Black" and "two or more races" into "non-white" may make less visible the differential treatment of these groups (thanks to Lionel McPherson for this point). See my new post for a disaggregated view. 

    Thanks are also due to Elizabeth (who I am not sure would want me to specify her last name), for helping me to realize I should be listing programs above the mean, rather than above the overall % (except in the case of the list of 11 programs, since there I am more concerned with the set of students as a whole than the individual programs).

     

    Here are philosophy PhD programs organized by % of women graduates, listing only those 39 programs with a greater or equal than mean percentage of women graduates for this time period (1973-2014). 

    Institution Men Women Grads 1973-2014 % Women
    University of Memphis 24 31 55 56.36%
    Georgetown University 107 77 184 41.85%
    Arizona State University 13 9 22 40.91%
    Binghamton University 78 54 132 40.91%
    Massachusetts Institute of Technology 87 49 136 36.03%
    University of Oregon 67 37 104 35.58%
    University of Minnesota 117 63 180 35.00%
    University of South Carolina 19 10 29 34.48%
    Stony Brook University 145 74 219 33.79%
    University of Washington 57 29 86 33.72%
    UC Davis 42 21 63 33.33%
    Temple University 101 49 150 32.67%
    University of New Mexico 29 14 43 32.56%
    University of Pennsylvania 106 51 157 32.48%
    University of Colorado 108 50 158 31.65%
    Bowling Green State 50 23 73 31.51%
    DePaul University 88 40 128 31.25%
    University of North Carolina at Chapel Hill 135 59 194 30.41%
    University of Florida 33 14 47 29.79%
    Harvard University 154 63 217 29.03%
    City University of New York 173 70 243 28.81%
    Rice University 50 20 70 28.57%
    Washington University in St. Louis 93 37 130 28.46%
    University of Maryland 66 26 92 28.26%
    Loyola University Chicago 106 40 146 27.40%
    University of Tennessee 80 30 110 27.27%
    University of Illinois, Chicago 73 27 100 27.00%
    SUNY Albany 30 11 41 26.83%
    Stanford University 175 64 239 26.78%
    Michigan State University 85 31 116 26.72%
    University of Massachusetts Amherst 127 46 173 26.59%
    Emory University 136 48 184 26.09%
    University of Nebraska-Lincoln 49 17 66 25.76%
    University of Utah 35 12 47 25.53%
    New York University 79 27 106 25.47%
    Rutgers, New Brunswick 120 41 161 25.47%
    University of Pittsburgh 200 68 268 25.37%
    Boston University 189 64 253 25.30%
    Claremont Graduate University 72 24 96 25.00%
    Overall Total Numbers and Mean Percentage 10,282 3,283 13,565 24.57%

     

    Here are philosophy PhD programs organized by % of non-white graduates (among permanent residents or citizens of the United States), listing only those 38 programs with a greater or equal than mean percentage of non-white graduates for this time period (1973-2014). 

    Institution # Non-White Citizens/Permanent Residents # Citizens/Perm Residents 1973-2014 % Non-White Citizens/Permanent Residents
    Arizona State University 6 20 30.00%
    Bowling Green State 11 54 20.37%
    Binghamton University 19 104 18.27%
    University of Hawaii, Manoa 15 90 16.67%
    Wayne State University 6 36 16.67%
    University of Miami 13 79 16.46%
    University of Memphis 7 45 15.56%
    UC Santa Barbara 16 103 15.53%
    New School 28 189 14.81%
    University of New Mexico 6 42 14.29%
    University of South Florida 6 46 13.04%
    University of Oklahoma 10 77 12.99%
    DePaul University 14 108 12.96%
    UC Santa Cruz 4 31 12.90%
    Stony Brook University 25 195 12.82%
    Stanford University 20 159 12.58%
    University of Oregon 11 89 12.36%
    Harvard University 19 155 12.26%
    University of Michigan 22 183 12.02%
    Brown University 16 134 11.94%
    UC Riverside 10 85 11.76%
    Boston College 22 193 11.40%
    Claremont Graduate University 10 89 11.24%
    UCLA 17 152 11.18%
    Baylor University 4 36 11.11%
    Duke University 9 83 10.84%
    University of Cincinnati 4 38 10.53%
    City University of New York 20 191 10.47%
    Columbia University 22 212 10.38%
    UC Irvine 12 121 9.92%
    Purdue University 16 162 9.88%
    University of Wisconsin 21 221 9.50%
    Indiana University, Bloomington 10 107 9.35%
    University of Illinois, Chicago 8 86 9.30%
    Yale University 17 183 9.29%
    University of Utah 3 33 9.09%
    Rutgers, New Brunswick 11 122 9.02%
    University of Maryland 6 67 8.96%
    Overall Total Numbers and Mean % 909 11,294 8.51%

     

    Overall, from 1973-2014, 96 philosophy programs had 24.20% women (among all graduates) and 8.05% non-white (among U.S. permanent residents and citizens) PhD graduates. The numbers for the programs listed at Brian Leiter's blog (Texas A&M's data were unreported due to low numbers of graduates) are 28.27% women and 9.73% non-white graduates, listed below alphabetically:

    Institution Men Women Grads 1973-2014 % Women # Non-White Citizens/Permanent Residents # Citizens/Perm Residents % Non-White Citizens/Permanent Residents
    Arizona State University 13 9 22 40.91% 6 20 30.00%
    Binghamton University 78 54 132 40.91% 19 104 18.27%
    Emory University 136 48 184 26.09% 8 162 4.94%
    Michigan State University 85 31 116 26.72% 7 99 7.07%
    SUNY Albany 30 11 41 26.83% 3 38 7.89%
    Tulane University 81 17 98 17.35% 5 88 5.68%
    University of Florida 33 14 47 29.79% 3 37 8.11%
    University of Kansas 82 17 99 17.17% 7 84 8.33%
    University of Oregon 67 37 104 35.58% 11 89 12.36%
    University of South Carolina 19 10 29 34.48% 1 24 4.17%
    Villanova University 39 10 49 20.41% 1 44 2.27%
    Subtotal 713 281 994 28.27% 82 843 9.73%
    Overall Totals and % 10,282 3,283 13,565 24.20% 909 11,294 8.05%

    What of the provocative rhetorical question provided above? Should these programs close, as is suggested by Brian Leiter? An argument against this is that these programs (as a set) host more women and other underrepresented minorities than philosophy as a whole, and so should be celebrated for that fact. Yet, one might be concerned that these graduates are not actually "better served" if one thinks these programs are less prestigious or have lower placement rates. I am not one to argue for the closure of PhD programs on the grounds of prestige, so I am going to set that aside. As for placement rate, two of these programs have very high placement rates into tenure-track positions (Oregon and Villanova). Although the data gathering efforts of the Academic Placement Data and Analysis project are not yet complete, I suspect that they will both land in the top 20 in terms of percentage of graduates between 2012 and 2014 who have by now found tenure-track positions. A few other programs from this list look to have tenure-track placement rates for this period that are just about the same as some top ten programs, according to PGR prestige. So I don't think that placement rates are likely to warrant the closure of many of these programs. Then again, I am not in favor of closing graduate programs (I would rather see philosophy aim for more transparency and relevancy than fewer programs!), so perhaps I am biased. As I say above, discussion is welcome!

    Also, feel free to take a look at the raw data here. 

    Update: Perhaps worth noting is that some programs listed here are newer than others, and so it is worth exploring whether their percentages are inflated, relative to other programs. For purposes of further transparency, I italicize above the names of any programs who had a smaller number of graduates 1994 or prior (a rough halfway point) relative to graduates 1994 or later than these 98 programs taken overall. (That is, I italicized all programs whose number of graduates 1994 or earlier was less than 44.4% of their total graduates for this time period, since these programs overall had 44.4% of graduates 1994 or earlier.) Around half of the graduate programs discussed above have this quality, which is what we would expect (22 of 41 in the first section, 24 of 41 in the second section, and 6 of 11 in the third). In other words, more recent programs do not seem to be especially over-represented in the above lists.

    Another Update: Thanks to Brian Weatherson's suggestion, I am updating the above for the non-white category to reflect the fact that race and ethnicity was reported for only permanent residents and citizens of the United States. Per his suggestion, I am now using the total number of permanent residents and citizens as the total for these calculations (but not for gender).

    Updates Galore! To check whether the difference between the 11 programs mentioned above and all other programs is statistically significant with respect to the percentage of women and non-white graduates, I ran a test in the second tab of the spreadsheet, linked above. As I say in comments below, the difference is highly statistically significant for the percentage of women graduates (p=.002) and very nearly statistically significant for the non-white graduates (p=.06).

    Update 1/21/16: I replaced overall % with mean % as the cut off for the above lists (thanks to a comment by Elizabeth), removed Oklahoma State University (thanks to a comment by Anonymoose), updating numbers where required, and added three notes of caution, above.

    Posted in , , , ,

    45 responses to “Women and Minorities in Philosophy: Which Programs Do Best? (Updated 1/21/16)”

    1. Bweatherson Avatar
      Bweatherson

      If I’m reading the tables right, the # Non-White column is a little misleading. It is, I think, a count of the number of non-white citizens/permanent residents who graduated. The program could have graduated any number of non-white temporary residents, but they wouldn’t show up here.
      If that’s right, maybe the percentages in the table should not be percentage of total degrees awarded, but percentage of degrees awarded to citizens/permanent residents. I don’t think that would change the rank order too much, but I know it would make some places look a little less lily-white.

      Like

      Reply
    2. Grad sockpuppet Avatar
      Grad sockpuppet

      Here’s a concern about the New School, at least. While it has (and graduates) an enormous number of students (82 currently enrolled PhD students; I can’t find the number of MA students), it only has ten full-time, non-emeritus faculty, and does not provide those students with full funding (and that information isn’t readily available on their website). So, while they may do fairly well in terms of percentages of women and non-white students, one wonders whether it actually does right by them.

      Like

      Reply
    3. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      This is a good point. I am updating the numbers now. Thanks for the suggestion!

      Like

      Reply
    4. Alison Stone Avatar
      Alison Stone

      How many of the programs are focused more on continental philosophy? I’m wondering whether that might be a reason they were viewed negatively by Leiter. But those might also have relatively great appeal to women and/or non-white people (might).

      Like

      Reply
    5. Chris Stephens Avatar
      Chris Stephens

      Hi Carolyn,
      Thanks for this data. I do wonder about your use of data that goes all the way back to 1973, as well as your focus on these programs as a “set”. If I’m following your suggestion, it is left open for someone to wonder whether a program that doesn’t score well on % of women or non-whites then doesn’t have this reason not to close (e.g., Kansas). At the same time, I’d be reluctant to put too much weight on this data, since it goes all the way back to 1973 – what if these programs have changed significantly in the last decade or two (e.g., Kansas had no women philosophy faculty until Ann Cudd was hired around 1987. But now they have more – how would this have affected the number of women graduate students before and after? Mutatis mutandis for when Derrick Darby was on the faculty there.
      I understand your wanting to get a bigger sample (going back to 1973) but I worry that what a department was like back in the 70s and 80s is not likely to be relevant today (again, to continue the example, NONE of the current philosophy faculty at Kansas were on the faculty in the 70s and early 80s. Times change, and so do faculty compositions.
      I know you’re not in favor of closing programs, but I think we should be very cautious about using this data in any arguments for or against whether a PhD program should be closed.

      Like

      Reply
    6. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Great points. Thank you!
      “If I’m following your suggestion, it is left open for someone to wonder whether a program that doesn’t score well on % of women or non-whites then doesn’t have this reason not to close (e.g., Kansas).”
      9 of the 11 programs mentioned by Brian have a higher percentage of either women graduates or non-white (citizen/permanent resident) graduates than the percentage for all of the covered programs for this time period. To me, this provides a special reason for their existence, beyond others. I don’t think that this reason should operate on its own, and the best data would probably be combined with qualitative information about graduate student experiences, retention rates, etc. I meant for this to be a conversational starting point. Should these factors be taken into consideration when we evaluate programs, and what are the limits of how we should take them into consideration?
      As for the time period covered: I went back to 1973 because that is how the demographic data were provided–as a chunk for that period of time. I plan to ask some day soon for a more recent chunk. My assumption is that if a program has a higher percentage than the overall percentage of women or non-white graduates for this time period, then that program is more reliably supporting those types of students than if we looked at a shorter time period. But I do think a shorter time frame would be helpful here. I will put up a new post if I manage to get the data.

      Like

      Reply
    7. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Thank you for this. This year APDA plans to put together a qualitative survey that may capture differences between programs not covered by the above. Information on funding and retention would also help. I will think about it some more.

      Like

      Reply
    8. David Wallace Avatar
      David Wallace

      “9 of the 11 programs mentioned by Brian have a higher percentage of either women graduates or non-white (citizen/permanent resident) graduates than the percentage for all of the covered programs for this time period. ”
      If you picked 11 programs at random, on average 8.75 would have a higher percentage of either women graduates or non-white (citizen/permanent resident) graduates than the percentage for all of the covered programs for this time period. So by itself that isn’t a persuasive argument.

      Like

      Reply
    9. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Here is another way of looking at this: 32 of the 97 (33%) programs have neither a higher percentage of women graduates than the overall percentage nor a higher percentage of non-white graduates, whereas only 2 of the 11 (18%) programs listed by Brian Leiter have this quality. Conversely, 67% of all programs have one or the other feature, whereas 82% of the programs from this list have one or the other feature. To me, this is a striking difference.

      Like

      Reply
    10. David Wallace Avatar
      David Wallace

      8.25, sorry, typo.

      Like

      Reply
    11. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      My guess is that this number is still not quite right, but I am not sure how you are calculating this, so I don’t know which assumptions might be off.

      Like

      Reply
    12. David Wallace Avatar
      David Wallace

      Replying to 9:
      “To me, this is a striking difference”.
      To me, it’s noise. Or to be more accurate, those numbers by themselves don’t obviously suggest anything but noise. (11 is a small sample.) You’d need to look at the actual scale of the deviation, institution by institution, to show there’s something more interesting going on.
      At this point I’d come back to my comment at Feminist Philosophers: this is driven by 3 of the 11 that really do have a statistically significant excess of women and non-whites. The other 8, collectively, are below average.

      Like

      Reply
    13. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      I don’t think it is noise. Check out the second tab of the linked spreadsheet above. If we compare the 11 programs mentioned by Brian Leiter to the rest of the programs (omitting the 11 from that set), then the difference is highly statistically significant for the percentage of women graduates (p=.002) and very nearly statistically significant for the non-white graduates (p=.06).
      As to the 8.25 number: if the whole has 67% programs with one or the other of these features, a sample of 11 should have around 7. This sample has 9.

      Like

      Reply
    14. David Wallace Avatar
      David Wallace

      To 11: (sorry, this Is getting a bit interleaved and hard to follow)
      8.25 is just 75% of 11, which is what you’d predict on the crudest of null hypotheses: mean and median coinciding, no correlation between number of women and number of non-whites. Both of which may be false, but I had the impression that you thought 9 out of 11 was striking in itself, not that it was striking given some particular features of the overall distribution – hence my reply.
      You’ve now given the actual population statistics, on which basis a randomly-selected 11 programs would have 7.33 programs with one or other feature, with a standard deviation of about 1.6. So 9 out of 11, by itself, looks pretty uninteresting – barely one standard deviation from the prediction. (All calculations done late enough at night – UK time – that I welcome corrections!)

      Like

      Reply
    15. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Ok, these assumptions make some sense, but I knew them both to be false from just a rough sense of the data. Although, I should add that I did not take the mean percentage. Mean percentage (24.71%) is a bit higher than overall percentage (24.24%), and median percentage is lower than either (23.53%). And, yes, I really ought also get back to work!

      Like

      Reply
    16. David Wallace Avatar
      David Wallace

      To 13 (and again, sorry this has got so interleaved!)
      – it is not statistically significant that 9 out of 11 programs in that list have above-average numbers of non-white or women students. The probability of at least 9 out of 11 programs being above average in one or other (given your 67% figure) is 23% (p=0.23, if you like).
      – it is a separate question whether the overall deviation from the average is statistically significant. It is. That’s entirely because of 3 programs of the 11 – Arizona State, Binghamton, Oregon – have strongly above-average numbers of both (statistically significant, in one or other category, for each institution individually). The other 8 programs do not deviate significantly from the overall figures (and are in fact slightly lower).
      To make the point vividly as to why I think it’s so misleading to describe this list of 11 institutions as collectively being above average in their representation of women and non-whites, consider this: fully 17% of the students in these 11 programs are based in upstate New York, as against only 6.4% overall. That’s a highly statistically significant difference and so technically it’s correct to say that these 11 programs, as a set, consist disproportionately of upstate New Yorkers. But it would be ridiculous to say that we’ve thus identified a general property of the 11, rather than just identified a feature of 2 of the 11 (Binghamton and SUNY Albany). There is no sense in which Oregon or Arizona State has disproportionately many upstate-New York-based students!

      Like

      Reply
    17. Derek Bowman Avatar
      Derek Bowman

      As Grad Sockpuppet has pointed out, enrolling members of underrepresented groups in PhD programs is only meritorious if those programs provide the students with good conditions while they study and good prospects upon completion. Merely graduating PhDs is not good evidence for this, given the oversupply of PhDs compared to academic jobs and the lack of well-established non-academic career paths for philosophy PhDs.
      For a discussion of the ethical issues in recruiting members of underrepresented groups into philosophy, see here:
      http://sgrp.typepad.com/sgrp/2015/09/bar-on-asks-should-we-occupy.html

      Like

      Reply
    18. Mark lance Avatar
      Mark lance

      I too would love to see this for more recent years. Many schools have changed a great deal since 73. GU certainly has. I was very surprised that we did so well going that far back, we do much better now.

      Like

      Reply
    19. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Ah, I see. I assumed that your criticisms were more general. Look, I didn’t choose the set. It would be arbitrary for me to remove three of the programs from the set in looking at this issue. If it so happens that 3 of the 11 mentioned are especially diverse, then focus on that when considering the substance of the post–should we be calling on Arizona State, Binghamton, and Oregon to close, given how diverse these programs are, and given how the discipline as a whole is not sufficiently diverse? Do you have any thoughts about the substance of the issue?

      Like

      Reply
    20. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      My perspective is that the PhD is worthy in itself and that supporting students through the dissertation is a service in itself. Moreover, doing so is not equivalent to recruiting underrepresented groups–programs with diverse students may simply be doing a better job of accepting underrepresented applicants and supporting them through the dissertation.

      Like

      Reply
    21. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      My supposition is that all programs are doing better now. But it would be good to see by how much and also how they compare on this point, so we are in agreement that more recent data would be an improvement on this.

      Like

      Reply
    22. Chandra Sripada Avatar
      Chandra Sripada

      The difference between the Leiter 11 and the remaining programs appears statistically significant for women (p=.03) though not for non-whites (p=.13). I used permutation methods as the data appear unlikely to conform to distributional assumptions for parametric tests (sample 11 of the 97 programs randomly, calculate proportion of women/minorities in the 11, repeat 10,000 times to get null distributions for the proportions). Despite the statistically significant difference, however, the actual magnitude of the difference is quite small. And there is striking heterogeneity amongst programs as David points out. Given the heterogeneity, a good use of this data is to motivate folks to investigate what a few programs (Az State, Binghamton, U. Memphis, Georgetown) are doing right.

      Like

      Reply
    23. C Marks Avatar
      C Marks

      Effect size matters more than statistical significance.

      Like

      Reply
    24. David Wallace Avatar
      David Wallace

      Thanks to Chandra for doing that – I’ve just been trying something similar but rather more crudely, and got a similar answer. (You can do an analytic approximation by ignoring the fact that departments vary in size and just working out the standard deviation in the percentages of women and nonwhites. On that basis you get a s.d. (in the 11 programs) of 2.1% for women, 1.4% for nonwhites. I don’t have much of a sense of how good that approximation is, though.)
      To Carolyn Dicey Jennings @13: I assume your “p=0.002” figure is based on a null hypothesis that the 994 students in the 11 schools are selected at random from the population of all students? That’s not the situation being considered here: the right null hypothesis is that the 11 departments are selected at random from the population of all departments. (I made the same mistake in my 16 when I agreed that the deviation from the average is statistically significant.)
      To Carolyn Dicey Jennings @19: I don’t have any general thoughts on the question of whether (and if so, which) US PhD programs should close; I don’t know the US system nearly well enough. But if one accepts Brian Leiter’s position that they should close except for being concerned about harming representation, the right conclusion would be to close the other 8 and leave those 3, not to reject the overall position and leave all 11 open. (If one rejects Leiter’s position on grounds independent of representation, then the data here is irrelevant in any case.)

      Like

      Reply
    25. Derek Bowman Avatar
      Derek Bowman

      The PhD may well be worthy in itself (I’m inclined to agree about much of the education behind the PhD, though not about the credential itself). That doesn’t mean a department hasn’t done its students a disservice if those students routinely graduate with significant debt and struggle to find employment. I don’t know if that’s true at these particular programs, and I don’t know that it’s any less true for many programs that – due to faculty prestige – are presumed to be fine.
      I agree that there is a difference between recruiting and “accepting and supporting” underrepresented applicants, but I’m shocked if you think that the latter doesn’t raise any of the same concerns as those raised by Professor Bar On in my earlier link.

      Like

      Reply
    26. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Thank you to Chandra and David for your insights, but I want to set aside technical discussion on whether the difference in diversity between the 11 programs singled out by Brian Leiter and the rest of the programs covered in this data set is statistically significant. I think this is distracting from the overall topic and message of this post. If someone thinks I have made an error in any of the above, please let me know. Comments and questions on the substance of the post continue to be welcome. I am particularly interested in whether and how we should compare and evaluate programs on the basis of support for women and minorities, and would really like to hear from more people on this. (I hope to respond to David and Derek’s ideas on this soon.)

      Like

      Reply
    27. Elizabeth Avatar
      Elizabeth

      It seems to me (from reading the post/comments but not looking at any of the calculations) that the difference between David Wallace’s interpretation of the data and Carolyn Dicey Jennings’s interpretation is that DW takes the unit of analysis to be the program and CDJ takes it to be the individual student. Hence, CDJ’s analysis weights larger programs more than smaller programs and DW’s weights all programs equally. Is that accurate?
      Different units of analysis will yield strikingly different views of data when the more aggregated units (e.g., programs) differ greatly in size (number of students). Choosing the appropriate unit of a statistical analysis is a subtle and difficult issue because it forces one to articulate precisely what question the analysis is intended to answer.
      For example, if the question is, “Does this list of 11 programs offer any prima facie evidence of bias against programs serving female students or students of color?” then it is hard for me to see how the student could be the appropriate unit of analysis.
      On the other hand, if the question is, “Would following a policy of closing these programs disproportionately affect women and students of color?” then I think the students are the appropriate unit of analysis. In that case it’s not directly relevant (though it presumably is important for other reasons) that all of the “action” in achieving that result may come from one or a few programs.
      The more general point is that, when offering analyses like these, I think it is very important to be explicit about what the implications of the analyses are intended to be, and to engage with whether the unit of analysis chosen is the correct one given that particular question.
      I have to say that I feel that some of the discussion here (and elsewhere commenting on this post) does implicitly slide between different questions, e.g., roughly those I suggested above.
      On a different note: Given the general time trend toward more women and more students of color over cohort, to make sense of these data I would also want to statistically control for year. This is true for either unit of analysis, but particularly for a student-level analysis. This would be relevant since some programs’ total students will be heavily from early years and others heavily from later years, and still others more uniformly distributed across time. I know that there is a rough version of this above in that programs are dichotomized into below-average or above-average numbers of their total graduates being before 1994, but this strikes me as a problematic way to do this control since it dichotimizes the data twice (before/after 1994, below/above the mean), which loses a lot of information from a continuous distribution.

      Like

      Reply
    28. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      I published Elizabeth’s comment because I suspect she started writing it before mine, but I do not intend to continue publishing comments on this issue, however interesting they may be, because I think it distracts from the substance of the post.

      Like

      Reply
    29. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      “Does this list of 11 programs offer any prima facie evidence of bias against programs serving female students or students of color?”
      If I understand the point here, it is not the question I am after. That is, I am not interested here in the question of whether Brian Leiter is biased against women or students of color. For me, this part of the post is not a personal matter but about what I take to be an important topic of discussion–whether these programs should close. I am sorry that Brian Leiter is taking it to be a personal matter.

      Like

      Reply
    30. LK McPherson Avatar
      LK McPherson

      “I am particularly interested in whether and how we should compare and evaluate programs on the basis of support for women and minorities, and would really like to hear from more people on this.”
      For “minorities,” any useful comparison and evaluation will require disaggregating the “non-white” categories. The historical and contemporary circumstances, stereotypes and biases, underrepresentation percentages, injustice dynamics, and corrective justice considerations are not roughly similar across the groups in question (even if the analysis is limited to US citizens, which is also a significant factor).

      Like

      Reply
    31. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Great point. I will aim to do this in future.

      Like

      Reply
    32. Ed Kazarian Avatar
      Ed Kazarian

      Thank you so much for putting all of this together, Carolyn. Your work on all of this is important and very much appreciated.

      Like

      Reply
    33. Eric Schwitzgebel Avatar
      Eric Schwitzgebel

      Yes, thanks Carolyn! Interesting analyses.

      Like

      Reply
    34. Anonymoose Avatar
      Anonymoose

      Oklahoma State University does not have a PhD program in philosophy so I’m not sure what data you are looking at when you include it on your list. And does Claremont Graduate University offer a PhD in philosophy? I couldn’t find any information on that. Forgive me if I’ve misunderstood your data.

      Like

      Reply
    35. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      I must have overlooked OSU, thank you for pointing that out. (The program list had many institutions that no longer offer PhDs, which I removed by hand.) As for Claremont: http://www.cgu.edu/pages/4359.asp .

      Like

      Reply
    36. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      “But if one accepts Brian Leiter’s position that they should close except for being concerned about harming representation, the right conclusion would be to close the other 8 and leave those 3, not to reject the overall position and leave all 11 open. (If one rejects Leiter’s position on grounds independent of representation, then the data here is irrelevant in any case.)”
      I am not sure if this accurately describes Brian’s view, since I am not sure if he has presented the full list of variables that could matter to him just yet. In any case, I would say that programs that especially support either women or racial/ethnic minorities have value, and so 9 of the 11 would have an argument in their favor on these grounds. They may also have others. I would want to see a systematic review of all programs based on all relevant variables to say more here.

      Like

      Reply
    37. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      I just want to point out that, as I mention in the post, 2 of the 11 programs have very good placement rates, and several others have placement rates that are competitive with top 10 programs (top 10 according to the PGR). As for Bar On’s article, I have not had time to study it with the care it deserves. One of my concerns about the conclusion is that “radical rejection” will favor confident, self-assured applicants over those who are less confident and self-assured, and I am not sure that is a good thing.

      Like

      Reply
    38. David Wallace Avatar
      David Wallace

      To Carolyn at 36:
      “In any case, I would say that programs that especially support either women or racial/ethnic minorities have value, and so 9 of the 11 would have an argument in their favor on these grounds.”
      For 6 of those 9, there is no reason at all to think their “support” is anything other than random noise.

      Like

      Reply
    39. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      On the one hand I am not invested in using statistical significance to determine the matter. On the other I would be more confident in agreeing or disagreeing with your statement if it came from a professional statistician with some neutrality on the issue and the time/energy to think carefully about it.
      EDIT: This could have been clearer. What I mean is that I would be more confident in agreeing with your statement if it came from a professional statistician with some neutrality on the issue and the time/energy to think carefully about it but ALSO that I would be more confident in disagreeing with your statement if this disagreement came from a professional statistician with some neutrality on the issue and the time/energy to think carefully about it.
      The point is just that I don’t think your claim has been established but this is also not a point that I find important enough to dig in to the details about right now. I think the way I wrote it above could come across as disrespectful, and I want to be clear that I mean no disrespect.

      Like

      Reply
    40. David Wallace Avatar
      David Wallace

      (a) I’m fairly sure I am neutral on the issue – and it does not take a lot of time / energy (let alone being a professional statistician) to do elementary things with the binomial distribution and its normal approximation. (The statistics component of my physics qualifications will do fine, if credentialing is really necessary here.)
      Look: of those 9, 3 clearly are statistically significant deviations by themselves. The others, collectively and individually, are not statistically significant at all. (p>0.1 in all individual cases that are above the average; the remaining 8 collectively are below average in both categories.)
      (b) You say “I am not invested in using statistical significance to settle the matter”. Are you seriously saying that it doesn’t matter to your view whether the effect you’re discussing is random noise?

      Like

      Reply
    41. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      I am just going to let David have the last word on this issue and set it aside, per my comments above.

      Like

      Reply
    42. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Thanks again for your suggestion–I put up a new post: http://www.newappsblog.com/2016/01/racial-and-ethnic-minority-graduates-of-philosophy-programs-a-more-detailed-look.html. Please let me know if you have further thoughts and suggestions.

      Like

      Reply
    43. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Thanks for your comment–I changed the cut off for the above lists to mean program % based on the ideas you present here.

      Like

      Reply
    44. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Ok, OSU has now been removed. Thanks again!

      Like

      Reply
    45. Carolyn Dicey Jennings Avatar
      Carolyn Dicey Jennings

      Thanks again for your input–I added a note of caution to this and a new post to help readers keep at least one of your concerns in mind.

      Like

      Reply

    Leave a comment