Where did CS PhD’s get their undergraduate degrees?

January 22, 2013 at 1:04 am 15 comments

The latest issue of Computing Research News has a report from CRA-E (their Education subcommittee) on where CS PhD’s come from.  Research universities, institutions that stop at Masters degrees, four year colleges, or top liberal arts institutions?  Turns out the answer is that the vast majority of CS PhD’s get their undergraduate degrees from research universities, but the sum of the PhD’s who get their undergraduate degrees from the top 25 liberal arts institutions is greater than any single research institution.  There’s also evidence that the research universities produce better graduate students, using NSF fellowships as the quality metric.  That was quite unexpected — I would have guessed that the four years and the liberal arts institutions would have played a much greater role.

In 2010, 1665 Ph.D.’s were awarded in computer science of which 714 went to domestic students.   Approximately 71% of the domestic Ph.D.’s received their undergraduate degrees from research universities, 15% from master’s institutions, 11% from four-year colleges, and 4% from other colleges.  These proportions have remained essentially unchanged since 2000 with all four types seeing similar increases since 2005.

via Computing Research News – Online – Computing Research Association.

Entry filed under: Uncategorized. Tags: , , , , .

First, Do No Harm: Inequality in American Education Will Not Be Solved Online Here’s the Real Reason There Are Not More Women in Technology – Forbes

15 Comments Add your own

  • 1. geekymom  |  January 22, 2013 at 6:14 am

    More than half to non domestic students? Wow!

  • 2. alfredtwo  |  January 22, 2013 at 6:36 am

    The more than half non domestic students sent a shiver though me for starters. Scary.

    But having gone to a smallish four year liberal arts school the rest of the story concerns me as well. I have seen a bias towards large research schools in industry hiring over the years. I have long believed that caused a lot of serious talent to be overlooked. I’ve heard some stories suggesting bias in entrance to graduate programs as well. The occasional story of “yes your record in industry looks good but we never heard of your undergraduate school” as a reason for not being accepted for a graduate program. How wide spread that is I have no idea but it would make an interesting story as well.

    What I see is a form of tracking. If you do really well in high school you can get into a really good undergraduate school which will get you into a good graduate school which will get you into a PhD program. If however you are seriously talented in CS in HS but ignore your other subjects (happens a lot more than any of use like) you’re probably not getting into a great undergraduate school which is going to keep you out of PhD programs. Fact or fiction? I’m not sure we know for sure.

    • 3. Mark Guzdial  |  January 22, 2013 at 7:35 am

      I suspect that grad student hiring is influencing these trends, Alfred. Faculty are now expected to cover a student’s Research Assistantship from the day that they walk in the door. We don’t accept someone unless they are good AND we have funding for the particular kind of work that they want to do. It’s not just grad admissions anymore – it’s literally a job interview.

      I don’t take on students without prior research experience, because of that. I want them to know what they’re in for, and I’m looking for a letter that says that they’re good at it. That experience and those letters may be easier to get at a research university where there is a lot of research opportunities to choose from.

      • 4. Cecily Heiner  |  January 22, 2013 at 9:15 am

        I think I agree with your assessment Mark, with an additional caveat. Not only is it easier to get a letter saying that they are good at research at an R1 school, it is probably a lot easier for them to get funding to get research experience as an undergrad at an R1 school. I teach at a primarily undergraduate regional/SLAC school right now, and although I probably have a lot of students who would like to be doing basic research(especially if it is that or work at Taco Bell) we simply do not have much funding for it available to them. I have at least a couple students who would probably be decent researchers(good at CS, excellent programmers) but who I do not have funding to support. If grad school is going to be a mid-level job interview, there ought to be a way for students at non-R1 schools to get some entry level experience that is paid.

        • 5. Krishnendu Roy  |  January 24, 2013 at 1:38 pm

          I think Cecily’s and Mark’s comments are spot-on. I also teach at a PIU (primarily undergraduate institution) and have similar experience. I just want to add that its not just lack of research-experience (as pointed out by Mark) and lack of undergraduate research-funding (as pointed out by Cecily). At times it is just the lack of exposure to a research environment.

          In my view it is similar to the problem that K12 student who do not know what computing is all about, face. The students at PIU do not know what is going on in a graduate CS program (especially is the school is far away from any R1), hence they do not see themselves as potential graduate students.

  • 6. dennisfrailey  |  January 22, 2013 at 7:51 am

    I would expect this result. 1) students with a long term interest in a doctorate are likely to choose research universities; 2) faculty at research universities would be much more likely to encourage their students to get advanced degrees. To me, the more interesting questions are what undergraduate majors lead to a doctorate in any particular field (for example, CS) and how the choice of undergraduate program relates to their later career choices and, ultimately, their career success. My impression, supported only by anecdotal evidence, is that faculty members at research institutions tend to be biased in favor of research careers for their students, even though the number of potential research jobs is quite limited.

    • 7. Mark Guzdial  |  January 22, 2013 at 9:57 am

      Dennis, you wrote “the number of potential research jobs is quite limited.” Where do you see evidence for that? You may be right, but when I look at the statistics for CS PhD’s from CRA, it seems that they’re all getting snatched up. When I look at BLS numbers, I see far more R&D slots than graduates.

      • 8. dennisfrailey  |  January 23, 2013 at 10:05 am

        My evidence is admittedly anecdotal, and it would be useful for someone to do solid research on this subject. My experience comes from having recruited in industry (for about 40 years) and from having served as an ABET program evaluator in CS and related disciplines since 1986. In my industry role I know that we usually had about 100 engineering/technical jobs for BS students for every 10 for MS students to every 1 for a PhD. (These numbers are imprecise, and varied from time to time.) We were a hi-tech company with a research arm, so we probably hired more advanced degree engineering and CS students than your typical industrial company.

        From what I’ve seen in government statistics there are a lot more jobs in industry than in universities and research organizations.

        We would typically end up with resumes from MS and PhD students for which we had no appropriate jobs, while often failing to get our quota of qualified BS applicants. A common scenario would be a suitable candidate who wanted only an internship because he or she wanted to go on to a higher degree. Sometimes I felt that they had not been given very good career advice.

        In my ABET role I sometimes observed programs whose “objectives” were to have their students be successful in research careers but whose students were, in large numbers, interested in getting their BS and getting a job. The curricula were sometimes better suited to the research careers.

        During my career I’ve also served as an adjunct faculty member (for 40 years) and this includes being on some doctoral committees and doing my share of student advising. I’ve encountered a significant number of individuals who had dropped out of PhD programs and who felt they had wasted years of their lives trying to get a research degree when they weren’t really cut out for it. I recall one fellow in particular who had two parents and three brothers, all with PhDs. He felt family pressure to get a PhD and had served as a low-salaried research assistant for many years before finally facing the facts.

        As I said, this is all anecdotal but I hate to see young people struggling in pursuit of something that isn’t right for them.

  • 9. richde  |  January 22, 2013 at 9:26 am

    This effort started when I chaired CRA-E, and one of the things we worried about was the vast disparities in size between research and liberal arts institutions. What happens when you scale for size? Are the research universities still proportionately more impactful?

    The quality result is unexpected and is hard to explain.

  • 10. Don Davis (@gnu_don)  |  January 22, 2013 at 9:58 am

    Wow. This quite the vicious cycle – How would / could Latina/o and other underrepresented student representation in CS be changed with such a system?
    1. Significant percentages of Latina/o students attend 2 year colleges.
    2. The pipeline to STEM studies from such colleges is problematically low.
    3. Underrepresented students’ choice of major is significantly impacted by the presence of ethnically similar faculty.

    This is further problematic as the number of Latina/o students receiving CS undergrad degrees is significantly bolstered by HSIs. (Overall representation would be even lower without HSIs.) Among the top CS producers listed, I recognized one school that qualifies as an HSI.

    • 11. gasstationwithoutpumps  |  January 22, 2013 at 10:52 pm

      UCSC is trying hard to get listed as an HSI. I think that the % Hispanics is 1% or 2% low, and that if the current admission levels continue, then UCSC will have HSI status in about 2 years. That will make UCSC the second UC to get HSI status, I believe. UCSC is not going to be a “top CS producer” by raw numbers, but the computer science and computer engineering programs (including the game design and robotics majors) are a big chunk of the Jack Baskin School of Engineering.

    • 12. dennisfrailey  |  January 23, 2013 at 10:11 am

      For what it’s worth, I know a number of minorities who entered the workforce with BS degrees in CS from non-research universities, rose through the ranks to high levels, and ended up on industry advisory committees for CS programs! The path to career success does not require a research university or an advanced degree.

  • […] released a report about where the CS PhDs in the US did their undergraduate work (thanks to Mark Guzdial for pointing me to it), and it is more lopsided than I […]

  • 14. Elizabeth Patitsas  |  January 23, 2013 at 12:51 pm

    It doesn’t surprise me that students from research universities are more likely to get grant money for grad school.

    In Canada, at least, the research universities get the most USRAs (undergraduate summer research awards) to give to their students and are the most forward about promoting these, since there are more research opportunities.

    The result is the undergrads at research universities tend to come out more familiar with applying for research awards (the application process for USRAs is not unlike the fellowships for grad students) and more aware that this is something they should do.

    • 15. Mark Guzdial  |  January 24, 2013 at 1:55 pm

      US grad students are rarely funded by fellowships, Elizabeth. CS PhD students in the US are funded typically by research grants to their advisor(s).


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trackback this post  |  Subscribe to the comments via RSS Feed

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 9,043 other followers


Recent Posts

Blog Stats

  • 2,021,363 hits
January 2013

CS Teaching Tips

%d bloggers like this: