The discussion on this infogram made me wonder about the number of PhD students that a full research professor successfully graduates in their entire career. By professor, I mean a full professor, not an associate or exclusively teaching professor or other positions referred to as professor depending on field and location. Of course, the answer is not a single number, but rather a probability density function that is a function of field, place, time, university, and probably other factors. To narrow the scope, I formulate the question as:

For selected fields and countries, what are recent figures on the mean and standard deviation (alternatively median and median absolute deviation, in case the distribution is non-Gaussian) for the number of PhD students successfully graduated per professor throughout their entire career?

  • 8
    In the US, I would bet heavily that the median is 0. There are lots of institutions with no PhD programs at all. Feb 26 '14 at 19:52
  • 3
    @trutheality such distinctions are not so easily made. In many institutions I have seen in the US a research professor is someone who does not teach anything. There are a very small number of those positions in the US. On the other hand what you call teaching professors still do conduct research. Gerrit, how would you count a student advised by a someone while they were an associate professor? Feb 26 '14 at 20:54
  • 3
    Why exclude associate and assistant professors ? at least in the US, these are "real professors" too :)
    – Suresh
    Feb 26 '14 at 21:52
  • 6
    @trutheality:The US makes no such distinction as far as titles: a senior faculty member at Harvard and one at Harvey Mudd both get the simple title "Professor". For the US, a better measure would probably be for the denominator to include all tenured faculty at PhD-granting institutions, and for the numerator to count all students graduates over the professor's career (including those advised as an assistant or associate professor). Feb 26 '14 at 21:54
  • 4
    @gerrit I graduated my first PhD student when I was still an assistant professor. Tenure status is simply irrelevant for PhD supervision in the US. (So getting stats about PhDs only from full profs in the US is going to be impossible.)
    – JeffE
    Feb 27 '14 at 9:17

It depends on the size and staffing needs of their lab. For example, theoretical computer science and mathematics professors may need no lab support at all. Thus, they are under no pressure to take grad students or post-docs and can choose just the ones that they want.

However, if you are doing work on stem cells, you may need a great deal of lab support. You would want a team of doctoral students and a couple of post-docs at any one time. In order to maintain continuity, you would want to accept at least one doctoral student each year. So if you had a 20 year career, you would have at least 20 students (or 20 - 7 = 13 given that it takes students 7 years to graduate and you don't want to leave students hanging at the end).

You'll need to narrow down what you mean by a "STEM" field in order to get a more precise answer.


My guess is that it probably varies hugely, by field, by department and then again by professor.

The fields might vary because of the different expectations about having co-advisors, the size of dissertation committees and so on.

Departments might vary based on their teaching needs. A department (Department A) that the university needs to cover lots of intro classes might be funded primarily by teaching, and so such a department is going to have a lot of graduate students. On the other hand, Department B that brings in tons of grant money might have more labs, but less teaching responsibilities, and therefore have a higher proportion of post-docs and lab assistants than grad students. Hence, profs at Dept B might have fewer students than profs at dept A, but that won't speak to the relative quality of the faculty at either institution obviously.

Finally, it might also vary just from faculty member to faculty member. Some people are jerks and nobody will want to work with them.

  • 3
    The question is very specifically asking for average numbers and statistics from different fields. I don't think this answer addresses the question.
    – ff524
    Mar 30 '14 at 7:06
  • I agree with @ff524; while I'm not disputing the factual accuracy of anything in this answer, I don't think it provides any useful information that goes toward answering the question.
    – David Z
    Apr 29 '14 at 1:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.