Female Science Professor (FSP) discusses the relative merits and difficulties of supervising PhD and Master’s (MS) science graduate students. I was going to write a comment about her posting on her blog, but it is sufficiently interesting a topic that I would like to throw in my two cents worth.
FSP specifically discusses the “cost effectiveness” of MS versus PhD students, from the standpoint of the individual faculty advisor. MS students spend roughly half of the time in graduate school, compared to their PhD counterpart; that time falls almost entirely in the steeper portion of the learning curve, when a student is less effective a research contributor. Financial support for MS students typically comes from the research grant of their advising faculty for at least half of the time; in principle, MS students require close to the same amount of training as PhD students, as the latter are expected to become fairly independent during their last two years; and, by the time MS candidates begin to be productive in research, they graduate and leave.
So, given the time and effort that a faculty has to put in, MS students seem “less bang for the buck” than their PhD colleagues. Although this belief is fairly widespread among science faculty, my own 10-year experience as physics professor has brought me to regard it as deeply flawed; the sooner we abandon it, the better.
Doubtless, considerations of student “cost effectiveness”  do influence, albeit to different degrees, our decision to take on or not a student under our supervision, as well as underlie the strong preference of many departments for graduate applicants who express right away their interest for a PhD, over those who wish to start out with the MS, possibly as a terminal degree. A graduate applicant who is interested in pursuing a MS rather than a PhD is often branded as “scarcely motivated” by admission committees, but I think it is mostly their predicted lower research productivity that makes them less appealing.
Now, don’t get me wrong: there is no question that PhD students are one the most compelling and enjoyable components of this profession. Indeed, it was mostly the lack of a PhD program in my discipline that eventually prompted my decision to leave my former institution. Any potentially significant and challenging research project requires time and motivation, and PhD students have more of each than MS students, almost by definition. Furthermore, not only faculty but also undergraduate and MS students themselves benefit from the interaction with PhD students, whose presence makes for more vibrant, intellectually stimulating and overall more fun departments.
I personally see nonetheless many good reasons to admit a much greater number of MS candidates in a science graduate program than it is currently commonly done; to me, a roughly 50/50 mix of PhD and MS candidate would be optimal. However, I wish to focus here on the very notion of “cost effectiveness” of a graduate student, because I think that its assessment, and the ensuing comparison of MS and PhD students based on it, are far more complex an exercise than suggested by FSP’s remarks.
I wish to offer the following comments:
0) PhD students, even advanced ones, are seldom as independent as the advisor may wish. The amount of time required to supervise their work, teach them how to give talks, edit their papers and dissertation (especially with foreign graduate students whose mother tongue is not English) etc., remains significant until the end.
Moreover, the commitment toward PhD students extends way beyond their graduation date. I do not know whether a faculty or a department should feel morally responsible for the success of their PhD students, but it is a fact that most do; departments advertise themselves to potential students largely based on the success of their graduates, and no faculty likes to be known as the one whose PhD students do not succeed later on. Because very often PhD candidates have academic ambitions of their own, typically an advisor will have to write letters on their behalf, make phone calls, hobnob, try to have them invited to give presentations at high profile events, push for their hire as postdoctoral associates and assistant professors, forcefully advocate their tenure and promotion, and so on, many years after they have graduated. MS students do not normally require the same type of commitment, for a number of reasons.
1) If we are going to judge cost effectiveness by sheer research output per unit time and effort (i.e., leave the all-important educational component temporarily aside, for the sake of argument), for many of us (especially on the theoretical side) periods of highest productivity occur when there are no students to supervise, either MS or PhD. Actually, places like national laboratories, where student supervision is absent, are often deemed superior to academia, in terms of maximizing raw research output.
While research projects assigned to MS candidates are typically relatively focused, low-risk and self-contained, we tend to give our PhD advisees projects with the highest potential payoff; these are big, challenging, risky projects, in the hottest research areas . In many, most cases, we would be better served research-wise pursuing those projects on our own, possibly in collaboration with other faculty, or postdoctoral associates, but without student involvement. For every PhD student who contributes with an original, brilliant idea to the successful completion of an outstanding research project, thereby gaining the spotlight in his/her research community, there is at least another PhD student in whose hands an interesting idea languishes, as (s)he fails to take the project at heart, spends a few years doing little more than routine work (executing almost blindly specific instructions from the advisor), taking far more time than it would have taken the advisor to complete the research work on his/her own.
2) An excessive focus on research productivity, at the expense of student mentoring, is not something that universities will normally reward. Academic faculty are expected to engage in individual student supervision . In practical terms, this means that any performance review will largely focus on the person’s success at doing that. Now, “success” is not merely measured in terms of number of papers. At many universities (including research ones), a tenure-track assistant professor with an excellent publication record, going up for tenure not having advised a single student, will almost certainly receive a weaker recommendation from his department or college than one with fewer publications, but with a more convincing track record of advising students. MS students are not necessarily regarded as any less valuable than PhD ones by the administration, and because they experience a lower dropout rate (see below), it may well be advisable for a tenure-rack faculty to supervise at least as many MS as PhD students, even from a merely utilitarian standpoint.
3) If one is really going to measure cost effectiveness, the relatively high dropout rate among PhD students, as well as the much greater likelihood of a late completion of their degree, should be factored in as well (will any graduate chair please raise his/her hand, whose department does not enlist at least a few PhD students who cannot get their act together and just finish the damn degree ?). Because a strong publication record is a must for a PhD graduate, incentives to postpone graduation if that may mean another paper or two, are much stronger than for MS students (in many of these cases, the extra year spent in graduate school results in nothing worthwhile).
And there are also many reasons why initially highly motivated PhD candidates, at some point decide to forgo the completion of their degree. Burnout, discouragement, fatigue, contentions with advisors, loss of interest (and/or concurrent development of different interests), life circumstances, realization that the career longed for a few years back is really not all that it is cracked up to be, financial needs, are the ones that come to mind. When that happens, not only can it stain a faculty’s career record (especially repeated occurrences), but, more importantly, the sense of waste and disappointment can be enormous, regardless of how many research papers the student has managed to produce during the time spent in the program. MS students are less likely to drop out, and when they do the emotional and time investment has not nearly been the same.
4) FSP argues that it is unrealistic to set the completion of a publishable piece of work as the target for a MS degree, without unduly extending its duration. I am not convinced; it does not seem unreasonable at all to set such a target.
Again, the underlying assumption seems to be that the MS candidate is “scarcely motivated”. Be that as it may, one does not see why academic standards should be set lower for MS than for PhD students. After all, most PhD candidates seldom publish their first paper before working at least one year under the supervision of their faculty advisors, whose involvement in both the work and its write-up is typically heavy. Assumptions such as “a PhD graduate must have co-authored several publications, whereas none is expected of a MS one” seem baseless, and contribute to perpetuating the “MS as a consolation prize” stereotype, which is factually wrong for so many reasons.
Finally, one could also reasonably argue, that even for PhD candidates setting the completion of the MS as an intermediate goal may be a very wise decision, an excellent way of giving both advisor and student an opportunity of re-evaluating their commitments, while giving both something to show for the time spent working together.
A more general remark, to conclude: at a time when many a science graduate program suffer from dwindling enrollment, insisting with snubbing self-declared MS applicants in favor of PhD may do more harm than good. Instead of regarding MS applicants as “less motivated”, why not take a more positive view, and appreciate the maturity, realism and conscientiousness of someone unwilling to make a 5+ year commitment upfront ? Why not give MS applicants who are academically strong, a chance to develop interest and enamoring themselves with the field and the project, in due process, eventually deciding based on experience that the PhD is what they really want, after all ?
 There is something inherently odd about talking cost effectiveness in relation to students of any level; it seems only slightly less bizarre than, oh, I don’t know, hospitals lamenting the cost ineffectiveness of short-term patients.
 Contrary to what many students believe, this is not done in hope that students will help us faculty advance our own research agenda. Rather, it is done for the student’s benefit; this day and age, a PhD graduate whose research portfolio consists of a number of self-contained, relatively simple, possibly disconnected pieces of research work, each one of somewhat limited scope, is not likely to elicit much interest on the part of a potential employer (especially an academic one).
 Anyone not enjoying this part of the profession is simply in the wrong line of work, if you ask me; sure, it takes time and effort, but the reward in human terms is priceless. Naturally, some students are easier to work with than others; so are some professors.