Thursday, December 13, 2007

MS v. PhD

My research group has always had a mix of M.S. and Ph.D. students, though typically Ph.D. students greatly outnumber M.S. students. Some of my M.S. students go on to do a Ph.D. (here or elsewhere), but most do not. Some of my former M.S. students teach at two-year (community) colleges or in K-12 schools, some work for government agencies, and some have better jobs in industry than they would have had with just a B.S or similar degree.

These are all good things, and I am glad that I was able to help these former students get a graduate degree that led them to career opportunities they wouldn’t otherwise have had. If I were to be totally selfish and pragmatic, however, I wouldn’t advise M.S. students who were clearly going to be ‘terminal M.S.’ (TMS) students. They are not cost effective. They require a lot of time and some require a lot of resources (depending on their research project). Rarely is that time/money well spent if ‘well spent’ is defined strictly in terms of tangible research products that typically indicate productivity (i.e., papers).

For this reason, some of my colleagues do not advise terminal M.S. students, or strongly prefer not to advise them.

I am not quite so severe in my TMS student avoidance, but I can certainly understand the philosophy. In my experience, it is extremely rare for a TMS student to produce a publishable paper or even to produce a high-quality dataset that can be used as a basis for a paper. In part, this is because I am rather strict about the amount of time a TMS student should take to get their degree. I think two years is enough for a TMS in most cases. Grad school is of course immensely fun, but why spend 3 or more years just to get an M.S.? I could insist that TMS students keep working until they produce a publishable paper, but in most cases, that’s not going to happen in any reasonable amount of time.

In the context of yesterday’s discussion, if a prospective student writes to me and says that they are interested in getting an M.S., I will certainly look at their application, but M.S. applicants are at a disadvantage relative to Ph.D. students in terms of being involved in grant-funded research. Teaching assistantships can fund M.S. students and therefore pay for salary/tuition/benefits, but someone still has to pay for the costs of doing the thesis research, and I am that someone if I am the advisor.

Even so, sometimes it is worth taking on a TMS student. One of my best Ph.D. students was initially an M.S. student who just wasn’t confident about doing research until she realized in the first year that she excelled at it and loved it. That example is definitely on my mind when I review applications for students who say they ‘just’ want an M.S. I think that if someone has even a remote interest in a Ph.D., it is good to express that interest in the application and in communications with a potential advisor.

Note: In some fields, the terminal M.S. degree is very useful for certain jobs, and good research can be done in the time frame of an M.S. thesis program, so my musings should of course not be universally applied to all advisors in all of Science. I am also not considering those Ph.D. students who fail at the Ph.D. and get an M.S. as a consolation prize.

I don’t want to give the impression that I regret the time I have spent advising TMS students. In fact, there have only been a few (in fact, 2.5) who were a total waste of my time and resources because these students were so unpleasant to work with. Any warm and fuzzy feelings I might have about creating employment opportunities for them is obliterated by memories of how awful they were to advise.

As for the others – those who spent a couple of years here, took classes, taught labs, and learned a few things about research – I am by no means disappointed that they ‘just’ got an M.S. It would have been great if they had actually produced a paper from their research, but I know it is unrealistic to expect that from a terminal M.S. student in the time frame of 2 years or so.


Anonymous said...

In the department where I was a grad student, students were only admitted to the Ph.D. program, but every year a handful opted for a terminal MS, for any number of reasons.

My advisor's most-often cited paper--the one which is often cited in introductions to any paper in its subfield, the one which really opened up this subfield, the one which set the direction for most of my advisor's subsequent research and really defined his career--was based on the work of a MS student, who was in fact the first author of this relatively famous paper.

Cherish said...

It's very interesting you mention this because in engineering it's a bit different. Almost everyone goes for an MS because PhDs are fairly useless. But then the job market for engineers is much better outside academia.

However, I can see that this is very frustrating for the profs in my department.

I actually intended to only get a masters in engineering while waiting for my husband to finish his PhD. I was aiming for more experimental background, since I had none, and then planned to go back to physics. I was even going part-time. I'm sure no one thought I was serious about finishing. Now I'm sticking around for my PhD.

Drugmonkey said...

"I am also not considering those Ph.D. students who fail at the Ph.D. and get an M.S. as a consolation prize."

I advise people in my area to forgo MS programs and try to get into the PhD instead. There are several reasons and the usual YMMV caveats.

-the "consolation" Master's is still a Master's. So if one is interested in the credentialing aspect, this is as good as any other way.
-often the "consolation" Master's is easier to come by in terms of years and effort.
-frequently the RA/TA support is better for PhD students where MS students may be actually paying tuition and stacking up more loans.
-in many PhD programs, the pre-existing Master's gets you exactly nothing in terms of speeding / easing the doctoral studies. So if one does come to the conclusion that doctoral studies are in order, one goes right back to the start.

Carrie said...

I was one of those students who came in as a MS student because I wasn't sure about the whole PhD thing, and then realized I loved research in my first year. I'm glad my adviser gave me a chance to figure that out!

Unknown said...

In my department we were somewhat encouraged to do an MS first, because it made preparation for quals much easier. Starting as an MS student didn't seem to have much effect on the time it took to get a Ph.D, as the master's thesis would be done on a subset of the same research that went for the doctorate.

Anonymous said...

It's not entirely clear that Ph.D. students are cost effective either, relative to post-docs. Howard Hughes used to (might still) refuse to pay for graduate students, presumably for this reason.

Female Science Professor said...

That's very true, but at least Ph.D. students have more time, and in theory each thesis 'chapter' is a published/publishable paper.

Ms.PhD said...

Cost effective, huh? That's kind of insulting, even though I can see where you're coming from.

I'm one of those students like thm mentions. My thesis advisor is still capitalizing on my chapters, more than I can, since he's the 'senior' author and I'm just the first author (who had all the ideas and did the vast majority of the work, but nevermind).

In my field (much like drugmonkey's, I think), I have a hard time telling students to go to PhD programs.

There just aren't enough jobs for them afterwards. I also have a hard time telling students to stay when they want to quit. I tell them, quit now, get a TMS, get on with your life. Don't stay here if you're miserable.

And if they're not sure, then I tell them to go work as a technician for a year and get sure, because grad school of any kind is not for the faint of heart.

But, I have a hard time picturing how I would handle this as a PI. I'd like to think there are plenty of good students and postdocs to go around, and I'd rather have the happy ones who know what they want to do and why.

I think it's really sad if students don't find out whether they like research or not until they're already in grad school. It's bad for them, and it's bad for the PI to have to give them the hard sell ... or risk wasting money by taking them and having them make no progress.

Gary Carson said...

Do you have the same self absorbed attitude about teaching undergraduates?

Female Science Professor said...

Do you always have so much trouble understanding the use of the past subjunctive tense?

Anonymous said...

Greetings from a consolation-prize MS-recipient! I'm curious about how you ended up with .5 of a waste-of-time TMS student. Were they a co-advisee, or did they drop out halfway through?

mentaer said...

3 questions (being from europe):
1) Are you allowed to make a Ph.D. without an MSc?
2) Is there a constraint on the number of years?
3) So why should one do a MSc if a BSc is sufficient in general for jobs & research?

EliRabett said...

The key thing is that I feel a lot less obligation to provide a stipend for an MS student (none if they are in engineering) than a PhD student

Anonymous said...

I was going to write a comment but it got out of control, so I decided not take up too much space here.
Here it is:

Anonymous said...

A wonderfully self-centered and capitalistic perspective on a student's degree! I don't blame you. Most graduate student advisors I have seen are like you.

(f.y.i I published two papers just with my masters.)

Anonymous said...

This is actually rather disturbing for me to hear - I'm looking to complete my education as an older student (25 is old enough to have people question your capability, apparently). To hear that my credibility and attractiveness are going to be fighting a two-front war because my academic goals are concise is very discouraging.

Anonymous said...

I think its great that you are honest. Frankly, in the experimental sciences, I would tend to agree with you--its hard to get anything out of MS students because they are on campus for such a short time.

However, in more maths heavy engineering, papers can be published in a very short time (6 mo. and less) because the experimental load is often less. If we use Parkinson's law (work expands to fill the time allotted), often MS students in engineering are a great value. They know they'll be here for a year or two so they'll compress their project, stay up late, come in on weekends, etc. The PhD student may be here for 5 years and will focus on classes, making social connections, throwing down roots, etc.