One of the major issues in universities with respect to PhD is the high dropout factor. Demotivated students cause a big loss to the university - a lot is spent on assistantships, but in the end, the students quit without making any meaningful contribution.

What steps do/should universities and faculty members take to reduce alarming drop-out rates? One step is obviously to choose the right person for the job, which we expect the admissions committee to do anyway. PhD involves years of wading through uncertainty and possibly it is in these years that students get distracted and unfocused. How should universities ensure the students remain unwavering in their efforts during these uncertain years?

  • If they did assistantships, then they have contributed something (and probably at quite low wages!), even if they never make a novel contribution with their dissertation. Mar 12, 2022 at 17:00

6 Answers 6


One difficulty is that it's far from clear what the ideal dropout rate should be. Probably not zero, for two reasons:

  1. Even the best students sometimes find that their interests change over time, in ways nobody could have predicted. Dropping out may become a quite sensible choice, in which case it's harmful for the university to discourage it.

  2. It's reasonable for a university to give someone a chance even if it's not certain that they will succeed. The only way to get a really low drop-out rate is to admit just the applicants that are obviously destined for success. The top few schools can get away with this, but if everyone tried it, then many talented candidates would be shut out from graduate school.

So some attrition is OK, and some is bad but may be a necessary consequence of policies that are on the whole good. The question then becomes how you distinguish these kinds from needless and damaging attrition, and then how you minimize that kind.

Advisors play a key role here, because some are much better than others at being supportive or motivational. However, at least in mathematics, most advisors don't supervise very many students, so the numbers often aren't large enough to see patterns clearly.

I'm sure people have studied this problem, and perhaps identified best practices for addressing it. However, in my experience any studies or solutions are not especially influential (at least in the few math departments I've been in). Most discussions are at the stage of trying to figure out whether there is a real problem and if so why, rather than what to do about it.

  • 6
    It's not enough for advisors to be supportive and motivational; they also have to be aware of their student's progress or lack thereof and willing to offer appropriate feedback/guidance/advice. My department's dropout rate (especially the late dropout rate) went down significantly when we started doing annual reviews of all PhD students. Also, I'd be surprised if one couldn't see patterns at the department level, even in fields like math where each advisor has few students.
    – JeffE
    May 3, 2012 at 9:57
  • 1
    @JeffE: Interesting, what do the annual reviews consist of (and how are they carried out)? As for seeing patterns, that's part of what makes this contentious. You can definitely see patterns, sometime even at an individual level, but the sample sizes are small enough that some faculty always argue that it's not statistically significant and therefore not fair to them. The net effect is that it's very difficult to target anything, even advice or guidance, to particular advisors, no matter how justified. These annual reviews sound appealing because they'd treat everyone on an equal footing. May 3, 2012 at 12:16
  • 6
    Every spring, students submit an online self-evaluation form, which includes classes taken, admin hurdles jumped, talks given, papers submitted/published, funding used, etc. since last year; a summary of ongoing work; and a few advising questions. Then advisors submit comments/feedback on the student's self-evaluation. Then (in principle) each research area meets to discuss all its students, and the area chair submits comments/feedback on each student's self-evaluation. Feedback is collected and sent to the students; other comments are saved for reference in future years.
    – JeffE
    May 4, 2012 at 5:04
  • 5
    And yes, this is also implicitly an advisor self-evaluation, with feedback from our peers. Mostly what we want to avoid is the answer "I don't know." (or worse, an unjustified "Just fine!") when asking an advisor "So how's your student X doing"?
    – JeffE
    May 4, 2012 at 5:08

I think this is asking the wrong question. To me, the right question is, what are the main reasons that students drop out, and what steps can be taken to address those reasons? To properly examine this data would need to be gathered, but anecdotally I can think of a few reasons:

  • Student gets bored/loses interest/experiences major life changes (marriage, death in family, midlife crisis, etc.). This will happen in any profession, and are essentially unavoidable. It can be mitigated somewhat by stressing at the outset the time commitment necessary to complete a PhD, but that can only do so much... a lot can happen in four (or five, or six, or...) years.

  • Student has major fallout with advisor. Again, this happens in every profession... people have personality conflicts with their bosses all the time. Communication, and lots of it, is often cited as the best tonic for this problem.

  • Students become disillusioned with academia. This problem often manifests later on in the career, as papers are rejected, grant applications are rejected, and/or the student has difficulty with his or her research. I would venture that this is simultaneously the most treatable problem and the most difficult to treat correctly; the advisor has to work very closely with the student and give them good academic and professional advice. From my experience, the success of the advisor in this task is often the main factor in whether the student remains in academia.

There are likely other reasons students drop out; if I think of more I'll add them, and please list others in the comments.

  • Barbara Lovitts' book Leaving the Ivory Tower presents a lot of interesting research on this topic. Your anacdotal summary does a good job covering several major findings in her research.
    – J.R.
    May 4, 2012 at 22:45

I think that the problem is that often many professors underestimate the PhD student's will and personal scientific aims.

Often it happens that a perspective PhD student apply for PhD admission to work on a project X, and then, once he/she starts the doctorate program, professors say to him or her "Surely you will work on X some day, but now you have to work on Y".

But, effectively, that "some day" never arrives. So the PhD student looses motivations, gets sick, and finally quits...

To answer the question, I would sat that PhD student supervisors should take more care of the PhD student will and expectations, and understand that letting him/her work on the topic he/she prefers will make him/her more motivated and finally get better results than working on someone's else objectives.


Possibly universities should be more careful about hiring PhD students in the first place. It's not enough that they are smart. Students need to be persistent and motivated and their motivations should be realistic. I think it helps a lot of if the students have had some time out of academia so they know early whether this really is what they want to do with their time. I've seen a lot of faculty try to talk smart students into staying when if anything we should be trying to talk them out of staying to make sure they really want to be here. (And "staying" is a bad idea anyway, it's better to change institution so it really does feel like you aren't at school anymore, because a PhD is a lot different from a college degree.)

I think the worst thing is students that hang around forever neither quitting nor working hard. It is the student's life and career and time, and they need to take it in hand and change supervisor if they need to, or change project, or leave. But wasting everyone's time including their own is almost criminal.

  • 1
    "I think the worst thing is students that hang around forever neither quitting nor working hard." – I'm one of those and I'm neither proud nor happy about it. My first project didn't go well (I'm still finishing it) and the increasing frustration end up tanking my motivation and productivity. I started working on something else two months ago and it felt like a breath of fresh air.
    – Daniel
    Oct 4, 2019 at 23:08

Pay, if you paid me what I would make working on an industry project I would have been a lot more compelled to stay.

At the moment, even looking purely at GRA hourly rate it's almost half the hourly rate of a junior engineer. The way current PhD system works is basically projects get subcontracted out to universities who then pay the people involved a fraction of their real worth in hopes of getting a degree afterwards.


As a person who dropped out from a part-time PhD programme in Engineering in the UK. I can add my reasons for dropping out.

  1. Family - being a father and husband changes worldview and sometimes it becomes apparent that there are more effective ways to further one's career than PhD if your aim is not to be an academic. But a university can't do much about this
  2. Lack of proper academic knowledge in the area of PhD project. For example my professional experience and academic background was in electronics, audio and acoustical engineering whereas PhD turned out to be be essentially a theoretical physics. Any physics bachelor would have easier time than me.
  3. Supervisor is not an expert in the area of PhD project. Especially difficult if also the situation in point 2. is present. The lack of appropriate guidance in this case is harmful.
  4. The experimental part is not within control of your department/unit in my case it was not possible to just pop down to the lab to do experiments. It was a cross faculty project and it was difficult to communicate with the other people from outside you faculty
  5. Lack of community support. As a part time student it was difficult to have other PhD students available as you colleagues. Also the project was so removed from anything that the other PhD student were doing so it was quite alienating project.
  6. If your goal is just to further your knowledge it actually may not be the best way to do it. As rigour required to prove your method would steal your learning time.
  7. Topic, it is true that even if you love the topic you will hate, but is also true you can grow into the topic and will start be passionate about it. But if your project has no purpose to you it will be extremely difficult to finish. You need to believe in it. In this case supervisor should at the beginning talk about possible applications or importance of the problem. But there is only so much they can do.

I need to add that the University and my supervisors were really helpful and nice, but the hurdles were just too much for me.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .