Wednesday, September 02, 2009

Moving Grads

Sometimes grad students change advisers. In some cases this is a brilliant move, saving the student's sanity and academic career, and in some cases this is one more step on the road to flaming out of graduate school(s).

I considered moving to another university partway through my graduate school days. My adviser was OK, but another professor who was on my committee was tormenting me, and I saw no reason why I should continue to put up with that for many more years. There were other reasons as well, all of which added up to motivation to move somewhere else, even if it meant losing time and effort. Then the evil professor died (he was ancient and had been quite ill) and I decided to stay where I was.

As a professor and adviser, I have taken on grad students who have had unsuccessful experiences with other advisers, and I have had a few of my own students switch to another adviser. In most of the cases involving intra-departmental moves, these adviser changes have not had good results for the students or for the advisers. In fact, the students who have switched from working with me to working with another adviser in the same department have ended up failing/quitting, and the one student I adopted from another adviser (at my previous university) also flamed out, after moving on to a third adviser. When I wrote about this topic before, however, commenters told stories of some successful switches to more compatible advisers.

The cases that tend not to work out are when students are looking for an easier adviser (i.e., one with lower expectations) or the students are not feeling motivated by a particular research topic and hope that by changing advisers they will find the motivation they are seeking.

More successful switches involve talented and motivated students who are seeking a better fit for their interests and abilities or who are leaving an unacceptable situation for an adviser who may be more sane.

Every once in a while I see a graduate application involving a student who wants to leave their current university and who indicate, either directly or by omission, that they do no want their adviser to be asked for a letter of recommendation. What to do? Call the adviser anyway (especially if it is someone you know) but keep in mind that you may not be getting a complete story, or just go by the student's record in the application?

In one case in which I was involved, I did not ask the adviser for an opinion but another professor at the other university made a very strong case for the student. This other professor was concerned that the student's application would be undermined by the lack of a letter from the adviser and wanted to make it clear that the student wasn't the problem, the adviser was the problem. The student was accepted, came to work in my department as one of my advisees, and did very well. He lost some time but said it was worth it.

I'm not sure what would have happened without that information from the other professor, but it certainly helped to get a strong positive recommendation. I think it is important that faculty do such things for students if they feel that the students are talented and would do well if given another chance with another adviser. It would also be a good idea for students to seek out such sympathetic faculty and request that the issue of student-adviser incompatibility by addressed proactively and directly in a reference letter or phone call.

Switching advisers is a difficult thing to do. Sometimes it is the best thing to do, but it should only be done if absolutely necessary, given the costs in terms of time, effort, and stress.

23 comments:

John Vidale said...

WRT asking a moving student's advisor for his/her opinion.

Unless I am confident in KNOWING the situation, I'd always ask for his/her opinion. The downside of taking on a serious problem is too great to leave that stone unturned. One might ask orally so that unfair and unfavorable opinions do not weigh in dept scholarship decisions, as they would with letters, and oral conversations can elicit freer opinions.

Of course, the former advisor's opinion need not be taken to be necessarily accurate.

Notorious Ph.D. said...

I contemplated switching advisers halfway through my M.A. -- I had one who was a bad fit for me (and I for him) for various reasons. Part of this might have been due to the fact that I was only 25 at the time, and had never encountered someone who thought I was so dumb and unpromising. But in the end, I stuck it out through another six months of agony, rather than disrupting my relationships and perhaps gaining a reputation as someone who was difficult to work with.

I did, however, choose a different adviser in that same department for my Ph.D., and ended up being very happy with the way that turned out.

Now that I have my own grad students, I give them a mini-speech to the effect that this is an elective relationship, and that if they ever feel like they need to switch advisers, they should do so without fear of retaliation from me. I've had one take me up on that -- for the wrong reasons, I think -- and I've respected that choice.

Anonymous said...

So no matter how good you are FSP, you still get BSs and students want to leave you. (why did you accept them in the first place?).
Are you bitter, do you write them a letter of recommendation regardless. They are just kids...

If there are BSs: Bad Students. How do you deal with BSs. Do BAs tend to produce more BSs.

And do GAs (good advisors) produce GSs (good students).

Finally, how does a BS become a GS if he/she is under a GA?

Anonymous said...

Based on my personal experience with being a grad students second advisor: always get the most detailed information possible from the first advisor, then filter this information through the possibility that that person may be sugarcoating things to unload an unwelcome student.

Sometimes the problem might be the advisor, but at least as often, the problems are with the student.

Anonymous said...

Sometimes, people change institutions for reasons that are completely unrelated to their advising situation or to their research.

Several grad students have left my (European) department during the last several years or so not because of advising problems but because they felt the department was treating the grad students unfairly in terms of pay and benefits.

I also know a student who came to our department after suffering harassment, sexual and otherwise, that the powers-that-be at her previous department refused to address.

As far as I know, all of these students eventually finished successfully.

Kevin said...

IF a student had not gotten a letter from his/her former adviser, I would probably ask informally, but if the student requested that we not ask the adviser, I would not do so, but I would try to find out more from other faculty in the department.

I have sometimes taken on students who were unable to find other advisers for personal reasons, and the results were mixed---some blossomed and some gave up. I've also "inherited" students from faculty who left for industry---they were somewhat less brilliant than students I would have chosen myself, but they finished perfectly fine theses, with less trouble than the brilliant but unstable students I usually end up choosing.

Anonymous said...

I am one the decent students (at least in my opinion :) that switched schools for a more sane adviser. It has set me back about a year, but has been one of the best decisions I have ever made in my life.

In this case, I also know that my current adviser also called another prof at my old university for the scoop.

mixlamalice said...

To anonymous 1:59.

I think a good advisor is someone who can deal with bad students at least to a certain point. I mean a Professor that is able to understand that not every student is a true genius which almost does not require any work, and that making (or at least trying) an average student better is also part of the job.

I guess it is less the case in the US because groups are usually quite big so Professors tend to be confronted to average students quite often, but in France a Professor usually has only a couple of students at the same time. I know a Professor that was really nice with one student that was very very good and pretty rude to all the other ones he had since: well, you can't expect to have someone as good as you (or as good as you think you are) every time. I guess it is the same in the US when big professors have 50 grad students and only focus to the two best ones...

Being able to see the flaws and the qualities and try to improve both (or point them to the student) is what makes a good advisor.
Really clueless students are, I believe, pretty unusual, so if you try hard (and well, I am not sure all Professors are willing to) you can almost always find something worth it and help a student become a better scientist.

NJA said...

My (UK) department has a bureaucratic arrangement to avoid this very problem: whether a student or advisor is mad, bad or sad, the other person has to have a clean way out of the arrangement.

All graduate students have (in addition to 1-2 research advisors) a separate academic advisor who's meant to act as a neutral go-between. Preferably, this academic advisor is in a different research area, shouldn't be a collaborator of the advisor(s), and have no vested interest in the research. The general idea is for the academic advisor to have a feel for how the PhD research is progressing and whether any problems have arisen.

I have this role for a couple of students - it involves one formal meeting a year plus some informal chats. If the circumstances ever arose, I'd be that extra faculty member who can write a letter of recommendation for a student if s/he doesn't want to ask the research advisor. Or I'd be that person who could verify that a failing student is a Bad Student and not the victim of a Bad Advisor.

It might be paranoid to have this arrangement officially in place, but it's not a bad idea.

Anonymous said...

FSP - I am struck that the majority of cases of switching you've seen have ended poorly. This is definitely thought provoking.

In my own experience, I waited until after passing my PhD qualifiers to switch to an advisor who was more in line with my research interests. I added ~6 months to my thesis work, but the change was very rewarding and resulted in a very positive experience for myself and (i hope) my advisor. My pre-qual advisor remained on my committee and engaged in my work, but it was awkward at first (for me, but probably not for him?).

I think your point about why the student is switching advisors is key in determining whether the switch will be successful. I will certainly keep this in mind for future advising...

Anonymous said...

Is it possible that some students have unrealistic expectations of their advisers? I'm not saying there are bad adviser but if your adviser isn't an awful person but you're still not getting what you want from them. . . why not talk to them?

It's your PhD, try and get the most out of it.

Anonymous said...

I'm struck that there is an assumption that the switch must be due to a bad student or a bad advisor. Also, I disagree with the final statement that switching should only be done if necessary.

I think that there are many graduate students who switch to follow their personal research interests, as was the case quite often among my fellow grad students (in a tier one science program). In fact, we were all encouraged to complete a secondary project (pre-qualifiers) outside of our main focus area so that we would have a well-rounded research experience going into the post-qual stage.

It is not always the case that a student was bad or the advisor was bad. The culture of the PhD program/department should also be taken into consideration.

Ioana said...

Assume that it's true, as FSP suggests it, that most students who switch advisers end up not getting a PhD. That doesn't prove that switching has a causal impact on PhD failure. Instead, it's quite likely that those students who switch have, on average, certain characteristics that lead to PhD failure. For example, they have an art for picking the wrong person to work with: they do it for their first adviser, and for their second too, which ends in failure. Or else they are people who are particularly likely to be dissatisfied and unable to adapt: they couldn't adapt to their first adviser, and their can't adapt to their second either. That's all to say that switching for the right reasons by the right person may lead to positive outcomes, even if on average switching is correlated with PhD failure (because most people who switch would have failed anyway).

Female Science Professor said...

I did not suggest anything of the sort.

FemgineerPhD said...

I have seen several examples of students switching after feeling miserable in the lab. Most of these people had advisors or labmates who set-up a very non-collaborative, competitive culture. In these cases the student was more successful after moving. I have also seen that some students who have a lot of promise work even HARDER when they get to their new lab because they really want to prove that they're not a failure. So win win!

I think someone should keep track of the number of switches into and out of labs. Some professors are notorious for having students who want to leave, and departments just continue to let it happen instead of at least attempting to counsel the professor on the root cause (eg, mentoring skills, clearer project definitions, etc.). This is bad for both the professor and the students since everyone ends up wasting time, funds, energy, etc. On the other hand, it would be very interesting to analyze any labs that have a high rate of student transfer IN. Are they just "easy" or is there something the rest of us can learn?

Anonymous said...

I agree that someone should track advisers who have a lot of students leaving the group or quitting altogether (or if the students take a long time to graduate) as this is an indication of something wrong with the adviser. If the department of my last adviser did that, maybe they would confront his bad advising. Whether he can actually become a good adviser is another issue, but at least someone could find out who the bad advisers are.

Actually I think there should be evaluations for advisers that are anonymous and that all graduate students who are looking for an adviser can have access to.

Some of the students who switch from the really bad advisers may have trouble in their new lab not because of their inherent promise (or lack thereof) as a student/researcher, but because of the trauma from their previous experience. It took me awhile before I could get productive again. Plus, since I switched advisers and have thus had a long and utterly mentally exhausting graduate school experience, I am feeling very burned out right now and having trouble staying focused through the end. Whether I would have been like this anyway is unknown (until someone invents a time machine), but I feel like if I had skipped the bad adviser, I would have a lot more momentum right now and a much better attitude about academia. It is very hard for me not to be mad at the whole system for letting people like that continue to do this to future students, especially since the ones he is currently tormenting are my old labmate buddies who are also being kept there much longer than they ought to be, despite many papers. Ironic that, not only does he make the entire process miserable but he drags it out so that most of his students take 8 years or more to graduate.

I know that the system doesn't care because I have some friends (3, all women) who filed an official complaint with the University about their previous adviser (different prof) and nothing really happened to him. He wasn't allowed to take new grad students for a year or so, but that's not much of a consequence. I'm sure he hasn't changed his ways.

Some adviser/student problems are just personality clashings, but some profs tend to clash with more students than others and some tend to clash with all students to some degree. Too bad they don't hand you a "Guide to avoiding advisers who will make you miserable and take away your love of science" manual when you come to graduate school.

John Vidale said...

Shifting to work with another advisor here is mainly presented as breaking with the old and then joining up with the new.

My route involved starting primarily with one simple project with Prof A and secondarily a canonical one with Prof B, then promoting a minor theoretical project with Prof C to primary for a while, then shifting to applying the methods developed with Prof C to attack a set of observational problems with Prof D.

Prof D wound up as my formal advisor, and Profs A, B and C remained on my committee. Technically, I switched advisors at least twice, but there was no retracing steps and no discernible friction. Perhaps I could have escaped grad school a year faster with a sharper focus, but the wider breadth from this path has served me well, and I was able to skip the post-doc stage.

This type of advisor switch has been missing from the discussion, but perhaps is the simplest way to do it, if one can incite some joint projects.

Female Science Professor said...

If you polled my current and former students about what kind of adviser I am, you would get the complete range of answers, from terrible/evil to nice/helpful. This is normal (I tell myself) as long as the ones who think I am evil are very rare.

Anonymous said...

so even good advisors make mistakes?

In terms of personality match, does gender make any difference?

How about Advisors who are Republicans and students who are Democrats and vice versa? (ex. lab work is so boring and there is nothing to talk about except news and politics).

How about students who can't work on a particular day due to religious reasons.

Anonymous said...

This is all fascinating, though a bit depressing to see how many people have experienced bad advisors. I've been very lucky in that all my advisors have been good (in different ways), though only one is still an important mentor.

I want to add one more datapoint in terms of changing advisors. Someone who is a great scientist and a great mentor (lets call her Prof A) had a grad student who just wasn't really able to work as independently as Prof A expected. So, by mutual consent, the student transferred to Prof B who was able to provide more structure (ie. tell the student exactly what expt to do and how). Now the student has successfully graduated with a reasonable PhD. This example shows that even great mentors can have students who don't work out, and that transfers can be successful.

And finally, on the point of penalties for bad advisors - here in the UK, advisors have to have monthly meetings with students and to keep written documentation summarising what was discussed, the students progress etc. If the student makes a complaint, the advisor has to be able to justify everything they have done. There are also BIG penalties if your students fail to complete their PhDs on time (ie. drop out or take more than 4 years). This doesn't eliminate bad advisors but does mean a student should be able to progress in research and finish on time regardless.

Devin said...

This is really great insight for grad students and professors alike. There are some similar discussions taking place over at GradShare. Here is an example: http://www.gradshare.com/question.html?id=145. Check it out, you may have some valuable advice to contribute.

hkukbilingualidiot said...

I have posted before about my 'trauma' when I was doing my placement in Japan, which I won't go into too much detail but of which involved sexual harrassment from a team leader. That experience seriously affected the first 5 month of my 8 month's grad year masters (it was an undergraduate masters degree in the UK). I've had to have serious counselling as the environment, though different, was similar due to the same field. As a result of my experiences what I had learnt from previous years had to be re-learnt as I had naturally removed them from memory due to the close emotional associations. All the while trying to complete my thesis within that 8 months which also affected my autumn exam results with it being the worst I've ever had and the summer ones only slightly better. Though it was a normal change a bad advisor CAN affect the future of a student. I almost lost my future in science as a direct result, but thankfully, my grad year supervisor was very good and nurturing to me so I've recovered soon enough to stay where I want to be. Though not everyone is that lucky. From my personal experience the second/third advisor usually have to do a bit more work as the student usually need a little more support to overcome whatever went wrong in the first place. Sometimes they drop-out not because they can't deal with the problems but rather they don't want it enough to pull through which requires a LOT of effort and more importantly, support, whether it'd be practical or emotional.

Anonymous said...

I like the idea of having a separate "career" adviser other than your research adviser. The problem, as I've seen it, with switching labs because of a bad PI is that it's unprofessional to badmouth the PI, even if it's the reason you're leaving. I had that problem when looking for my second Postdoc. I was VERY fortunate in that two other faculty members knew about my lab problems and both had counseled me. When I decided to apply elsewhere, they provided references. Even though I hadn't asked, at least one told my current (ie, new) PI about the problems I'd been having. Since it wasn't ME telling, I felt like it was a huge burden off my shoulders that it was told. However, I never would have asked that of him.

Rock and hard place. If you had asked my former PI, he would have denied denied denied. As it was, he didn't know I was even looking for a job, which is why I asked that he not be contacted.