Friday, July 03, 2009

Loose Ends

It is in the best interests of advisers and graduate students that students graduate in a timely way and move on to another scientific (or other) adventure, but in some cases a 'timely way' means that a student leaves before completing all the things that need completing (i.e., papers). What to do?

It depends a lot on the (ex)student whether the post-graduation tying-up-of-loose-ends occurs efficiently, if at all. A common complaint of advisers is how difficult it is to get thesis-related papers completed and submitted after a student has moved on to another place or position. Here are the scenarios I have experienced:

- A student graduates with one or more unfinished papers but finishes them in a timely way and submits them. These students are exceptional and are therefore (by definition) the exception, alas.

- A student graduates with a few -- or maybe more than a few -- unfinished papers and needs to finish them (for their own career) but never seems to find the time. As adviser, I might be able to finish the papers if provided with the necessary material, but this only works if I have been closely involved in all stages of the research and if the ex-student is willing to relinquish some control over the paper(s). Lingering unfinished papers are a major problem for projects funded by an (ex)adviser's grant and cause a lot of stress for all concerned. In these cases, I need to get the papers out because I am ultimately responsible for the results of the research (grant), and my former students may also need the papers for their own career advancement, but my former students may have no time for 'old' projects now that they are working on new things.

A student graduates, leaves academia, and has no interest in writing papers. This is more clean-cut than the previous example, as it is clear that the ex-student will not be writing the paper(s), but this situation has its own challenges. For example, I am still limping along trying to finish some papers from grad students who got jobs, graduated, and left me to finish their paper(s). It is really hard to write a paper entirely for someone else, no matter how involved you were in the research. These were mostly MS students. I didn't want to delay their graduation and therefore their employment options, but there was no way the paper(s) were going to be completed before the student graduated. In some cases I have given up entirely and just let a project drop, unpublished, but in other cases I don't want to (or can't) do this.

If I refused to let my students graduate until all the essential papers were submitted, it would add years to the graduate program for some of them and they would run out of funding. I don't think that is a good option for anyone, no matter how many problems it causes for me, as adviser, as I tie up loose ends and try to get thesis-related papers finished and submitted.

If a student is out of funding and/or has a post-graduation job offer, they should leave as soon as they have completed a thesis that is deemed acceptable for the degree. Ideally, the thesis will consist of published and/or submitted manuscripts, but it is rare for the entire thesis to have reached this stage by the time a student departs.

This post probably sounds like a complaint -- and it sort of is -- but I also accept that this is just the way things are going to be with some (many) students and I need to find the best way to deal with it. I will still continue to encourage my students to graduate in a timely way, but I also hope I can find better ways to get manuscripts completed and submitted even when the students don't or can't help much with this.


Kevin said...

The "best" solution for papers by students who have left varies. Here are some ideas:

1) if you have a big group, set up multi-student collaborations. If you can get several students working on a project, then it is easier to finish the papers, even if one student leaves. It helps if at least one person on each project is capable of writing quickly.

2) hire an undergraduate "reporter" for each grad student. The reporter's job is to learn everything about the project and produce written reports that can be turned into journal submissions. This is an excellent introduction to research for the undergrad, and helps get things documented. (I know of one case where this worked very well---a brilliant grad student who generated tons of great ideas was unable to get anything written. An undergraduate co-author helped him to get out a decent, published paper---the grad student never managed to write his thesis, even though he had at least 3 completely separate research projects with enough novelty and results.

3) Require detailed quarterly (or semester) reports from all grad students. Grade the students based on the reports. Explain to them that the reports should be more complete than a journal paper, so that if necessary they can be edited down to a journal submission. This doesn't entirely solve the problem, but I've found it very useful for my students---those that have been really conscientious about it have found it a great help in getting their thesis written also.

A Life Long Scholar said...

As one who has just submitted her thesis and is in the process of moving to a new country for a post-doc position, this post surprised me to read--it never occurred to me that I could expect much help from my (soon to be former) advisor in writing up the results of my work for publication (other than sending him drafts for comment, of course).

Granted, it has been nearly two weeks since I turned in the thesis to the binders and I've not yet made much progress towards a draft of a paper written, but my advisor warned me that he didn't expect me to accomplish much while visiting my family I'd not seen in four years...

Tara said...

A common enough problem in my university is students accepting a job offer and never even submitting a masters thesis. As a masters student myself I would consider this a disrespectful act, especially when the supervisor left empty-handed chose to allocate their funding to you. The unfinished-papers usually fall into the hands of the supervisor, other research group members or even undergraduate summer students. Then of course the question arises about what order the names go on the paper!

Phys Student said...

How long after graduation is it still acceptable for the student to finish those papers? 3 months? 1 year? 3 years?

I was exactly in that position when I finished an MS in Engineering and then decided to switch to Physics for the PhD. I had 3 months in between on and the other and wrote the drafts (2). It has been already > 4 years and they haven't even been submitted. The students from his group that knew me, when they graduated (after me) were asked the same question and they just didn't do write them anymore.

I am in a completely different area now so I don't care about those (could've been) papers. The part that bothered me was that he made me lose my time in something he probably knew he wasn't going to care about.

Also, do you see this happening with papers that could be submitted to good journals just as with those that would go to not so good ones? Or is it only the weakest papers that don't get written? That might explain why some students don't want to spend their times on that.

Alyssa said...

It does seem to be a widely held problem. It's interesting though, because it is different from the student perspective (at least in the situations I've seen):

I've heard of many (ex)-students wanting to publish their thesis work, but their previous supervisor is so busy with current work that they let the papers slip through the cracks.

For example, I still have a draft of a paper from my masters (4 years ago!) that's ready for comments. In fact, I gave my MSc supervisor a number of drafts - the last one about two years ago - and it is still in the writing stages. I'm actually taking a trip back to MSc city after I finish my PhD so we can crank it out and finally submit it!

My husband also had this problem - he had two papers to submit after his PhD, but his supervisor never replied to his drafts. Eventually, a year after, DH just told the guy he had a week to look it over and he was going to submit no matter what.

I think it's an "out of sight, out of mind" thing, especially when supervisors have so many current students/projects to worry about.

So, kudos to you for caring enough about your students to see their publications though!!

Anonymous said...

I had last summer to submit papers before taking my post-doc. I defended in March and left in July. I had three manuscripts on old PIs desk and one I am holding for ransom for some of his other works.

It has taken old PI now over a year to read them. One was submitted and accepted with minor revsions. The second took six months to read. He submitted it after he read it-we're waiting to hear on that.

It's now been a month. Has he read the third manuscript? No. Now, I'm writing a fellowship application.

So, sometimes, it's not the graduate student. I wrote the thesis and he did his job (but did I mention it took him three months to read chapter 1?)The manuscripts were modified chapters out of the thesis.

Not always the graduate student..

Anonymous said...

I think you're right if they are out of funding AND have a job offer... but what about the situation where they have funding/offer for a postdoc who's acceptance can be delayed a year, but they also have graduate funding to finish their thesis (TA or RA)? This seems to be relatively common among many grad students in my department.

In that case, I'd think it would be in the student's interest to spend 6 mo to 1 year publishing those papers without the distraction of new research / new adviser's demands... I think it would ultimately improve their career prospects to delay graduation a little bit. Then, they'll also be able to really focus on their new project(s) once they do move.

John Vidale said...

A couple of disjoint thoughts:

This is a reason to prefer short papers to long papers. If the student can get out the first short paper, the second short paper or the longer, more comprehensive paper has less necessity.

Ideally, one doesn't start the students on a paper whose completion will be problematic - hard to predict, I know, but people often do that to concoct enough results to fill progress reports rather than that the results are valuable. Filler papers easily run out of steam before completion - if the research project didn't find anything, don't write about the "nothing" that was found.

I think writing material just for a thesis is a waste of time, even spending much of one's time close to graduation to assemble a thesis is a sign of remedial work. The last interval in grad school should be wrapping up papers and starting projects of relevance to the next job. No one reads theses any more.

A corollary - it is irritating when the first 6 months of a post-doc I pay is spent finishing papers for an old boss that expects free labor from a newly-minted PhD to justify his/her grants that aren't paying for the time.

Anonymous said...

What Mrs. CH and Anon said!! I totally agree. FSP said, "A common complaint of advisers is how difficult it is to get thesis-related papers completed and submitted..." but I'd have to add:
A (very!) common compliant of students/postdocs is how difficult it is get their former PIs to sign off on and submit papers after they have moved onto a new place or position. (In many cases, as has already been said, the paper *is* written in a timely manner, but sits on the PI's desk. I think that when the postdoc/student is no longer in the lab everyday to ask, "...did you get a chance to look at that manuscript?" it falls to the bottom of the to-do list...)

Anonymous said...

Our university recently created a new policy regarding this issue. 1) In order to graduate the student must submit all data sets, programming code, etc in addition to their thesis. 2) Post-graduation, the student has one year to submit their research manuscripts or they give up their rights on the research and the PI, or other students, can write up the manuscript for publication w/first author going to the person who writes the paper.

In the academic world, it is publish or perish and there is too much risk to the PI to let work go unpublished so this protects their work. Plus, students who are serious about staying in academia should be well aware that publishing is critical to success and this policy re-inforces that point.

John Vidale said...

A more specific comment - collaboration is much more effective when the authors are is daily proximity, as anon@10:18am stated, which is why, in general, papers based on a thesis should be submitted by graduation or abandoned.

Advisors seem to forgot their responsibility to cut the umbilical cord and let graduates escape their advisors' shadow. It's not always that simple, but that should be the goal.

neurowoman said...

I think post-PhD paper submission issues is probably one of the top reasons it has become quite common in my field for graduate students to stay on in their PhD lab as 'postdocs' for up to a year after defending.

Ms.PhD said...

Many relevant points here! I'm bookmarking this post in case I ever find myself in FSP's position.

Personally, I like Kevin's suggestion to require regular reports (although I think "grading" is a bit stupid and demeaning at the graduate level).

I also like the suggestions about smaller projects for some students to ensure that they will be writing something up, most likely smallish papers.

I think it's right on to point out that it's always better training for students to have several small papers than one big, splashy one (although not as marketable in our current system, and therein lies the conundrum). Then they've hopefully been through the process of publishing something at least once before they're anywhere close to graduating and leaving you in the lurch.

In the cases where I've seen students not write up, it was always due to the same few factors:

1. fear of writing (you've blogged about students with this problem before)
2. slow PI or non-responding PI (i.e. student is in the mood and available to work on it, but you're not; then when you're ready, the student has long since gone past flabbergasted to exasperated to "I give up, who cares?")
3. project was way over student's head, and student is smart enough to know that it's not likely to be accepted anywhere without more experiments (which now-former student won't do, so they figure it's your problem now)

I guess my point being, you're clearly aware of #1, but maybe not always as aware of #2 and #3, especially if the student thinks they're communicating loud and clear and you're so busy, you didn't hear them.

Many of my friends who left science left for the same reason: they resented having to repeatedly ask and remind their PIs to do their jobs. We've come to expect it in academia. I'm guilty of this- if someone asks me to do something I'm not excited about, I often wait until they ask again. It's sort of like the over-politeness clause. Asking once might just be to raise the issue; asking again means it's a serious request. However, the business-minded students and postdocs HATE this, and for good reason. It sends the message that we think our time, as experienced scientists, is worth more than theirs.

And it's really not. Their lives are passing by just as fast as ours are passing us.

Anonymous said...


I'm up for tenure this year and submitted an unpublished chapter from my dissertation just last month. My goal is to get my second unpublished diss. chapter (yeah, that's right, I have another one) submitted before I hear back about tenure. Then, even if I don't get tenure, at least I'll be done with my dissertation!

All the rest of my dissertation was published before I finished. These two chapters were based on my own ideas, funded with little grants I wrote myself, so I don't feel like I've done my advisor wrong on this one. In fact, I suspect the delay in submitting them was due to some rather negative, dismissive comments my advisor made about the work during my defense. He also said he didn't want to be an author on them. In retrospect, as an experienced, soon to be (un-)tenured prof, I realize they are great papers.

[Expletive] advisor!

Anonymous said...

I think it's the PI's responsibility to make sure that everything is published/submitted prior to approving a graduation date for the student. If this is not possible (i.e. the project is not finished within ~5-6 years) then it's the PI's responsibility to put another student/post-doc on the project a year or two in advance so that the project can be gracefully transitioned to the other person when the student leaves. That's probably not always easy, but this sort of high-level, long-term management of projects is one of the primary jobs of the PI, in my opinion.

[full disclosure: I'm a PhD student that is a few months away from graduating, and have a post-doc lined up that I'm eager to start]

Emily said...

Interestingly, my advisor just re-submitted the journal article version of my last thesis chapter this week - two years after I graduated, submitted it the first time, and left the field.

In my defense - I did stick around the summer after graduation to get it submitted. On the other hand, by the time reviews came back three months later, I had neither time, interest, nor motivation to address them. It took my advisors and I a while to figure that out, though...

scatterplot said...

These are all interesting comments - does anyone have any advice for a student who would like to be one of the exceptional timely ones?

Space Prof said...

FSP: great post.

I agree with those who say the dissertation should not contain unique material but rather a collection of papers, but I also feel the dissertation should be a cohesive book that transitions well between topics/chapters. I have seen several dissertations that were simply "papers stapled together" and this greatly annoys me. I know professors who advocate this to their grad students. By I disagree. In my view, a dissertation should be greater than the sum of the parts, presenting an in-depth, multi-faceted, original analysis that proves that you are ready to be given the license for independent research (the PhD). This means that the material it contains for specific papers is often spread throughout the book and integrated with similar sections from other papers (published or virtual).

Another anonymous one said...

This one hits a nerve and I have experience on both sides.

A dissertation is more than a collection of articles. It's hard evidence the candidate can think in fundamental ways about one problem, collect data, and write it up -- not necessarily for publication -- for a dissertation which includes ramifications and tying in to other scholars' work. Decent dissertations are worth copyrighting in themselves and are bigger intellectual accomplishments than collections of papers.

For advisees who gripe that your adviser does not turn around your after-thesis paper on a time line you want... Hello, and welcome to our small Planet Earth. Your needs are real. And so my fellow scholars ask not what your adviser can do for you but what you can do for your adviser. Maybe you need to say that last sentence out loud a few times. Glue it to your computer screen. Never mind what it should be, this is what it is.

Rules change for papers after the adviser has approved your thesis. If your adviser doesn't look at your thesis for months, you are a victim. If your PI doesn't look at your manuscript draft for months, start thinking ¿how you can help the PI to help you?

The PI is responsible for quality control. That means the PI has freedoms you don't. Often something good enough to pass the defense needs more work to get published in the PI's opinion. Maybe you could ask the PI what you can do to help the work get publishable. Don't be too sure you know what the PI does.

Maybe the PI needs to prioritize the PI's own first author papers, proposal deadlines, etc. You could try to negotiate how you will help with those, so the PI will help you with your needs.

In practice your PI doesn't owe you. You owe him (or her).

Apparently this is as hard to accept as it is necessary. Problems are everywhere.

Kevin said...

@Ms. Phd
"Personally, I like Kevin's suggestion to require regular reports (although I think "grading" is a bit stupid and demeaning at the graduate level)."

By "grading", I meant the local satisfactory/unsatisfactory grade, with a paragraph of comments in the permanent record. I agree that A/B/C or numeric grades are a bit meaningless for on-going research projects. I have never given an "unsatisfactory" for a grad student research quarter, but I have given "incomplete" for students who did not turn in their reports on time.
This automatically turns into an "unsatisfactory" if the student does not clear the incomplete within a year.

Unbalanced Reaction said...

I'm experiencing similar frustrations to Mrs. CH (and a few anons): can't get The Boss to focus on my manuscript from over a year ago. On one hand, it doesn't really matter since it doesn't count towards tenure.


It's *my* freakin' paper! Submit it already!

(Okay, I feel better now. Nothing like a Sunday night mini-rant to kick off the week!)

Female Science Professor said...

I have no doubt that slow responses from advisers is a very real problem for students/postdocs for getting manuscripts submitted in a timely way, but I suspect that the preponderance of blog complaints about advisers vs. complaints by advisers is a reflection of blog demographics.

Anonymous said...

in my grad department one of the requirements for when you are allowed to even schedule your thesis defense is that you needed to first have at least one journal paper published or at least submitted. that was many years ago, I don't know if times have changed.

Female Science Professor said...

That's fine, but what about the other papers that can be written from the thesis work? Those are the loose ends to which I referred.

Woodstock said...

I'm much delayed in responding to this, and I'm not sure how much I will contribute, but this post has been percolating in my brain since it went up and I finally decided why.

I'm in category 2.5. I have a science job outside of academia which leaves me with little time to finish papers, but I am still interested in doing so and have been working on a draft. I cannot, however, actually get my former PI (hereafter FPI) to respond to me in any way because FPI views leaving academia as leaving science entirely. FPI has been known to say that there is no point in training female grad students because we all leave academia and thus science. Don't get me started on THAT. :)

I would like that first author credit even if I don't need it for tenure, so I am doing an end run around FPI. I know I will see a committee member at a conference and plan to discuss the situation with the committee member. If he can provide me with the feedback I know I need to get this draft up to snuff, I'll work with him on getting it submitted. Of course, FPI will still be the last author because he was the PI.

Sneaky? Maybe. Good for everyone involved including FPI? Definitely.

(My word verification is "a grant", which seems amazingly appropriate for this comment.)