Wednesday, October 05, 2011

Dissertation Co-authors

Somewhere, somehow, I must have touched on this topic before, but maybe not lately, and it came up in a recent discussion with a colleague, so here goes:

In some fields, research is highly collaborative, and, as a result, publications have many co-authors. Therefore, in places where the doctoral thesis is a collection of papers resulting from the PhD research, there are 'co-authors' of the thesis, although of course only one official author of the thesis itself.

If the person whose name is on the thesis played a major role in the papers included in the thesis, I don't think having 'co-authors' of a thesis is a problem, as long as it is clear that the thesis chapters are papers and the co-authors are clearly indicated.

In some cases, formatting rules of the institution allow for reprints to be bound together to create the thesis, so there is no question about the relationship of chapters to papers. In cases in which the thesis must be formatted following strict rules of font, font size, margin size, pagination etc., the title of each chapter needs to be very clear about the relationship of the chapter to a paper, including listing co-authors of the paper. [Most, but not all, students are aware that they should list co-authors or, once told, have no problem with doing this and realize it is the right thing to do.]

The role of a PhD student may vary from paper to paper, and therefore from chapter to chapter in the thesis. In some cases, the PhD student may not have been the primary author in a paper that is included in the thesis. What then?

I remember one potential problem with this particular scenario once, when there was a question raised about a student's including a chapter that was a paper on which the student was not the primary author. It was not my student, or even my field, but I was nevertheless asked to weigh in. So, I did a bit of scouting and learned that, at many institutions, it is considered OK to include a minor-authored chapter/paper if the majority of chapters (papers) represent major contributions by the thesis author. Again, the authorship of each chapter/paper needs to be spelled out, but if that is done, then there should be no problem.

There are certainly fields in which it is common for a doctoral student to be just one person in a very large group of collaborators, all of whom are included as co-authors on papers. Are any of you concerned about the amount of research done by PhD students -- specifically in a collaborative project -- with respect to what ends up in a thesis?

In cases in which a thesis is a bundle of co-authored papers, I think it is good if the student writes an introductory chapter (and possibly also a concluding chapter) that gives a broad view of the body of work and gives the student an opportunity to put their own stamp on their own thesis, without any co-authors. It may be that no one will ever read the thesis itself, but, if time permits the writing of such 'extra' chapters, the exercise of being sole author on at least one part of a dissertation can be very useful.

How much of your thesis did you write (yourself)? (please note your field, if you are willing)

43 comments:

SocSciProf said...

I'm in anthropology, so my dissertation was 100% my work (from obtaining funding to data collection to writing). But my husband's in computer science, so his dissertation was a bundle of papers, and I think all but one he was first-author on.

MathTT said...

I wrote the whole thesis myself.

Every single paper I've written since then has been co-authored. I wonder what that says about me?

(Oh, and math... if that wasn't obvious from my handle.)

mOOm said...

One of four main chapters was based on a coauthored paper (with my primary advisor). My PhD was in human geography.

Amy said...

I'm in biology and was first author on all the papers that made it into my thesis, but would happily have included a further-down-the-list paper if the part that I contributed was still relatively major (e.g. if it was a consortium paper and I was in charge of Analysis X, then that could just as well have been its own mini-paper to be published on its own once the dataset was out).

One of the first pages of my thesis was a list of the manuscripts that each chapter was based on, with a full list of author contributions in the same style that you see in many journals. As far as possible I also removed what other people had done from my results and put it into the discussion (this worked well for me, but of course depends on the nature of the collaboration). In addition to my four experimental chapters I also wrote a relatively substantial introduction and discussion, linking the different aspects of my work together.

In my view, students should be encouraged to write as much as possible for their thesis, at least if they haven't had the opportunity to write any papers of their own (not just providing comments on a pre-existing draft). I thought writing my first paper was pretty tough (and I love writing), so if you come out of your PhD without having written anything and then as a postdoc are suddenly expected to start whipping out papers and grant proposals, then you're in for a ride.

Anonymous said...

I work in probably the most collaborative field in science (experimental particle physics, typically >500 authors per paper) and except for the description of the theoretical background and the used detector equipment the thesis was all my own work, I wrote the papers myself and had no substantial help from my 500+ best friends.

But - to be honest - what I have seen since I became faculty is that this is indeed a problem, particularly the weaker Ph.D. candidates struggle to get a published result that is their own, in this field the most of the competition/peer review happens before the paper gets out and this can be really difficult for a weak result or even a solid result with an inexperienced author.

muddled postdoc said...

About 95% of my thesis was written by me - the other 5% to corrections made by my advisor/coauthors in papers that were reproduced. I am in experimental physics/material science. All my "results" chapters and even some of my lit review have been published before but with me as the first author. Some parts in preliminary work sections were done together with a senior grad student, but the section was rewritten though we did use some similar figures. Although he graduated 2 years ago I couldn't avoid putting this bit in as there would be a bit of a disconnect in how did you suddenly get to point B.
In my thesis I listed out the papers that my chapters are based on, with the co-author contributions.

I did have 3 papers where I was a 2nd author and had done significant work and analysis and writing, but these were different from my thesis topic and could it was difficult to tie it in. Additionally and more importantly this work was part of the thesis topics for other junior Masters and PhD students so if I had a detailed chapter on that work could they include it later as well? One of the papers could have been tied in loosely but I never did because of this concern rather just listed it under my publications.
What happens in such a scenario where coauthors are all grad students?

Anonymous said...

I'm in physics, and wrote the whole thesis myself. The only place I used any information from a paper where I wasn't a primary author was in a small section of one chapter, and there I rewrote the material because I didn't want to get into a lot of details of the paper, I just wanted to use my involvement in that project as one example of how an experimental technique could be used. In general, I had the issue that my advisor was unwilling to give many thesis comments, so it felt like my thesis was too much of my perspective rather than not enough.

I can't imagine not writing your own thesis, but then the whole graduation process was really anticlimactic for me so maybe it doesn't matter. I doubt anyone will read the thesis I put so much work into anyway.

Anonymous said...

2/3 econ
One of the three papers was coauthored with a classmate.

It is common for 1/3 of the three papers in top programs in econ to be coauthored either with a professor or another student (I did have a coauthored paper with a professor, but it did not go in my dissertation). If it's not coauthored, then the third paper is often much weaker than the other two.

nordicTT said...

I wrote all the chapters of my thesis (in physics) but a small one (applications of the technique developed). The papers have been mainly written by my supervisor because we could not wait very long (the topic was hot at that moment). But I wish I could write more papers from scratch, since this is completely different as contributing with data, plots, comments... and it's a skill that one can improve only by doing it.

zed said...

my father has an actual co-authored PhD thesis, from MIT in the 1970s. He and a fellow student co-wrote the thesis.

Yael said...

I wrote the whole thing. In the place I went to grad school (a private med school), all the data presented in the thesis must have been the work of the author (hence there is no notion of co-authored dissertations). In cases where students co-authored papers with other people, those data must be edited out and cited, so as not to give the impression that the student did the work (hence, these students tend not to use the compilation format). Some students staple a few papers together and write and intro/conclusion--those are almost always student/advisor only authored manuscripts.

Unknown said...

I didn't bundle papers per-say but I did take the existing papers and augmented them. Basically, the paper formed a backbone to the chapter and then I filled in the flesh around it.

When I wasn't the primary author, I put a footnote at the start of the chapter stating the primary author and what my contribution to the was.

Mechanical Engineering

queenrandom said...

I'm in molecular biology. My thesis included a revised/resubmitted coauthored paper as one chapter, although I had written the paper and had performed the vast majority of the work therein. I had several other chapters as well, including an intro and discussion, which I did 100% of the writing for.

My school required students to write intro and discussion chapters even if the thesis was a compilation of papers. We were also expected to include a short (~1 page) explanation of coauthor contributions in coauthored chapters. My school did many things bass ackwards, but I agree with their handling of the thesis in this respect; it ensures the committee that the student can write scholarly material and that s/he has indeed done the required benchwork.

Female Genetics Professor said...

My dissertation was based primarily on work that I did alone but included some data from a collaboration with another student. This was many years ago, before it became common to combine papers in a bundle for a thesis. Students in the program that I am currently affiliated with are required to have at least one first-author publication accepted before receiving their PhD's. They often bundle papers but write an overall Intro and Discussion to pull everything together.

Anonymous said...

My thesis was a bundle of 1st author papers. Field=ecology. I developed the research project, did all sample collection and analysis, and wrote the papers. Coauthors were major advisor (who revised manuscripts, made suggestions, usual advisor-y stuff), people whose labs/equipment I used to run samples, and who helped me with fieldwork.

Anonymous said...

I am in neuroscience and wrote the whole thing. I was first author on all the papers that went in the thesis and spent a good deal of time writing a historical overview of the field for the intro (this was actually a lot of fun - "modern" sub-field that I'm in dated back to ~1850's). Co-authors, were, of course, acknowledged.

Anonymous said...

my dissertation was written 100% by me and it included all papers that I have written so far ( and I was the first author on all of them). My area was computer science and engineering

Susan said...

I am in engineering. In my department the thesis was not a compilation of papers. I worked ina group and we often worked with the same data but did different analysis. So my thesis was 100% mine.

Anonymous said...

I'm in chemical engineering, and if I recall correctly from the scant six years ago, my dissertation was a review-style book chapter (me + advisor) and three first author papers (2 x two authors, 1 x 4 authors). I didn't have any non-first-author papers to worry about.

Anonymous said...

I was sole author on my computer science PhD thesis, though there was one result in in that was published as a co-authored paper with 4 authors (though I did about 80% of the work).

Students in bioinformatics often have papers that have multiple authors. Our department requires that they rewrite the papers to become thesis chapters, making it very clear what their personal contribution to the research was. It doesn't matter where on the author list they appear. Even sole-author papers usually need to be rewritten, as a thesis chapter should go into more detail than most journals in our field allow.

A simple "staplegun" thesis is not enough, as the main point of a thesis is to establish the individual's ability to do research, while the point of a journal article is quite different—the competence of the author is assumed.

Anonymous said...

In my field, a collaborative physical and biological science, it is most common for a dissertation to be a bundle of papers. Our institution requires an introductory and conluding chapter that links the papers together as well as a statement near the acknowledgements that says where things were published, who the co-authors were, and what the student's contribution to each paper was. Sometimes papers that a student was not first author on help to link various ideas together and thus are included. It's expected that at least 3 of the papers in the dissertation will have been led by the student. Each chapter also gets a footnote on the first page with the original publiction information if it is already published. Having the student have to explicitly spell out their role makes it easy for their committee to determine if they have done enough of the work and if they still have weaknesses to address.

Anonymous said...

Chemistry.

My thesis was based on four papers: two published, one submitted, and one still in preparation (I finished it right after the thesis was written, but it has not been submitted because my ex-advisor prefers to sit on these things).

For one paper, I was formally second author (against my wishes) even though I wrote 75% of the paper and 75% of the data was mine (some of this data was redone - only 30-40% of the data from me was "new"). I essentially reworked a skeleton of a paper by an earlier student in the group. I chose to rewrite the chapter based on this paper to include nearly exclusively my data and my analyses. My advisor actually counseled me against this, but I found it made a more coherent contribution to my thesis as a whole. I also wrote a lengthy introduction chapter that ideally would have formed the basis for a review articles (I did a really good job on it and it was somewhat creative), but again, my ex-advisor would not let me submit it as sole or corresponding author, and that advisor would sit on it forever and then take all credit, so I dropped the idea and moved on with life and a beautiful thesis.

However, I AM struggling with this issue now in attempting to mentor grad students as a post doc. What do you do with the less productive/less competent grad students that are not first authors on (m)any papers and have not written any papers themselves (because they cannot write/think coherently enough). Simply not allow them to write a thesis and get a PhD? Or allow them to rehash everybody else's work in their thesis? The first option might seem the best, but as usual - easier said than done...

Anonymous said...

My dissertation in analytical chemistry was 100% written by me. There was some data included that was done on a collaborative project where I ended up being first author on the paper when it was eventually written. All of the text was mine and we weren't allowed to bundle papers though that approach makes tons of sense to me for chapters where the paper is done or close to done.

Anonymous said...

earth sciences, i wrote the thesis 100% myself and even the first paper with 8 co-authors i wrote about 95% of.

Anonymous said...

All the theses that I've seen in astronomy are stapled papers/manuscripts that were written, first-author, by the graduating student. There usually are co-authors, but the student is the first author. They all have separate Introduction and Conclusion chapters to turn the stapled contents into a coherent whole. This applies to both European and US theses I've seen.

But what counts as "how much did you write yourself"? I certainly know of theses where the advisor wrote much or most of these papers that were ostensibly written by the student...

Anonymous said...

Maybe the question isn't 'how much did you write?', but how much of the new research represented by what is in your thesis (that is, excluding citation of the work of others) is yours: ideas, data, writing, everything? If the question is written that way, I don't believe most of these "100% of my thesis is mine" claims for many of these fields.

Alex said...

A valuable chapter in the thesis is the chapter on the project that didn't really work out, but did lead to something that was useful for another project and snagged the student a second-author paper.

Or, as I call it, "Chapter 2."

Seriously, I think that's a valuable chapter. It's the one where you learn how to take a very sour lemon and still make at least a little bit of refreshing lemonade. Not as valuable as the chapters describing work in first-author papers, but still valuable as a learning experience.

Anonymous said...

This is an interesting discussion. I work in the library of a major research university and provide workshops for graduate students on managing copyrights (often in relation to their thesis or dissertation). I recently had a student ask me to comment on a situation where his advisor, a co-author on an article, was refusing to allow the student to include it in his dissertation on copyright grounds. It's not something that I could (or would necessarily want) to comment on specifically (not a lawyer, etc etc and can't provide specific advice on this), but to my understanding this is not a valid argument and that the co-author wasn't addressing the real issue. I wonder if others have seen this argument made by co-authors in relation to dissertations or theses?

Doc said...

Interesting discussion.

Field: Chemistry/Biochemistry
Ph.D dissertation: 5 papers (first author) + intro + conclusion.

I actually had one co-author paper that we left out and just cited in the intro. That was because I provided the wet-lab work for a computational group to check their model, and I didn't feel qualified to defend their methods.

Interestingly, I helped mentor a grad student at my post-doc institution, and she is graduating. She has 2 papers where I am first author and I wrote the paper and only 1 of her own (where I am still co-first-author, but listed second and I wrote only small amounts of the paper). I'm kind of curious about my old PI's intentions there...

ChrisMath said...

I'm in math as well and wrote the entire thesis myself. I was fortunate to choose an area to work that was not in my advisor's main area of expertise, so I did all of the background research, problem selection, and whatnot.

Anonymous said...

At my institute compilation theses are mandatory for all science PhDs - based on a strong belief in the peer review system and the belief that results need to be in the public domain to be valid, I suppose. I'm in an interdisciplinary field (cross-appointed in physics and chemistry) and we typically have 6-12 authors on papers. Most students in my department include 8-10 papers in their thesis, where most of them are first author on 4-5 of these (I was first on all 9 papers I included, but this was never done before and also not in the decade since).
As some have noted, compilation theses typically include an introduction section which serves as a review of the field and puts the student's work in context (the best of these can be published after as reviews with the former student as sole author), as well as a list of papers that should clearly state the student's individual contribution to all multi-author papers (although most students exaggerate their contribution, in my opinion). Anything the student claims credit for is fair game for the defense (which is public), so it usually becomes clear very quickly how much a student has actually done themselves.

phd biology said...

biology.

I had four published papers "in" my thesis, 2 first-author, 2 second-author (collaborative). The thesis in itself was written by me only - as in a "text in front of the papers, with background and history of the field, questions/aims to answer and putting the research in context and future thoughts" with the papers as appendices to my written part.

That said, in my defense I got questions on every single paper (as I should) and had to be able to answer for the whole paper since I did put it as "mine" when I added it into my thesis.

i.e. if you put it in your thesis, at least I don't accept "I didn't do that step so therefore I can't answer that question" since you should know inside and out what's going on in the paper when you claim it as yours in the thesis...

Anonymous said...

Biology
My thesis: 5 chapters that led to 6 first author papers - plus short intro/summary chapters to provide a coherent framework (not published but useful as the basis for a dissertation summary). One paper was co-authored but for the other 5 I was sole author. I wrote everything in the thesis independently and the co-authored paper was later contributed to/edited by co-authors. I had several collaborative projects that came out of grad school research but these weren't part of the thesis, mostly because they were usually side projects that were somewhat peripheral to the main thesis topic. I think this structure is pretty normal for my field.

Anonymous said...

econometrics
My standard model for supervision of a student would be 3 papers.
paper 1: coauthored strong input from supervisor
paper 2: coauthored smaller input from supervisor
paper 3: sole authored
This way they get 2 publications to submission during the candidature or shortly after and learn how to structure a paper in a supported way. Seems to work so far but its not a hard and fast model, some students have chapters which dont work out but are worth including, some write more papers. The field has moved from coauthoring with supervisor indicates a weak student to coauthoring with supervisor indicates strong student that supervisor is willing to devote time to working with for an important publication. Note in our field it is not uncommon for a publication to take 4 years to come to fruition.

Anonymous said...

Biochemist. My work was really collaborative and though I was first author on all but one of the papers that came out of the thesis, I explicitly credited who did what in my main text. In one chapter, though, I also included figures that weren't mine (and said so in the captions and text). This annoyed my committee as they could not legitimately make me try and defend them so, in my final revision, I was ordered to remove those figures and tone down my discussion of that data. Being more than a little burnt-out, I complied...if they'd asked me to pick up thumbtacks with my tongue, I would have complied. I just wanted it to be over so I could finally get some rest.

BTW, at my institution, we can't just cut and paste papers because our dissertations themselves get sold to publishers. Once a paper is published, the journal holds the copyright and so a copy-paste job is both copyright infringement and self-plaigirism. There's nothing quite like re-writing what you've already polished up. Really makes you want to send you face through a wall.

Anon Y Mous said...

I am currently a grad student in biology, so I cannot confirm how I will write my thesis, but my PI's philosophy is that first-author publications where the clear conceptual and experimental work came from the student are included. Anything else (non-first author or first-author but not the primary driver of the work) would go in the appendix as a way to show that you also did some other things.

I plan to include an intro chapter that will be an expanded version of a short review article i wrote (w/ PI as co-author), 3 or 4 first-author pubs (all w/ co-authors), a discussion chapter, and then put the 1 out of 2 2nd-author pubs I have that relates to my thesis the best in the appendix. At my institution, you make it clear at the beginning of each chapter, if there are co-authors and/or the work is published, what the role of each co-author in the work was (so that the student's role is clear). I think the first-author pubs are pretty clear in my case since we always have to write the first draft of a paper ourselves before my PI will edit/suggest changes.

Anonymous said...

100% and computer science

Old Biddy said...

Chemistry, ca. 17 years ago. I was on mostly solo synthesis/reactivity projects, so I wrote 100% of my thesis myself. While I was waiting to do my defense I turned the chapters into papers. A side project which did not make it into my thesis was then picked up by a first year.

Pagan Topologist said...

My field is topology, as can be discerned from my handle here. This is a branch of pure mathematics. My dissertation, completed in 1968, was entirely my work: I chose the problem, over objections from my advisor, and did all the work and wrote it up. My advisor asked a lot of questions, but he did not contribute any of the actual work.

Anonymous said...

My thesis was (eventually) 6 published papers. Two single author papers (published prior to graduation), two with young gung ho advisor (submitted prior to thesis) and two papers wih old grant rich advisor (submitted verbatim years after graduation).

I am now a tenured applied math prof and never publish single author papers but rarely work with people in my area.

None of my students has ever written a single author paper, they do the mathematics in interdisciplinary projects. Our department considers it solo work though because no mathematician co-authors the papers that go into the thesis.

Anonymous said...

PhD thesis (History) was 100% me, as were the sole author publications that resulted from it.
Masters thesis (Genetic Epidemiology - post history phd) - I didn't do the data collection, but I did all the analysis and writing, and was first author on the two publications from it.
In either of these (very different) fields, it would be odd to include substantial chunks of someone else's writing.

Anonymous said...

Sociology. 100% me.

Which didn't prevent one of the male senior statesmen in my field to cite the work and attribute it 100% to my advisor. Classy.

Anonymous said...

In my biomedical science Department we encourage folks to publish their work, and are fine with published manuscripts as chapters. The expectations are that each student will produce :

1. At least one first author paper

2. An introductory chapter and Discussion chapter written by the student for their thesis.

3. A one-page preface to each data chapter explaining who did what, if not all the data is from the student.

Back in the stone age, I wrote my entire thesis, both published and unpublished chapters--half of the work was mine alone and the other half of the chapters were a collaborative project with a postdoc who wrote less well than I. However, the era of two author papers is sadly almost at an end.

Mark P