The ethics of co-authored doctoral dissertations

Every thesis is, to some extent, a collaboration between the supervisor and the student. A student will have a research topic or an idea. The supervisor can say, "this is an excellent data set," "have you read this paper?" or "try taking this approach." A meeting between supervisor and student can turn into an hour long session at the chalkboard - "O.k., let's try it this way and see if we can work something out" "what you're saying is essentially the relationship between X and Y is like this."

Most supervisors are more experienced writers than their students, and correct minor grammatical errors. And there is a slippery slope. At first, it's just putting on "track changes" and fixing misplaced apostrophes. Then one begins adding long editorial comments that can easily be incorporated into the student's text.

The next logical step is to take the professor-student collaboration and turn it into a co-authored paper, suitable for publication. The professor, of course, is better at writing introductions, literature surveys, and conclusions, so takes care of those parts of the paper. And why not include that co-authored paper in the PhD thesis - after all, it is the student's work?

And what is wrong with a student co-authoring a doctoral dissertation with his or her supervisor when most academics co-author most of their research? The myth of the brilliant genius working alone is just that, a myth. Ideas are tested, refined and sharpened through discussion with other people. Even the words we use are not our own. Our heads are filled with words and phrases that we have heard elsewhere - definitions drilled into our brains, lyrics of a favourite song, a catchphrase from the Simpsons. Those words and phrases find their ways into our fingers and onto the screen.

One problem is, when research is co-authored, the supervisor is in a conflict of interest position with regard to that research. Suppose, for example, in an otherwise excellent paper, one comes across a turgid and overwritten section obviously authored by the supervisor? Suppose one criticizes that section during the thesis defence, and the supervisor strongly defends it. One could say, "of course you're going to defend that section, you wrote it, but it's still wrong." But generally one doesn't.

Another problem is that, when research is clearly co-authored, it becomes difficult to assess the contribution of the student. This is a particular problem if the student is a brilliant, diffident and modest woman, taught from a young age to say, "oh, it was nothing" - even if it was all her own idea. It is  crucial for PhD students to understand the importance of protecting and defending their own intellectual property rights.

There are ways of changing the status quo.

I was recently external examiner on a doctoral dissertation that contained a statement of authorship, outlining the contribution of supervisor and student to each chapter. I would like to see such a statement of authorship on every thesis.

Supervisors should also be required to refrain from the evaluation of any co-authored research, and allow the evaluation of that work to rest entirely on the hands of the rest of the committee and, most importantly, the external examiners.

 

Comments

[Trevor]

I'm just finishing up my PhD, and plan to include a co-authored Chapter within my dissertation. Here's my thought on the subject: Dissertations do not matter; only job-market papers matter. Determining the specific contribution of each author within a particular dissertation chapter is not necessary, since the vast majority of job-market papers are single-authored. Thoughts?

[Frances Woolley]

Trevor, will the co-authored chapter be identified as co-authored and, if so, how? How do you know that the external examiner isn't going to assume that it's all your supervisor's idea/work? How is your supervisor going to deal with the potential conflict of interest that he/she faces when it comes to examining that chapter? Are your single authored papers 100% your own work?

[Chris J]

A thought from the realm of physics. The collaboration on my PhD experiment was small by standards of my field: about 100 people. I needed to describe the detector and how the electronics worked (for example) and I had nothing to do with that - I wrote the words, but didn't do the work. But without that information, my work would have been impossible to understand. So I included a statement of originality in the front matter which stated chapter-by-chapter what work was mine, what work was done in collaboration with colleagues, and what was simply my description of what my colleagues did.

In physics dissertations matter alot. When hiring a recent PhD into a postdoc position we want to know what they did on their PhD experiment, what they did not do, and how they found the opportunity to be scientifically creative and novel. In short we want to know what their thesis was. One of the best seminars I attended of someone applying for a job was stressful for the recent graduate because he lost his powerpoint! He had his thesis on pdf and went through in order showing us the figures and discussing what they meant. What I came away with was the knowledge that this person could think in stressful situations and he wrote a very good thesis because in going through the thesis in order he was able to present a beautiful, logical narrative.

[Mike Moffatt]

Great stuff!

One thing I've always found particularly interesting is when either the student or the supervisor is a terrific writer and the other one, well, isn't (or is ESL). There are more than a few theses floating around out there in which some sections are easily readable while other sections are incomprehensible (and not because they're overly technical, either).

[Trevor]

Frances: Those are excellent questions, but do not apply in my case. Without going into particulars, the co-authored chapter is not with anyone on my supervisor committee, so no conflict exists. But, you raise an interesting issue. If a supervisor does co-author with a student on a chapter, wouldn't it be incumbent on the supervisor to think about the conflict-of-interest issues that might exist?

In any case, my feeling (at least at my school) is that only one paper matters, the job-market-paper. "You only need one good paper" is repeated constantly. The dissertation is put together only after a successful job-market. Is this different at Carelton?

... it certainly is different from the physics community, described by Chris.

[Frances Woolley]

Trevor: "Without going into particulars, the co-authored chapter is not with anyone on my supervisor committee, so no conflict exists."

If, as an undergraduate, you handed in a paper that had partially been written by someone else, you would be sent down to the Dean's office, read the riot act with regard to academic integrity, and (probably) given an automatic F in the course.

Why doesn't the same thing happen with respect to a PhD thesis?

In your school, what does the rest of the thesis consist of (other than the job market paper)? If you can say so without revealing too much about yourself, are you in the US, in Canada or elsewhere?

Mike: I think it's easy to give ESL speakers a hard time on this because the extent of collaboration is much more obvious when the supervisor and the student have vastly different writing styles. And, yes, it's tempting to say "step aside, it would be quicker if I just wrote this for you." But as Trevor and Chris J's comments reveal, collaboration and co-authorship is a big issue even when people write English fluently.

Chris J - it's interesting that in physics - which economists often hold up as a model of "this is what real science is like" - originality statements are (based on what you say) standard practice and the nature of collaboration is dealt with directly and openly.

[Jon]

If you have a document of original work and the student can articulate the idea contained and cogently answer questions beyond the four corners of the text, that is enough. The dissertation itself is basically for the committee alone to test whether the defense is original and that the understanding is complete.

There is also the issue that your basically introducing a risk of taint. Certainly when people present their work in group meeting, there are and ought to be discussions. I'd hope those discussion cover critical points, otherwise what's the point? Are we to speculatively keep our mouths shut? You could kill four years of effort by adding the critical piece at the end.

I find your argument incoherent as well within its own terms. If students are modest, they will understate their authorship--a point which makes that section useless.

[adjacent / q]

i did my work in mathematics, and i don't think my advisor read my thesis, much less wrote it.

[Frances Woolley]

Jon:"There is also the issue that your basically introducing a risk of taint."

Yes, work that is obviously co-authored is tainted - and as Mike points out, ESL students are particularly vulnerable here, because co-authorship is easier to spot. Without a statement of originality/authorship, the entire thesis is compromised.

Jon: "The dissertation itself is basically for the committee alone to test whether the defense is original and that the understanding is complete."

I don't understand what you mean. The dissertation is the written product, which may be written by the student, or may be written jointly by the student and someone else. Without a statement of originality, how does the committee know what the student's contribution to the written product has been?

The defence (Canadian spelling) refers to a final examination where the thesis is evaluated by the members of the committee and (usually) an internal external examiner (e.g. a colleague from the physics department) and an external examiner. I would certainly not want to base my assessment of the student's understanding of the work entirely on the oral defence - students not uncommonly become inarticulate and crack under pressure.

A supervisor who has any concern at all with his or her students' well-being will make sure that the statement of authorship is accurate. Moreover, it is much easier to write a statement of authorship (I did the econometric work and wrote it up, my supervisor wrote the intro and conclusion) than claim ownership of something in public. If someone says to me "these are delicious cranberry orange muffins" I'll say something like "oh, we just had some cranberry sauce left over from Thanksgiving, I wanted to finish it up" (note to American readers: Canadian thanksgiving is in October). Women are deeply conditioned to understate their contributions - but a supervisor can make sure that the statement of authorship on the thesis is accurate.

Moreover, a statement of authorship is going to have to be written at some point anyways. Tenure and promotion files commonly require a co-authorship statement (e.g. "equal contribution from both authors"). I've known of friendships being severely tested when co-authors couldn't agree on the division of credit on a co-authored piece.

[Frances Woolley]

adjacent / q - sounds like you had a bad experience.

I don't want it to sound as if supervisors are angels and students are spongers - PhD Comics has a brilliant cartoon describing "Profzi schemes". If a student comes up with a really brilliant idea, of course I want to get my name on it.

I've heard from other people that co-authorship is an issue in math, as well, because otherwise brilliant people might not know what to put between the equations.

[Trevor]

Frances: "If, as an undergraduate, you handed in a paper that had partially been written by someone else, you would be sent down to the Dean's office, read the riot act with regard to academic integrity, and (probably) given an automatic F in the course." -- Maybe I've misunderstood what the issue is, but this comparison is not valid at all. You only get sent to the Dean's office if you pretend the work of someone else is your own. In co-authoring a paper, both names appear - no one is hiding anything. Group work is common in undergraduate studies and is essential in frontier research as well. The larger and more complex datasets become, the more essential teamwork will be in our profession. Graduate students will eventually become colleagues of faculty at another university and should be expected to function productively with others. I'm not going to actually suggest this, but... maybe... co-authoring at least one chapter should be required of a graduate student to actually demonstrate their ability to function collaboratively within the profession. ... lol... well, maybe not required... but I don't see the problem with it.

:) As for revealing more information about me... lol... I am starting to think I'm wildly off base in my arguments, so I'll just stay anonymous... especially since Carleton is hiring this year, haha.

[Trevor]

... a follow up from an earlier post of Frances I didn't read through completely. "Without a statement of originality, how does the committee know what the student's contribution to the written product has been?"

I completely agree. I think I may have been misunderstanding the post. So, my apologies. Of course I agree that credit and contributions should be made clearly in writing, especially for the committee. So, I think we're on the same page... other than my again reiterating that co-authoring papers should be encouraged for graduate students.

[Frances Woolley]

Trevor, I think I might have been unclear also. My concern is about the when only one name appears on the thesis, but there is, say, a working paper in circulation, identical to the thesis chapter, which is co-authored.

I agree with the points you made about group work and the value of collaboration. There isn't an academic integrity issue when the co-authorship is clearly acknowledged - or at least it's a much much more marginal concern. But isn't there still an issue that the person who writes three essays including one or two co-authored chapters does less work than the person who writes three single-authored essays? So allowing the thesis to be (partially) a group project is, effectively, a reduction in standards?

Good luck on the job market - please feel free to email with suggestions/requests for job market related blog topics.

[Mike Moffatt]

"Mike: I think it's easy to give ESL speakers a hard time on this because the extent of collaboration is much more obvious when the supervisor and the student have vastly different writing styles. And, yes, it's tempting to say "step aside, it would be quicker if I just wrote this for you." But as Trevor and Chris J's comments reveal, collaboration and co-authorship is a big issue even when people write English fluently."

Agreed. The difficulty then is it is difficult to see who wrote what.

And it's not just ESL speakers - there are a lot of native English speakers who cannot write coherently.

[pat]

Frances:
I am not sure if 3 average quality single authored chapter is more work than 3 quite good (and more ambitious) papers: lets say 1 single authored and 2 co-authored.

Especially in the applied field where I think an experienced economists could sometime do in maximun one year the 3 chapters...Moreover, I believe the student might learn more if he coauthors... I think his future work might be better than someone who worked alone on the 3 chapters.

[Frances Woolley]

Pat: "I think his future work might be better than someone who worked alone on the 3 chapters."

The thesis is a demonstration of a student's ability to produce original research. If the thesis is written jointly by Person A and Person B, it is very hard/ impossible to use the thesis to demonstrate the competence of Person A, because all of the substantive content may have come from Person B.

Inferring Person A's competence from a joint Person A/Person B product is easier if:

a. All co-authored material is identified. ****This is not at present standard practice within the economics profession.****

b. The precise contribution of each author to co-authored material is clearly stated. ****This is not at present standard practice within the economics profession.*****

[pat]

Frances:

I agree about the importance of identifying the contribution of each co-author. Eventhough I didn't make it clear above.

I was just making a point in favor of co-authorship as a learning tool for a grad student. I believe it personally help me a lot back then. My advisor was more involved that way...

[Frances Woolley]

Pat, your point in favour of co-authorship as a learning tool is a good one. And it's true for more senior scholars too - I've learned a great deal from various co-authors over the years.

Just out of curiosity, did your thesis contain a statement of authorship? How specific was it? Was this standard practice where you got your PhD?

[Jack]

I haven't seen discussion of recommendation letters. These letters are crucial in the job market, and can be used as a ''roadmap'' or legend for the PhD dissertation, in light of this discussion thread. If the letter writer says, This is a bright kid, but (s)he needed a lot of help, so it isn't more than 70% original work, it would be a weak endorsement, even if all chapters in the dissertation are in the student's name.

However, I suppose most advisers want their students to succeed, and there is a perception that 99% of recommendation letters are glowing... maybe untrue?

[pat]

Frances, no my thesis didnt contain such statement. But in my case, I came up with the ideas and did quite everything. I just felt free to go ask my advisor multiple questions...

The co-authorsip is very common where I studied and still is. In one field in particular (not mine) profs even give research idea to their student (at least) for the first 2 papers. They all manage to publish in top journal field and keep working together after graduation... Those student usually get the better jobs...

I think it is quite common and not specific to where I studied. I know that a lot of prof give subject to their student to help them do a good second year paper... That's my understanding of it... I think it is very hard for a student who doesn't know much the litterature to come up with a good subject in a short period of time...

[Jon]

The defence (Canadian spelling) refers to a final examination where the thesis is evaluated by the members of the committee and (usually) an internal external examiner (e.g. a colleague from the physics department) and an external examiner. I would certainly not want to base my assessment of the student's understanding of the work entirely on the oral defence - students not uncommonly become inarticulate and crack under pressure.

Thanks, but so that you don't stick your foot in it further, I have a PhD from Caltech (in engineering, not economics). I know how the process works.

But I disagree with your emphasis here. Of course the student is expected to write the thesis, but I don't think over-editing is a concern. The student is responsible for the content, however it came into being, and if they aren't willing to argue it as their own, its their risk for including it.

You seem to conflate who did the work with who writes it down. I'm not sure how you manage your students, but ownership of ideas and insights is hardly clear-cut.

For instance, we were required to regularly present our work in progress to the entire group. That time was spent with vigorous debate, questioning, speculation, and brainstorming, much of which infects the entire process of inspiration.

This point, you concede forth right but then switch to discussing authorship, but everything is entangled, and it does not seem separable to me as you claim. These are not distinct issues, and I do not think the origin of ideas is necessarily on the whole traceable.

Then you allude to a conflict of interest. Either the student believes in something or they do not. If they do not, they should not be bullied into including it as such.
Perhaps you argue that your adviser would be too forgiving to the idea if he originated it, but so? Surely the adviser believes in the thesis, or he would never have granted consent for the student to go forward.

Here perhaps our experiences diverge. The median year of graduation in my group was the seventh, and this is generally true at the top schools in science and engineering. Advisers won't let their students defend (or even start writing!) until the material is sound, although you might make it out half-assed by Mercy after your tenth year.

Regardless, my point was focused upon what should it mean to deserve a PhD. I don't find the authorship of the thesis itself to be meaningful--and indeed in some disciplines the thesis is little more than compendium of prior publication. Indeed, what matters is this:

- Ability to organize the investigation of topic: Here the thesis shows nothing because there is to varying degrees mentoring and guidance from the faculty. This is more or less assumed to have been met by having your adviser's grace to go forward.

- Further that the investigation results in ideas or knowledge of value: Here the thesis serves as an exposition on what is claimed to be novel and of what use that novelty is. The authorship of a passage changes neither and the preface or introductory remarks along with citations serve to frame what is preliminary and what is being claimed.

- A thorough understanding of the matter at hand: Here is where the defense is essential. This is quite simply a matter of convincing the committee that your work (and your understanding) is robust.

As a student, I read, commented, and scrawled ideas and notes upon too many drafts of theses to believe that an authorship summary is the lest tractable or relevant.

[Frances Woolley]

Jon - perhaps we don't disagree much at all, and it's just a matter of different experiences. You wrote:

"Of course the student is expected to write the thesis,"

I agree 100%, but this is no longer a generally accepted proposition.

My concern is about situations where parts of the thesis are not written by the student. For example, one commentator above frankly acknowledged that one of his chapters was co-authored (and presumably co-written) with a fellow student, and that everyone on the committee was perfectly o.k. with this.

[Lindsay]

Frances

I agree so much that your post has turned from a long time lurker to a first time commenter on this blog (which I and my students use all the time in a course I teach on Economic Policy Analysis).

The dissertation, as a package, is absolutely vital, regardless of comments here to the otherwise. The package demonstrates the student’s ability to engage in a lifetime of researching and writing. And that is a big part of what is being evaluated, or so I hope. The dissertation also serves as the students first three publications. One of my committee members, who was always came up with excellent words of wisdom throughout my studies, said that in his experience those who struggled to obtain tenure were those who were unable or not willing to get all papers in their dissertation published. The job market paper matters to get the job, the dissertation matters to get tenure.

And this all relates to who is contributing what to the dissertation. As we all know, co-authorship can vary from the student being more than just a glorified RA to the student being the main author. Distinguishing where the student falls on the scale is vital to the external examiner being able to do the job they are being asked to do (for free I might add – thank you to you Frances and to all the others out there who agree to take on this task). It is also an important task for the student since it will outline the roles and tasks they will have to take on as they move into a tenured position and start doing research on their own.

But I think there is also more to that. I must admit to being one of the individuals who requires my students to include a detailed statement of co-authorship in their dissertations. My first PhD student, with whom you are familiar, did the vast majority of the work, made a lot of crucial decisions, and wrote the majority of the dissertation. While the idea may have originated with me, the execution of that idea was a partnership vastly weighted on the students end. I was proud of the role the student played and it was important for the student to take the time to articulate that not only to themselves but also to committee, particularly the external examiner.

I recall not too long ago a process where a hiring committee was trying to narrow down their selection to the 3 candidates for a fly back. During this process, it became unclear if their job market paper was coauthored or sole-authored and the candidate was not overly forthcoming. A bit of leg worked turned up that the paper was coauthored with another student who was also on the job market who was also presenting the paper as though it were sole authored (there was not clear indication that it was coauthored). Needless to say the fact that the applicant was not forthcoming in what would normally be a trivial fact resulted in the candidate being screened out. I wonder if there was not more stress on documenting co-authorship of the dissertation if such a situation would have been avoided.

I fully agree with you and am behind you as you lead the charge.

[Frances Woolley]

Lindsay, thanks so much for your comments. Great news that you and your students find the blog useful.

A reaction... Your story about the student whose unacknowledged co-authorship cost him/her a flyback is fascinating. One thing I think students don't realize when they first hit the job market is how small the academic community is, especially the Canadian academic community. What goes around comes around - and often much more quickly than one expects.

And a question... It sounds as if that excellent detailed statement of co-authorship that I saw was your idea. You've been associated with several universities - and none of them require or suggest including a statement of co-authorship?

[Lindsay]

Frances

I did my dissertation at McMaster but none of my papers were coauthored so I did not come across it there. Here at Uvic, I do not see anything in the guidelines that address this issue.

It was when I was at U of Manitoba where it came up and on my first supervisory experience. There the guidelines when writing a sandwich thesis are "The nature and extent of the student’s contribution to each of the papers or articles must be specified. This is particularly important when the papers or articles are multi-authored." Why it just applies to sandwich theses is not clear to me.

When I followed up with more senior faculty at UofM to get a better feel for what was expected to adhere to the guideline, it appeared as though this guideline as not really followed and if so, the statement boiled down to something like "paper 1 and 3 were coauthored with C and paper 2 was sole authored." I did not think that a simple statement like that was in the spirit of the guidelines and opted for something more detailed (the one which you saw). It really is what "felt" right to me. Now it is a part of what I require in the thesis document.

[Daley]

UofT School of Graduate Studies has a note about co-authored papers as part of a thesis:

Joint (co-authored) papers may be included as part of your thesis if intellectual property issues are worked out in advance and approval is sought from your graduate unit. In all cases of joint publication, there should be a statement in the thesis explaining the nature of the collaboration and the contribution of the thesis author.

I suppose that "should" does not mean there must be a statement because I have read a few that explicitly state chapters are published/in press, but do not clarify the writer's contribution.

[Joseph]

In Epidemiology we would never have PhD students if co-authored work was not permitted (or at least PhD topics would become a lot more minor). On the other hand for my PhD, I had to submit a very detailed statement of authorship (section by section) and this required a fairly extensive justification of my contribution (including at my defense).

[Just visiting from Macleans]

FW: I thought you might enjoy this column (Doug Saunders, G&M tweeted the link)

The Shadow Scholar
The man who writes your students' papers tells his story

http://chronicle.com/article/The-Shadow-Scholar/125329/