Poll: can a thesis or dissertation comprise multiple unrelated projects?

Traditionally, a thesis or dissertation comprises a unified body of work. The chapters are like chapters of a book or monograph–each tells part of a single story.

But these days it’s fairly common for a thesis or dissertation in ecology (and probably many other scientific fields) to comprise multiple projects that are only loosely related. My own dissertation comprised two chapters that were part of one integrated project,* and a third chapter that was originally a side project.** I still wrote “bookend” chapters that presented the whole thing as an integrated body of work, and my supervisor and committee were fine with it.*** But honestly, I didn’t see my dissertation that way, and still don’t. Don’t get me wrong, I think it was a useful exercise to have to write the bookends. They forced me to think about the big picture of my developing “research program” and identify the common threads that tie together all the work that I do. But a research program comprising separate projects linked by common threads isn’t nearly as unified as a traditional dissertation. And so those bookend chapters felt like a bit of a fib.

Can a thesis or dissertation comprise multiple unrelated projects? I ask because, once you’re prepared to allow theses and dissertations that are “unified” only in the sense that mine was, it’s not clear where you stop or why. If all you need is some sort of common thread, well, those aren’t that difficult to come up with. For instance, Meg’s talked about how for a while she had two totally separate projects going, which she called “thesis 1” and “thesis 2”. But had she wanted to, and had her committee been ok with it, I’m sure she could’ve written them both up as a single thesis, unified only in the sense that mine was. Heck, they were both about Daphnia, it wouldn’t have been that hard.

To be clear, I and most everyone whose views I know is fine with theses that are only unified in the loose sense that mine was. But I can imagine that not everyone is fine with this, feeling that something important is lost if a student can get a graduate degree without ever producing a large, tightly-integrated body of work. So as a conversation starter, here’s a little poll:

Looking forward to your comments.

p.s. After I wrote the above, it occurred to me that even if a thesis doesn’t necessarily have to be a single tightly-integrated project, arguably the proposal should still be for a single tightly-integrated project. That was the case for me, and for every case I know of. There’s a lot of value in being forced to develop a proposal for a master’s or PhD’s worth of tightly-integrated work. Even if you later end up writing a loosely-integrated thesis that’s somewhat different than what you proposed, because s**t happens.

*Those chapters eventually became Fox 2002 and Fox 2007. See if you can guess what Fox 2007 was originally about! (I ended up having to reframe it to address a completely different question in order to get it published.)

**That became Fox and Morin 2001.

***I wrote that my dissertation was about “causes and consequences of community structure”. The first two, tightly-integrated chapters were about the causes, and the third chapter was about the consequences. The “causes and consequences” framing let me present it as an integrated body of work.

23 thoughts on “Poll: can a thesis or dissertation comprise multiple unrelated projects?

  1. As a grad student, when I worried that the work I was doing might not be sufficiently unified, Tom Getty said he knew someone (or maybe knew of someone) who had a dissertation entitled something like “Some Aspects of the Biology of Organisms”. When my students are worried about if their dissertation is sufficiently unified, I always tell them that they’re far from needing a title like that one, and that they’ll be fine. (In all cases, their dissertations so far have focused on Daphnia-parasite interactions, and have been pretty conceptually unified, in my opinion.)

    That said, if it’s too disparate, it can become hard to sell yourself to search committees. So, even if your committee is okay with “Some Aspects of the Biology of Organisms”, there are other reasons to try to avoid that.

    • Yes, I’ve had two students whose theses were quite unified and one whose thesis wasn’t (because he injured his ankle and could no longer hike to his original field site). But even for that latter student, everything in his thesis had something to do with plant-pollinator interactions. You’re right that it’s pretty rare for anyone to need a title like “Aspects of the Biology of Organisms”. Which is fortunate, because yeah, that that sort of thing is hard to sell to search committees. And to your supervisory committee, for that matter. Though its sellability to your supervisory committee likely depends on the details of your circumstances.

  2. My own PhD thesis was, I guess, very unified. All five articles/manuscripts were about tritrophic interactions between birds, caterpillars and trees. But now looking back, I feel that I was lucky that (most of) my experiments were successful. I know that is not the case every time and not for everyone. In biology nature happens. So I think there needs to be more space in how unified a thesis must be. It is better to have good manuscripts/articles in a thesis even though the topic in those is not close to each other. It is always good to plan a unified topic for the project but leave margin for changes.

  3. There’s a lot to be said for some diversity within a thesis; after all, a PhD is meant to be a training in doing research, so exposure to a range of approaches/techniques is a good thing in my view. In my own work I did some soil nutrient analysis as well as working with flowering time, pollinators, and seed predators.

  4. After my proposal #1 failed and then my proposal #2 failed, my dissertation became 4 completely unrelated chapters, all resulting from side projects. (Yes, I had to write an intro that claimed to tie them all together — multi-trophic communities and human influence on them, or something similar — but like Jeremy, pretty much a sham to satisfy the requirements of the grad school.)

    I feel really torn about the *should* of the poll. On one hand, I think I would have felt much more satisfied with my dissertation had it all been related. I would have personally preferred it, but I had to finish up because I was running out of funding after 6 years. It would certainly have been easier to write the chapters themselves, because I essentially had to know 4 different sets of literature to write my chapters. (This carries over into getting them published, too.) It was pretty much impossible to come up with a coherent story of “what do you study?” when talking to prospective post-doc advisors.

    On the other hand, my goal for my Ph.D. was to *learn* things. I could have easily put together a modeling dissertation with my comp sci background. But I wanted to learn. So, yes, I have one modeling chapter. But I also have one chapter on a field experiment I designed and carried out myself. Another on a long-term field experiment, with all the accompanying messiness. And another as an environmental economics analysis. Because of the variety of my topics, I learned a ton — not only about four different systems, but also about tools and methods. I guess you can say I learned broadly. But, I certainly did *not* come out of my dissertation program as the single most knowledgeable expert in a tiny part of science, which I think is the standard model. Other pluses include the broad professional network I built, since each chapter had its own experts that I consulted with. And never being “stuck” — if I was waiting on model simulations to run for three days, I could easily work on another chapter, since my chapters didn’t depend on one another. And I’ve seen connections between methods and between disciplines that I don’t think are commonly recognized when everyone silos themselves into tiny pockets of science.

    And then: none of my four chapters led to me getting my postdoc! In fact, it was a(nother) side project that landed me that position.

    So then “should” to me depends on one’s goals. If one wants to learn broadly, then using the dissertation period to do so makes a lot of sense to me. If one wants to get a faculty position as quickly and efficiently as possible (and there are very good reasons to do so), then I think the traditional tightly related dissertation is what one should aim for.

    I also think the culture should change to better recognize that a lot of science these days needs to be done in teams, with *some* of the participants domain experts, but others who can talk across disciplines (and subdisciplines), address problems with a diverse toolbox of approaches, and appreciate the challenges faced by domain experts without having to be one.

  5. When I was writing up, and stressing about how little my chapters had to do with each other, I’d always look at Alan Hasting’s thesis title: “Some models in population biology”. I figured he seems to have turned out well, so creating a single monolithic whole couldn’t be that vital.

    I agree with Jeremy and Margaret here; having a single cohesive proposal is important, but much less so than for the thesis. Part of it I think is that often parts of the proposal turn out to be not as interesting or insightful to work on as they first appeared, and it ends up being better to focus your limited attention on a few really interesting (but disparate) projects, rather than filling in smaller holes of a single big cohesive unit.

  6. I suspect the answer to this question of whether you *can* do this varies a lot by institution, even if the question of whether you *should* can be answered more broadly and with less attention to the institution. At Wisconsin, my PhD was a series of only loosely connected projects. At McGill, my students are absolutely required to have a single, strong theme that unites the thesis. Although I was annoyed by this at first, I now think it makes sense to have a more closely themed thesis, especially in Canada where basic NSERC funding is based on having a research ‘program’ (not a research ‘project’). The thesis, and statements about the ‘contribution to knowledge’ of that thesis then becomes the student’s first chance to practice declaring their research program.

  7. Voting so far: 173 votes. 57% say it’s fine not to have a tightly-integrated thesis as long as there are some common threads . 20% each for “yes, ideally it should be tightly integrated, but there are exceptions” and “no, it’s fine to have a bunch of unrelated projects”. Only 2% said yes, a thesis should be a tightly-integrated body of work. That’s about what I expected, though I thought we might have more than 2% of voters stumping for tightly-integrated theses.

    I voted for “no, as long as there are some common threads”, but there are rare situations (like Margaret Kosmala’s, above) where I’d be fine with a lack of common threads.

  8. You need to be careful when using asterisks both for notes and to blank out profanities. Unless of course you want to provide a subtle hint that Fox and Morin 2001 is the result of “s**t happens”.

    For the actual topic, I voted that unrelated works are fine, but I’m inbetween “unrelated” and “some common”. I’m a bit biased though, since my own thesis was a bit disjointed. Personally, I would maybe have liked a bit more cohesiveness, but its not a big deal. Its nice to see some diversity as well and it forces you to think about your projects in broader terms.

  9. If a PhD is training for academic research, then desperately attempting to throw scaffolding around five unrelated chapters was excellent preparation for writing selection criteria, explaining my research to new audiences, and writing both grant applications and the final reports of grants.

  10. When my project(s) started to diverge a bit, I asked one of my committee members about how I should tie my dissertation chapters together to form a cohesive unit he said (literally) “Get a stapler”. But I agree with Jeremy’s note about proposals – students should think they’re going to have a cohesive project, even if it doesn’t turn out that way, for good or bad reasons. One of my chapters was based on data I never could have hoped to get when I was writing my proposal, and it replaced a chapter that failed miserably.

  11. Thank you Jeremy for this post. As a PhD student, I started my project which eventually did not go well with the field work. I modified original plan and now includes 3 different case studies. In the beginning, I was a bit uncomfortable to join those case studies. I often discussed with my supervisor on what about of a PhD thesis-who suggested as majority agreed on this post, now I am more convinced that my case studies can be combined as a PhD thesis.

  12. A slightly different perspective on this.

    In many UK universities it’s now possible to obtain a “PhD by Publication”. This involves binding together a set of papers that have already been published (typically between 6 and 10, depending on length) and spending up to a year writing new introduction and conclusion chapters, guided by a supervisory team. It’s becoming popular with researchers who work in disciplines where obtaining a PhD has traditionally not been common, or who may be mid- to late-career, and have worked in non-academic public and privately funded research environments.

    In this case often the papers presented will have little in common other than the fact that they all relate to the same field. Nonetheless they provide evidence that the “student” can conduct independent research to a high standard.

    Is this route to a PhD available in other parts of the world?

      • It’s part of a process broadening the scope of doctoral education, including things like Professional Doctorates which include an element of taught classes.

      • I presume that part of the motivation is that it’s a moneymaker for universities? People are willing to pay more for certification than it costs the uni to provide it?

      • How very cynical of you Jeremy 🙂 But yes, that’s part of it of course (as with all students at all levels). It’s also the realisation that there are many routes up the research mountain and the traditional PhD route (which actually has not been “traditional” for very long compared to how long we’ve had universities) is only one way to do it.

      • In France this is an option too – in fact I believe it’s quite common in European universities. The PhD program is also only 3 years long in France, after that you are out. Quite different from the US programs, where I once knew someone who was in his 17th year of being a PhD student (he had lots and lots of side jobs 🙂 ). Some American students who come here have trouble accepting this limit, but the “école doctorales” are quite strict about enforcing it these days.

    • University of Chicago Physics used to allow a single-author paper to take the place of a thesis. Most people who did this also had several multiple-author papers along the way, but these did not serve a formal function in regards to their dissertation. Starting a couple years back, the administration phased this out of the program.

  13. Pingback: Friday links: Oregon Trail vs. peer review, the Mr. T test, and more | Dynamic Ecology

  14. Pingback: Linkfest: 9 November 2015 | Tea 'N' Mango Juice

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.