Science, discussed.

Poll: Have dissertations outlived their usefulness in science?

Several friends of mine have been recently finishing their Ph.D. dissertations, which of course were the cause of much stress: they tend to be quite long (100+ pages usually) and cover most or all of the work one did in grad school, not to mention the symbolism of the dissertation as the physical embodiment of 5+ years of labor.  But underlying all the effort and angst that goes into a dissertation is a cold, hard truth: almost no one will read it.

Lenny Teytelman over at the Spectroscope recently addressed this reality, prefaced by his story of pulling a pretty fantastic prank to test his advisor’s scrutiny of his dissertation (of course his advisor failed).  I don’t have definitive proof, but I suspect no one on my committee read my dissertation in any more detail than maybe flipping through it — they certainly said not a word about it at my defense.  I know a few people whose advisor or one committee member did diligently read their dissertations, but anecdotally the majority of people I know didn’t get anything more than a cursory reading and few minor suggestions.

But besides negligent committees, I think the truly important issue here is whether writing a dissertation is a good use of a grad student’s time at all.  The fact is that almost all significant scientific work is published as papers in peer-reviewed journals, and these papers are not only the primary currency of a research career, but they are also the canonical scientific record that people will read and cite in the future.  Dissertations, on the other hand, are almost never cited in most areas of science I’m familiar with; I’ve only ever referred to other people’s dissertations a few times, and maybe cited one once.  This is of course compared to the thousands of times I have referred to or cited papers.  So if no one’s going to read or cite your dissertation, and it’s redundant with your papers anyway, is there really any point to writing one?

In the comments of the aforementioned blog, Nikolai Slavov noted that in some fields (namely certain areas of experimental physics), dissertations are actually read and cited since they often contain detailed descriptions of experimental or computational methods not available elsewhere.  I heard something similar from a friend, who said her research group used a former member’s dissertation as a standard reference for new members to learn a particular method used by the group.

Nevertheless my personal opinion on the matter is that dissertations are just a relic from a different era of science (and science publishing), and that senior grad students would be much better off spending that time doing more research, writing another paper (which might actually get read!), or maybe just basking in their newfound PhD-ness.  To the extent that some people or fields use dissertations to provide technical methodological documentation, I think it might be better anyway to put such material in standalone methods papers that are probably easier for readers to find anyway.

What do you think?  Have dissertations outlived their usefulness in science, or are they still an important part of the Ph.D.?  Please vote in the poll and leave a comment below!


12 comments on “Poll: Have dissertations outlived their usefulness in science?

  1. Andrew Hardaway

    My thesis was over 300 pages long and took a solid month to write and compile. Noone will ever read it at this point. My time would have been better spent compiling data and text into manuscript form. This is an archaic exercise that many attempt to justify by saying its part of the scholarship of getting a Ph.D. Maybe, but shouldn’t we be behaving scholarly all along?

  2. Christian

    Dear Michael,

    Thank you very much for writing this interesting opinion. I think the issue you raised is an important one.

    I would like to add that there may be lots of results within a PhD research time which for some reason cannot be published, e.g. experiments with negative outcome which are important (i.e. for not trying again) and are usually mentioned in “ordinary” theses, but much less in publications.

    Hence in my opinion, before omitting the thesis-writing, possibilities to e.g. online-depositing these results (maybe even in a publication-like manner) need to spread first.

    Best regards,


    • Michael Manhart

      Thanks for your comment, Christian, and yes, I totally agree that we need more options (or at least more incentives) for publishing negative results and methodological details!

  3. Melinda

    As a professor, I feel that dissertations are a critical part of both the students training (it is different to put together a large document than a small), and they are also critical for my lab as they provide continuity. Dissertations by students in my group are read by newer students in the group and I also look at them regularly. Writing is fundamentally a framework for data analysis and the dissertation writing focuses that. I also agree that this is a wonderful place to analyze work to date in partially complete studies (no grad student can finish everything) and it documents a student’s contributions to a work that will be completed by others which is valuable in assigning later authorship credit.
    As for length, honestly, they are no where near as long as graduate students like to imagine. They are double spaced, with huge margins and lots of images/figures usually. In a career, one will likely write many things that are actually longer in word count (even the typical book chapter can be longer). It is really important to know how to write larger documents and to not find the process intimidating going forward.

    • Michael Manhart

      Thanks for your comment, Melinda. I agree about the importance of writing in grad student training, especially the importance of writing long documents. However, for many people I know (including myself), most or all of the work in the dissertation had already been (or would soon be) published as individual papers, and so “writing” the dissertation was really just an exercise in reformatting and decorating existing material. So there wasn’t much new writing here. Of course, my perspective on this is from a physics background — I am sure it varies in different fields, which is something that will be interesting to learn from the poll.

      • Melinda

        Actually, I would think the dissertation would be even more critical in physics, a field where many publications in certain sub-fields have 10s or 100s of authors. Even in the life sciences, many papers have five or more authors so publications do not actually map easily to a graduate student’s accomplishments. Also, dissertations should have much more extensive literature reviews and methods sections, and can tolerate way more speculation than most papers in my field at least. Thus, it is not just making publications prettier, it is definitely a bunch of extra writing over and above publications, particularly since publications must be more terse for most fields. A doctoral student of mine who is defending soon wrote a data chapter, converted it to a paper, then had to cut the text by 1/3 because the journal did not take papers over a certain length. This took the student and I over a week of word smithing/rearranging to get this done. The cutting would be even more dramatic if the work was submitted to PNAS or even as a science report…..

  4. Ngayu

    Allowing presentation of publications (as suggested in the survey) instead of a thesis/dissertation is an interesting idea. It would change the focus of students’ research to work that will quickly yield publishable results. This may be no bad thing, but it may discourage more exploratory research where “sexy”, publishable results are less likely. And in the context of the well-publicised problems in scientific publishing (see “Trouble at the Lab” in the Economist), indoctrinating students in the “publish or perish” culture may not be necessarily be good for science in the long run, at least not until those problems are fixed.
    Also, I deeply envy the writer & his friends whose theses/dissertations weren’t read by examiners! In my viva, one examiner’s copy of my thesis was covered in dozens of little fluorescent tabs, each marking (I assumed) a reason to fail me. (Thankfully it wasn’t that bad.)

    • Michael Manhart

      Thanks for your comment, Ngayu. I agree there are problematic incentives with regard to publishing, but aren’t those already a problem, with or without dissertations? I don’t see abandoning traditional dissertations as exacerbating it. Interesting to hear that your dissertation *was* carefully scrutinized — perhaps we should’ve had a poll question on that as well…

      • Melinda

        I will say something about reading dissertations as well, over the past twenty years, I have graduated 10 Ph.D. students of my own and read and scrutinized at least five drafts of each dissertation. I have also been on the examining committee of about 40 others, all of which I read and marked typos and other broader improvements. In my experience, the vast majority of examiners read dissertations carefully and make comments on them. My program has even had a few students pass their defense but never graduate because they do not edit the dissertations, get the final signatures and turn in their paperwork showing that they really are incapable of taking a large project to its completion. I personally feel that someone who can not complete such a task does not deserve the degree…….

  5. Bàrbara

    I agree that many dissertations are too long an somehow redundant. However I also agree with some of the above comments that “dissertations are a critical part of students training”. In my field nowadays dissertations have acquired a new format: a compilation of the papers published (at least 2) form the core of the dissertation and the student only need to write a global introduction and discussion. Each paper is a chapter of the dissertation, and unpublished or submitted data ca form additional chapters.
    I think this is the way to go. It avoids the need to rewrite already published data. And it requires students to put their work in context. Specially important when some of the papers have been written with purposes slightly different than the thesis. This writing step is even more important when the published work may belong to large projects or consortia and maybe the student only wrote part of the manuscript.

    • Ben_Chaffey

      I think this is the ideal format, Barbara. The only issue comes with time-lines; following the completion of my lab work I had 12 months to write up my dissertation and submit it, otherwise I would be failed on a technicality (these were the rules of the University and funding body). As many of the critical results of my experiments were only generated in the last few months of my research, the resulting papers were still in the process of being published long after this deadline had passed.

  6. Mary Ellen

    I believe that dissertations in life sciences (and possibly other fields) are more critical today than ever. It is increasingly difficult to discern, from publications alone, whether the student has actually succeeded in four important objectives of doctoral education:

    a) identification of a significant research area
    b) identification of critical knowledge gaps that must be addressed to advance understanding in this area
    c) formulation of a scientific question(s) to address those knowledge gaps
    d) construction of an evidenced-based argument that integrate newly discovered knowledge into the intellectual framework of the research area

    Admittedly, a, b, and sometimes c are accomplished by the thesis mentor prior to trainee’s arrival in the lab. That makes it all the more important that we have some measure of the extent to which they have accomplished d. Reports in the primary literature and even reviews that bear the trainee’s name amongst the many authors are not, by themselves, a reliable indicator that the trainee has met these objectives.

    In the current era, where faculty members at research universities (the kinds that have doctoral students) are under pressure to meet enormously high standards for productivity and impact, and where journals have a preference for work of a scale and scope that can only be produced though coordinated efforts of teams, it can be very difficult to discern what an individual trainee has accomplished experimentally or analytically, let alone how independently she or he acted or the extent to which he or she can claim intellectual ownership of “her/his” work. One only sees this if the trainee is required to produce an integrated and original document that encompasses most or all of the work performed during graduate training.

    Those arguing that the thesis is unnecessary appear to consider the thesis a summary of the trainee’s work. It is not a summary, it is an argument. Or at least it should be. That some trainees and faculty members view the dissertation as a summary is of concern and should be addressed.

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s


This entry was posted on 15/07/2015 by in Science, Science and Society.


Creative Commons License
This blog is licensed under a Creative Commons Attribution 3.0 License.


Enter your email address to follow this blog and receive notifications of new posts by email.

Join 1,559 other followers

All Categories

%d bloggers like this: