4

I just don't understand why the overall findings & implications of a PhD cannot be communicated in the style of a research paper?

Succinctness is a virtue in a field characterised by time pressure.

The research output is equal to or less than that of a 3 year postdoc - which is always communicated as a research paper.

Just seems like added embellishment for the sake of embellishment.

What am I missing?

  • 20
    What am I missing? A lot actually. The length of a PhD dissertation varies greatly by field, university, and department. My dissertation had 5 chapter: an Intro, 3 manuscripts chapters, and a conclusion. Also, during my postdoc I produced far more than a single publication. You're making field specific assumptions and generalizing them – Richard Erickson Jul 22 at 18:14
  • 4
    The nature of a PhD thesis varies wildly from field to field and also country to country. Please make your question more specific. – Thomas supports Monica Jul 22 at 18:23
  • Systems Biology in the UK – Asymptotic Tri Jul 22 at 19:06
  • 1
    Out of curiosity, I looked for the largest digital dissertation file I have a copy of (something I could do easily and quickly), and David E. Rowe's dissertation Felix Klein, David Hilbert, and the Göttingen Mathematical Tradition is xxviii + 637 [= 665] pages (26.56 MB). Here's one I missed a few moments ago that is v + 632 pages (29.31 MB). – Dave L Renfro Jul 22 at 21:20
  • 3
    Quick answer: Because they are single column, double spaced with big font and big margins – axsvl77 Jul 23 at 2:27
13

I don't know where your 150 page idea comes from since I have never heard this as a guideline. However, in my field, "staple theses" are common, composed of roughly 3 papers stapled together (either already published or publishable drafts), with an added introduction and conclusion that tie the works together and may get a bit deeper into background than is acceptable in a published manuscript.

Therefore, they must be at least the length of three research papers, and aren't typically all that much longer than that. However, theses are also typically formatted in a longer form, more similar to the original word-processed manuscript, whereas journal articles are composed in a dense, newspaper-like format. Figures may have their own pages. In sum, a manuscript that is 10-15 published pages can easily be 30-50 pages in this format.

No embellishment necessary to get to 150 pages.

I'd also add that a post doc in my field who only produced 1 paper in 3 years would probably have their next job in industry; a graduate student with only one paper will have been quite disappointing (it may happen that only one paper is publishable due to being scooped or unexpected failures, but even those outcomes can be part of a thesis). Not necessarily, it's possible that paper could be highly impactful, and might be reasonable in some subfields, but generally that would not be a suitable output for an academic career.


Edit: decided to pull up my own thesis from the archives. 230 pages. Of those, 53 pages are "thesis-only" pages; the other 177 pages are verbatim copies from 2 published papers and a third draft manuscript which was later revised and published, all reformatted to fit the thesis formatting requirements. Of the remaining 53 pages of "fluff," 6 pages are title/contents/acknowledgements/abstract, and about 19 are references for the introduction/conclusion chapters, so about 28 extra pages of generously spaced writing, to be treated as "embellishment" if you wish, or alternatively, to provide enough background and context for the work for someone familiar with the field but not previously familiar with my research area, such as members of my thesis committee besides my advisor.

In summary, even if my thesis had nothing but 3 published/publishable manuscripts, it would already be >150 pages.

  • I recognise that PhD theses vary in length; however in general theses have a substantially higher word count than a research publication. – Asymptotic Tri Jul 22 at 19:14
  • Number of papers as a marker of progress is annoying too. It is possible to publish 100 papers without changing the world in any meaningful way. – Asymptotic Tri Jul 22 at 19:16
  • 3
    @AsymptoticTri Yes, they do, and my answer explains why, at least in my area, they would typically have at least 3 times as many words as a research publication, plus a few extra pages of context. I am also in biology. – Bryan Krause Jul 22 at 19:16
  • 12
    It is possible to publish 100 papers without changing the world in any meaningful way — If changing the world in a meaningful way were a requirement for a successful PhD thesis, there would be very very very few PhDs, almost all awarded posthumously. – JeffE Jul 22 at 20:46
9

Succinctness may be a virtue, but page limits means that papers are often overly terse. They also commonly leave out important information, helpful steps, and additional but less eye catching results. The presentation and style of a paper is also often geared towards experts in a field, while you have more freedom to be pedagogical and explicit in a dissertation. Factor in a double-spaced one-column format and some front matter, and it's easy to get to 150 pages and beyond.

In fact, one of my papers during the PhD (physics) was five pages long when published, plus references (not counted in page limit). Six pages of supplemental material was deposited along with the paper, for a total of 11 pages. The corresponding chapter in the dissertation is 46 pages long, with some details still relegated to appendices... The main difference in presentation, however, isn't the length or layout - it's that I attempted to make the description more self-contained. A new graduate student may actually get a reasonable idea of the background and calculations from reading this chapter and the introduction. In contrast, there's no way the same grad student could get all details from the paper and be able to reproduce the calculations without following a bunch of references, well unless they already were an expert. And if you do read those papers, the amount of material to go through would quickly exceed 150 pages anyway.

Yes, many dissertations could be made more compact, but why? The dissertation is a rare opportunity where you don't have to compromise style or contents. It's an opportunity to provide an alternate description than already exists in preprints or published papers that can be useful to another audience. It's also a place where you can make deposit certain results, derivations, proofs or procedures that don't fit elsewhere in explicit form for later use by yourself as well as others. So if you want to write a shorter dissertation yourself, that's alright, but avoid judging others for trying to be more pedagogical than the journal paper form allows.

  • [quote]there's no way the same grad student could get all details from the paper and be able to reproduce the calculations without following a bunch of references, well unless they already were an expert[/quote] Thanks Anyon, I found that explanation helpful. – Asymptotic Tri Jul 24 at 11:19
  • @AsymptoticTri I'm happy it helped. – Anyon Jul 24 at 12:36
8

Your question includes a number of assumptions, which are not universal and may be simply wrong for some areas.

  1. You probably know that John Nash had one of the shortest PhD theses with only 26 pages and 2 references. PhD theses of 300+ pages are also not unusual, particularly in humanities.
  2. I would expect a postdoctoral researcher in Mathematics to participate / contribute to at least 1 paper per year. I know of colleagues in Engineering / Computer Science for whom this expectation seems very low - a postdoctoral researcher in their group would put their name on 3-5 outputs per year.

Number of pages, as well as the number of outputs, is an very poor metric of research effort and performance. Some universities may put an upper boundary on the number of pages to make sure that their supervisors are not overwhelmed with the amount of material they have to read and comment on during the supervision (after all, Universities want their staff to do a lot of work apart of supervising the particular student). I don't think that the lower boundary is very usual.

Having said this, there is definitely an expectation that a PhD theses should contain a detailed introduction and thorough literature review. After all, a PhD student should study the area and then train themselves to become a professional researcher in this area. In contrast, a postdoctoral researcher is already expected to be trained and their work is only to do novel research. For a PhD student, a lot of emphasis is on studying the methodology, and writing about it takes time and a significant number of pages.

Finally, compare the number of references: a typical journal paper could have 30, while a PhD theses can easily have 300. This clearly shows that the scope of a PhD theses is much wider than the scope of an academic research paper, while a paper may be more focused and deep.

  • 7
    @AsymptoticTri The content in the intro and lit review of a paper is nowhere near as detailed as those in a thesis should be. That's the whole point: the candidate needs to be able to show that they understand the breadth of the field and how their research topic fits into that, as well as give an introduction to the broad picture so that a non-expert could understand the thesis. Instead, papers show the depth of knowledge in a very narrow area. – astronat Jul 22 at 19:51
  • 5
    Successful PhD student aren't merely "trained to become" professional researchers. They are professional researchers. – JeffE Jul 22 at 20:45
  • 1
    @JeffE From the moment when they start their program? --- no, not necessarily. When they successfully complete it? --- yes, sure. – Dmitry Savostyanov Jul 23 at 17:09
  • 1
    @DmitrySavostyanov By the time they become successful. That's what success means. – JeffE Jul 24 at 0:03
  • 2
    @AsymptoticTri Because, as many answers have already stated, their PhD work is usually more than can be effectively communicated in a single paper. Or perhaps more to the point, because a single paper is usually considered insufficient evidence of being a successful researcher. – JeffE Jul 24 at 0:05

Not the answer you're looking for? Browse other questions tagged or ask your own question.