Dis the Diss

What is the size/scope/style/length of PhD dissertations in your field? Do they have a lot of background material or focus on the student’s original contributions? Are they comprehensive (everything the student has done, even if fairly disjointed) or a coherent body of work?

A typical dissertation in my group is between 100 and 150 pages, with most 120-130, and this appears to be pretty common in my department. I like to keep them very lean, focusing only on what the student has done and published as first author or as the lead theory student within a collaboration with experimentalists; basically, a dissertation is very close to papers “stapled” together. There is a common introduction and a common conclusion/outlook, but the actual chapters inside are the students’ original work. If a student has published well, writing the dissertation should be quick and painless IMHO.

Also, I like the dissertation to be coherent, i.e., basically the largest interconnected body of work that the student has done and that can be placed under a common umbrella.  On occasion, I have a very productive student who has taken part in several very different projects and published well on all; in that case, we pick the project of which they feel the most ownership as the one to write the dissertation on, and while the satellite work doesn’t make it into the dissertation, the student definitely mentions it in the defense and even very briefly discusses the highlights before diving into the dissertation project.

Why do I ask?

A colleague from another department who will be on one of my student’s dissertation committee came to ask what the norms in my department were in terms of dissertation length. If it were me, I would have left it at that and waited to hear what the norms were, but the colleague had to add “because this wouldn’t pass muster in my department.” (Do dudes ever get told condescending stuff like that, seriously?) I got very ticked off, because this is one of my best and most versatile students, and is worth a PhD twice over. His dissertation work was extremely complicated and he did a great job, which he published in several papers, and then wrote up for his dissertation. He also did work on three completely different side projects, all with papers published, which did not make it into the dissertation.

The implication of “not passing muster” based on document length was really irritating. So I went and looked at some of the colleague’s group dissertations, and they were in the 300-400 page range. Maybe it’s the colleague’s subfield norms (e.g., dump all from lab notebook in dissertation) or there is a lot of filler text (since we are being condescending here), because I have been on committees for other folks in the colleague’s department and the dissertations look about the same as in mine.

So what say you, blogosphere? What is the size/scope/length of PhD dissertations in your field? Do they have a lot of background material or focus on the student’s original contributions? Are they comprehensive (everything the student has done, even if fairly disjointed) or a coherent body of work? (Please state your field.)

 

 

 

24 comments

  1. Biomedical Sciences. We tend to go for about 4 chapters of “data” papers (cushioned in between an introduction that could be review-style article and a general discussion). On average, 1-2 of those tend to be (derived from) first author publications (having 1-2 first author publications as the output from your PhD is more or less our quality control check). All chapters in a thesis are “journal article style”, so not everything from the lab notebook ends up in there unless there’s a beginning, middle and end to the substory.

  2. Physical Science. Dissertations in my department vary a lot, depending on the group and the sub-field. My dissertations sound a lot like yours. The meat of the dissertation is 3-4 chapters of published work or work in preparation for publication that has not quite been submitted yet, with some connecting text to kick off each chapter. There is an introduction chapter, which could be a review I suppose, that explains some background and mostly situates the work in the broader field. There is a conclusion chapter that discusses future directions and how it all ties together. My average is around 150 pages.

    All of my students so far have done work that didn’t make the dissertation for various reasons (doesn’t fit the narrative of the dissertation, didn’t work well enough, incomplete story, work not finished in time, minor contribution as a collaborator, etc). I have never heard of using page count to judge the quality/quantity of work. Does your colleague also use the stair method to grade for classes?

  3. Related question, which ties in to this issue, fellow academicians: is it mandatory that
    (a) the dissertation include all the raw research data (usually as an appendix/appendices), so that their data doesn’t have a chance of disappearing with the student etc etc etc., and in the lines of supplementary data that is now generally required with published papers in many journals, and/or by departmental/institutional policy: or
    (b) the raw data is NOT included (as appendices) with the dissertation, out of concern that others could “poach” it, and to ensure that all the doctoral candidate’s data remains in-house (since PhDs are generally public documents)?

  4. 100-200 pages seems pretty normal in the group where I did my degree. One introductory chapter with 3-4 chapters on the research, also with each chapter roughly being equivalent to a paper, and also with a short conclusion chapter. I would say one difference is that the group culture assumes that the most likely person to read the dissertation is a future group member intending to build off the work, so there tends to be more methods detail than a paper would typically include. A brief (paragraph or so) discussion of the blind alleys and pitfalls of a particularly challenging technique is common. Appendices including code are also common. In this group I think it’s more common for the dissertation to include disparate chapters for students whose work wanders around while the defense presentation is focused on a cohesive story. This usually still leaves work out of the dissertation, but I think there’s an effort to capture the breadth of work, usually for people who have a few false starts and end up completing a project that might seem smaller in scope.

  5. We have had a few students in my department (a physical science field) who themselves are researchers or professors at institutes in other countries, but with only Master’s degrees, who basically take sabbaticals (often funded by their governments) to go back and earn a doctorate (often required for them to keep their jobs or get promoted). (Usually, these are great people to have join one’s lab.) In some of those cases, it’s required by their bureaucracy back home that the dissertation be “old-fashioned,” in the form of a “book”, basically, and not in the form of papers stapled together with an introduction and conclusion chapter (which is the ideal for most of our traditional doctoral students).

    I have found that the “standard” length of a dissertation varies wildly by discipline, just as does the standard length of a published journal article.

  6. In bioinformatics and biomolecular engineering, the dissertations vary a lot, because of the very different cultures of the faculty from biology, computer science, statistics, …

    I generally see dissertations in the 100-300-page range. Some are very coherent, others are one project per chapter. I generally like a first chapter which is an overview of the goals of the thesis, a second chapter which is a detailed literature review of the specific topics (what has been done before), then as many chapters as are needed to cover the different aspects of the project.

    Data for bioinformatics is usually in the gigabyte to terabyte range, so is not included in the thesis, but I do expect to see an appendix that says where it can be found, preferably in a public repository. In biomolecular engineering theses, I expect to see any non-standard protocols detailed in appendices, but a lot of the bio facullty are really sloppy about letting their students just wave their hands at the methods.

  7. Economics phds from the top schools. 3 publishable papers (or two publishable papers if you don’t have a paper coauthored with an advisor). 1 publishable paper may pass muster from lower ranked schools or if the person promises not to try to get an academic job (these will sometimes have chapters that are lit reviews, but that doesn’t happen at the top schools which are more about efficient use of committee members’ time). Length is not a concern. In my particular subfield they probably range from 90 to 240 pages, but I’m pretty sure theory papers can be a lot shorter.

  8. One of the purposes of a thesis is to memorialize in one place all the details that anybunny in the group following on will need to know, Thus simply to replay the published papers does not do the job. This argues for a great deal of detail, but 300 pages is crazy.

    Back in the days of typewriters 80-100 was more than enough

  9. Let me guess … this condescending comment is typical of what science faculty make against their engineering colleagues. I have been on both sides of the aisle and I can tell you this kind of comment comes mostly from those who are intellectually challenged. Don’t worry and shrug off the arrogance!

  10. @Eli, my thesis was barely out of the typewriter era (1983) and was 167 pages. That was about typical for an engineering thesis in those days at Stanford.

    @physicsprofessor, engineering and science faculty are equally likely to be convinced that only their discipline is sufficiently rigorous and upholding academic quality. Engineering faculty on average may be slightly more interested in thorough documentation than science faculty, but the distributions overlap enormously.

  11. Physical geography. I would guess in the 200 page range, but our papers tend to be in the 10-15 page range, so formatted back into manuscript format they’re closer to 50 pages each. I ask for 4 first author papers to graduate. Whether those actually follow a ‘theme’ or have to be creatively made into a theme after the fact depends on the path (i.e. funding) of the student.

  12. I’m in a neuroscience department in a technical university (in Europe, if that matters), and at least in our department the norm is thesis by publication: a minimum of 3 first author papers with some 40-80 pages of general introduction and discussion to tie the thesis together. If student happens to have some second author papers that are on the same topic, those might be also included in the thesis, but they are not required. And similarly to pyrope, some theses have a clear theme and some have a bit of creative writing in the introduction to bind the publications together into a whole.

  13. In Chemistry, in Israel. What you describe is exactly typical for what we had in my department.

    > Do dudes ever get told condescending stuff like that, seriously?
    A friend of mine just (barely) passed his Masters defense, with one of the committee members telling him something close to these lines (“where I come from, it won’t be considered a Masters. Is this all you’ve done?!”) in front of the entire committee. Both student and advisor defended the work, but the committee member was still sceptical. All involved were males.

  14. Also a physical scientist here. We definitely do a “staple thesis” in our sub-field: ~3 chapters of published or nearly submitted work, an intro that covers the basics for the physicists outside our discipline, and a conclusion to pull it all together. I also had some appendices on details that I had written out since it was helpful to me along the way, but weren’t things that are worth publishing (simple procedural things, equations I used, a few derivations, etc.). Came out to ~150 pages.

  15. Biological sciences here.
    Our PhD students’ dissertations are around 300 pages, and consist of comprehensive introductory and concluding chapters, and 2-4 middle chapters which are each typically a published paper. Sometimes all the papers are written by the student, but it is also quite common that some papers included the student’s research results were actually written by a postdoc or another student since we do a lot of long-term and collaborative projects. However at least one of the published chapters/papers in a PhD thesis must list the student as the first author, which in our field indicates they generated the bulk of the research results included in the paper and also wrote it. We require the first and last chapters to be included in the dissertation as a demonstration that the student has learned scientific writing, and can actually present a comprehensive literature review on their topic and discuss their overall hypothesis and results as a scientist should do. If these chapters are too short or poorly written, we will hold up the student’s degree until they rewrite them. Sometimes one of the middle chapters is a set of unpublished data generated by the student that hasn’t been included in a publication yet, but this is up to the student and mentor. The students’ lab mentor thus usually needs to spend a painful amount of time editing and advising the students’ writing of the dissertation–but a scientist who cannot write is essentially unemployable no matter how brilliant they may be otherwise, so we consider writing a suitable dissertation a critical part of our students’ training.

    Our MS students’ can choose between 2 formats for their dissertations: 1) writing a 6 page original research project grant proposal on their topic (does not need to actually be submitted for funding, but often is), or 2) writing a review/book chapter on their topic (does not need to be published, although sections often are).

  16. PhD dissertations are a pointless anachronism. In my opinion, the PhD should be awarded on the basis of publishing papers as a lead author. I tell my students to put as little effort into their dissertations beyond “stapling papers” and writing intro/conclusion as possible.

  17. A student of mine recently submitted their dissertation (Engineering discipline, R1). 125 pages including bibliography, ToC etc.. The committee had absolutely no issues with the length. There were some constructive suggestions from committee members on content, that the student incorporated prior to final approval. Only the student’s original (first author paper) work was included.

  18. Lone humanist here (lit studies). A diss is usually 200-300 pages, completely original material–usually conceived of as a unified book with a single argument though more recently could be a series of linked essays. A lot of conversation in the field about this, given market problems.

  19. My PhD thesis was around 250 pgs, in biomed. Contained an intro chapter (lit review), 3 data chapters that reflected my papers, but with way more methods, supporting data, and some fun speculations, and then a concluding chapter discussing the significance of the body of work I had produced in the context of the field, kind of like an uber discussion section.

    My little brother’s masters thesis in geology was 200 pgs.

  20. Very sad that November is over and we are not going to get daily posts any more. Your posts brightened up an otherwise depressing month.

  21. anon, thanks for reading! I can’t commit to daily posts indefinitely, but I will try to post more often than before. There should be one up coming later today.

  22. Dissertations in my field were about 100 ish pages. I think mine was too?

    I don’t know how these monster dissertations happen, if people are writing papers of reasonable size, surely it can’t be more than ~10-20 pages x the number of papers plus a little for intro etc?

  23. Biomedical sciences.

    280 pages including bibliography/appendices which means 200 pages really.

    I had a fairly comprehensive introduction (review) followed by 6 data chapters (1 on human data, 3 on cell data, 1 on short-term animal work and 1 on long-term animal studies). Didn’t include 3 ‘spin-off’ projects/papers which have been published separately as did not tie into main thrust/idea.

Leave a Reply

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 )

Google+ photo

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

Connecting to %s