Monograph or Sandwich Dissertation?
I have an unofficial rule that whenever I find myself explaining something to a student more than twice, I should write a blog post about it. Well, I’ve given my spiel on different kinds of dissertations far more than twice, so it is high time I collect my thoughts once and for all. So, what’s the deal with dissertations again? Here is my attempt at an answer, but, as always, YMMV.
The traditional form of the Ph.D. dissertation is the monograph: a single-authored long-form piece of writing on a specialized topic. In the olden days, when you were writing a dissertation, you were writing a monograph; essentially, a book on your research. You would spend your time writing this book locked away in your ivory tower (okay, okay, I admit that I am oversimplifying and stereotyping here), and only once you were done with it and had defended it would you think about how to publish it: either as a single book, or chopped up into articles.
In many fields, however, this practice has gone out of fashion in favor of collection dissertations: a dissertation made up of a collection of previously peer-reviewed and published articles. This form of dissertation mirrors the modern approach to science in many of the engineering, medical, and natural sciences where no Ph.D. student can afford to be locked away with their research until it is finished, but rather will publish it in article-sized chunks as they go along.
The primary difference between these two types of dissertations is that the monograph, by virtue of being a single and unified piece of work, is likely going to be more coherent and focused than the collection dissertation. In contrast, the collection dissertation consists of some more or less related articles stapled together and sandwiched between an introduction and a discussion and conclusion. For this reason, such dissertations are sometimes called “sandwich” or “stapler” theses.
I personally tend to think that you can go to a little more effort than this to reach a point halfway on the continuum between a monograph and a collection thesis. More specifically, even if your Ph.D. work is really a collection of articles, you can still make some effort to turn these articles into something resembling a monograph to create a more unified and satisfying reading experience. To achieve this, there are several things you should do:
- Use a unified dissertation format throughout your entire dissertation (in other words, don’t just use the disparate formats used for each of the different articles).
- Turn each of your articles into a chapter. (Sometimes you may want to chop up and move around material even more than this.)
- Create specific introduction, discussion, and conclusion chapters and accordingly reduce the corresponding material in each article chapter to minimize redundancy.
- Unify your references and your discussion of the related work into a single chapter and move all of the references to the back of the dissertation. Of course, you may still need to cite specialized literature in specific chapters.
- Move all of your appendices to the end of the dissertation.
- Replace all occurrences of “this paper” and “this article” with either “this dissertation” or “this chapter”.
Finally, one of the most common questions that my students ask me is how many articles need to go into the collection dissertation. In other words, when do you know that you are done (Ph.Done, as it were)? Obviously, this exact number depends on your discipline and institution. For computer science and informatics, the number is typically three for a Ph.D. and one for a masters’ thesis. All of the articles don’t have to be published, but the more articles have been accepted for publication, the less quality control has to be done by your committee for your defense. However, a more meaningful (but potentially more confusing) answer is that your dissertation should contain just the right amount of articles to answer your research questions: no more, and no less. Ideally, your work has a nice and symmetric shape to it with no major loose ends.
A final word on your dissertation: once you are done and your committee has signed off on your work, you will likely be asked to upload a final version for archival. Remember that dissertations are forever. Take an extra spell checking pass and think about future-proofing the writing. For example, if you very ambitiously submitted a paper to an aspirational journal or conference, you may want to leave that information out lest it become a frozen testament to your embarrassment of rejection for all time. (No, I’m not speaking from hard-won experience, why do you ask?)