There is one thing that a PhD student fears more than their research failing, exploding or becoming sentient and that is the oncoming dread of having to write up several years of hard work into a thesis at the end of their project. In the UK, where I’m based, a PhD thesis typically consists of 40,000 to 100,000 words describing the research area, methods and results. I’m approaching the end of writing mine and it’s been a particularly gruelling experience.
I’m writing this blog to consider if there is an alternative way to assess PhD students after completing their research. I’ve spoken to a lot of academics about thesis writing and I’ve heard it described as outdated and Victorian but also as the only effective way to assess how much work a student has done and a necessary evil that makes achieving a doctorate suitably challenging. Once written and submitted the thesis is typically assessed by an interview with examiners, known as a viva. I’ve attached two poll questions to this blog post as I thought it would be interesting to see where people’s opinions lie and whether a thesis should be viewed as the gold standard of PhD assessment or whether alternatives should be considered.
Being in the middle of the writing process my opinion is somewhat biased but I feel the thesis is unnecessary and somewhat damaging to PhD students in the 21st century. The career market for new doctors is now highly competitive and some of the best weapons you can have in your arsenal while job hunting are published papers. Published work demonstrates an ability to research novel ideas and communicate them effectively. Job hunting itself is often a stressful and demanding experience and it can be hard to find the energy when you’re writing 1,000+ words for your thesis every day.
Published papers that have been through the hurdles of peer review are far more useful to science than a PhD thesis as the latter has not been as rigorously examined. If a PhD student could dedicate more time to paper writing I feel it would lead to more publications and important data reaching a wider audience. Like many things in academia I feel we are living in a system that is outdated and hurts both PhD students and the communication of scientific research. Though many institutions are considering alternatives, for now thesis writing remains the standard method to assess a PhD student. In my opinion a system that is more suited to the nature of modern academic life needs to be brought in as soon as possible. Getting a doctorate should still be an immense challenge but the output should be papers and presentations that benefit science and the student and not a lengthy tome that exists merely to pass an examination
Fourth rock from the Sun 
Im a PhD Student at Karolinska Institutet (Stockholm, Sweden) and here our doctoral thesis consists of our published papers (approx 4) with a short introductory chapter. It makes so much more sense to me, than the UK style! Good Luck though!
I’m rather jealous ha ha
This is an interesting question. I think that the thesis as a way of demonstrating very detailed, very specific knowledge about a particular topic is useful, but only as a means to explain that knowledge so that it can be useful to others. If you write a huge thesis only because that’s what’s expected and it doesn’t make sense for the research/work, then it’s a waste of time. Developing writing skills remains important for researchers, but maybe content from the thesis can also be used in papers along the way?
I don’t think there’s a reason why paper- and thesis-writing should be separate. My thesis was composed of slightly altered versions of my published papers. It took maybe a month to write a general methods section including details the papers didn’t have, and a general discussion to link them all together. Still a “thesis”, still focused on publishing and papers. I have heard of institutions (and supervisors) that don’t like papers being used as chapters without a lot of alteration, though – that definitely seems outdated to me. But generally, I think learning to write cohesively is one of the most underrated skills you *should* pick up during your PhD – it would be a shame to abandon that entirely.
It’s a good question, and even though most theses/dissertations consist of what gets published, there is a mountain of other data, other experiments, that perhaps wasn’t publication ready or didn’t quite fit the story of a manuscript submitted for publication. The formality of a printed and bound hard-cover thesis (to sit on a shelf, never to be read) could probably be relaxed. And as a measure of “are you ready to graduate” likely they are redundant at best. But for our lab (microbiology/immunology) theses are in-house resources for compiling techniques and unfinished threads of experiments, should someone down the road want to pick them up again.
While I think that the traditional thesis is outdated, there may be many reasons why it is not possible to submit published papers instead. For instance, if your research is held under a confidentiality agreement you may not be able to publish the work before you finish your PhD. I think the format of a PhD thesis should be in the form of written up papers instead of the standard format, but I don’t think you should be required to publish them or even to submit them to journals before submitting to an examiner.
It’s a good point but I wonder what proportion of PhD students are subject to confidentiality agreements. In those cases I agree it would make sense to have a thesis of written up (but not submitted) papers.
The number of papers you can get out of a PhD, and how easy it is, varies a lot by discipline and even by project. I got my name on four papers (one as first author) by the end of my PhD, which is fairly unusual in my field, because I worked in a tight-knit research group and was lucky. Other people I know in my department haven’t got any papers, because they worked pretty much on their own.
I agree that paper-writing is a better use of a student’s time, but I think assessment needs to be based on other criteria as well to level the playing field a bit.
Interesting and debatable questions. In my opinion, a PhD thesis serves to both evaluate the writing skills of the candidate (particulatly on compiling/showing that (s)he accumulated lots of information — intelectual evaluation) and allow particular ideas and speculations about the field to be recorded. Otherwise the “technocratic” scientific trend would be established, which is not good. A good solution would be to have a worldwide centralized thesis depository, with online search engines for key words, so that the written info would’nt be lost in insulated university depositories. Concerning the short time to struggle with the job search, I believe that the system is currently indeed stressful and a mandatory additional period of 3 months could be formally established for both writing the thesis and looking for a job.
I am currently working toward my PhD in mathematics in the US. I am still required to submit and defend (basically option 2 of your alternative methods only with my entire dissertation) a dissertation in order to receive my degree.
I understand your desire to have PhDs awarded based upon published works, but I disagree with its practicality. Many math PhDs have published at least one, but usually multiple papers stemming from their dissertation. However, given the length of time much mathematical research takes in order to prove a new result, requiring the publication of multiple papers in peer-reviewed journals before receiving a PhD is, in my opinion, a waste of time and funds. Yes, publishing results is imperative, but the process of getting even just one paper published can take years, let alone multiple. Also, a majority of math graduate students who succeed in attaining their PhDs are funded either by fellowships or assistantship positions so requiring publishing that takes more time is an inefficient use of those funds when a person could move on and work on publishing his or her work once the degree is attained, allowing more funding for those beginning their degrees. I feel as if requiring the publication (when really it usually happens afterwards anyway and is arguably more stressful than writing/defending a dissertation) will just draw out the process simply to produce more frustrated and tired PhDs rather than those who want to go forth and continue doing wonderful work.
I agree, I think a pressure to publish too much too soon would just as stressful as thesis writing if not more so. When I was phrasing that optionI deliberately phrased it as manuscript rather than paper but may have gone too far with saying they should be submitted.
All of the comments I have received have been very thought provoking and I feel a good balance may be a chapter 1 that reviews the research area and provides an overview for the work but subsequent chapters are predominantly composed of manuscripts/drafts rather than a single body of work.
While I agree that a thesis-based evaluation of PhD work isn’t always the right way to go, I have issues about framing this as a “21st century problem.” I think the thesis model should be questioned when it comes to experimental methods, projects that highlight and depend on public engagement, and any other situations where expository writing is not the suitable medium to represent a doctoral work. I think replacing a dissertation with publications regardless of the project, just to meet the standards of the current academic climate, runs the risk of reinforcing the product-oriented approach to knowledge-making in academic publishing today. The open source trend in journals has opened up a whole new set of barriers for authors: you need to pay hundreds to thousands of dollars to have your work published in these “minor” journals. In the current academic job climate in which there are so many PhDs stranded outside of university departments, with no access to grant funding, this trend can be dangerous. And indeed, there are more and more of these popping up. I work between borders of disciplines in the humanities and the social sciences (both theoretical and applied), and the pressure to publish – not to put your ideas out there, but to demonstrate to funders and employers that you are (economically) productive – is extremely high. THIS is the academic climate that I see… and while I don’t agree with defending the dissertation on the basis of tradition, I do want to be careful about the implications of some of the alternatives presented. I also agree with the points presented that the dissertation is a valuable exercise in thinking, reading and writing. We need to be careful about the role and function of academic activities, including the dissertation. We need to think beyond the product-oriented notion of knowledge-making, because knowledge isn’t just an article in peer reviewed journal or a book by an academic publisher. If we think in terms of these equations, we are headed down a very dangerous road. I know we all lament the dust collecting on our bounded theses in the dark corners of university libraries, but if we are talking about readership and impact on the world, an article in one of tens of thousands of peer-reviewed journals isn’t much better. Honestly, how many of us will produce an article that will be read by thousands of people? I think there can be support for people working on their PhDs to publish during their doctoral studies in order to help with their careers, but this is a separate thing from restructuring the dissertation itself. I’m not against a rethinking of the dissertation. But I think we need to be careful about the reasons for rethinking it and implications of its alternatives.
I think the standardisation of the thesis-format across disciplines is rather unfortunate. It makes sense in the humanities, where the monograph is still the blue-ribbon research-output, but seems a bit off-kilter in the sciences (and indeed in the social sciences), which revolve around journal-articles. A bigger issue is the assessment system: examination by two academics chosen by the PhD’s supervisor. Both relative seniority of supervisor and examiner and favours owed can become major factors, so that often obtaining permission to submit from the supervisor is a much bigger hurdle than the viva. Quality-control varies wildly from supervisor to supervisor. I’m not that this issue is practically fixable. Double-blind peer review increases the chance of (e.g.) a reviewer being a dick and rejecting something because it disagrees with (or fails to cite enough of) their own work or collides in some trivial way with one of their pet hobby-horses. That’s fine with peer-review for journals, because it’s not a one-shot system, but would be a nightmare for PhDs.
I could write my own comment, but I am simply going to copy and paste what kjmacleod said:
“I don’t think there’s a reason why paper- and thesis-writing should be separate. My thesis was composed of slightly altered versions of my published papers. It took maybe a month to write a general methods section including details the papers didn’t have, and a general discussion to link them all together. Still a “thesis”, still focused on publishing and papers. I have heard of institutions (and supervisors) that don’t like papers being used as chapters without a lot of alteration, though – that definitely seems outdated to me. But generally, I think learning to write cohesively is one of the most underrated skills you *should* pick up during your PhD – it would be a shame to abandon that entirely.”
Exactly that. The dissertation should simply be an accumulation of high quality papers. That’s how it is in the Netherlands. I am surprised this is apparently really different in other countries?
In that regard, perhaps you should investigate *in which country* the dissertation process may or may not be sub optimal. For me the situation is that I get paid, have a lot of freedom, can do cool research, choose my own (sub)topics and study design, etc. What’s not to like?