Friday, October 22, 2010

The final stages: An account of thesis submission and review

My PhD thesis was compiled one rushed November. I plonked together some papers with a linking page between each and wrote my introduction and discussion sections. “Don’t spend too long on it – only three people are ever going to read it”, intoned my advisor. “Keep it simple, don’t waffle on. Your papers are what count.”

Waffle I did. This was my Magnum Opus, was it not? As far as theses go it was still on the short side, but not as succinct as it could have been.

It was submitted in December. I knew it wasn’t the most carefully pruned piece of writing in history, but I was proud of my achievements. Now a move overseas. The thesis was gratefully out of mind.

It took until the end of May to receive my reviews. One assessor liked it, thought it could have been more fleshed-out, but was happy with it overall. The other assessor, a preeminent scientist in my field (assessed 50 theses, supervised 23)… hated it. Dismay. Some choice quotes from the review include:

“It wanders, it addresses important points then slides into other things. It’s a mess.”

“In conclusion, the thesis is poorly organized and self-contradictory.”


Perhaps arrogantly, this was not what I had been expecting. The positives noted were for “a substantial body of innovative experiments… that lead to provocative results”. However, upon reading the assessor’s comments, it became clear that his reading was… less than complete. He pointed out that I had not done an important control experiment, rendering some of my main conclusions invalid. However, I had indeed done this very experiment. More broadly, it seemed like this assessor had missed the central point of what I was trying to say. He saw contradictions where I did not, and seemed to miss paragraphs of my discussion where (I feel) his concerns were addressed.

The lesson I’ve learned from all this is that, just as for journal articles, it is important for a thesis to be clear and concise. Don’t feel that your introduction and general discussion must provide some great new insight, or be an epic work of deep scholarship. Summarising what you did and saying what you think it means is sufficient (even if this feels on the short side).

It also helps to consider who your markers will be, and what academic system the likely markers work in. One of my assessors (the first) was Australian – hence the feeling towards a longer thesis and “more evidence of traditional scholarship”. The second assessor worked in the US. I am told that my 103-page thesis (including preambles, but with 26 of those pages formatted as two-column journal articles) is a little on the long side for a US audience – perhaps this is why it was only skimmed?

That said, don’t worry too much about the actual thesis write-up: if the chapters are all manuscripts (hopefully submitted / accepted), you can’t go far wrong. It’s worth noting here that even though the second reviewer believed I’d missed a crucial control experiment, his recommendation was still “pass with minor revisions”. At the end of the day, my thesis passed, and three people have read it (well, some at least). Focus on turning out a couple of manuscripts during your candidature and the actual thesis becomes a formality.

Tom Wallis


  1. Great post. I'm all about concise... End comment.

  2. I think the most difficult part in a thesis by publication is the last part – how to write an intro that’s different from the ones you already have in your papers? How to summarise the findings without becoming totally redundant? One advice from my supervisor that helped me a great deal was: Write it like a review paper.
    In the process, I came across 3 major problems, which resulted in 20-fold re-writing of my intro and GD. The first trap is describing extensively what’s already widely known: You can safely assume that your markers know the most important findings in your field of research, and deviating from the Honours Thesis, you don’t have to prove to your markers that you’re familiar with seminal papers. So try to avoid regurgitating well-known findings. Also, don’t fall into the trap of just summarising your thesis: The markers won’t be thrilled if they have to read the whole thing twice.
    The hallmark of a good review is that it sheds new light on findings, e.g., by providing a new interpretation of some findings, by pointing to inconsistencies in the literature, or by showing how the evidence conflicts with a widely accepted doctrine. Restrict your synopsis to issues where you have an interesting contribution to make.
    Finally, my greatest problem with the synopsis (which is however probably too idiosyncratic to be a real trap) was that I erroneously believed I had to come up with a wholly new theory or model: Fortunately, the complexity of the undertaking killed that madness. In the end, I just briefly described a collection of hypotheses that would render my results compatible with previous findings, and simply mentioned that and how these ideas were still incomplete. And I think that was a winner - for the thesis, and more importantly, for my own sanity. :-)

  3. I'm definitely going to book mark this page to come back to before I start writing my thesis.