Things I learn from my PhD

12.13 AM in Bangkok, I was suffering from jet lag as my body was wide awake although it was late at night here in the capital of Thailand. I was finding something to do to pass time (besides fixing my thesis which I will start tomorrow) and was stuck with some thoughts left in my mind from the viva (i.e., the PhD defence) …

They were the last two questions from the examiners:

“If you can do it [the PhD] again, what do you think you could do better?”

“What are the things you think you did best during your PhD?”

I did not prepare for them during the exam, but I did give some answers. Some of them were trivial, and one of them was something I personally felt.

Now, since I took a break after the exam, and currently comfortably stay at home and not in that distressing situation, I can think of more substantial things that I did regret, experienced, or was proud of during my doctoral journey.

And I think I better sum it up here as a note. Nonetheless, instead of answering the two questions separately, I will combine them into a list of “things I learn from my PhD”, both good or bad. Just to remind me to avoid the bads and to repeat the goods. Please note that it is nowhere near a comprehensive guide to a PhD. Just my two cents from my own experience. This list is not finished and I may add something later :-)

Here it is.

  1. TRANSPARENCY — I believe one of the most important traits of a researcher is being honest about what you find and report. And I’m still struggling to be as transparent as I can about my work. My PhD work is about software analysis, and I write code to analyse code (confusing, huh? :-)). I am not a very good programmer, but I can program and I love to. Many times due to the close-to-submission time constraints or lack of testing, I made mistakes, which resulted in incorrect results and another (or many other) reruns of the experiments. They were a few times that after I discovered the issues and was so tempted just to let it go. “Nobody would ever know this except you.” But what is the point of publishing erroneous results that other people might read and build upon? Isn’t it better to fix it now? This is one of many things I learn from Jens Krinke, my supervisor, and I really appreciate it. I am also glad that, every time, he carefully checked what I did and reported to him. Several times he spotted issues just by looking at the results I showed to him without asking how I did it. Luckily, I always received comforting responses from him such as “well, that happens” or “no worries. let’s do it again” And I found that really encouraging. At the end, I felt much better after going back to fix the issues even it took days or months. A prime example of this happened during the time I prepared for my viva. I and my supervisor discovered two issues in one of the experiments in Chapter 7 in my thesis. They affected the whole Chapter 7 and also partially affected Chapter 8. I got a chill over my spine when I discovered the issues and was reluctant to tell him the truth. What if the PhD examiners failed me because of this? But I did tell my supervisor eventually, and we decided to fix it. This resulted in fixing the bug, rerunning the whole experiments again, writing a new version of the two chapters (I really appreciate my supervisor’s understanding and support in all of this). I was asked to explain the issues during my viva, and we discussed them for more than an hour. Interestingly, even the examiners did not spot the issues. This resulted in a change to Chapter 7+8 by replacing them with the new results I got and I also still passed my viva. I might kept it with myself and possibly slipped away without anybody knowing about the issues, but I felt much better to accept it and did it the right way. However, spending the time to double (or triple) check the results and reduce mistakes are something I have to improve though …
  2. PRESENTATION — How useful is good work if nobody understands it? I am very fortunate to join CREST which emphasises a lot on how you present your work. Thanks to Mark Harman, who lead CREST during the first three years of my PhD, I learned a lot about how to write good papers and to make nice presentations. Mark himself was a very good example for us. All of his talks were great. They were easy to understand, and always with nice graphics. During our monthly group meeting, the PhD students and the research assistants had to prepare a single slide to explain what they did during the past month. We always had a cash prize of £10 (or £20 depending on which banknote Mark had in his wallet at the time 😂) for the “best slide” award. He hated bullet points, so you would never win if you had even a single bullet point on your slide. I found this practice very useful throughout my PhD. I went to a conference and found the talks that the slides were cramped with text really boring. Although the work itself might be interesting, I lost my attention very quickly. On the contrast, good talks were memorable and stuck in my mind much longer. They triggered me to search for the presenters’ websites and learned more about their work. Similarly, nicely written papers kept me reading longer and understood better. Anyway, after 4 years, I think I’ve improved quite a lot on presenting my work, but still has a lot more work to do especially in writing …
  3. REJECTION — I got several rejections during my PhD and did not have good publication records compared to many of my PhD peers. My first publication came in my second year (some of my friends got their first publication in a few months!). Two chapters of my thesis are still waiting to be published. It hurt every time I submitted my work to a conference or a journal, and it got rejected (although, sometimes, I kinda knew it had a high chance of being rejected). It also hurt when I read the reviewers’ harsh comments. At the very first time, I treated getting a publication as passing an exam. When I got rejected, that also meant I failed the exam; my work was not good; it was not interesting, and so on. This gets better over time (also thanks to a lot of positive comments I received from academic friends after whining about it on facebook :-). I am now trying not to expect much when submitting a paper and think of the publication as a repetitive process to improve the work. Interestingly, when I look back, every time a paper got rejected, it got better for the next submission. This is always true. The reviewers’ comments became valuable assets. With their careful scrutiny of my submissions and their discovered flaws, I kept improving the paper. So, rejection is normal. Turn it around and make a good use of it.
  4. PERSEVERANCE — I think this follows the previous point very nicely. During the course of my PhD, I kept trying and failing. Both in submitting a paper for publication or trying out new ideas. And it is one of a very few things I think I did best: keep pushing (it was my answer to the second question during the viva). Most of the times my first attempt at something would fail or had issues. There was one time I spent a month developing an idea just to completely threw it away because it didn’t work. I also spent a lot of time repeating the experiments to fix implementation issues or to address reviewers’ comments. Luckily, I never thought of giving up, possibly thanks to the family in Thailand. I knew that they were waiting for me to go back and I could not fail them. I just had to “keep calm and carry on.”
  5. LESS IS MORE — This is one of my personal problems. I tend to overly complicate things with tiny details. The very first paper I submitted was a mix of multiple experiments and they were not nicely combined into a coherent story. Surely it got rejected. We later split the paper in half and submitted the first half (with some improvements) to another conference and it was accepted. All the authors agreed that the second version, which was only a half of the original paper, was much better. It was easier to understand, with deeper analysis and gave a nice coherent story throughout the paper. After that, the paper lead to an invitation to a journal extension which we put the other half back in :-). Another paper I submitted got rejected and a review said we were “sausage making.” We did read the paper again after not touching it for a few months after the submission and also found that it was too complicated with too many details. We decided to simplify the methods and completely redid all the experiments. The simplified version looked much better (but still struggling to get accepted anyway :-)) So, if you can, keeping things as simple as possible is the key.
  6. PHD IS NOT THE END, IT’S THE START — When I started doing the PhD, I always thought of it as an ultimate goal, something to be achieved and, boom, done. Along the way I found that it was not true. The PhD is more about training yourself to become a good researcher and to equip required skillsets to conduct impactful research (this was a word of wisdom from my PhD friend DongGyun Han actually). Although some of the work that I (and other PhD students) did during the PhD might not get published, or did not attract citations, it was ok. The point is you’ve already improved along the way, and that is the most important. I believe the actual journey starts after you finish the PhD and works by yourself without your supervisor. My supervisor suggests me that I should find “another research topic” that does not relate to my thesis. It is a test whether you can make another PhD by yourself and whether you are ready to supervise someone to do research. That is something I am very much looking forward to …

That’s it for now. I’m getting a bit tired and will try to go back to sleep. I’ll keep adding more things to the list in the future if anything else comes to mind :-)