I’ve started to reflect on my EngD, thinking about the PhD Journey as a whole, and the people I’ve interacted with over the past 4 years. I’ve experienced the ups and downs, the observed PhD curve and I’ve grown a lot.
The biggest thing I’ve learned though, over the past 4 years, is about research attitudes. Knowing that there are unknown unknowns and having the humility to accept that. When I began my research, I felt overwhelmed: I tried and tried to review fields and areas where ‘more research was needed’ to no avail. I became frustrated, anxious, and felt hopelessly lost in a sea of ‘potential areas of interest’ and began questioning my own ability. Not from the healthy perspective; questions like “Am I right?” Or “What if there is another way?” These are good questions, and fall under my “humility” heading. Instead, I was asking “Can I do this?” and “How do I know where to begin?”
I recently completed an application for an RA position at Bath, and when I was writing my personal statement, I thought about my own approach to my research and realised that an interesting theory applied….
The Big Five
How would you describe personality? Sure, you can to great lengths, with use of colorful adjectives and poetic prose to describe personality. This is great if you’re writing a book, or a reference for someone, but scientifically its way too much. Its difficult to compare and discuss personalities from a scientific point of view with such a high dimensionality or degrees of freedom. Rather, some clever psychologists attempted to reduce this dimensionality to a set of 5 core traits, and claimed that personality can be described by plotting across these five dimensions. Sometimes called the OCEAN model (scientists love acronyms, especially when those acronyms equate to actual words with ambigiously related semantics to what the acronym itself represents), it is widely applicable to many domains, and is useful as a measurement tool of personality. Once you can measure something, you can cluster data into categories, and then develop around these clusters.
From a HCI perspective, one of the most novel applications I’ve come across of the OCEAN theory is the thory applied to gamers (i.e people who play games often). Jason Vandenburghe gave a talk at the 2014 CHIPlay conference in Toronto. Sadly I wasn’t at the talk, but I’d seen Vandenburghe’s earlier GDC talk in 2012. Vandenburghe applies OCEAN as a way of modelling player behaviour and their expectations from a game. The gist is that, if you can identify personality traits of players, for example those who are more open to exploration versus those who are more introverted, perhaps less keen on direct competition, you can adjust the game mechanics and dynamics to suit that player. Obviously, this is of direct interest to the games industry, and is useful for game designers to take into consideration. In my personal statement, I applied OCEAN to a PhD candidate, explaining the sides of the spectrum I feel a good candidate would lie. You can read my statement here, but I’ll elaborate on my choices here. I’ve also kept them discipline agnostic; while I’m sure this can be further refined to fit specific disciplines (a model is a model though; always be weary of overfitting), I’ve opted to keep things general. These are qualities of traits I feel any PhD student should possess.
An OCEAN PhD
I guess at this point, I should probably elaborate a bit on the OCEAN dimensions. The first, Openness relates to a person’s apetite for new experiences. Rather than sticking to what works, a good PhD candidate is someone who embraces new opportunities and experiences, open to failure and a tenacity to take risks. A PhD is ultimately about learning how to deal with failure. All the paper rejections, deadlines missed for various reasons, and feelings of complete inability to complete anything is part of the learning process. My advisor said to me “There’ll always be another deadline”. It took me a while to fully appreciate the gravity of that sentence. It wasn’t just to pick me up, but to really drill home the notion of research as an endless cycle. There will always be unknown unknowns, therefore there’ll always be more research to do. If theres always more research to do, there’ll always be another conference venue or journal deadline.
The second trait is Conscientousness. This relates to ones ability to be resourceful and self sustainable. Professor Philip Willis once sent a departmental email around about how in the US they use the term PhD ‘advisor’ rather than the term used in the UK and Ireland, ‘supervisor’. This stuck with me as I really saw it as a more appropriate term. A PhD should not require supervision; when you apply for a PhD in computer graphics, I would assume you have working knowledge of vector math and some experience with practical rendering techniques in theory and code. The same applies for any discipline be it music, geography, history, politics. The point is that this is already assumed: I would expect you to be capable of studying and working without supervision. However, at time what you will need is guidance. You may strike a mental wall in your research. Advice in the form of guidance is more effective here, perhaps suggesting related avenues that you may have overlooked, or indirect research areas that, with a bit of creative thinking, may lead to breakthrough methods just waiting to be applied to your research.
Up next is Extraversion. A PhD student needs to be extraverted. Gone are the days of solo PhD theses; massive texts stuffed full of secluded work done in a lab or an office over the course of 3 years. The world has shifted, with globalisation catapulting us into the age of fast communication of ideas, and an ever expanding space of research domains opening up. Its impossible to do good research alone. Yes, I understand that a PhD is your work, but that work is done in a context of people, other disciplines, and more and more information on a daily basis. Some people even consider doing a PhD is irrelevant. My point here is that, tying in with Openness, a good PhD is someone who embraces methods and approaches from related research fields, and adopts an extraverted mind in approaching skill development. Complex problems in your field may only be complex because the right tools either do not exist, or do exist, but are applied in other disciplines where they also are approriate. Be adventurous!
Agreeablenss is an interesting one, as a good PhD is somewhere in the middle. You should agree with data that opposes your initial hypotheses and beliefs, as painful as it may be to swallow. However, this does not mean you do so blindly. Ask why this data is the way it is? What does it mean for your hypothesis? Is it wrong, or just misguided? How can you learn from it? What if your hypothesis was originally based upon previous studies, a la Bayes Theorem? Following Bayes logic, new information should not alter your view so much if you already have a substantial background in the opposing direction. The point here is to take in this new information, reflect on it, and then feed it back in to your new thinking.
Finally, we have Neuroticism. I’m not saying a good PhD is a never neurotic; you’d have to be a sociopath to not feel upset at times during the course of your PhD. However, it is important to constantly reflect on your situation and understand that it is a bump in the road, not a road block. Remain calm in the face of opposition. Rather than a fit of hysteria, react to oppposition in a way that is productive. Process it, then fire some questions back at it. Don’t take it at face value, but rather assimilate it and refine it. You may find that what was initially the polar opposite may actually only have issue with certain aspects of your standpoint. Be dynamic and invite opposition. It’ll make your PhD richer in the end.
I understand this post turned into a self help piece. I hope you found it enlightenting and perhaps even learned a thing or two. I have found The Big Five highly applicable to PhD candidates, and looking back now, as pretentious as this may sound, I do see myself fitting nicely into this model. Lets just hope I got the directionality right; that I fit into the model rather than the model fitting around me!