Way back in the middle of July, I finally completed the fieldwork for my dissertation exploring the environments of walking simulator video games. Armed with data gathered from interviewing 11 game developers, playing 12 games, and a wealth of secondary sources, I somehow had to make a path towards fulfilling my research aim. This was to understand the extent to which virtual video game environments can be experienced as places – spaces we come to know through the human meanings and feelings we associate with them.
I had a great deal of data to work with, but unfortunately not a lot of time. My fieldwork finished later than originally planned, due mainly to a couple of game developers having to postpone interviews because of their own commitments. Although this couldn’t be avoided, I knew I would have to make up the time by being efficient with my analysis. After all, the final deadline for submitting my dissertation wasn’t moving, and soon enough I would need to get words on paper.
I had already transcribed all of the recordings from my interviews and typed up my handwritten research diary to make the data as accessible as possible. To make it valuable for my research, though, I needed to somehow translate these texts into key themes and points that have something interesting to say for the questions I’m asking as a researcher. In qualitative research, this process is known as coding.
The only way to get started is to dive straight into the data. I started off by reading everything as closely as possible, noting down details that came across as significant in the context of my research aims, the rest of the data set, and prior reading on the subject. Creating a list of these topics as they were encountered, I’d then underline each one once for each new instance of that topic in the transcripts and research diary entries I read. By the time I’d read through the ~200 pages of data, I was left with 5 sides of topics, underlined to different degrees, with lines between them to show their connection to other topics.
So what do we learn from this early stage of coding? Well, already we can begin to see which observations were more or less prevalent during the research, simply by looking at the number of lines underneath each topic. We can also start to assess how different topics fit together by looking at the network of lines going between them on the list. It was a messy document, but within it were the foundations for the series of points I would go on to make in my dissertation.
To make a coherent argument in a piece of academic work, though, you can’t simply reel off a list of points. These points need to be placed within a framework that guides the reader towards understanding how they contribute to a wider discussion of the subject matter. In dissertations, each chapter usually provides a pivot for the discussion, so the path you make through the material often becomes a question of how you organise this chapter structure.
For subjects with lots of interconnecting factors, as are typical in human geography, themes are normally the best way to approach the research subject, which have distinct characteristics but also fluid boundaries that are able to flow into each other freely. This is the approach I took to the complex array of topics that my dissertation presented, condensing my various data into groups of ideas that provided the basis for my dissertation chapters.
I can’t claim to have any particular words of wisdom for identifying themes, especially for this project. It’s like alchemy. You collect together your thoughts and magically new patterns begin to emerge, and everything starts making sense. It’s as if your subconscious has been slowly weaving and winding the different bits of information throughout the research, and only by thinking deeply can you access this strange fabric that connects everything together. Like Sherlock’s ‘mind palace’ that he uses to thread evidence together until he reaches a moment of insight.
It started at the tired end of a busy day when I was considering getting an early night. I started to look through the list of topics I’d made, when suddenly things started to click. I picked out one key theme, then another appeared, and eventually two more to form a group of four themes that each nicely divided the range of points I’d identified during the initial coding stage. These themes stayed effectively the same right the way through the analysis and writing stages, and were as follows:
Agency and interactivity: relating to lack of mechanical interactions in walking simulator games (i.e. button pressing, complex and precise controls), and instead the loose interactive framework that designers use to encourage players to explore the game world mindfully.
Immersion and believability: how the game worlds feel authentic and visceral, as if you are really experiencing them – but also how they fail to do this.
Navigation and narrative: how game developers design powerful experiences in worlds, and the ways players navigate these to form stories as they explore.
Emotion and subjectivity: how game worlds are designed to make players connect emotionally with them, and how players’ own identities become part of the world itself through this process.
Once these four themes were solidified, it was time for the highlighter pens – probably the most important tool for coding analysis. Using four different colours, I turned back to my list of topics and highlighted each one according to the theme(s) they were associated with. By the time I was finished, I was left with a document that not only told me the most prevalent topics and how they joined together, but also where they would ultimately fit into the chapters of my dissertation text.
My dissertation was taking shape.
The final stage of my analysis was one I’d never properly done before in any research project, but turned out to be the best course of action I could have taken. I was lucky enough to have a fantastic, enthusiastic supervisor for my dissertation, and he gave me a gold nugget of advice for the final phase before writing. He suggested writing abstracts for each chapter – short summaries of around 200 words or less – that clearly set out exactly what I was arguing for each one. This helps you to tie together all the points in each theme to make a coherent argument for the chapter; taking the reader on a determined, purposeful journey towards conclusion, rather than a sprawled collection of ideas that prevent clear understanding. Furthermore, by knowing exactly what you want to say in each chapter, it helps you to decide how you’ll order them so that the reader’s journey is logical, flowing from one argument to the next.
For my abstracts, I went deeper and further than my supervisor suggested. For each chapter, I also wrote down three key pivots of the argument, which formed the chapter’s subsections. This then helped me to categorise the different pieces of data I’d gathered more precisely to save time. At the larger scale, I also decided to write a rough abstract for the whole dissertation. In the end, a dissertation is marked as a single text, and therefore it has to read well as such. It’s by having a consistent line of reasoning flowing through the chapters that the reader will get a strong impression of your research findings, and their place and impact within the field.
There needs to be a clear take-home message, otherwise the research will be wasted no matter how well it was completed.
Like the winding paths of the walking simulator games I played, my analysis was also about treading a path through a world of ideas, objects and feelings, and threading them together to make a story – a series of events that the reader must encounter to eventually reach conclusion. Hopefully this post has revealed how you can transform a big bunch of wordy data into a story, but also how this can be done concisely, wasting as little time as possible.
If writing an academic text is about taking the reader on a journey, then good analysis gives you efficient tools to break through the unknown and make a path towards understanding.