Step by Step, Line by Line

Introduction

Beware reader! this is a very lengthy and dry post, but it feels necessary. So make yourself a gin and tonic, settle down, read on, and try not to get too bored.

As CGT is new to me I need to practise the skills I've been reading about over the past year, so  I decided to use the results of the interview with my second guinea pig (I'm going to call him 'Bradley') as my first foray into initial, line by line coding. 

The Interview

I prepared five questions based on the responses Bradley had given to the Call for Participants survey. I looked for repeated patterns in Bradley’s responses, and in doing so, discovered the following themes: 

• Learning 

• Emotions and emotional attachment 

• Completing a task 

• Distractions: ‘good’ and ‘bad’ 

These themes led to the development of the five  questions.  I admit to there being an element of using both my instinct and shared experience and understanding as a gamer that came into play as I developed Bradley's questions, which were as follows: 

1. In the survey, you mentioned how you have learned things relating to the real world. Could you elaborate on this for me? 

2. You mentioned that completing a task gives you satisfaction. Can you give me an example of one of these tasks? 

3. You talk about how gaming distracts you, and what interested me about this was how you seem to have experienced ‘good’ and ‘bad’ distractions. Would you be able to tell me more about this?  

4. You’ve experienced many emotions when playing, and these look to be at both ends of the 'happy-sad spectrum’. Can you talk me through what led you to feel any of these emotions? 

5. Finally, and linked to the last question, I’m keen to know more about becoming emotionally attached to your avatar or an NPC. Would you be able to tell me more about this? 

The interview was carried out in person, in a neutral setting (a frighteningly quiet on-campus coffee shop in the middle of the summer recess), and I recorded the interview as an MP3 file, using a digital voice recorder.  

After confirming with Bradley that there were no trick questions, that there was no requirement to answer any question he felt uncomfortable with, and that it was okay to criticise or disagree with any questions I asked or comments I made, I confirmed with him that I had permission to record the interview, and began the interview-proper by asking if he was happy to tell me a little about the games he enjoys. I planned to use an example of a game we both play: Assassin’s Creed Odyssey as a prompt if required, but this was not necessary. I hoped that talking in a positive way about a shared interest from the start would begin to establish a rapport and frame my role as a gamer and therefore a part of Bradley’s world. This, I theorised, would manage any perceived power imbalance between researcher and participant, and give Bradley with the time he needed to relax into the interview. 

I went on to ask Bradley the five open questions highlighted above, remembering that the interviewer must carry out something of a discreet juggling act while interviewing, involving many balls: 

• looking at the participant's body language

• listening out for anything of note or something that may be worth asking for expansion as something of possible interest

• keeping an eye on the time without making it look like you’re keeping an eye on the time in case the participant assumes the interviewer is bored 

• making notes / memos of anything of note that occurs (remembering Charmaz’s question: ‘what is happening here?’), and that strikes the interviewer as being particularly of note. Again, this must be done in such a way as to not appear bored or disengaged to the participant. Brief, ‘automatic’ writing in a small notebook, and in as subtle a way as possible was the only way to make notes.  When you have handwriting as illegible as mine, this is particularly difficult. 

Finally, by way of an interview 'debrief’, I told Bradley a little more about my research, and my interest in the positive, negative and neutral unintended consequences players experience when playing open world RPGs, before asking him whether he had experienced any unintended consequences that he would be happy to share.  

Below are images of part of my ‘script’, along with a selection of notes and memos made during Bradley’s interview: 

After the interview concluded, I uploaded the resultant audio file to Panopto to generate a transcript. Then began the laborious task of listening to the audio and editing the transcript for accuracy. I then stripped out pre- and post-interview chat, along with my own voice, finishing with a text file that, very much like the final chapter of James Joyce’s Ulysses, was nothing more than one continuous paragraph of text solely containing Bradley's responses. This was copied and pasted into a Word document and formatted into a continuous column of text with 11-13 words per line, before being printed out. I ensured there was space on the eft hand side of the hard copy for me to be able to add handwritten initial codes. As a left-hander, it was important that I did not cover the text with my hand when developing codes as not being able to see the transcript meant that I would lose the connection I had with the contextual aspects of the text, hence the rather ‘back to front’ looking document that can be seen below.

The irony of academia is that the very skill one is encouraged to develop and use - critical reflection - means that my instinct is to question every...instinct. I am concerned that I was overthinking and over- analysing the words I used to compile a code, and where one or two words would be sufficient, I found my codes were entire sentences or phrases, so almost memo-like. I need to look at further examples of line-by-line coding to examine their construction and content, with a focus on deciphering the links previous researchers have made between each line of a transcript and its meaning. I will also need to learn how to switch off my tendency to analyse and question every choice I make and learn to trust my instinct. 

After repeating the line by line coding process in NVivo, I examined my codes to look for common themes and started to categorise them around these themes as they emerged. Each theme became a parent code; a ‘folder’ of sorts, into which I dragged and dropped corresponding child codes. Any extraneous codes left at the end of the process were deleted. The resultant parent codes are as follows:

1. Achieving is important

2. Adhering to a moral code

3. Being immersed in games

4. Connecting with avatar

5. Experiencing emotions and situations in games

6. Impact of gaming on wellbeing

7. Learning through gaming

8. Playing with others

9. Remembering gaming moments

10. Responsibilities as an adult

11. Socialising in real and digital worlds

12. Using time well

The example below displays the child codes belonging to the parent code: ‘Adhering to a moral code':

I wanted to capture some initial thoughts around each top-level code as memos, so, as recommended by Charmaz, I wrote the first thing that came into my head as a stream of consciousness as I examined the child codes grouped with each parent code. These memos were then linked to each parent code, and colour coded accordingly:

An example of one of these memos: Adhering to a moral code can be seen below:

I then ensured that child codes were aggregated to parent codes. This allowed me to see the number of references attributed to each parent code. I also colour coded previous memos written when transcribing the interview to fit in with these parent codes. This was when I noticed that errant child codes that could not fit into a single parent code were uncategorisable because they fitted into more than one parent code. This also linked to the content of my memos, which indicated, along with the number of references aggregated per parent code, that learning through gaming, the positive and negative effects of gaming on wellbeing, playing with others, adhering to a moral code, using time well, and responsibilities as an adult were intrinsically linked to one another:

I ran a codes comparison using NVivo, comparing all parent codes against all child codes and memos. This generated the following ‘Sunburst’ diagram, and provided a useful, visual way to see the size of parent codes based on their child code and memo content: