The goal of analysing the qualitative data from the interviews was to find and confirm patterns within the data, interpret those patterns and identify behavioural goals that inform the creation of a persona which in turn lead to actions taken in the design of the end product, Goodman E., Kuniavsky M., Moed A., 2012.
The first step was to gather all the relevant data, text, video, audio, etc. into a format that the team could work with and understand. Often the best way to do this is by “making the data visual and physical”, as Dan Saffer recommends in Designing for Interaction, Goodman E., Kuniavsky M., Moed A., 2012.
After the interview process was finished each team member transcribed their interviews so that they could be analysed more easily.
Using the transcriptions I started to break-up large chunks of data into smaller units so that it could be recombined and more easily segmented, Goodman E., Kuniavsky M., Moed A., 2012. I gathered statements made by the participants and organised them into different rows according to a particular heading or denominator. These segmentations were based on each participants actions during the interview and divided between their overall behaviour and their behaviour as related to the specific tasks.
Designers need to reconstruct goals from observed behaviours, answers to questions, verbal and non-verbal cues and from the environment. One of the most critical tasks in the modelling of a persona is to identify user goals. Don Norman’s book Emotional Design (Basic Books, 2005) introduced the idea that product design should address 3 different levels of cognitive and emotional processing; visceral, behavioural and reflective.
1. Visceral is the most immediate level of processing where we react to a products visual and other sensory aspects so that we can perceive things before any significant action is taken.
2. Behavioural is the middle level of processing and lets us manage simple everyday behaviours. Norman states that historically, interaction design and usability practices have nearly exclusively addressed this level of cognitive processing.
3. Reflective is the least immediate level of processing and involves conscious consideration and reflection on past experiences, Cooper A, Rerimann R., Cronin D. 2007.
Although these 3 processes gave me a good insight into how a user interacts at a cognitive level with a system I found it difficult to separate the qualitative data under these heading. So I looked again at how I might segment the data into different groups to find underlying patterns of behaviour.
As mentioned personas are largely derived from qualitative research, especially the behavioural patterns observed during interview and observations. Creating believable and useful personas requires an equal measure of detailed analysis and creative synthesis. (p81 About Face). Listing the distinct aspects of observed behaviour for each participant produced a set of behavioural variables. By focusing on the list below one can produce a distinct set of behavioural patterns, Cooper A, Rerimann R., Cronin D. 2007.
Activity – what the user does, frequency and volume
Attitude – how the user thinks about the product domain and technology
Aptitude – what education and training the user has, ability to learn
Motivation – why the user is engaged in the product domain
Skill – user abilities related to the product domain and technology
Using the above variables I separated the data for each user into these new behavioural segments.
Once I had each of the 5 participants set to this new format I then did a final table with the same heading and put a colour coded reference for each name under the headings. This way I was able to see quite clearly the common patterns of behaviour between the participants as they interacted with the GoBus site. Also the how and why they were interacting with the interface and where they were having the most difficulties, especially during the tasks.
Using these patterns of behaviour along with the summary comparisons findings “Persona Segmentation” the team was able to accurately extrapolate the goals for the persona. Further to this, although the survey and interview data showed that most people made their bookings via a desktop computer they did so because they were required to enter a lot of data and they didn’t like doing this on their phone. However, they said they would be open to using an app and doing it via there mobile if the process was made easier. The data also showed that when surfing the internet using a mobile was the most common devise used. So to fulfill a major finding from the research and make it more convenient and hence easier for the user to use we decided to develop a mobile app to make it more accessible for users – where booking via desktop was more restrictive.