Had the participants tagged, described and catalogued the images, the next step was for me to follow this up with interviews to provide the opportunity for them to flesh out any descriptions and explain some of their choices behind the tagging and categorising. Since this aspect had not taken place, greater importance became vested in the interviewing process. The choice was open to meet the participants as a whole group (only five of them in total) or individually. Each has its own merits and problems but in the end the choice was one of practicality; it was just too difficult to find a time during which they were all free. So the interviews took place individually (the two Y7s as a pair) at times to suit them. A few questions were prepared (see Appendix E), though the format was to be much more loose than even a semi-structured might be, enabling the respondents to drive the direction wherever possible, thereby hopefully imbuing them with a greater sense of ownership and responsibility over their data and in an attempt to reduce the power imbalance described in Section 2.4). In developing a structure of sorts, care must be taken to avoid inflicting the participants with my preconceived notions, yet avoid confusion which could lead to meaningless data. As Goulding (1999: 8) observed:
"The art lies therefore in finding a balance which allows the informant to feel comfortable enough to expand on their experiences, without telling them what to say."
A challenge to be faced at this point was how interviewer and participant could view the images together. Although a monitor displaying the computer output and showing the folder containing their images could have been used, it was felt that projecting these images onto an interactive whiteboard and using the whiteboard software package would bring an added dimension.
Figure 5 – Using an IWB to support discussion
By using thumbnails of the images in this way, all the images captured by the participant could be gathered onto a single screen. This allows them to easily focus on a single image if necessary, to group them together, to order them, to filter them and perhaps most important of all to annotate them, thus providing an opportunity for coding. Helpfully as groups are formed, it is easy to amend their composition just by dragging the images elsewhere, or if an image needs to exist in more than one group, it is easily copied. The act of dragging an image around the screen and adding a verbal description and explanation to provide context at the same time seems to provide a much richer experience. Physically being able to drag images into groups and see the group starting to swell in numbers whilst other images appear isolated also sets a different (more informative?) scene than if they had been tagged then filtered to be viewed as a coherent group.
Once the participants had all been interviewed, it became clear that the advantages of interacting with an extensive catalogue of images in this way were being outweighed by disadvantages. The time required to mentally process such a large range of images and reflect on the meaning of them, important though this process obviously is, meant precious interview time was being absorbed with little additional data being generated. However this initial input from participants began to filter and categorise the images, suggesting areas of interest for further study. In the next phase a smaller subset of images was selected (see Section 3.3.5) to be shown to a second cohort of students known as respondents, since they weren't involved in the image generation process, but did provide their thoughts on the images generated by the previous set of participants. The interactive whiteboard could be left behind in favour of a monitor as a reduced number of images were examined and discussed, with less manipulation, selection and filtering being required.