During the first project conference, the partner organizations discussed the merits of the PhotoVoice methodology and chose to employ this technique as part of the research process.
Each local organization identified and invited male and female former underage combatants and community members to participate in the study. Between four and ten people from each community participated. A total of 50 PhotoVoice essays (916 photographs) were collected. To see a selection of photographs, see Appendix 2.
After selecting interested individuals, each organization hosted a training session for participants that covered the purpose of the study, how to use the cameras and elements of visual storytelling.
The session moderator explained that every photograph should represent something about each person’s life and experiences. As part of the training session, participants practiced composing photographs that told a specific story, using a piece of cardboard with a viewfinder cut out. Participants were also trained to use a script to obtain consent from individuals included in their photographs, prior to taking the photograph. At the beginning of the process, each person created a list of photographs they wanted to take, but were told that they could change their plan as they wished. Cameras were distributed at the conclusion of the training session.
Participants were given two days to complete their photographic essays before returning their cameras to the project staff, who then developed the film. Data were de-identified using codes. Each local organization hosted a follow-up session with participants the day after the film was developed. During this session, participants captioned each of their photographs. Moderators were available during the session to help participants write their captions. After participants completed this process, each person presented their photographic essay to the group. The moderator then facilitated a group discussion about similarities, differences and major themes that appeared in the photographs.
Members of the research team who are native to eastern DRC translated the captions from Swahili and French to English. Three team members independently generated lists of themes found in the data after looking at both the photos and their captions. Through triangulation, the research team then created a codebook of defined categories and sub-categories. Using this codebook, the captions were coded independently by two people using NVivo 8 (QSR International, Cambridge, Mass.).