M.MAS 7/22/15 or geographic origin. We also asked questions about use of mobile phones and where they access news, out of interest as well as to establish a connection between our questions and our backgrounds.25 The second set of questions asked people to share their concerns for the whole country. After working with them to brainstorm between three and five concerns, we then asked them to choose their primary concern and then probed to encourage them to explain their concerns more fully. People required various amounts of prompting, from extensive follow-up questioning to none at all. Typical follow-up questions included ‘what are you worried might happen?’ ‘how long have you been worried about this?’ (or, ‘when did you start worrying about this?’), and ‘where did you learn about this?’ or ‘with whom do you discuss this?’ After engaging in this discussion, we then repeated the same process, but positioned the questions as about the narrator’s local area rather than the whole country. The third set of questions were about the narrator’s hopes for the future (‘in the next three years, what would you like to change?’) and open time for them to add anything else they felt to be important. In making decisions about who to interview, we set a quota to ensure that interviews included a diversity of perspectives that accounted for gender, age, religion, and class. Ethnicity was felt to be too diverse to include, but we maintained a generally even division between ethnic Bamar and non-Bamar narrators. Class was also difficult to define practically, and so we agreed to seek interviews with a mind to diversity of educational backgrounds26 as a rough proxy for class.27 However, in practice we found it difficult or disruptive to ask people questions about their educational backgrounds, particularly for those narrators who had had limited experience with schooling, and so in the end this information was not consistently collected. The training participants assisted with approaching or introducing us to narrators, and where educational background was not recorded we used other locally-understood markers of class position to assist us in including a diversity of perspectives. Individuals interviewed by the authors are summarized in Table 1 below; an additional 96 interviews by training participants are not cited because they have not been analysed for this paper, though these informed group reflection discussions during the Mandalay meeting that, in turn, informs this paper. Table 1. City Lashio # Interviews Women Men Over 40 Under 40 Buddhist Christian Muslim Hindu 13 8 8 4 12 7 2 3 4 8 4 7 4 7 6 2 1 2 13 8 5 3 10 5 2 3 3 9 6 5 7 4 4 1 4 2 Sittwe 16 13 5 11 7 10 1 2 5 Pathein 9 5 4 4 5 7 0 1 1 Totals 68 44 34 33 45 39 8 14 17 Mandalay Mawlamyine Meiktila 25 Part of the informed consent process included introducing ourselves, which entailed explaining the ‘Myanmar Media and Society Research Project’ as well as the ‘Myanmar ICT for Development Organisation.’ In light of this, we decided that, in the interview interaction, it would help make narrators more comfortable if we began the interview with questions that were both easy (age, work experiences, etc) and easily contextualized in light of how we had introduced ourselves. Later, we would ask follow-up questions about where the narrators may have heard a specific anecdote or piece of information. Such questions were better framed in light of an earlier conversation about information sources; if we had asked them only later in the interview, they may have been taken as an accusation or implying that the narrator was otherwise ill-informed. 26 Three categories were defined: high school or college graduates; those with education up to some high school but not graduated; and those with no education or education below high school level. 27 Varshney adopts a similar approach, using literacy rather than the level of formal education reached. See, Varshney, Ethnic Conflict and Civic Life, 19. 5