52 > Next Generation Myanmar APPENDIX: METHODOLOGY AND APPROACH Data was collected for the Next Generation Myanmar study between August and November 2018. Data collection was staggered between the study’s quantitative and qualitative methods throughout the data collection period, allowing initial survey results to inform the content of subsequent focus group discussions and key informant interviews (see overview in Figure 19). Each stage of the methodology following the literature review is detailed below. For an overview of the data collection stages, see figure to the right. Data collection 1: survey As a first step, a household survey was conducted. Demographics The survey universe – or, less formally, the group of people that the survey aimed to say something about – included youth aged between 18 and 30 throughout Myanmar. This age range is more narrow than the Myanmar government’s definition of youth. The government, as well as many Myanmar CSOs/CBOs, define youth as young people aged 15 to 35. The researchers chose a truncated range of age 18 to 30 in order to concentrate data collection on a smaller share of the population, thus allowing a higher level of research validity. However, in rare cases when conducting key informant interviews and focus group discussions, interview subjects aged 30–35 were included in this study, recognising the fact that the age range 18–30 is not universally accepted and is less common than the age range 18–35 among Myanmar organisations. As such, a small number of 31-year-old informants appear in the report. This is not believed to have affected findings. Any research in Myanmar must consider ethnicity. A total of 135 ethnicities are officially recognised in Myanmar, and these have long been invoked as constituent demarcations in Myanmar politics. A possible sample design might allow for disaggregation by, and comparison between, respondents of each of these 135 ethnicities. However, critical limitations make such a design undesirable. First, despite its ubiquity in Myanmar politics, the extent to which this frequently invoked idea in fact depicts Myanmar demography remains a point of ambiguity. Ethnic groups listed among the 135 may not exist, may no longer exist, or may have merged or divided. Further, respondents may belong to more than one listed ethnic group, or none at all (as is the case with the Rohingya-identifying populations from northern Rakhine State). A survey design built from the ‘135 nationalities’ framework would assume the accuracy of the framework before it can be empirically validated. Second, a sample design of 135 nationalities would be costly. Any statistically valid sample would require an absolute minimum of 30 respondents per ethnicity, resulting in a total minimum sample size of at least 4,050. This amount is a minimum, and ideally many more respondents of each ethnicity would be included (135 × 200 = 27,000 respondents would make for a quality grade sampling size). In the case of the Next Generation Myanmar study – as with most studies – such a large sample size and its associated cost was prohibitive. The Next Generation Myanmar study therefore does not include membership in one of Myanmar’s alleged 135 nationalities as a variable in its survey. However, this does not mean that the study does not explore the role ethnicity plays in the lives of youth, as told by youth in today’s changing Myanmar. Figure 19: Timeline of study Literature review Data collection 1: survey Analysis 1: stock-taking Data collection 2: focus group discussions and key informant interviews Analysis 2: synthesis Data collection 3: youth policy workshops Final analysis