MOBILE PHONES, INTERNET, AND GENDER IN MYANMAR MOBILE PHONES, INTERNET, AND GENDER IN MYANMAR Table 1 2.2 Qualitative research Focus group discussion sample table This section summarises the methodology used for the qualitative research module, which was designed to explore the research questions identified in Section 1 in depth, based on the baseline survey findings. Protocol Respondents FGD 1 6 FGD 2 6 FGD 3 6 A total of 25 home visits (involving in-depth interviews with selected respondents) and 11 focus group discussions were conducted in one urban location (Yangon) and one rural location (Pantanaw) in Myanmar. Protocols lasted approximately two hours each. Focus group discussions involved six to seven participants who fit the sample criteria. Fieldwork was conducted in July 2015 by MSR (Myanmar Survey Research), with the participation of LIRNEasia researchers for some of the protocols.The sample included women and men, urban and rural, and working and non-working respondents, ranging from age 18 to 65 and socio-economic classification (SEC20) groups B–E. FGD 4 6 FGD 5 6 FGD 6 6 Informed consent for participation was obtained from all research subjects. FGD 7 6 FGD 8 6 FGD 9 6 FGD 10 6 FGD 11 6 Source: MSR Figure 5 Qualitative research locations Yangon Pantanaw • Urban • Rural • Commercial centre of Myanmar • Typical rural and traditional small town, approximately 96 km from Yangon by road • Widest availability of modern infrastructure and services • High penetration of mobile phones • Lower penetration of mobile phones • Mobile internet has been available since early 2015 • Fast mobile and internet service; wi-fi access available The key limitations of the survey data are that the data is based on respondent recall; questions on usage patterns, spending patterns, and household expenditure are therefore subject to recall bias. In an attempt to reduce this bias, questions were phrased to capture information about the most recent occurrence or time period (e.g. their last phone call or expenditure in the last month). The key limitation of the qualitative data is the inability to generalise findings, due to the non-random sampling method used. Respondents were purposely selected to fit certain criteria (e.g. mobile-owning woman, aged 18–29, who works and also uses internet services). The sample was designed to capture as much diversity as possible (gender, age groups, SEC categories, usage habits, etc.), while ensuring there were enough sample points to identify patterns. Given our resources, we were not able to include equal numbers of men and women or equal numbers from the three age groups (18–29,30–45, 46–65), but best efforts were made to balance the sample based on the research questions. Protocol Internet and app use (via any device) 18-29 Owner Yes 46-65 Owner 50% Yes, 50% No C-D 18-29 Non-owner No B-C 30-45 Owner 50% Yes, 50% No C-D 18-29 Owner 50% Yes, 50% No Working B-C Non-working Female Working D-E Rural Non-working Urban D-E 18-29 Non-owner No 30-45 Non-owner No 46-65 Non-owner No Owner Yes Owner 50% Yes, 50% No Non-owner No B-C Working Male C-D Non-working Location Gender Occupation IDI 2 D-E IDI 3 Working IDI 4 IDI 6 SEC B-C IDI 1 B-C D-E C-D Urban B-C IDI 7 D-E IDI 8 Non-working B-C IDI 9 D-E IDI 10 D-E Female IDI 11 C-D IDI 12 D-E Working IDI 13 C-D IDI 14 D-E IDI 15 C-D IDI 16 Rural C-D IDI 17 D-E IDI 18 Non-working IDI 19 IDI 21 IDI 24 IDI 25 C-D D-E IDI 20 IDI 22 12 Mobile ownership SEC Urban Rural Age group Occupation D-E 30-45 18-29 Home visit sample table Urban Male Rural Age group 18-29 30-45 46-65 18-29 30-45 46-65 18-29 30-45 46-65 18-29 30-45 Mobile ownership Internet and app-use (via any device) Owner Yes/No Non-owner No Owner Yes/No Non-owner No Owner Yes/No Owner Yes/No Non-owner No Owner Yes/No Non-owner No Non-owner No Owner Yes/No Non-owner No Owner Yes/No Non-owner No Owner Yes/No Owner Yes/No Non-owner No Owner Yes/No Non-owner No D-E 46-65 Non-owner No B-C 18-30 Owner Yes Non-working B-C 30-45 Owner No Working D-E 18-29 Non-owner No Working C-D 18-29 Owner Yes Non-working C-D 30-45 Owner No Working IDI 23 20. SEC categorises households into groups A (highest) to E (lowest) based on the socio-economic characteristics of the chief wage earner (education, occupation, etc.) and/or household (presence of durable goods, housing material, income, etc.). In Myanmar, the most commonly used SEC is based on monthly household income. SEC D and E (those earning less than MMK 500,000, or USD 289, approximately) are considered low-income households in this research. Gender Table 2 IDI 5 2.3 Limitations Location 13