Mixed-methods facilitaters/barriers RCB UB - Dataset
Description
Ulaanbaatar, Mongolia, is in the midst of a winter air pollution crisis driven by the combustion of solid fuels in the peri-urban Ger districts to which young children, fetuses and pregnant women are particularly vulnerable. To address this, the Mongolian government banned the sale and use of raw coal in May 2019 and has subsidized and promoted refined coal briquettes as an alternative fuel. This mixed-methods study utilized semi-structured interviews (n=30) and a questionnaire survey (n=369) to identify facilitators and barriers to compliance with the ban amongst Mongolian mothers living in Ulaanbaatar. Four main themes were identified that affected compliance: knowledge about air pollution, information sources, initial policy impacts and governance. Facilitators of participants’ ban compliance included awareness of the severity of air pollution’s health impacts, support for the policy and governmental enforcement. Concerns regarding fuel alternative safety and cost implications, as well as uncertainty about the policy (inadequate information) were found to be barriers. Incorporation of these insights in future air pollution mitigation strategies could be beneficial for strengthening their effectiveness, both in Mongolia and settings that face similar challenges throughout the world, which in turn could be a crucial step in meeting the 2030 Sustainable Development Goals agenda.
Files
Steps to reproduce
Data collection: The study was conducted in the first year after implementation of the RCB and utilized a mix-method study design, comprising semi-structured interviews and questionnaires that were piloted prior to the study. Women who were currently pregnant and/or mothers, who resided in Ger districts of Ulaanbaatar, and who used a solid fuel stove fwere included in this study. Recruitment of participants from major Ger districts in Ulaanbaatar took place in family health centers and district hospitals during attendance for routine check-ups. For the semi-structured interviews, a minimum of 30 participants, including at least 15 mothers and 15 pregnant women, were targeted. All interviews were conducted in Mongolian via telephone. Groups were not treated separately in data collection and analysis. Face to face recruitment using purposive sampling by local medical staff initially occurred for one week at a district hospital in Chingeltei Ger district. Snowballing from initial recruits, where participants recommended individuals they thought would be interested in the study and who fit the study criteria, was then utilized. Both in parallel and sequential to the interviews, quantitative data was collected using face-to-face questionnaires at a medical facility setting. Women attending antenatal check-ups or other routine doctors’ appointments in the five major Ger districts (Bayanzürkh, Byangol, Chingeltei, Songino Khairkhan, and Sükhbaatar) were recruited by fieldworkers from the Mongolian National University of Medical Sciences using convenience sampling and an opt-in approach. Data analysis: Audio-recordings were transcribed verbatim in Mongolian before the transcripts were translated into English. The research assistants who carried out these tasks were given training and a sample of translations was checked by a second translator to ensure accuracy in words and meaning. Data from the first interviews were analyzed before later interviews were carried out, allowing the research tools, such as the topic guide, to be iteratively adapted and particular lines of enquiry to be pursued as the study progressed. This also allowed the practice of reflexivity to be applied, as the positionality of the research team members was evaluated. A constant comparative method was employed in order to perform a comprehensive and critical thematic analysis and to assess for saturation. The data were coded and the Braun and Clarke’s six step approach was used to generate themes and organise ideas. 2.3.2 Quantitative analysis Responses were compiled in an excel spreadsheet and descriptive statistics (frequencies, means and standard deviations) were used to summarize demographics and outcome variables. Chi squared tests were performed to assess differences between categorical variables. P-values of less than 0.05 were considered statistically significant and all tests were two-tailed. All data were analyzed using Stata/SE 16.1 and Prism GraphPad.