Blog by SPARC
SPARC and its community federations have, for a long time, used research tools such as surveys to learn more about access to basic services in their own and neighboring communities. This is an SDI-wide methodology called ‘Settlement Profiling’, which is now subsumed under a larger campaign called ‘Know your city’. In ARISE, we applied the same methodology to collect data around health and wellbeing in the slum relocation colony action sites in Mumbai and Ahmedabad, India..
As art of the ARISE project, the SPARC team developed a simple survey questionnaire to collect quantitative and qualitative data around the health and wellbeing (H&WB) of respondents. In our sampling strategy, we chose locations where residents of slums displaced by city infrastructure projects had been relocated by SPARC in collaboration with state institutions. The sample covered about 30% of households in four locations in Mumbai and one location in Ahmedabad, covering a total of 3,450 household surveys.
While the federations in Mumbai have used surveys for several years, the H&WB survey was new to them. This was particularly because the H&WB survey process required collecting personal information and because the concept of wellbeing is difficult to articulate in the local language spoken in the communities. Therefore, while they knew that they would be trusted by the survey participants, it was important to follow proper ethical guidelines and anonymize data to the best possible extent.
In Ahmedabad, such processes have rarely been undertaken in relocation sites. Any such attempts of data collection are met with numerous doubts and elicit a sense of anxiety among the residents. People often mistake surveyors for state actors, such as those who notified their houses while living in slums before being relocated. Considerations of privacy and disclosure in this context were paramount.
In this blog post, we share our experiences of the survey process, and those of the federations, particularly the women. who undertook the surveys in their respective localities.
Sampling method and the fun with every ‘third’
In Mumbai, the Mahila Milan members – ARISE co-researchers who carried out the survey – found the idea of surveying every third house as a sampling strategy very humorous. Co-researchers and residents were amused as to why we would drop households in between. In fact, during surveys, the neighbors would peep into the discussion and would express disappointment if their house was dropped. At first, the surveyors felt perhaps that this stemmed from the anxiety around missing out if the survey data was to be used for any immediate benefit. However, over time, it became clear that many neighbors were genuinely interested in participating and wanted to be asked about their health and to share their experience with treatment, safety, and support from state and non-state actors.
In Ahmedabad, the absence of community-based organisations such as the Mahila Milan meant that researchers from SPARC and prominent community members from the relocation sites had to develop a different approach, where community members were collaborators rather than co-researchers. While influential community members were unfamiliar with surveying tools, they were able to facilitate the process successfully, helping to recruit participants and organising meetings where they explained the purpose, intent, and potential benefits of the surveys.
Resistance and survey response fatigue
Communities living in informal settlements or in poverty are often subjected to repeated surveys. While the communities at action sites had not been surveyed on issues of H&WB before, they were initially quite reluctant to participate as they saw no value in it. Repeated surveys, especially those for research, have not previously produced tangible outcomes for them. It was not uncommon to find people expressing their perceived futility of these processes and raising issues, resulting in other households turning down the surveys as well.
In Mumbai, the surveyors, being largely people from their own communities, could discuss with leaders and were successful in explaining the benefits of such activities. They gave leaders formal letters explaining the process, intent and future benefits, which were modelled after the information in the participant information sheets, and sought their agreement. Eventually, most households participated and shared information freely.
In Ahmedabad, the process was similar, although the barriers were steeper; the lack of community organisation meant that a few meetings with the leaders did not suffice and needed much stronger and repeated efforts to elicit participation.
Research ethics and remaining anonymous
One important element of the H&WB survey was to maintain the anonymity of the participants and their families. Therefore, a two-step process was designed to tackle this – each household was assigned a code, and the code and associated respondent name was recorded on a separate sheet for data verification purposes. Data would then appear on the survey forms with the code, which is developed from the initials of the surveyor and the date of survey, ensuring that the participant would remain unidentifiable from the survey data.
However, it was challenging for the surveyors to develop a code and then noting the health data of different family members being shared by the respondent without writing names. It was also difficult to record any action that would require follow-up or re-consultation with individual households. Though this proved to be quite challenging as their first experience, over time the surveyors learned how to overcome it.
Survey questions generating more questions
While a survey question seeks an honest response to the question, with some participants in this process, it raised more questions, and it therefore became a tool to generate a ‘mini discussion’ during the survey. For example, we included a question on mealtimes and what foods were largely consumed in their meals. This was to understand if people were consuming adequate amounts and getting all major sources of nutrition. However, some respondents became inquisitive and asked why we were asking about mealtimes and if the meal times they followed and the food they consumed had any bearing on their health. It generated a small discussion around current eating practices, the change in lifestyle where meals are now consumed at irregular times, the inability of families to consume meat due to its high price, etc.
This also brought forth questions and concerns regarding the quality of produce, and the uneven benefits offered by public distribution schemes and their relation to their current living circumstances. That said, as food habits are deeply cultural and embedded in religious beliefs, survey questions pertaining to this were met with apprehension in some instances.
While the process of survey was not new for the Mahila Milan women, the experience of surveys as part of a research methodology was. Firstly, surveys that elicit personal information require trust between the surveyor and the participant. While ethical processes do safeguard respondents, community members have no prior knowledge of formal safeguarding processes, and mere sharing of information via participant information sheets and having them give written consent does not always ensure that a participant feels confident of data confidentiality. In these circumstances, personal trust becomes most important and therefore good quality data on health and health practices were elicited only from a small proportion of the sample.
The Mahila Milan are important collectives in their localities, and their support groups provide advice and assistance with everyday issues. The survey experience helped the Mahila Milan and the residents to feel that they can also look at health, health access and other social determinants of health within their support work. In addition to the survey being useful as a tool to collect data, the interactions during the survey also improved everyone’s understanding of health as a subject.
Surveyors also unexpectedly found the indirect survey questions helpful to elicit answers on difficult questions. For example, we added in a question about safety at home. While everyone felt it was obvious for anyone to feel safe at home, the surveyors soon found it an easier way to ask about domestic violence.
The Mahila Milan was also initially resistant to ask people about their personal health, as they felt it was ‘not useful’ and ‘nobody shares health data’, and they were concerned about the length of the survey. But their opinion changed when they saw that the first respondents were actively interested in sharing, and the nature of questions and the reasons for them being asked began giving a clearer picture. A senior Mahila Milan leader stated:
“We used to giggle thinking of the kind of questions in the survey. Who would ask such questions, why would anyone answer, and what are we even going to do with this kind of data? But we now realise how the sections and questions are interlinked and come together to tell us a story about health and wellbeing among the people in our localities.”
At SPARC, we fully acknowledge the challenges of data quality issues, as it is unlikely for any data to be 100% accurate. The data being collected over a year on paper, entered into Excel, and then analyzed through manual calculations will seem like any research programme’s nightmare. However, for us, it was important to engage the Mahila Milan actively into the process, to provide this as an avenue to develop thinking around health and wellbeing within their communities, and to help them take actions based on the findings of the data. Processes encouraging community participation in pushing for health systems improvement also need to acknowledge the long and painfully slow process of understanding the challenges faced by community residents. But it is imperative to drive motivation and collective action towards long term and sustainable change. To us, surveys as a process are an important tool in the hands of communities that enables their active participation in the process of learning. The data collected from it is only one of the byproducts that enhances that learning and pushes for action.
Image credit: SPARC