Beyond Surveys: Mixed-Methods Surveying for Better Data Quality in Remote Indigenous Communities

The final speaker in this mixed-methods session at the ACSPRI 2024 conference is Alexandra Gregory, whose work is on mixed-methods research with Indigenous communities in the Northern Territory. In the first place, her work uses surveys with Indigenous populations in remote areas, but this has significant limitations: such populations are unlikely to be included in survey design, misalign with culturally appropriate communication styles, and require further qualitative elements and adaptation.

Remote communities in the Northern Territory are small communities with a mostly Indigenous population; one example of a survey with such communities surveyed local mothers about a nurse home-visitation programme, and included formal surveys as well as open-ended questions. Survey questions were also read out in order to address possible literacy limitations, and the study also included recorded interview conversations.

Alexandra’s project examined 27 interviews to explore departures from the script, further dialogue exchanges, and the contexts surrounding them; some of these were to ensure accurate data (specifying participant responses to survey items, and facilitating participants’ understandings of items), and building connection and understanding (exploring participant context, or spontaneous sharing of participants’ stories).

Specifying responses was especially important in determining the numerical response to an item which needed to be answered on a Likert scale, for instance; building connection and understanding saw interviewers ensure that answers were appropriate to the question, and participants express through stories how they would answer a survey question. This is at odds with traditional survey design and method, but such conversation at a qualitative level helps ensure the immediacy and reciprocity in discussion, improves the accuracy of survey responses, and creates a deeper understanding of numerical results.

All of this is in aid of meaning-making and social interaction, and reduces the risk of missing data through missing or invalid responses. It does require greater resources, timelines, and interviewer skills, however, and a consideration of new approaches to doing surveys in remote communities.