Appendix G: Methodological learnings
Reflections from the research team
The research team at Ipsos MORI identified a number of lessons from this study that could be applied to future research:
1. The SHS recontact database is an effective source for identifying participants living in households categorised as being in fuel poverty. It is the only source of contacts for which the fuel poverty estimates have already been carried out, therefore by using this sample source we knew that everyone we contacted had already been categorised as living in fuel poverty. If using other methods such as face-to-face free find recruitment, from research panels, or via gatekeeper organisations, we would not know if potential participants fitted the fuel poverty definition. The SHS database also allowed us to effectively target different groups to match the groups we wanted to include in the research (e.g. those categorised as in extreme fuel poverty, those with long term health conditions, those with children aged 5 and under, etc.)
2. While the SHS was an ideal sample source, time had passed between participants taking part in the SHCS and taking part in this research. This therefore presented the risk that their situations may have changed, meaning they may no longer have been in fuel poverty. At recruitment stage it was therefore important to clarify if their circumstances had changed significantly since they took part in the SHS. If they told us they were better off since they took part in the SHS, they were screened out and not asked to participate.
3. In designing the research materials, it was helpful to refer to previous research (as outlined in the Evidence Review) to identify the types of topics that we would want to cover in these interviews, including those where there was an evidence gap (e.g. in relation to use of smart meter and in relation to policy ideas).
4. Conducting the interviews over two stages was helpful for this type of lived experience research. The first interview helped to establish a rapport with participants, which was particularly important as we were asking them to discuss relatively sensitive topics such as their financial circumstances, their health, and how these made them feel. It also helped to shape the focus the content of the second interviews, where we could pick up on points raised in the first interview.
5. The use of a heating diary did not provide the level of detail we expected it would. Not all participants completed the diary (mainly because they simply forgot to) and several of those that did provided the same information for each day in the diary (e.g. the content of each day simply said that the heating was on all day, and that they felt comfortable). While it helped provide a stimulus for some of the discussions around heating patterns, it did not necessarily add much to the findings. More in-depth and engaging methods, such as mobile app diaries which ask participants to upload content daily including photo and video content, may have provided richer findings.
6. We had a research team of four, which felt the right size. With 40 interviews being completed, having four team members meant that interviews could be shared in a way that did not overly burden any member of the team, but also meant that the findings were not influenced by the perspective of one or two researchers.