I'm conducting a comparative case study of a community health worker pilot in three sites in England. I want to identify barriers and facilitators to the implementation through interviews with professional stakeholders and focus groups with service users and the community health workers. I'll collate routine quantitative data from the pilot and local sociodemographic data to contextualise the cases.
I've found a survey instrument that measures the appropriateness, acceptability and feasibility of an intervention (Weiner, et al., 2017. Psychometric assessment of three newly developed implementation outcome measures. Implementation Science, 12(1), pp.1-12) and I'm trying to decide whether to use this alongside the stakeholder interviews. I'll seek to qualitatively assess appropriateness, acceptability and feasibility through the interviews, so I'm concerned that would I be duplicating myself for questionable added value. Doing the survey runs the risk of wasting time that could be better spent talking to the interviewees, and increasing my workload on what is already a big project. On the other hand, the data might be useful for triangulation.
More generally, I'm wondering what the value of a survey like this might be in a case study. The data wouldn't be generalisable as the sample isn't probability based and the sample would be very small (in this case, perhaps about 30 people if everyone who consents to interview agrees to complete the survey too). I've read elsewhere that surveys are useful in case studies for triangulation, but I'm not convinced!
If you want to use the scores to make statistical comparisons across the three sites, then you will need to determine the number of interviews that you need to achieve this purpose. The preferred way to do this is to assess the N needed for your tests to have sufficient "power" to detect a significant difference, and the tool to do this G*Power, which you can find online. If your N is too small to detect an effect (which is often the case in qualitative research), then there probably isn't any reason to collect this data.
Andrew Riley
An excellent survey can assist you in developing a picture of the clients' experience working with you. During the interview, you can ask appropriate questions to assist fill in the specifics. The information you hope to gather through your survey is a mix of qualitative and quantitative data.
The goal of case study survey research is to determine the characteristics of a group by asking a series of questions about a topic to people in that community. The veracity of the information acquired, on the other hand, is dependent on participants' honesty and desire to participate.
Its a great idea to use survey within case if the research questions demand such data and method
Thank you David L Morgan, Fatemeh Khozaei and Ronald Yesudhas. I really appreciate your input on this question. I won't be able to make statistical comparisons between the cases as the surveys are not sufficiently powered. However, as Fatemeh suggests, they will be a useful companion to the qualitative data, and will perhaps usefully indicate differences in attitude between participant groups (implementors, service users, community health workers).
- Badges
- Science topic
- Similar topics
- Statistics
- Survey