As a survey methodologist who spends a lot of time working in the “real world,” my philosophy is always: Respondents first. Respondents’ comfort and satisfaction with the survey experience should be a priority; after all, without them, we would have no data! Below, some points to keep in mind when conducting survey research in the health care arena.
Who has the answers you want? Think long and hard about this. You might believe you want to talk to doctors—but maybe nurses, or pharmacists, have a better understanding of the “front line” issues. A focus group with patients who had a great experience will be an easy experience; but if you want to learn about areas your hospital needs to improve, the harder work with unsatisfied patients is imperative.
Choose your methodology wisely. Think about who you are talking to you and what you want to know. Do you want busy nurses to show you what their day is like? Consider a mobile phone survey, where on-the-go respondents can include pictures and videos in their responses. Are you interested in the experience of late-stage cancer patients? Recognize that they may be too sick to respond, and you could do better with a proxy audience, such as caregivers.
Keep it short. This is a recommendation for almost all surveys, but especially when thinking about busy healthcare professionals or patients who may be juggling many competing demands on their energy. Prioritize what you need to know and leave the “nice to know” items on the cutting room floor.
Speak your respondents’ language. Use the words they use. Don’t speak a layperson’s language if your audience will relate better to medical terminology. Pre-testing a survey is a great way to find out if your language is appropriate.
Consider cultural contexts. Really think about who your respondents are and how they will relate to your survey. The survey methodology literature has documented that Hispanics tend to have an extreme response style when compared to whites, and there are best practices for question and scale design which can enable more valid interpretation of results from surveys among this group.
