Tips for formulating interview questions

Asking questions of interviewees in ways that help them tell their stories is something of an art. It goes without saying that it is good practice to be well-prepared for interviews. This includes thinking about the physical setting for an interview and the technology one will need to record an interview. And there are so many choices: what will be used — a digital recorder, cell phone, tablet, pen and paper, telephone or synchronous online meeting room? If you are using a digital technology for the first time, it is always useful to test it, and check that the file form can be easily downloaded and made accessible for transcribing. This is because some digital recorders do not have a means to download files. Yet even before setting up the interview and asking the first question, one needs to have thought about the topics one wants to learn about, how these relate to the research questions posed, and the kinds of questions that might elicit information about those topics. In this blog post, I discuss tips for formulating interview guides.

When planning a study, I find it useful to draw up a matrix in which I relate the research questions to what I hope to learn, i.e.,

By listing the topics that relate to a particular research question, I am able to do some brain-storming about the kinds of interview questions that might be asked. Also, I have found that reflecting on what it is that I hope to learn from asking a particular question helps me think through what it is I really want to learn from each research question.

There are many kinds of interview questions that might be posed. Michael Q. Patton’s matrix of interview questions is helpful too (Patton, 2002, p. 352). Let’s explore a simple example based on Patton’s question matrix. Let’s say I wanted to examine the following research topic:

The purpose of this study is to examine the experiences and perspectives of online teachers in higher education with respect to course design and implementation.

In conjunction with reviewing previous literature, I will need to formulate research questions. In this example, these questions have not been informed by review of the literature – this means that if I was actually going to conduct this study, these questions might not prove to be the most pressing questions to ask given what is already known from prior research. This speaks to the significance of the study, or put simply the “So what?” question. Nevertheless, for the purpose of showing how interview questions are formulated, here are some hypothetical research questions that might be posed:

• How do online teachers in higher education describe their preparation for learning to teach online?
• How do online teachers in higher education describe their course development and implementation?
• What kinds of instructional support do online teachers in higher education perceive to be beneficial?

Using Patton’s question matrix, with different areas of focus, I can draft questions that relate to past experience, present experience, and future plans. This matrix then, provides all kinds of options for how questions might be posed about different sorts of phenomena (i.e., peoples’ behaviors, experiences, opinions, values, feelings, emotions etc.). A study might focus on one aspect of lived experience over another (e.g., past experience).

You will notice that these questions frequently invite participants to tell a story, through the use of “Think of a time…” prompts. When conducting qualitative interviews, it is important to ask open-ended questions. With the exception of the “background information” questions, I have avoided asking closed questions that seek factual information, or yes/no responses. If I wanted that kind of information, a survey might be a more efficient means to generate data.

You’ll notice that the “background” information sections seeks specific information. Sometimes I have used a single sheet to gain this kind of information (e.g., demographic information, educational qualifications, work experience etc.), and have asked participants to complete that at the end of an interview, rather than asking interview questions about that. This will ensure that the same type of information about each participant is collected.

You will also notice that these questions ask only one question at a time. For example, if one were asking opinion questions about the “best” or “worst” experience in relation to a phenomenon, it would be preferable to ask separate questions about these rather than ask: “Tell me about your best and worst experiences with respect to x“.

In these sample questions, I have tried to avoid embedding assumptions in interview questions. Yet, you can see that several questions above do include assumptions in them. Sometimes this can’t be avoided, since the research topic focuses on a particular experience (here, online teaching in higher education). For example, the question “With the increase in use of mobile technologies, what might a new online teacher need to know about how to integrate these?” assumes that a participant would have some knowledge on the topic of the use of mobile technologies. Depending on how participants for the study are recruited and selected, this question might be irrelevant. Thus, if this kind of question were posed to interviewees who had been recruited without respect to having had experience in the use of mobile technologies, the potential response might be, “I don’t know.” Of course, responses of “I don’t know” are also data, and a research study might be interested in learning about online teachers’ knowledge and expertise of mobile technologies. Again, this indicates how the questions posed in an interview relate directly to the kinds of research questions asked in a study (which relates in turn to what is already known from prior research on a given topic).

Of course, even when interview questions are carefully thought through, an interviewer will not really know how these will actually work in practice until one has completed some interviewers. Even when interview protocols work well with some interviewees, for all kinds of reasons, some individuals might not interpret questions in the way that a researcher has envisioned. These sorts of misunderstandings can be worked out within the interview itself. I also like to try my interview guides out in a pilot interview or study prior to using these, since some questions might need to modified.

I hope these tips help in thinking about ways to draft interview guides.

As always, all the best with your research.

Kathy Roulston

Reference

Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.