Before conducting interviews, you need an interview guide that you can use to help you direct the conversation toward the topics and issues you want to learn about. Interview guides vary from highly scripted to relatively loose, but they all share certain features: They help you know what to ask about, in what sequence, how to pose your questions, and how to pose follow-ups. They provide guidance about what to do or say next, after your interviewee has answered the last question.
A good interview guide also acknowledges four important facts of human social interactions that influence what people are likely to say to you. These four facts are: (1) Research questions are not the same as interview questions; (2) People's espoused theories differ from their theories-in-use; (3) Interviews are social occasions; and (4) Testimony by itself is relatively weak form of evidence. This guide to interview guides offers some techniques for accommodating these four important facts.
1. RESEARCH QUESTIONS AREN'T INTERVIEW QUESTIONS
The first important fact of interviewing is that research questions are not the same as interview questions. Your research question describes the issue you want to learn about, but you rarely can learn about that issue by asking others that literal question. If you want to learn why students bully one another, you can't just ask them, "Why do you bully him," or "Why do you think he bullies you?" Research questions are usually too broad to serve as productive interview questions. Once you have a research question, you must devise a data collection plan that will help you gather credible evidence, or clues, that are relevant to your research question. Your interview guide is your data collection plan. My human caretaker has posted a sample of an interview guide to show you how you move from research questions to interview questions.
2. IF YOU ASK A QUESTION, THEY WILL ANSWER IT
The second important fact about interviewing is that people will answer the questions you ask them, even if they have never really thought much about your topic. If they agree to be interviewed, they will continue to try to be helpful by offering whatever they can about your topic, even if it means inventing answers or exaggerating how much they have thought about your question. This means that the "evidence" you are gathering may not very accurately reflect real views. So you need to think about ways to pose questions that don'telicit overly-helpful responses. One strategy that helps a lot is to have a collection of probes ready to use as needed. A probe is a follow-up question, designed to get the interviewee to clarify or elaborate what he or she has just said. I have another page that gives several ideas about types and formats of probes.
Approach your question from the side
One way around this is to approach your topic sideways. For instance, suppose you are interested in whether or how teachers use research. If you ask them to speak specifically about this issue, they will try their best to come up with some uses, refer to studies they have read and talk about how they have responded to those studies. These responses will give you an inflated notion of the degree of reliance on research. An alternative approach would be to ask how they make decisions about particular things-things where research could have been useful--say, deciding whether to promote a student or retain him in grade, or whether to use cooperative groups or not, and then probe extensively for where they got their ideas when they made that decision. If you are lucky, they will refer to research studies when they answer these probes.