Research Tips: Interview Data Collection

Abstract

Interviewing is a popular way of gathering qualitative research data because it is perceived as talking, and talking is natural. Interviews do not presuppose any statistical knowledge, and persons to interview, called respondents, might be close at hand and willing. On the other hand, Flinders (1997) offers several limitations to interviews: People interviewed may not be able to say what they think, may not have an opinion, or may not be able to state their opinion in a clear way; individuals available for interviews may not have the desired information; and respondents may be unwilling to discuss what they know. In addition, interviewing requires high level questioning skills and an active interpretation.Issues and DecisionsHitchcock and Hughes (1995) list eight interview types, but this column will discuss only the type most often used in educational evaluation: the semistructured interview. A semistructured interview means questions are predetermined, but the interviewer is free to ask for clarification. When conducting interview evaluation, several issues and decisions need to be made.1. Decide whom to interview. Spradley (1979) suggests a person with a history of the situation, who is currently in the situation, and who will allow adequate time to interview them. Begin by explaining the interview's purpose, for example, I'm interested in X, and I'd like to know your opinion.2. Choose when to stop a particular' interview, as well as how many interviews are enough. For an individual interview, Wolcott (1995) suggests stopping when the data desired are elicited.3. Select a place for the interview. The final report should describe the location in enough detail for readers to get an idea of the interview context.4. Decide which questions to ask. Spend time on question formation, and even write them down beforehand.5. Consider how the data will be collected: for example, listen only, listen and take notes on the spot, listen and take notes later, or tape record the interview.Piloting and FeedbackA pilot interview is a practice interview that can serve many purposes: getting started, practicing interview questions, and getting feedback on the topic as well as the interview method. The resulting feedback may be positive or negative. After a pilot interview, for instance, notes can be read to decide if a tape recording is necessary, and locations can be changed if the noise level is too high.Analyzing Interview DataThe interview data are not only the literal words from a respondent but include evaluator assumptions, biases, and questions. The interview is cocreated between the evaluator and the respondent. It is the job of analysis to give concrete form to the meaning of the interview which is called the interpretation.The words of the interview constitute raw data, somewhat like the numbers resulting from a test. Raw data does not in itself reveal its meaning; rather it must be interpreted. Hitchcock and Hughes (1995) describe two main strategies to analyze interview data. The first strategy is to become very familiar with the data; the second strategy is to create meaning using analytical categories.In the first strategy, the evaluator becomes familiar with the data-depending on how it was collected-by going over notes many times, listening to tapes repeatedly, or constantly reading and rereading the interview transcripts. The idea is that, as the evaluator becomes familiar with the data, slowly but surely categories or become apparent. The data are reviewed, and the evaluator begins to see that the respondent has been talking about theme A, theme B, and so on. Pondering on these themes, the evaluator finally comes to understand (interpret) that the respondent is talking about X. In this way of looking at analytical categories, the categories are grounded in the data. Grounded means they emerge from the data and reflect the data. …