Speaker 1: In this episode of Off the Shelf, I'm going to talk about interviewing in social science. In addition to some definitions and theory, I'll give you a brief step-by-step guide to how researchers organize, conduct, and use interviews in qualitative research, and talk about some skills that you can practice to become a more competent and confident interviewer. I'm trying to tell you I'm not your enemy, I'm a scientist. You know, I'm something of a scientist myself. Social scientists study the beliefs and behavior of human beings in a number of different ways. We use methods like participant observation and covert ethnography, as well as studying history, examining personal correspondence and administrative documentation, using questionnaires and surveys, or occasionally through forms of social experimentation. But when our goal is to understand how people construct or articulate meaning, and make sense of the world around them, interviews become a necessary, if not always completely sufficient investigative tool. Unlike surveys, which are more quantitative, the goal of most qualitative interviewing is not to extract information by questioning your participants, but instead to better understand how your subjects view the world and themselves by engaging them in dialogue. It's very common in classroom teaching to group every type of interview into three general categories—structured, semi-structured, and unstructured—and the differences between these help us articulate the roles that interviewing can play in social scientific research. Structured interviews are typically used in surveys and employ the research logic of questionnaires. Researchers ask the same questions in the same way and in the same order in every single interview. This allows for a quantitative—that's to say, more mathematical—comparison of each subject's responses. At the other end of the spectrum, there are unstructured interviews, which are open-ended conversations often without predetermined questions that focus on a specific topic but give the participants a great deal of freedom regarding what they discuss and how they represent themselves. And finally, semi-structured interviews fall somewhere between these two poles. Here the researcher does have a series of prepared questions they'd like to discuss, but the sequence and wording of those questions can vary, and the researcher is encouraged to ask follow-up questions and use forms of active listening to more deeply explore specific topics of neutral interest. Now there are some problems with this taxonomy that I'll discuss in the video description below, but taking this at face value, semi-structured interviews are by far the most common method you'll encounter in qualitative research. They're also the method that I have the most experience with personally. So for the rest of this video, we're going to focus on how social scientists prepare, conduct, and use semi-structured interviews in their research. In my experience, students sometimes come to interviewing because they think of it as an easy and quick solution to the rigors of qualitative research, which is a bit misguided. To use interviews as sources of social scientific data, we have to, minimally speaking, go through these steps. The first thing we do is conceptualize and design a research project with a specific set of goals. Whatever your project, you have to be able to articulate the role that interviewing will play in your research and why interviews are an appropriate method of collecting data for your stated research objectives. And it may sound obvious, but you would be amazed by the number of students that jump into fieldwork without being able to fully articulate the goals and parameters of their project, and that can have a very negative impact on the quality of their data. Once we're able to do that, we go on to establish access and make contact with our participants. Now, I talk a lot about gatekeepers and establishing access in my Outlander Ethnography series, which I'll link in the video description, but making contact is quite a bit more complicated than it sounds, and there are a few general guidelines here that you should follow. To begin with, theorists and educators often advise that you avoid interviewing people with whom you have a close personal relationship. And it's not that you can't interview those people under any circumstances, but it's much harder to maintain an edict or outsider's scientific perspective when you're speaking with people that you already know. You might, for example, avoid a follow-up or slant a question, or in some way distort the interview process because of your concern for the relationship or what the interview might reveal. And that's what we call a potential research confound, something that introduces new variables to the research situation, the impact of which can be difficult or even impossible to measure. A related point is that when you do make contact, whether by phone, email, or really preferably in person, do it yourself. Beyond the most basic introductions, avoid having other people make contact on your behalf. According to the author Irving Seidman, one of the most important parts of making contact is establishing the groundwork for mutual respect. By taking the time to speak with them and introduce yourself and your study, you can demonstrate that you take both your participants and your research seriously, which is a valuable first impression to make. And as part of those preliminary conversations with your participants, it's also very important to begin the process of informed consent, which is vital to maintaining ethical research standards and to describe in the broadest terms possible the general subject and goals of your research. At this point, students sometimes ask an important question. If I'm using interviews, how many people should I interview? How do I know when enough is enough? And faced with that question, there are a number of different methods that we use, some of which I'll discuss in the video description below. But for the purposes of this video, I'm recommending two basic criteria as an answer to that question, sufficiency and saturation of information. Sufficiency is more quantitative. Are your numbers sufficient to reflect the range of participants, the diversity of opinions and beliefs, and the sites that make up your research environment? And that's a question that you'll have to answer for yourself, but you don't have to do it in a vacuum. You can and should read other successful studies that employ similar methods and use those publications as a watermark against which you can measure your research design and results as they begin to emerge. By contrast, saturation of information is more qualitative. Researchers like Jack Douglas, for example, have noted that at a certain point, interviewers begin to hear the same information reported again and again by their participants. So you can reach a point where expanding the scope of your interviews doesn't necessarily provide additional usable data. Now neither one of these is entirely adequate on its own, but keeping sufficiency and saturation of information in mind, you'll be better equipped to determine when you have enough usable data for your project. Ultimately, though, as Irving Seidman writes, the idea of enough is an interactive reflection of every step of the interview process and is different for each study and each researcher. So as you prepare a project, keep in mind that you have to be able to articulate where you draw that line and why. And if you're ever in doubt, always err on the side of too much information rather than too little. With research design and preliminary contact out of the way, let's talk about interviewing itself. To begin with, qualitative interviews are often recorded. There are some authors that caution against recording, but the process of transcription, taking the oral interview and transforming the spoken conversation into a written, word-for-word record of what was said, is an important part of qualitative interviewing. This not only preserves what was said, but gives us a source of original data that we can return to to check for accuracy or use to defend ourselves if we're ever accused of misrepresenting our material. So as a rule, record your interviews. And always, always ask for permission from your participants before you begin. I mentioned that semi-structured interviews make use of both pre-prepared questions and more free-form follow-up questions, and there are two rules that apply to both of these that we have to learn. The first is to avoid leading questions. These are questions that, based on how they're asked or their wording, will influence the direction, tone, or word choice of the response. You see that type of thing in gotcha political interviews or in legal cross-examinations on TV, where the goals and stakes are very different. Again, in qualitative interviewing, we're not trying to provoke a specific response. Our goal instead is to better understand the subject's perspective on a topic by getting them to speak openly, with minimal influence on how they choose to respond. With that in mind, the second rule is to ask open-ended questions. These are questions that don't presume an answer when they're asked, but instead establish the territory to be explored while allowing the participant to take any direction she or he wants in their response. For example, if you're interviewing students, avoid asking something like, How happy were you with the results of your exam? This pushes the participant to interact with the term happy as something to be accepted or rejected and may not reflect terminology or ideas that are shared by your subject. Instead, try to ask something like, How do you feel about the results of your exam? This gives the participant space to answer and opens up the conversation to follow-up questions as they elaborate in their own terms. And the key to asking follow-up questions is to let them develop naturally, as much as possible, from the conversation rather than from your preparation. Remember, the structure of interviews is cumulative. As an interview progresses, your questions should build upon what the participant says and how they say it. For instance, if your participant says that they were fascinated by or regret an experience, if it suits your purposes, these are both excellent points for a follow-up question asking for elaboration. The term fascinate, for example, often works to communicate an interest without revealing the specific nature of that interest or its relative value to the speaker. Seidman makes a point of investigating that type of terminology. By taking the subject's language seriously without making them feel defensive about their remarks, which is very important, we can deepen the conversation and implicitly encourage the subject to speak more openly on a particular topic. And when you're asking follow-up questions, the golden rule is don't interrupt. Your subject should always feel that what they have to say is important, that they are valuable. So if they go off-topic, try not to interrupt them. Instead, in your research journal, write down a few keywords, and then follow up on the remarks that you want to discuss in a way that doesn't break the participant's train of thought. Finally, while conducting interviews, always consider the larger context in which the conversation takes place. So if you're meeting right before an important election, for example, or if the participant has recently undergone some type of personal tragedy, those things may influence their responses. Additionally, pay close attention to any extraverbal cues that seem pertinent to the topic of the interview. The context and cultural significance of things like a wink, a whisper, or an emphatic gesture are easily lost in transcription and can potentially provide important data. As Studs Terkel writes, a laugh can be a cry of pain, and a silence can be a shout. And part of our job as qualitative researchers is to have the knowledge and cultural sensitivity to understand the difference. So let's recap. The first thing we do is design a research project with a well-articulated set of goals. We then reach out and make contact with our participants, and when we do that, we make an effort to do it ourselves. During those preliminary conversations, we also begin the process of informed consent and make an effort to describe our project in broad terms so that our participants can understand what it is that we're doing and why their perspective is valuable to our research. As for the interviews themselves, it's important to prepare questions in advance, but focus on asking follow-up questions while you're speaking with your participants rather than sticking to a predetermined interview structure. Ask open-ended questions instead of leading questions, and always take the broader social context of the interview and immediate environment into account. And finally, assuming that we're recording our interviews, we then transcribe the conversation. We take the entire spoken interview and word-for-word transform it into a written document that we can then use as a source of qualitative social scientific data. And that's the interview process, rinse and repeat. It's time-consuming, labor-intensive, and sometimes also requires a fair amount of money to complete, but for the better part of 100 years, it's been one of the backbones of social science and is a method that's shared across numerous academic and applied social scientific fields. If you found that helpful, please do like and subscribe to the channel. It's one of the best ways you can help us grow, but we do also have a Patreon page where we publish scripts, reading lists, and additional written commentaries if you're interested in pursuing the topic of this video more deeply or any of our videos more deeply. I would also like to thank my current patron, UMSingular, and if you would like your name to appear here as well as a patron of independent social scientific education, come on over to the Patreon page and help us build our community. And until next time, never stop learning.