Interviews

What are interviews?

• Social surveys in the form of questionnaires are a popular means of collecting information , however, face-to-face or telephone interviews are also a widely used alternative.
• Both questionnaires and interviews gather information by asking questions, but interviews have the added feature of a social interaction between the interviewer and interviewees.
• This social interaction allows the interviewer to have an influence upon the answers the interviewee gives.
• 

Types of interview

• Different types of interview allow the researcher and the interviewee more or less freedom in the way in which the questions are asked and answered, as well as exploring additional areas of interest.
• Structured or formal interviews are very similar to a questionnaire: the interviewer is given strict instructions on how to ask the questions. The interview is conducted in the same standardised way each time, asking each interviewee precisely the same questions, word for word, in the same order, tone of voice and so on.
• Unstructured or informal interviews (also called
• discovery interviews) are like a guided conversation. The interviewer has complete freedom to vary the questions, their wording, order and so on from one interview to the next, pursuing whatever line of questioning seems appropriate at the time, asking follow-up questions or probing more deeply.
• Semi-structured interviews lie in between these two extremes. Each interview has the same set of questions in common, but the interviewer can also probe for more information. For example, Aaron Cicourel and John Kitsuse (1963) always followed up their questions with ‘How do you mean?’ as a way of gaining more information. Additional questions can be asked where the interviewer thinks it relevant.
• Group interviews are where up to a dozen or so people being interviewed together, instead of being one-on-one. Paul Willis (1977) used group interviews as part of his research into the ‘lads’ and schooling.

Group interviews

• Focus groups are a form of group interview in which the researcher asks the group to discuss certain topics(such as how pupils feel about the school) and records their views.
• Group interviews have their own particular strengths and limitations:
• Strengths:
• participants feel more comfortable in a group setting, and are more likely to open up
• in discussion, participants are likely to throw around ideas, which stimulates each other's thinking, which in turn, produces richer and more reflective data
• group interviews are useful way of generating ideas and then following up with further research
• researcher can combine questioning with opportunity to observe group dynamics and norms
• Limitations:
• one or two individuals may dominate the discussion, inhibiting others from contributing
• much depends on the researcher's ability to keep the group focussed on the research topic
• peer group pressure may coerce individuals to conform with group norms and not say what they really think
• data generated by group interaction is more complex and difficult to analyse

Structured interviews

• Structured interviews are like questionnaires: both involve asking people a set of prepared questions. In both cases, the questions are usually closed-ended with pre-coded answers.
• The main difference is that in the interview, the questions are read out and the answers filled in by a trained interviewer rather than by the interviewee.
• This basic similarity between structured interviews and questionnaires means that they share many of the same advantages and disadvantages.
• Where there are differences, these often come from the fact that structured interviews involve interaction between researcher and interviewee.
• 
• 

Practical Issues

• Advantages:
• Training interviewers is relatively straightforward and cost effective.
• Surveys that use structured interviews can cover quite large numbers of people with relatively limited resources because they are quick and fairly cheap to administer.
• Structured interviews are suitable for gathering straightforward factual information such as a person’s age or job.
• The results are easily quantified because they use closed-ended questions with coded answers. This makes them suitable for hypothesis testing.
• Disadvantages:
• More expensive to train interviews than to just send out a questionnaire, which both provides the same information.
• Structured interviews cannot reach as many participants as questionnaires.

Response Rate

• The large numbers who can be surveyed using structured interviews increase the chances of obtaining a representative sample of the population.
• Although the numbers that can be studied are lower than for questionnaires, structured interviews generally have a higher response rate .For example, of the 987 people Young and Willmott approached for their main sample, only 54 refused to be interviewed. This may be because people find it harder to turn down a face-to-face request, and some may welcome the opportunity to talk.
• Response rates can be increased if the interviewer can make several call backs to pursue those who fail to respond initially. However, this increases the cost of the survey. High response rates help to produce a more representative result and therefore a better basis for making generalisations.
• On the other hand, as with questionnaires, those with the time or willingness to be interviewed may be untypical (for example, they may be lonely or have time on their hands). If so, this will produce unrepresentative data and undermine the validity of any generalisations made from the findings.

Reliability

• If a method is reliable, another sociologist could repeat the research and get the same results.
• Structured interviews are seen as reliable because it is easy for the researcher to standardise and control them. They can ensure that each interview is conducted in precisely the same way, with the same questions, in the same order, with the same wording and tone of voice.
• If each interviewer conducts every interview in exactly the same way, then any other researcher following the same interview procedures should get very similar results.
• The structured interview provides a ‘recipe’ for repeating the research: as in cookery, anyone who follows the recipe ought to get the same result.
• The fact that all interviewees are asked exactly the same questions also means that we can compare their answers easily to identify similarities and differences.
• 

Validity

• A valid method is one that provides a true, authentic picture of the topic being researched.
• Critics of structured interviews argue that, like questionnaires, they often produce a false picture of the subjects they are trying to study.
• For instance, structured interviews usually use closed-ended questions that restrict interviewees to choosing from a limited number of pre-set answers. If none of these answers fits what the interviewee really wishes to say, the data obtained will be invalid.
• Structured interviews also give interviewers very little freedom to explain questions or clarify misunderstandings. For example, they may be given one alternative form of words to use if the interviewee doesn’t understand the question, but if this fails to do the trick the interviewer usually has to move on to the next question.
• What tends to happen if the interviewee doesn't understand a question is that they will give an answer anyway despite the lack of clarity. People rarely like to say that they do not understand something, so some of the answers given may be invalid for this reason.
• People may also lie or exaggerate, which produces blatantly false data.
• The interview is also a social interaction and so there is always a risk that the interaction between interviewer and interviewee will influence the answers given.
• For example, gender and ethnic differences can affect the answers, as can the interviewee’s desire to be seen in a favourable light.

Inflexibility

• Like self-completed questionnaires, structured interviews suffer from the inflexibility that comes from having to draw up the questions in advance.
• In doing so, the researcher has already decided what is important – yet this may not coincide with what the interviewee thinks is important.
• As a result, the findings may lack validity because they do not reflect the interviewee’s concerns and priorities.
• In particular, establishing the questions beforehand and then sticking to them rigidly will make it impossible to pursue any interesting leads that emerge in the course of the interviews, thereby losing valuable insights.
• Also like questionnaires, structured interviews are merely snapshots taken at one moment in time, so they fail to capture the flowing, dynamic nature of social life – unlike participant observation, for example.
• 

Feminist Criticisms

• Hilary Graham (1983) argues that survey methods such as questionnaires and structured interviews are patriarchal and give a distorted, invalid picture of women’s experience. She argues that:
• The researcher, not the female interviewee, is in control of the interview and decides the line of questioning to be followed. This mirrors women’s subordination in wider society.
• Survey methods treat women as isolated individuals rather than seeing them in the context of the power relationships that oppress them.
• Surveys impose the researcher’s categories on women, making it difficult for them to express their experiences of oppression, thus concealing the unequal power relationships between the sexes.
• These feminist criticisms are similar to those put forward by interpretivist sociologists, who argue that structured interviews fail to reveal how the interviewee sees their situation.
• Graham argues that sociologists need to use methods that allow the researcher to understand women’s behaviour, attitudes and meanings. She therefore advocates the use of direct observation instead of structured interviews.
• Other feminists favour unstructured interviews, which enable the researcher to build a more equal and collaborative relationship based on trust, empathy and support.

Young and Willmott

• Used structured interviews in their research into the extended family in east London, 1962;
• ‘The general sample being much larger – 933 people – we could not do the interviewing ourselves; these interviews were carried out by other interviewers we employed for the purpose. The interviews were formal and standardised, the questions precise and factual, with a limited range of alternative answers, on straightforward topics like people’s age, job, religion, birthplace etc. The interviewers’ task was to ring the appropriate codenumber opposite the answer they received or, at a few points in the interview, to write in a fairly short and simple reply. Each interview took between about ten minutes and half an hour, depending on the number of relatives possessed by a particular informant.’
• Of the 987 people Young and Willmott approached for their main sample, only 54 refused to be interviewed.

Unstructured Interviews

• Whereas a structured interview follows a standardised format, in an unstructured interview the interviewer has complete freedom to vary the interview. Supporters argue that this brings a number of important advantages.
• While structured interviews are criticised for their lack of validity, unstructured interviews are widely seen as a way of gathering valid data, enabling researchers to get a deeper understanding of the interviewee’s world.
• However, despite their strengths, using unstructured interviews as a method of collecting data has a number of disadvantages.
• 

Rapport and Sensitivity

• The informality of unstructured interviews allows the interviewer to develop a rapport (relationship of trust and understanding) with the interviewee. This is more likely to put the interviewee at their ease and encourage them to open up than a formal structured interview.
• A good example of this is the work of William Labov (1973). When using a formal interview technique to study the language of black American children, Labov found that they appeared to be tongue-tied and ‘linguistically deprived’. However, adopting a more relaxed, informal style – the interviewer sitting on the floor, the child allowed to have a friend present – brought a completely different response. The children opened up and spoke freely, showing that they were competent speakers.
• Unstructured interviews are particularly useful when researching sensitive topics.
• For example, Dobash and Dobash used them to study domestic violence. The empathy and encouragement of the interviewer will help the interviewee to feel comfortable discussing difficult or personal subjects such as abuse.

The Interviewee's View

• Because there are no set questions, unstructured interviews allow the interviewee more opportunity to speak about those things they think are important.
• This contrasts with the structured interview, where the researcher decides in advance what questions are worth asking and limits interviewees to a fixed range of possible answers.
• By allowing them greater freedom to express their views, an unstructured interview is more likely to produce fresh insights and valid data. Similarly, the interviewer’s probing can help formulate and develop interviewees’ thoughts more clearly.
• In their study of claimants’ experiences of unemployment, Hartley Dean and Peter Taylor-Gooby (1992) used unstructured tape-recorded interviews, lasting up to 90 minutes, with 85 claimants. In their words:
• ‘Questions were not put in any set order; the wording of questions was adapted to fit the circumstances of the respondent and/or the interview situation; questions which were evidently inappropriate were omitted altogether; additional questions or prompts were used at the interviewer’s discretion to clarify or develop themes as they emerged.’
• This approach gives interviewees the freedom to talk in their own terms about the issues that concern them.

Checking Understanding

• In structured interviews, there is a great danger that the interviewee misunderstands the question, or the interviewer misunderstands the answer.
• A major advantage of unstructured interviews is that they make it much easier for interviewer and interviewee to check each other’s meanings.
• If the interviewee doesn’t understand a question, it can be explained.
• Similarly, if the interviewer is unsure what the interviewee’s answer means, follow-up questions can be put to clarify matters.
• 

Flexibility

• Unstructured interviews are highly flexible.
• The interviewer is not restricted to a fixed set of questions in advance, but can explore whatever seems interesting or relevant.
• The researcher can formulate new ideas and hypotheses and then put them to the test as they arise during the course of the interview.
• There is no need to go away and draw up a new interview schedule, as there would be if using structured interviews.
• 

Exploring Unfamiliar Topics

• With structured interviews, researchers need to have some knowledge of the subject and preferably also a clear hypothesis before they start interviewing; otherwise, they will have little idea of what questions to ask.
• However, where the subject is one that we don’t yet know much about, unstructured interviews may be more useful, precisely because they are open-ended and exploratory.
• As with an ordinary conversation, we can start out knowing nothing and, by asking questions, learn as we go along.
• Some sociologists use unstructured interviews as a starting point to develop their initial ideas about a topic before going on to use more structured methods of investigation such as questionnaires.
• 

Practical Problems

• Time and sample size - Being in-depth explorations, unstructured interviews take a long time to conduct – often several hours each. This limits the number that can be carried out and means that the researcher will have a relatively small sample compared with the larger numbers who can be studied using structured interviews or questionnaires.
• Training - Training needs to be more thorough than for someone conducting structured interviews. The interviewer needs to have a background in sociology so they can recognise when the interviewee has made a sociologically important point and so they can probe further with an appropriate line of questioning. All this adds to the cost of conducting unstructured interviews.
• Interpersonal skills - Interviewers also need good interpersonal skills so they can establish the rapport that is essential if interviewees are to answer fully and honestly.
• Because unstructured interviews use mainly open-ended questions, the answers also cannot be pre-coded.
• This makes it very difficult to count up and quantify the numbers of interviewees giving this or that answer. In turn, the lack of quantitative data makes unstructured interviews less useful for establishing cause-and-effect relationships and hypothesis testing that positivists prefer.

Theoretical Problems

• Representativeness
• The smaller numbers involved mean it is more likely that the sample interviewed will not be representative. This means that it will be harder to make valid generalisations based on the findings of the interviews.
• Reliability
• Unstructured interviews are not reliable because they are not standardised. Each interview is unique: interviewers are free to ask different questions in each case if they feel it is relevant to do so.
• This makes it virtually impossible for another researcher to replicate the interviews and check the findings or compare them with their own.
• Validity
• Unstructured interviews are generally seen as producing valid data.
• However, critics argue that the fact that they involve an interaction between interviewer and interviewee inevitably colours and distorts the information obtained.
• As we have seen, structured interviews also face the same problem, even if not to the same extent. We examine the problems of interviews as interactions next.

Positivism and Interpretivism

• Positivists favour structured interviews because they achieve the main positivist goals of reliability, generalisability and representativeness:
• Standardised questions and answers produce reliable data because other researchers can replicate the interview.
• Pre-coded responses allow us to produce quantitative data, identify and measure behaviour patterns, and establish cause-and-effect relationships.
• Structured interviews are often large scale and thus more representative.
• Interpretivists reject structured interviews because they impose the researcher’s framework of ideas on interviewees.
• Interpretivists favour unstructured interviews because they achieve the main interpretivist goal of validity:
• Absence of a pre-set structure means interviewees can discuss what is important to them.
• Open-ended questions allow interviewees to express themselves in their own words, thereby producing qualitative data that gives us an insight into their meanings.
• Positivists reject unstructured interviews because each one is unique and cannot be replicated.