Interactive Data Collection 1: Interviews PDF

 4 Interactive data collection 1: interviews OVERVIEW...

 4 Interactive data collection 1: interviews OVERVIEW What are qualitative interviews? When and why would I use interviews? Designing and piloting the interview guide Issues to think about in relation to participants Preparing for the face-to-face interview Conducting the face-to-face interview Preparing for and conducting the virtual interview What to do when interviews go badly Interviews are everywhere. On any given day, we might listen to an interview with a politician on the radio while eating breakfast, read an interview with a celebrity in a magazine on the bus to university or watch members of the public being interviewed on the local evening news. The social prevalence of interviewing (according to US sociologists Gubrium & Holstein, 2002, we live in an ‘interview society’) means that interviews are perhaps the most familiar data collection tool both for new qualitative researchers and for participants. Interviews are certainly one of the most common meth- ods of data collection within the social and health sciences (Briggs, 1986), and the most common qualitative method of data collection. There are a number of different styles of qualitative interviewing, including narrative (Mishler, 1986), active (Holstein & Gubrium, 1995, 1997), grounded theory (Charmaz, 2002) and feminist (Oakley, 1981) approaches. In this chapter we outline a generic approach to qualitative interviewing (that can be adapted according to specific requirements). We define interviewing as a ‘professional conversation’ (Kvale, 2007), with the goal of getting a participant to talk about their experiences and perspectives, and to capture their language and concepts, in relation to a topic that you have determined (Rubin & Rubin, 1995). 04-Braun & Clarke_Ch-04-Section 2.indd 77 28/02/2013 7:35:21 PM  78 Successfully collecting qualitative data WHAT ARE QUALITATIVE INTERVIEWS? The semi-structured interview (sometimes called the interview guide approach; Patton, 2002) – our focus in this chapter – is the dominant form for qualitative interviews (see Box 4.1). In this approach, the researcher has prepared an interview guide before the interview, but does not rigidly adhere to it, either in terms of the precise word- ing of questions, or the order in which questions are asked. For example, compare a planned question from Victoria’s research on lesbian and gay parenting (‘What do you think about the argument that children need male role models?’) with the actual question she asked a lesbian couple who talked about their decision to co-parent with a gay male friend (‘Is it important for you for your kids to have male input? Is that one of the reasons why you chose a donor?’). Question wording and order are contextual, and responsive to the participant’s developing account. In the words of US sociologists Rubin and Rubin (1995: 42), the ideal qualitative interview is ‘on target while hanging loose’. Participants are given the opportunity to discuss issues that are important to them and that the researcher hasn’t anticipated, and aren’t on the interview guide, so the researcher needs to be flexible. BOX 4.1 STRUCTURED, SEMI-STRUCTURED AND UNSTRUCTURED INTERVIEWS Interviews are often divided into three types: Structured: the questions and the response categories are predetermined by the researcher; this is the commonest type of interview in quantitative research. Semi-structured: the researcher has a list of questions but there is scope for the participants to raise issues that the researcher has not anticipated; this is the commonest type of interview in qualitative research. Unstructured: the researcher has, at most, a list of themes or topics to discuss with the participant, but the interview is strongly participant-led; this type of interview is used by some qualitative researchers. What this tripartite typology overlooks is that all interviews – indeed all social encounters – are structured in some way. Even in supposedly unstructured interviews, the interviewer (mostly) asks questions and the interviewee (mostly) responds to them. Some researchers have instead distinguished between standardised or closed (quantitative) interviews (where the response categories have been pre-determined by the researcher) and reflexive or open (qualitative) interviews (where the responses are determined, to a greater or lesser extent, by the participant) (Hammersley & Atkinson, 1989). The distinction between these is shown in Table 4.1. 04-Braun & Clarke_Ch-04-Section 2.indd 78 28/02/2013 7:35:22 PM  Interactive data collection 1: interviews 79 Partly because the qualitative interview arose as a method in response to cri- tiques about the ‘depersonalisation’ of (then) standard social scientific methods of data collection (Oakley, 1981), face-to-face contact between researcher and partici- pant has typically been viewed as the ideal way to collect interview data; the ‘gold standard’ (Novick, 2008). Virtual interviews are often viewed as a (poor) substi- tute for face-to-face interviews. However, increasingly, telephone, email and online interviews are regarded as different types of interview method, as extensions of the traditional method rather than substitutes for it, which have particular strengths and weaknesses in their own right (Sturges & Hanrahan, 2004). They are seen to provide the qualitative researcher, in the correct circumstances, with effective data generation tools. We discuss both face-to-face and virtual modes of interviewing in this chapter. In a qualitative interview, the researcher asks the participant a series of (ideally) open- ended questions, and the participant responds using their own words. When interviewing face-to-face, the researcher and participant have a spoken conversation, which is typi- cally audio-recorded and the recording is transformed into written text, ready for analy- sis, through a process of transcription (see Chapter 7). The same basic process applies Table 4.1 Comparison of standardised and qualitative interviews Standardised interviews Qualitative interviews The ideal interviewer is a ‘robot’ Interviewers, like participants, are individuals with their asking each participant exactly the own particular interview style; question wording and the same questions, in exactly the same order in which questions are asked vary according to the way, in exactly the same order personal style of the interviewer and the responses of the participant All of the questions are prepared in An interview guide is prepared in advance, but the advance ideal qualitative interview is flexible and responsive to the participant; good interviewers follow up on unanticipated issues and ask spontaneous and unplanned questions Closed (yes/no) questions are Open-ended questions are preferred to encourage widely used participants to provide in-depth and detailed responses and to discuss what is important to them Response categories are determined The goal of an interview study is to capture the range in advanced by the researcher and diversity of participants’ responses, in their own words Every effort is made to minimise the The interviewer plays an active role in the interview, impact of the interviewer on the co-constructing meaning with the participant. It is participants’ responses neither possible nor desirable to attempt to minimise the interviewer’s role. The interviewer should reflect on how their practices and values may have shaped the data produced 04-Braun & Clarke_Ch-04-Section 2.indd 79 28/02/2013 7:35:22 PM  80 Successfully collecting qualitative data for telephone interviews (see Burke & Miller, 2001; Miller, 1995); the possibilities of software like Skype for conducting spoken virtual interviews are also being explored (Hay-Gibson, 2009). Virtual written interviews can be conducted synchronously (e.g. using instant messaging or chat software) or asynchronously, meaning there is a variable gap between question and response (e.g. via email) (Mann & Stewart, 2002). Online/ email interviews require no transcription as the software preserves a record of the inter- view dialogue. Some of the guidance we provide in this chapter will seem like rules that must be followed. However, ultimately you should aim to develop your own interviewing style (Rubin & Rubin, 1995). A qualitative interviewer is not a robot, precisely programmed to conduct every interview according to a set of inviolate rules. Rather, a qualitative interviewer is a human being, with a distinctive personal style, who uses their social skills, and flexibly draws on (and, in some cases, disregards) guidance on good inter- view practice to conduct an interview that is appropriate to the needs and demands of their research question and methodological approach, the context of the interview and the individual participant (Table 4.1 provides a comparison of qualitative and quantita- tive interviews). A well-conducted qualitative interview can generate amazing data, and you’ll get rich, detailed and quite often unanticipated accounts if you use the method well (Table 4.2 discusses the strengths and weaknesses of qualitative interviews). Table 4.2 S ummary of the strengths and limitations of (face-to-face) qualitative interviews Strengths Limitations Rich and detailed data about individual Time consuming for researchers to organise, conduct experiences and perspectives and transcribe Flexible: you can probe and ask unplanned Lack of breadth because of smaller sample sizes questions (compared to a qualitative survey study) Smaller samples: you often need only a Not necessarily ideal for sensitive issues: some people small number of interviews to generate feel may more comfortable disclosing sensitive adequate data (see Table 3.2 in Chapter 3) information in a group setting or in an anonymous survey Ideal for sensitive issues: a skilled Time consuming for participants: an interview often interviewer can get people to talk about takes at least an hour to complete sensitive issues Lack of anonymity: may be off-putting to some Accessible: can be used to collect data participants, especially those who are ‘hard to engage’ from vulnerable groups such as children in research (see Chapter 3) and people with learning disabilities Not necessarily ‘empowering’ for participants: Researcher control over the data produced participants have less control over the data increases the likelihood of generating produced (compared to qualitative surveys and email useful data interviews) 04-Braun & Clarke_Ch-04-Section 2.indd 80 28/02/2013 7:35:22 PM  Interactive data collection 1: interviews 81 WHEN AND WHY WOULD I USE INTERVIEWS? Interviews are ideally suited to experience-type research questions (see Table 3.1 in Chapter 3). For example, British feminist psychologists Celia Kitzinger and Jo Willmott used interviews to research the experiences of women with polycystic ovarian syndrome (see Illustrative Research Example 4.1). Interviews can also be useful for exploring under- standing and perception- and construction-type research questions – such as (supposedly) healthy men’s constructions of their health-promoting practices (Sloan, Gough, & Connor, 2010). In this case, they are best suited to exploring understandings, percep- tions and constructions of things that participants have some kind of personal stake in – people without a personal stake in a topic are unlikely to generate the rich and detailed responses you want from interviews. Focus groups (FGs; see Chapter 5) or qualitative surveys (see Chapter 6) are better methods when people don’t have a personal stake in the topic. Interviews can also be used to explore practice-type research questions, such as the clothing practices of fat women (Colls, 2006). Grounded theorists have used interviews to answer influencing factor-type questions, such as the factors that influ- ence people’s decisions to continue regular genetic screening (Michie, McDonald, & Marteau, 1996). DESIGNING AND PILOTING THE INTERVIEW GUIDE Good preparation is the key to the successful use of interviews in qualitative research. Your first task is to design your interview guide – the series of questions that will guide your ‘conversation’ with the participant (see Material Example 4.1; the companion website includes other examples of interview guides from our research). A good guide will enable you to build trust or rapport with the participant – a key com- ponent in interactive data collection (Reinharz, 1993) – so they feel comforta- ble disclosing personal information to you. Rapport and well-planned questions are important for generating rich and detailed accounts relevant to your research question. Start developing your guide by brainstorming a list of questions relating to the areas you are interested in (Smith, 1995); if relevant, some or all can be informed by, or adapted from, previous research. Then reflect on the following issues: Opening and closing questions: start the interview with an introducing question (e.g. ‘So erm why don’t we start by you telling me something about yourselves and your family?’; see Kvale & Brinkmann, 2009); end the interview with a closing or ‘clean-up’ question. Clean-up questions allow the participant to raise issues that are important to them that haven’t already been covered (e.g. ‘I think that’s basically everything I had to ask you to talk about, um have you got anything else you’d like to say or any kind of final thoughts or any things you’d like to follow up that I haven’t asked you?’). Sometimes clean-up questions trigger really useful unan- ticipated data. 04-Braun & Clarke_Ch-04-Section 2.indd 81 28/02/2013 7:35:22 PM  82 Successfully collecting qualitative data ILLUSTRATIVE RESEARCH EXAMPLE 4.1 Women’s experience of polycystic ovarian syndrome British feminist psychologists Celia Kitzinger and Jo Willmott (2002) noted that research has failed to explore women’s experience of polycystic ovarian syndrome (PCOS), a condition marked by symptoms such as infertility, ‘excess’ hair growth, irregular or no menstrual cycles, weight gain, acne, male pattern hair loss and excess androgen production. Jo Willmott conducted 30 in-depth, semi-structured, audio-recorded interviews with women with PCOS, all recruited through a national PCOS self-help group (she shared her experiences of living with PCOS with the participants). The participants were largely white (24 white, six non-white) heterosexual women aged between 21 and 42, with a mean age of 29. Interviews were mostly conducted in the women’s homes and lasted 45 to 90 minutes. The interview guide ‘was deliberately broad based and wide-ranging, and was designed to allow women to tell their own stories, rather than adhering to a strict structure’ (p. 350). Prompts, probes and follow-up questions were used to ‘elicit breadth and depth in responses’ (p. 350). The guide was piloted on two women, and the final version directed women to talk about: how they came to be diagnosed with PCOS; what their symptoms were and how they dealt with them; how they felt about having PCOS and how it affected them; issues around disclosing their condition to others. The authors used thematic analysis (see Chapter 8) to analyse the data and were interested in ‘women’s qualitative experience of their PCOS, and in the issues they themselves raised in relation to it’ (p. 351). They let the data suggest names for themes, and took direct quotations from the transcripts to illustrate the kind of data classified by each theme. The analysis was based on the assumption that what women say is evidence of their experience. Their key finding was the women experienced themselves as ‘freaks’ and as not proper women. The women’s concern with ‘normal’ femininity often focused on hair, periods and infertility: ‘Bearded ladies and hairy monsters’ – women were upset and embarrassed by their facial and body hair and took steps to remove it; they contrasted their own (perceived) hairy, monstrous bodies with the imagined bodies of normal women. ‘Irregular women’ – women often experienced their periods as ‘freakish’ or ‘abnormal’ because they were irregular; regular periods were desirable either as a sign of womanhood or fertility. ‘Infertility:“My whole purpose of being a woman was gone’’’ – women expressed very strong feelings about actual or possible infertility. Although ‘excessive’ hair growth and lack of periods was distressing, infertility was ‘crushing’ and added to women’s feelings of ‘freakishness’ and of not being ‘real’ women. Kitzinger and Willmott concluded that the women in their study were challenged in their perceptions of themselves as feminine and as women and called for further research. 04-Braun & Clarke_Ch-04-Section 2.indd 82 28/02/2013 7:35:22 PM  Interactive data collection 1: interviews 83 MATERIAL EXAMPLE 4.1 Interview guide from Victoria’s research on lesbian and gay men parenting (see Clarke, 2006; Clarke, 2007) To start with, I’d like to know more about having children as a lesbian/gay man. Can you tell me what it is like for you? Can you tell me about your family? (Do you use the word ‘family’?) Who counts as your family? For couples: How do you divide up child care? (Is one of you the main carer?) Did you choose to be a parent or did it just happen? If you chose it, why? (Is your choice related to your sexuality?) Have you always wanted to have children? If it ‘just happened’ – what do you think/feel about this? Do you think your family is ‘different’ in some way? Why? In what way(s)? Why not? Are you raising your children differently? Why? In what way(s)? Why not? What differ- ence (if any) do you think it makes for your children to have two female/male parents/a lesbian parent/a gay parent? Do you think your family/your having children challenges any stereotypes? Why? Which one(s)? Why not? Can you tell me about any challenges/issues you have faced? Any challenges/issues specific to being a lesbian/gay parent? How have you resolved them? (Have you resolved them?) What is the most positive thing about raising children as a lesbian/gay man? The most negative? Have you encountered prejudice (as a lesbian/gay parent/family)? Have your children? How have you/they dealt with it? Do you discuss your sexuality with your children? How have they reacted to your sexuality/‘coming out’ as a lesbian/gay man/_________? What do you think about the argument that children (especially boys/girls) need male/female role models? What are your hopes for your children’s future? How do your relatives feel about you being a lesbian/gay man (raising children)? What impact (if any) does lesbian and gay parenting have on the lesbian and gay community? What impact (if any) does it have on the wider society? Do you go out on the ‘scene’; spend time in lesbian and gay community spaces, etc.? Have you ever encountered any hostility in lesbian and gay community because you are a parent? How do you think your life is different from that of lesbians/gay men who don’t have children? What advice would you give to a young lesbian/gay man thinking about whether or not to have children? Is there anything else you would like to add? Ask me? 04-Braun & Clarke_Ch-04-Section 2.indd 83 28/02/2013 7:35:22 PM  84 Successfully collecting qualitative data Sequencing of questions: organise your questions so that they flow logically and cluster into topic-based sections. Early questions should be less probing, sensitive and direct than later questions. For example, in Victoria and colleagues’ research on money man- agement and commitment in first time heterosexual marriage, the interviews started with ‘gentle’ questions about how the participants met their partners, how and when they decided to get married, what marriage and commitment meant to them, before moving onto to more sensitive questions about the couples’ money management practices (Burgoyne, Clarke, Reibstein, & Edmunds, 2006). Another useful hint is to funnel questions: the ideal interview guide is often conceived of as an inverted triangle, mov- ing from the general to the specific. Structuring questions (Kvale & Brinkmann, 2009) can be useful to signal a shift to a new topic area (e.g. ‘Er you’ve already started talk- ing a bit about money and that that’s kind of what I’d like to talk about next’). These features can be seen in the interview guide provided in Material Example 4.1, from Victoria’s research into lesbian and gay parenting. Constructing and wording questions: wording is vital for developing effective interview questions. Poorly worded questions (see Box 4.2) can damage rapport and subsequent data collection. Expect to draft and redraft your interview guide: in first drafts questions are often too direct (Smith, J.A., 1995), too closed and too leading. Questions should be redrafted until they are ‘gentler and less loaded but sufficient to let the respondent know what the area of interest is and recognize that he or she has something to say about it’ (Smith, J.A., 1995: 15). Trying out ques- tions on someone else and asking for advice on level of difficulty and tone (Smith, 1995) is very useful. Prompts and probes: in addition to your main questions, prompts and probes encourage par- ticipants to open up, expand on their answers and provide more detail. In the example inter- view guides on the companion website, the main questions are often bolded and the probes aren’t. Notice that the probes are often more explicitly worded than the main questions, but they wouldn’t be asked in this way. Probing can also take the form of ‘an expectant glance’ or an ‘um hm, mm, or yes, followed by an expectant silence’ (Fielding & Thomas, 2008: 251) or a specifying question (Kvale & Brinkmann, 2009), which spells out the type of further detail required. You don’t need to prepare probes for every question, just when you think it would be helpful (Smith, J.A., 1995).  esearch questions are not interview questions: in Victoria Clarke and Kevin Turner’s R (2007) research on non-heterosexual visual identities, the research question was ‘What role do dress and appearance play in the development and maintenance of lesbian, gay and bisexual identities?’ They didn’t pose this question to participants, but asked more focused and concrete questions, such as ‘Do you think there is a stereotypical look for gay men?’ Participants’ responses to such questions allowed Victoria and Kevin to answer their research question. S ocial desirability: finally, think about whether you are likely to only get obvious, socially desirable responses to particular questions; these may not provide useful data. 04-Braun & Clarke_Ch-04-Section 2.indd 84 28/02/2013 7:35:22 PM  Interactive data collection 1: interviews 85 Once you have a polished draft of your interview guide, review it by asking yourself the following questions: What am I trying to find out with this question? Will it generate that information? Don’t make the mistake of confusing opinions and feelings; don’t ask ‘How do you feel about that?’ when you want to know about the participant’s views and opinions (Patton, 2002). Does this question help me to answer my research question? Are there (problematic) assumptions embedded in this question? What would it feel like if I was asked this question? How are participants from different backgrounds likely to feel if asked this question? Is this question likely to be meaningful to my participants? This is really important – for instance, the first question on Virginia’s sex in long-term relation- ships interview guide (on the companion website) wasn’t meaningful to partici- pants, and failed to generate any useful data. If your interview questions don’t stand up to this scrutiny, you’ll need to rework them. Time spent honing and refining your guide is important. As British health researcher Jonathan Smith (1995: 12–13) noted: Producing a schedule beforehand forces you to think explicitly about what you think/hope the interview might cover. More specifically, it enables you to think of difficulties that might be encountered, for example, in terms of question word- ing or sensitive areas and to give some thought to how these difficulties might be handled. Having thought in advance about the different ways the interview may proceed allows you, when it comes to the interview itself, to concentrate more thoroughly and more confidently on what the respondent is saying. Testing out your interview guide on a trusted friend or colleague is useful, but in small projects there’s limited scope for formal piloting – where interviews are con- ducted to test the guide but no data are collected. Instead, we recommend thoroughly reviewing your guide after the first few interviews – are you getting the kind of data you need to address your research question? As a qualitative interview guide doesn’t need to be treated as fixed at the start of data collection, it can evolve across the entire data collection process if new issues arise (Charmaz, 2002). Questions might be reworked, or removed, or new questions added to the guide during this review; the entire guide might be reorganised. ISSUES TO THINK ABOUT IN RELATION TO PARTICIPANTS Different groups of participants can raise particular issues that need to be considered for interviews: Interviewing people you know: in qualitative research it is perfectly acceptable to interview someone you know, such as your housemate, a friend or a work colleague; 04-Braun & Clarke_Ch-04-Section 2.indd 85 28/02/2013 7:35:22 PM  86 Successfully collecting qualitative data BOX 4.2 DESIGNING EFFECTIVE INTERVIEW QUESTIONS Ask open questions: this is the most important guidance for designing effective qualitative interview questions. Open questions, which avoid yes/no answers and encourage participants to provide detailed responses and to discuss what is important to them, are key. Try starting questions with phrases like: ‘Can you tell me about …?’, ‘How do you feel about …?’, ‘What do you think about...?’ At the same time, avoid questions being too open and not providing participants with enough direction. For example, from Victoria’s lesbian and gay parenting interview study (e.g. Clarke, 2006): Victoria: Erm I don’t know anything about you at all so erm why don’t you we start by you saying something about your family and your situation Mary: Me or Jane ((inaudible)) Victoria: Both of you Mary: Oh both of us [...] Mary: Erm what sort of things do you want to know personal things or how long we’ve been together Victoria: Well cos I- I don’t know if you’ve got who’s got kids or ((inaudible)) Jane: Right we’ve got five Ask ‘non-leading’ questions: you are interested in participants’ perspectives, so avoid putting words into their mouths by asking leading questions. Hypothetical questions (Patton, 2002) that invoke other people’s views (‘Some people think that...’) or imagined scenarios that project into the future or invite participants to view an issue from someone else’s standpoint can be a good way of asking about controversial issues. Ask singular questions: questions that ask about multiple things can be confusing for participants (which part of the question do they answer first?), and you can miss collecting important data because participants only answer one part of the question. For example (from Virginia’s research on sex in long-term relationships, Terry & Braun, 2009): Virginia: Mmm so what were you, what, when you felt bad what was, what sort of things was it, how did it make you feel, like was it worried about the relationship or about him or about yourself or Ask short questions: long and complicated questions can be confusing for participants. Don’t ask questions with double-negatives: these can also be confusing (‘So you’re not not getting married?’). Ask clear and precise questions: avoid ambiguity (especially when asking about intimate and sensitive topics such as sexuality); don’t assume that participants will share your understanding of widely used and assumed-to-be commonly understood terms (such as ‘sex’). 04-Braun & Clarke_Ch-04-Section 2.indd 86 28/02/2013 7:35:22 PM  Interactive data collection 1: interviews 87 Ask linguistically appropriate questions: ensure the wording of questions and use of terminology is appropriate to your participant group. Avoid using jargon and complex language (you don’t want people to feel stupid); also avoid overly simplistic language that leaves your participants feeling patronised. Rubin and Rubin (1995: 19) advise that ‘by being aware of your own specialized vocabulary and cultural assumptions, you are less likely to impose your opinions on the interviewees’. Ask non-assumptive questions: avoid making assumptions about your participants. For example, imagine you are interviewing ‘average-sized’ women about their feelings about their bodies. If you ask a question like ‘Can you imagine how you would feel if you were a size 20?’ you indicate that you assume your participant has never been a size 20 – because you ask her to imagine what that would be like. Although such presumptions can potentially be damaging to rapport, and result in poorer data, assuming shared knowledge or asking ‘presumptive’ questions works sometimes (Patton, 2002). But it should not be done unintentionally. For example (from an interview with a Muslim couple): Victoria: Mm-hm what about Christmas? What did you do about Christmas with presents and stuff like that buying presents for each other and? Farah: Erm ((pause)) we don’t actually celebrate Christmas Victoria: Oh of course of course you don’t then Ask empathetic questions: avoid questions that overtly or covertly criticise or challenge participants (‘You’re not into country music are you?’), and questions that can be perceived as threatening: ‘why’ questions can be interpreted in this way (instead of ‘Why do you...?’ try ‘What are your reasons for...?’). these are known as ‘acquaintance interviews’ (Garton & Copland, 2010). But you enter into a ‘dual relationship’ with that person (e.g. they are your friend and a par- ticipant), which raises some additional ethical considerations and some inflexible ‘dos and don’ts’. Don’t use your pre-existing relationship to pressurise someone to participate in your research or to disclose information in the interview. If a friend discloses something in the interview that is new to you, that information should remain confidential to the interview (unless your friend raises it again later). At the same time, don’t gloss over (relevant) information that isn’t new to you; only the audio recorded information counts as data, not the things you happen to know about your friend from other sources. If the person you are interviewing is someone you have a hierarchical relationship with (e.g. you’re their therapist or line-manager), you need to be very sensitive to the ways in which your position could be experi- enced as potentially coercive, even if you don’t think it is. The delicacy of this issue is such that ethics committees often require information about pre-existing relation- ships with participants. 04-Braun & Clarke_Ch-04-Section 2.indd 87 28/02/2013 7:35:23 PM  88 Successfully collecting qualitative data Interviewing strangers: in some ways interviewing strangers is easier than inter- viewing people you know because you don’t have to manage a dual relationship (it may also feel easier for the participant to disclose personal information to a stran- ger). However, it can be more difficult to establish rapport and feel comfortable having an in-depth and possibly intimate social encounter with someone you don’t know. Building rapport and putting the participant at ease are a priority; make sure the participant has more time than the interview will take so you don’t have to immediately jump into it, without rapport-building pre-interview chit-chat. Finally, safety is a greater concern when interviewing strangers, especially if interviewing people in their homes (see Chapter 3). Interviewing across difference: some researchers argue that some participants feel more comfortable disclosing (sensitive) information to someone who is broadly similar to them, meaning that for an effective interview, it is important to ‘match’ the major social characteristics of the participant and the interviewer (see Sawyer et al., 1995). So, a woman researcher will interview female participants (DeVault, 1990; Reinharz, 1992) and a black male researcher will interview black male participants (obviously, this conceptualisation of ‘matching’ overlooks the ways in which two women, or black men, can be different; see Riessman, 1987). Sometimes a par- ticipant may request, or only agree, to be interviewed by someone from a similar background (see Kitzinger, 1987). This type of interviewer–participant ‘matching’ may be possible in large, well-funded interview studies; it is rarely practical or possi- ble in smaller-scale research. This means that most qualitative interview researchers have experience of interviewing people who are different from them in important ways. However, ‘interviewing across difference’ is a complex issue, and you should undertake further reading in this area if you plan to do this (e.g. Blee, 1998; Dunbar, Rodriguez, & Parker, 2002; Reinharz & Chase, 2002; Riessman, 1987; Schwalbe & Wolkomir, 2002; Scully, 1994). In interviewing people who are socially margin- alised in ways that we are not, we are researching and representing the experiences of the ‘other’ (see Wilkinson & Kitzinger, 1996) and this requires extra caution and sensitivity (discussed further below). On a practical level, communication can pose challenges: ‘even when the interviewer and interviewee seem to be speaking the same language, the words they use may have different connotations’ (Rubin & Rubin, 1995: 18). Furthermore, the liberal view of ‘pretending difference doesn’t exists’ is not always helpful: people are different, and it is important not to assume that other people experi- ence the world in the way that we do. At the same time, it is also important not to assume that our differences shape every aspect of our lives. Power in interviews: the notion of ‘interviewing across difference’ draws our atten- tion to the power relationships that shape interviewing. The relationship between researcher and participant is typically conceived of as a hierarchical one (this is more explicit in quantitative and experimental research, where participants were tra- ditionally called ‘subjects’), with the researcher in control of the interview. Others argue that control isn’t inherent in the position of the interviewer but is something 04-Braun & Clarke_Ch-04-Section 2.indd 88 28/02/2013 7:35:23 PM  Interactive data collection 1: interviews 89 that is achieved between the researcher and the participant in the interview itself (for an excellent discussion of this, see Russell, 1999). Participants may perceive you as an ‘expert’ and for some participants your status as a researcher will override other aspects of your identity and experience, such as a shared experience of an eat- ing disorder or a shared sexual identity (Clarke, Kitzinger, & Potter, 2004). Some researchers seek to challenge the interviewer–participant hierarchy and empower participants through a process of empathic interviewing (see below, and see also Oakley, 1981; Reinharz, 1992). At the very least, it’s important to be aware of the exploitative potential of interviews. Interviewing people who occupy societal positions of greater or lesser power than you adds another dimension to the power relationship between interviewer and participant. You hold a position of power as the researcher, but if you are a (young) student interviewing high-status individuals such as consultant surgeons or heads of large companies, there is potential for you to feel vulnerable, for them to domi- nate the interview (in ways that are unhelpful) and for you to lose control of the interview (see Odendahl & Shaw, 2002). A number of feminist researchers have shared their experiences of interviewing men and boys and feeling vulnerable as women in the process. For example, Kathryn Lovering (1995: 28), an experienced British school teacher, youth worker and educational psychologist, has written about her experiences of losing control of FG discussions with adolescent boys when the boys’ discussion became ‘smutty, sexist and oppressive’; ‘all I wanted to do was get out of the room as fast as possible!’ There are no easy or obvious ‘solutions’ to such dynamics – simply trying to avoid any power differences is not a solution, and would limit samples to those who reflect the characteristics of researchers (which is already a problem with the dominance of the ‘usual sus- pects’ in psychological research; see Chapter 3). The key is to be prepared, to have read about other researchers’ experiences (e.g. Edwards, 1990; Lee, 1997; Thomas, 1995; Willott, 1998) and to have given some thought to how you will manage the power dynamics of the interview in advance. Participant distress: it’s not unusual for participants to become distressed when discussing sensitive issues, so don’t be too anxious if it occurs. Manage participant distress effectively by acknowledging people’s distress (‘Are you okay? Would you like to stop the interview for a while?’) and allowing them to express it, but also ‘containing’ it within the context of the ongoing interview. Don’t stop the interview and start thrusting tissues at people at the merest hint of tears; in our experience people are usually happy to continue the interview after taking a moment to collect themselves. It may be appropriate to back off from a topic that is causing distress, and tentatively returning to it later in the interview (‘Would it be okay to talk a bit more about X, or would you rather move onto talking about Y?’). Interviewing vulnerable people: certain groups of people are recognised to be more vulnerable than others, and we recommend that new interviewers only interview groups such as children (see Docherty & Sandelowski, 1999; Eder & Fingerson, 04-Braun & Clarke_Ch-04-Section 2.indd 89 28/02/2013 7:35:23 PM  90 Successfully collecting qualitative data 2002), people with learning disabilities (Swain, Heyman, & Gillman, 1998) and older people with dementia (see Russell, 1999; Wenger, 2002) if they have professional experience with the participant group, because interviewing vulnerable groups requires additional skills and experience and has different ethical requirements. PREPARING FOR THE FACE-TO-FACE INTERVIEW First, if you’re new to interviewing, it’s really important to test out and practice your interview technique: a trusted friend or colleague can help. Interviewing is one domain where practice makes perfect, so don’t be put off by critical feedback. Interviewing is challenging. It involves lots of multi-tasking: listening to what people are saying, being attentive to their tone of voice and body language (‘Do they seem uncomfortable with this topic; should I back off?’), mentally ticking off questions on your guide as the participant speaks (‘They’ve already addressed question seven, so I don’t need to ask that, but I must follow-up on this issue …’), spotting relevant information in participant responses and asking unplanned follow-up questions (Kvale & Brinkmann, 2009), keeping an eye on the recording equipment – the list goes on! Practise your opening blurb (about your study, ethi- cal issues and consent; see below), test out your recording equipment, so you are familiar with exactly how it works, and try to memorise your interview guide as much as possible. That’s one less thing to worry about! PLANNING THE DATA COLLECTION PROCESS EFFECTIVELY Once you have designed your research materials (interview guide, consent form, participant information sheet [PIS], demographic form; see Chapter 5 and the companion website for examples), it’s time to plan for data collection. The participant is being very generous with their time and is willing to share with you intimate details of their life, so make participating as easy as possible. Negotiate a time for the interview that is convenient for the participant and be considerate of them in other ways. For example, if you have a hor- rible cold but feel well enough to conduct the interview, you probably want to give them a chance to decide if they want to be interviewed by a sniffing, coughing researcher! Give participants a clear idea of how long the interview is likely to last so they allow plenty of time. Most qualitative interviews last around an hour, plus about 30 minutes for pre- and post-interview chat, negotiating consent, completing demographic forms and so on, or even longer if you are meeting in the participant’s home; we have on occasion spent over four hours with a participant. Think about scheduling. We do not recommend doing more than one interview in a day for a few reasons. Conducting a good interview requires intense focus, which is tir- ing, and if you do more than one a day, you can miss asking questions, or following up on points, because the content of different sessions mix in your mind (Rubin & Rubin, 1995). Furthermore, this sort of research can be emotionally draining (Hallowell et al., 2005); not over-doing data collection is part of maintaining your own wellbeing as a researcher. 04-Braun & Clarke_Ch-04-Section 2.indd 90 28/02/2013 7:35:23 PM  Interactive data collection 1: interviews 91 Ideally, avoid scheduling lots of interviews very close together. If you are transcribing your own interviews, schedule your interviews so that you have time to transcribe each interview as soon as possible after conducting it – interviews are a lot easier to transcribe when they are fresh in your mind (it generally takes about eight–ten hours to transcribe one hour of audio recording thoroughly; see Chapter 7). This also allows you to reflect on, and adjust, your interviewing style and questions before your next interview (Rubin & Rubin, 1995). LOCATION, LOCATION, LOCATION Select or negotiate a location for the interview in which the participant feels comfort- able and you feel safe (some ethics committees may require certain participants to be interviewed on university grounds). It should be as quiet as possible, with little or no background noise (so you both can concentrate and the recording is crystal clear), little in the way of distractions (e.g. windows with lots of people walking past, posters on the walls), comfortable seats, a table/surface to place the recording equipment on, and con- venient for the participant. If participants have concerns about anonymity, and/or don’t want anyone they know to find out that they are participating in your study, you need to give extra consideration to a suitable (private) location. It’s often easiest to book a room at university; the advantages of such a space are that it’s an environment you choose and control, and it’s safe. Disadvantages include that it’s a somewhat formalised/intimidat- ing setting that some participants may feel uncomfortable in, it can be sterile, it may be hard for some people to find, and parking/access may be an issue for people outside of the university (provide really clear instructions). If you are interviewing people you know, it’s often easier to pick a suitable location and safety can be less of a concern (it’s okay to interview people you know well in your own home, if this suits the participant). If you are interviewing people you don’t know in their homes or workplaces (or in a community centre or very quiet café), you need to think carefully about your safety. We recommend the ‘safety buddy’ procedure for off-campus interviews (see Chapter 3; see also Boynton, 2005). If you’re going into someone else’s space, think about practical issues like pets, chil- dren, televisions and housemates. We have encountered aggressive Staffordshire bull ter- riers, wheezing asthmatic Yorkshire terriers who insisted on sitting on the microphone, house rabbits, housemates who were planning to stay in the interview room to watch television, noisy televisions, screaming toddlers, and so on. If you feel anxiety about dogs or other animals, or have animal-related allergies, check out the pet situation. If possible, schedule interviews for when small children are in day care or in bed. Don’t be afraid to politely ask participants to turn off the television for the sake of the recording, or if the interview can be conducted in a private room – if you don’t, those ‘distractions’ may hamper your data collection. Finally, in some cultures such as Māori or Pacific cultures in Aotearoa/New Zealand it is important to take a small gift of food; in other cultural con- texts, such as white British culture, this is not expected. If you are organising the interview space, a simple way to help participants feel wel- come is to offer refreshments on arrival, and during the interview: tea, coffee or a cold 04-Braun & Clarke_Ch-04-Section 2.indd 91 28/02/2013 7:35:23 PM  92 Successfully collecting qualitative data drink, and some sorts of (quiet) snack – like soft biscuits or grapes (avoid noisy snacks; crisps, crackers, apples and food with noisy wrappings are a no-go for interviews; with drinks, although disposable cups are less environmentally friendly than ceramic ones, they don’t create a ‘bang’ when put down beside a microphone). Always provide drinking water. Always have tissues in case the participant becomes distressed. Think about where you position the chairs in the room. Facing each other at an angle so the participant is not sitting directly opposite you is best. Use chairs that are roughly the same height – it can feel rather awkward if you’re sitting on a chair that’s much higher or lower than the one the participant is sitting in. If there is a table in the room, consider whether it would help or hinder the interview to have a physical ‘barrier’ between you and the interviewee. Either way, a table is useful for positioning the recording equipment. This might seem like a lot to consider, but these are all aspects of making the participant comfortable, and therefore conducting the most successful interview. AUDIO-RECORDING VERSUS NOTE-TAKING Because most qualitative researchers are interested in the detail of participants’ responses, and the language and concepts they use in talking about their experiences and perspectives, it is important to have a precise record of the interview. This is best achieved by audio-record- ing the interview and then producing a transcript of the audio-recording (see Chapter 7). Audio- recorders are now almost all digital, ranging from basic MP3s through to professional-level devices. Quality can be significantly impacted by the microphone. A built-in one may not be high enough quality, so test it and consider an external microphone (if your device allows it). Good quality devices will often have a light to indicate that the device is recording – keep an eye on it throughout the interview. Carry spare batteries and/or plug the device into the mains supply if possible. We recommend using two audio recorders so you have a back-up. The recorder should be positioned close to the participant, without being intrusive – and if interviewing in their space, ask permission before putting your equipment down, especially on an expensive looking coffee table! Always pre-check the equipment. If the interview is in your space, test the audio pick-up from different points in the room and with different speech volumes. When recruiting participants, ask them to read the PIS before agreeing to be inter- viewed (the companion website has an example of an interview study PIS); it’s important that they understand that their interview will be audio-recorded, and that they are both consenting to participate in an interview and consenting to being recorded. Virginia has interviewed people in different places around the world who, when she got there, refused to be recorded (saying they hadn’t been ‘warned’ – they had, in the PIS; they just hadn’t read it carefully enough). At best this can result in a poorer interview and data, at worst, if audio-recorded data are crucial for your analysis, wasted time – both for you and the participant – and expense. If we rely on written notes as the record of the interview, much of the richness and detail of the interview will be lost. Furthermore, it is difficult to develop rapport with a participant and conduct a successful interview if we are looking down and scribbling away in a note- book rather than focusing on them. But it’s wise to make brief notes as you’re going to keep 04-Braun & Clarke_Ch-04-Section 2.indd 92 28/02/2013 7:35:23 PM  Interactive data collection 1: interviews 93 a track of things to follow up on, or new questions to ask (tell the participants you might do this so they’re not thrown when you start making notes). It is also very useful to make ‘field’ notes after each interview, where you record details of the participant’s self-presentation and surroundings, and reflect on your personal reaction to the participant, how you think the interview went, important features of the participant’s responses, ideas for data analysis, additional questions to ask in subsequent interviews, things you need to work on with regard to your interview technique, and so on (see Box 3.8 in Chapter 3). PERSONAL DISCLOSURE The final thing to consider before the interview is your strategy around personal dis- closure. Personal disclosures during recruitment can encourage people to participate in your research. If you are a member of the group you are researching and feel com- fortable about ‘outing’ yourself, this can be important information to include on a PIS; likewise, it can be equally important to disclose that you are not a member of the group you are researching (see Asher & Asher, 1999). But what about personal disclosure in the interview? A research methods text from the 1960s gave the following advice about personal disclosure in (quantitative) interviews: ‘If he [the interviewer] should be asked for his views, he should laugh off the request with the remark that his job at the moment is to get opinions, not to have them’ (Selltiz, Jahoda, Deutsch, & Cook, 1965: 576). This strategy is not appropriate in contemporary qualitative interviewing! Some researchers think that personal disclosure is important to establish rapport and to chal- lenge the researcher–participant hierarchy (see Oakley, 1981, for a classic discussion of self-disclosure in interviewing); others think that personal disclosure on the part of the interviewer can create a false sense of intimacy and encourage the participant to over- disclose (Finch, 1984). Whatever your position, this needs to be considered in advance. See Box 4.3 for an example from one of Virginia’s PhD interviews where personal dis- closure was handled poorly, through lack of preparation (see Chapter 7 for a discussion of transcription notation). If you do make disclosures, though, keep them in check. The interview is not an opportunity for us to talk about ourselves or to form a new friend- ship (Cotterill, 1992); the disclosure must happen in the framework of a professional interview. Related to personal disclosure is the issue of personal presentation. Prior to the interview, give some thought to the image you present to participants – for most inter- views a relaxed but professional image is appropriate, but this will largely depend on the participant group. You should dress up for some groups (business people) and down for others (students) (see Boynton, 2005). CONDUCTING THE FACE-TO-FACE INTERVIEW With everything planned and prepped, you’re ready to start interviewing! Needless to say, there are many more things to consider during interviewing, not least managing your own nerves and monitoring potential participant distress. 04-Braun & Clarke_Ch-04-Section 2.indd 93 28/02/2013 7:35:23 PM  94 Successfully collecting qualitative data BOX 4.3 EXAMPLE OF AN INTERVIEWER RESPONDING BADLY WHEN ASKED A QUESTION BY A PARTICIPANT From the first interview Virginia conducted for her research on women’s experiences of their genitalia (Braun & Kitzinger, 2001; Braun & Wilkinson, 2003, 2005): Kim: What d’you think tell me some stuff about what you think Int: About Kim: ((laughs)) About on the subject of vaginas tell me what you know you’ve been asking me a lot a questions I’m wondering kind of Int: I have um (.) I’m not actually su- I’ve been thinking about this and I’ve been trying to clarify my own thoughts Kim: Mm Int: and in terms of doctors I see it very clinically Kim: Mm yeah Int: the whole thing an an an an and it’s quite interesting I I mean I would only (.) have a woman doctor and I don’t know why that is but it’s just what I’ve always Kim: Yeah Int: always wanted and I’ve always done and um (.) so it’s quite it’s quite interesting um (.) in in that the Kim: Mm Int: I mean I d- I would there very much kind of clinicalise it but in terms of other things like um (.) awareness or (.) they’re things I’m struggling all these things ((laughs)) I’ve been asking you are things I’m struggling with myself Kim: Mm Int: and I haven’t (.) um got any answers ((laughs))...[Virginia’s response continues in this vein for quite some time] OPENING THE INTERVIEW Once you have greeted your participant and thanked them for agreeing to take part, explain (again) what your research is about, its purpose and why you are conducting it (participants are often interested in our personal motives), and give them the opportunity to ask questions. Emphasise that there are no right or wrong answers to your questions and that you are interested in their views; they are the expert on their experiences. Next, negotiate consent – we generally do this by giving the participant a copy of the PIS (if they don’t have one to hand) and consent form, talking through the key information on the PIS, and giving them an opportunity to read the PIS and consent form, and ask us any questions. Only then do we invite them to sign the consent form. We give a copy of the PIS and consent form to the participant in case they want to refer to it after the interview. 04-Braun & Clarke_Ch-04-Section 2.indd 94 28/02/2013 7:35:23 PM  Interactive data collection 1: interviews 95 We would usually then ask the participant to complete the demographic form (remind- ing them that the provision of such information is voluntary), which gently gets them in the mode of ‘being the interviewee’ and gives us a final opportunity to re-check our audio-recording equipment. When we’re ready to start the interview, we usually ask the participant if they are happy to begin, and for us to switch on the audio-recorder. Once recording, we ask our opening question. It’s important to treat your interview sheet as a guide for conducting the interview and asking questions, not a recipe to be followed to the last gram! As we noted and demon- strated earlier, you tailor the wording of questions for individual participants, and the con- text of the interview; the ordering of questions should be responsive to the participant’s developing account. For example, if a participant raises something towards the start of the interview that you planned to ask about later on, it is appropriate to discuss it earlier; simi- larly, if something has already been covered by an earlier response, don’t feel you have to ask the question you have prepared on this topic just because it’s on your guide. You should treat interviews as a flexible tool, which are partly planned and partly spontaneous. Some ‘rules’ for question and guide design can be broken: some planned questions may not be asked and some unplanned (but nonetheless highly relevant) issues may be discussed. FOLLOWING UP AND GETTING PEOPLE TO TALK New interviewers need to be wary of the rushed and overly rigid interview. One of the trickiest aspects of interviewing is asking unplanned and spontaneous questions and getting people to open up and talk at length. If the person doesn’t seem uncomfortable talking about the topic, but has provided only a very short response to a question, don’t auto- matically rush on to the next question on your guide; stay with the question you have just asked. As well as using silence effectively (see below), to generate richer, more detailed responses, you can ask for examples (‘Could you give me an example...?’), clarification (‘What do you mean by...?’), specific details (‘How did that make you feel?’) (Rubin & Rubin, 1995), or simply ask for more information (‘Can you tell me more about that?’). If necessary, reassure the participant that they are the experts on their experiences, and that their views and opinions are interesting to you. If the participant talks about something that you hadn’t planned to ask about but is potentially relevant to your research question, ask more about it. Tailor subsequent questions to the specific circumstances of the par- ticipant and to their developing interview narrative. The key to all of this is memorising your interview guide and, more importantly, having a clear sense of your research ques- tion, so you can make on-the-spot decisions about whether information is potentially relevant to it and if you should ask for more or move onto the next question. Our students’ interviews often start out fairly rushed and rigid (sticking closely to the question wording and order on the interview guide) and become more relaxed and fluid (and produce richer data) as they gain more experience and confidence. INTEREST VS. EMPATHY For many qualitative researchers, the key to successful interviewing is to show inter- est in, and to appear non-judgemental about, what the participant is saying. Gaze, body 04-Braun & Clarke_Ch-04-Section 2.indd 95 28/02/2013 7:35:23 PM  96 Successfully collecting qualitative data language (a relaxed stance) and non-evaluative guggles like ‘mm’, ‘mm-hm’, ‘ah-ha’ can all be used to convey to the participant that you are actively listening to them and you want them to continue (equally, lots of ‘mm’s in quick succession can convey that you want them to stop). Some qualitative researchers advocate the value of personal disclosure and the use of evaluative comments like ‘Yeah, I know what you mean’ or ‘Yeah, something similar happened to me’ to signal that you agree with what the participant has just said, or that an experience is shared, and to build trust and rapport (see Oakley, 1981). But such an empathic approach assumes participants talk about things you ‘agree’ with (not all par- ticipants do, e.g. Scully, 1994), and becomes tricky if the participant comes to expect your affirmation. For example, Victoria once interviewed a woman who expressed the view that the Harry Potter novels were anti-Christian and damaging to children – at the time of the interview, Victoria was reading and loving one of the Harry Potter novels. It would have threatened rapport (and thus data collection) and been disre- spectful, for Victoria to say what she really felt in this moment; instead she showed non-judgemental interest in the participant’s perspective. This response was appropri- ate; she was conducting the interview as Victoria-the-researcher, rather than Victoria-the- person. Whether you show interest and/or empathy can depend on the topic and purpose of the interview, the participant group (empathic approaches are best used when you are an ‘insider’ researcher or interviewing someone you already know; see Garton & Copland, 2010) and your particular interviewing style (see Rubin and Rubin, 1995, for a detailed discussion of the pros and cons of interest versus empathy in interviewing). AVOIDING ‘DOING EXPERT’ While we are ‘experts’ in the sense that we are trained researchers and know a lot about our research topic, the participants are the experts on their experiences, views and prac- tices. Yet they may look to us for expertise; a very effective way to close down an inter- view is to assert such authority and ‘do expertise’ about an issue. As an example, in her lesbian and gay parenting research, Victoria conducted FGs on the meaning of family; in one, a participant expressed uncertainty about whether gay couples could marry. Victoria found herself giving an (unsolicited) mini-lecture on gay marriage – the discussion effec- tively ground to a halt, and instead of some potentially interesting and useful data about the participants’ knowledge of gay marriage, Victoria had her far less useful mini-lecture instead! USING SILENCE As any journalist will tell you, a great way to get people to talk is to simply remain silent when someone has finished speaking; they will usually start speaking again, often expanding on what they were saying before, to fill the gap. By remaining silent you effectively give the interviewee permission to continue, and subtly encourage them to do so. We have noticed that many interviews conducted by students (including the first ones we conducted as students) are rather ‘rushed’ and wooden. Questions have been asked in quick succession, and no silences have been allowed to develop. If you are feeling nervous, the risk is that you will rush the interview to get it over and done with 04-Braun & Clarke_Ch-04-Section 2.indd 96 28/02/2013 7:35:23 PM  Interactive data collection 1: interviews 97 as soon as possible, and a rushed interview is rarely a good interview. It’s really useful to learn how to tolerate silence – it might feel like a silence is stretching on for ages, but it will only be a second or two. Of course, don’t take it to extremes or it may make the participant uncomfortable. MANAGING NERVES It is natural to feel nervous when conducting an interview – it’s a new (and rather odd) experience that might involve meeting a new person (will you be able to establish rap- port?) and has high expectations associated with it (getting good data). Even though we have both been conducting interviews for over 15 years, we can still find it anxiety pro- voking. But we have strategies for managing this. Very basically, give yourself enough time to prepare and ‘calm’ yourself: avoid rushing to an interview, all the while wor- rying about being late (and invariably getting lost); take time to catch your breath and focus before starting the interview. Practising your interview technique will help with nerves; we have become more confident and less anxious over time. But when we have been visibly nervous, interviewees have generally been very kind, patient and under- standing. To a certain extent ‘interpersonal’ nerves can be alleviated as you can perform (and hide behind) the role of a professional researcher and interviewer, with its script and specialist equipment and materials. It also helps to remember that most participants have little understanding of the interview process, and although you may not feel like one, you are an expert by comparison. Even if you are meeting a stranger in their home, you are ‘in charge’ of the interaction and you guide the interviewee through the process. CLOSING THE INTERVIEW Once you have asked all of your questions and you have given the participant the oppor- tunity to add any other information they think is important, it’s time to close the inter- view. A clear end to the interview avoids the participant saying lots of interesting things after you have switched off your audio recorder. When the interview feels like it has come to a natural end, check (again) whether there is anything else the participant wants to add, and then ask specifically if they are happy for the interview to end and for the recording equipment to be stopped. Finally, thank the participant and give them another opportunity to ask you questions about the research. Participants might ask if they will receive the transcript of the interview, a copy of your report, or a summary of the results. Think very carefully about what you can realistically offer and avoid making promises in the ‘warm afterglow’ of the interview that are difficult to keep. We usually offer to send participants a short (two-page) summary of our results, and this satisfies most people. PREPARING FOR AND CONDUCTING THE VIRTUAL INTERVIEW As noted above, telephone, email and online interviews are no longer regarded as (poor) substitutes for face-to-face interviews but as different types of interview method, with 04-Braun & Clarke_Ch-04-Section 2.indd 97 28/02/2013 7:35:23 PM  98 Successfully collecting qualitative data BOX 4.4 ADVANTAGES OF VIRTUAL INTERVIEWS Convenient and empowering for participants People can participate in the comfort of their own homes or in a location of their choosing (and email interviews can be completed in a participant’s own time); these can also be more convenient for researchers. Participants may feel a greater sense of control and empowerment (particularly in email interviews) because they can reply to questions when they are ready and have time to reflect on and edit their responses (they are not sitting opposite a researcher who is waiting for a response). Accessible and (more) anonymous Not limited by geography (and email interviews not constrained by time zones); accessible to geographically isolated and dispersed participant groups. More accessible for some people with physical disabilities and mobility issues. Participants with concerns about anonymity and ‘hard-to-engage’ groups may be more willing to participate in a virtual interview than a face-to-face one. Facilitates the participation of shy people and those who lack confidence in talking face-to- face, and people who feel they express themselves better in writing. Potentially ideal for sensitive topics People may feel more comfortable disclosing sensitive information in virtual interviews because of partial (telephone) or complete (email/online) anonymity and because they are confiding in a non-judgemental machine rather than directly in another person. their own advantages and disadvantages. Boxes 4.4 and 4.5 summarise some of the key advantages and disadvantages of (different types of) virtual interviews (Bowker & Tuffin, 2004; Chen & Hinton, 1999; Evans, Elford, & Wiggins, 2008; Hamilton & Bowers, 2006; James & Busher, 2006; McCoyd & Kerson, 2006; Meho, 2006; Murray & Sixsmith, 1998; Opdenakker, 2006; Sturges & Hanrahan, 2004). There are a number of important procedural differences between face-to-face and virtual (especially online and email) interviews, so before you begin collecting data you need to think about the following issues: Length of virtual interviews: Email/online interviews tend to take longer to complete than face-to-face and telephone interviews. Online synchronous interviews can take around 90 minutes to two hours (compared with around an hour for a face-to-face interview); email interviews can unfold over several days, weeks or months depend- ing on the number of questions you have, whether you email your questions all at once or in batches – see below – and on the participant’s schedule and other com- mitments (Meho, 2006). British-based employment and education researchers Nalita 04-Braun & Clarke_Ch-04-Section 2.indd 98 28/02/2013 7:35:24 PM  Interactive data collection 1: interviews 99 There is less social pressure and no visual cues to inform judgements about the researcher/ participant (and potentially fewer of the complexities associated with interviewing across difference); however, remember your use of language can reveal a lot about your social background, educational experiences, and so on. Relatively resource-lite No need for transcription and no loss of raw data for online/email interviews. For this and others reasons (e.g. no time and money spent travelling to the interview location), virtual interviews are considerably more time- and cost-effective than face-to-face interviews (and with email interviews, it is possible to conduct more than one interview at a time). Larger samples are possible (because of the time and costs saved from no travel and, in the case of online/email interviews, no transcription). Potentially allow more engagement with data during data collection Because of the extended time-frame of email interviews, questioning and starting data analysis can co-occur, and the developing analysis can inform the interviews; participants can be invited to comment on the developing analysis (see Chapter 12 for a discussion of respondent validation or ‘member checking’). In online and especially email interviews, researchers have time to formulate prompt and follow-up questions, so the interview can be more responsive to the participant’s developing account and there is less chance of missing useful information. James and Hugh Busher (2006: 414) found that interviews they expected to take two or three weeks ‘eventually extended in many cases over several months, because this speed of responses suited participants in the busy press of their daily lives’. For email interviews, it can be helpful (for you and the participants) to set a deadline for completion of the interview (this is particularly important if you are working to a relatively tight schedule). Participants will be typing rather than speaking: In online/email interviews, reas- sure participants that correct spelling and grammar are not important and use, and encourage participants to use, acronyms or abbreviations (LOL for laugh out loud, ROFL for rolling on the floor laughing), emoticons (☺ to signal smiling), and underlining and italics (for emphasis) as a substitute for non-verbal cues (Murray & Sixsmith, 1998). But be aware that the language of emoticons is not universal; for instance, the Japanese have a different emoticon language from that used in the west, and some countries don’t use them (Opdenakker, 2006). The type of 04-Braun & Clarke_Ch-04-Section 2.indd 99 28/02/2013 7:35:24 PM  100 Successfully collecting qualitative data BOX 4.5 DISADVANTAGES OF VIRTUAL INTERVIEWS Less accessible to some groups Potential participants are limited to those with access to a networked computer or mobile device (or telephone); such people tend to be more affluent. With online/email interviews, participants need to have a certain level of competence in reading and writing (these types of interviews are potentially less accessible to participants with limited literacy skills); more challenging for participants who type slowly. Some people feel they express themselves better when speaking rather than writing (so prefer face-to-face or telephone interviews). Less convenient for participants Online/email interviews are potentially more time consuming than face-to-face and telephone interviews; writing responses takes longer than speaking and for this reason some participants may perceive online/email interviews as more onerous than face-to-face and telephone interviews. The researcher has less control over the interview In online and especially email interviews, participants can edit their responses so data are less spontaneous and ‘natural’. The context in which the participant is typing their responses, and the ways in which this context shapes their responses, is unknown (e.g. other people may be present; the participant may be switch-tasking and doing other things at the same time as completing the interview). encouragement that we give participants in face-to-face (and telephone) interviews is also important in online/email interviews (e.g. there are no right/wrong answers; we are interested in your experiences and perspectives). Negotiating informed consent: It is not possible to hand ‘virtual’ participants a PIS to read or a consent form to sign in the interview. However, this information can easily be emailed or posted to participants. Participants can be asked to (print off and) sign and return a consent form via post, or use an electronic signature and return as an email attachment. For email/online interviews, it can be easier to ask participants to read the consent form and then write or cut and paste a statement of consent into a response. For example: ‘I have read the informed consent and have had the opportunity to ask questions. I understand that I can withdraw from the study at any time with no negative effects. My responses confirm my ongoing consent’ (McCoyd & Kerson, 2006: 394). Establishing trust and rapport: It is often assumed that without access to visual cues