Read Understanding Research Online
Authors: Marianne Franklin
Getting people to agree to be interviewed, if not in person then to respond to an email question list, can be harder than you think; people in businesses, NGOs and governments, and media professionals are often very busy. A refusal can be outright (here, ‘no’ usually means ‘no’) or conveyed by silence. If the latter, there is no reason not to try one more, if not two more times. An initial silence can simply indicate oversight. But consecutive silences should settle it for you, so move on.
Remember that these people are granting you their time and knowledge; if no answer comes straightaway this need not mean a refusal; ask again as above. The main thing to note is to keep the initial request short and to the point; make quite clear what your topic is (in accessible language – no Big Thinker name-dropping!) and your aims for the interview.
Once a person agrees in principle you need to then gain informed consent; a signed form (adapted to your needs) or as a clear acceptance in an email if this suffices (see Appendix 1). People need to be fully informed of the intent and purpose of an interview; unlike some experimental situations, deception is not an option.
Once again, draft and redraft (if you have to) the questions. It is also customary to offer to send the questions to the interviewee beforehand. Keep the list short (6–10 questions maximum, of which some are basic information and where the key questions are clear and distinct from one another). Too many questions overwhelm your respondent and often indicate a lack of concerted thinking on your part, i.e. avoid sending your first brainstorm list to anyone agreeing to speak with you and, under advisement, before they have accepted or not.
If you only get one or two positive responses, take these on and readjust the interview’s role in the overall design if you need to. For BA and MA level projects, be content – and work with what you get. The main rule of thumb at all levels is to
note that interviews require time and preparation. They can flow easily and they can be difficult. People can be all too ready to tell you anything they want you to know but others can be evasive and hostile. Sometimes this is for reasons beyond your control.
Berg has an overview of ways to troubleshoot if things start to go off-course during an interview (Berg 2009: 117–21). As noted above, interviews can start well and end less happily, and vice versa. Points to note, however obvious, include:
How you start an interview, and how you bring it to an end also requires care, a sense of timing and an awareness of fatigue (on your interviewee’s part) and the other person’s time restrictions. What was the arrangement? Check with the other person what their timeframe is and do not take too long to get things off the ground.
As noted above, all interviews (and focus-group situations – see below) need to warm up. This is where the information-type questions play a role. It helps both of you settle down. Do a sound check if the interview is being recorded (and check that the interviewee is happy with this – not all people are).
The biggest pitfall is letting an interview go on too long, past the point the other party wants, or can sustain. Set your watch, or ascertain which questions you really want to cover before the end, prior to starting (see Berg 2009: 141–3). You can always approach them for a follow-up or ask supplementary questions by email or phone if need be.
Do your homework well; if this interview is with a government official or someone with management responsibility, be clear as to their area of expertise and portfolio. If you find yourself interviewing someone about an area in which they are not qualified, or permitted to comment upon, both of you will be frustrated.
Here attitudes, including underlying assumptions or perceptions from both parties, can lead to tensions as well as open doors. For instance, some activists and communities where research does not have a good ‘street credibility’ can be guarded in their responses. Probing questions are important (as journalists know), however, your job is to encourage responses, not to grill the interviewee as would an investigative news journalist.
When interviewing people from other cultures, backgrounds, or language groups, your ability to negotiate the linguistic and cultural mores underpinning good exchanges is paramount. In some cases you may need to do interviews by proxy. Male researchers interviewing women in some parts of the world may not be welcome, unmarried female researchers may have restricted access, homeless people may not feel comfortable with a stranger, and so on (see inter alia Berg 2009: 136–43.
At the end of the interview do not forget to thank your interviewee for their time, no matter what you may think of the usefulness of the material at that point. More importantly, be sure to send a follow-up thank you, again no matter how well things went or not. If you have arranged for the interview subject to consult the transcripts or eventual use of their material (whether they are named or not, depending on the terms of the consent they grant) be sure to fulfil this obligation.
Final point
: sometimes we can exit an interview convinced that it has been no use at all. Given that working with human subjects and the sorts of inquiries conducive to this sort of data-gathering, often our first impressions can be incorrect. An interview can feel like a great success – flowing, affable, lots of material. But on looking more closely we may notice that there is less substance than we imagined. And the converse is true, so note how emotions and interpersonal dynamics, contingencies like hold-ups, the time of the day, or interruptions, can affect our initial judgment. All the more reason for taking time to absorb and analyse the interviews at a later date, and more than once if this material is important to your investigation.
6
When considering whether interviews will allow you to investigate your topic and elucidate your research question, note that not all inquiries lend themselves to interviews: for instance, policy analysis or various forms of textual analysis (for example, in film studies, audio-visual media) in which first-hand information about human subjects’ views and experiences are not the main focus.
If interviews do suggest an avenue then you also need to consider what sorts of interviews are best suited. Perhaps a survey/questionnaire can precede this element, or perhaps fewer, more in-depth interviews are better. Perhaps focus groups (see the following section) are a more productive way forward. Neither do all sorts of research with human subjects need to be completed by setting up a face-to-face, or email-based interview. Ask yourself, particularly if you have previous interview experience from another research project or in journalistic work, whether this sort of data-gathering is best-suited, indeed relevant to your current research question (see Berg 2009: 110).
That said, if interviews are going to be the main data-gathering mode for your project there needs to be a clear reason for embarking upon this course; a time-consuming one which requires preparation and due care with the material even in the most unstructured settings. Interviews for research purposes are, by definition, directed and directive; they are intended to flow from the researcher to the research subject, more informal and narrative formats notwithstanding. Interviews are social acts, intersubjective and thereby often unpredictable. They are comprised of a particular sort of listening (from the researcher), speaking and narrating (from the interviewee), and post facto analysis and representation (as cited/verbatim material in the role of substantiating evidence in the findings). For these reasons they provide rich and diverse material, challenge the researcher to reconsider their own operating assumptions (for example, about what people really think or experience as opposed
to what we as researchers would like them to tell us!), and create productive tensions between research aims, and research findings; if, and only if interviews are relevant to the research on hand. The ways, and challenges of analysing these data (transcripts, email texts, video footage if pertinent) is the next task, for
Chapter 7
.
The working premise of semi-formal to informal interviews is that the researcher
expects
a wide variance in response – in form, delivery, and substance. For more narrative modes of research, the open-endedness and individuality of the responses are integral to the research question; the ensuing analysis will then emphasize and tease out these nuances and inconsistencies rather than seek to eliminate them as bias or overstate their importance.
Successful interviewing at the more informal end of the spectrum does require an in-depth knowledge if not empathy on behalf of the interviewer; being responsive to the situation and language/idiom of those interviewed.
Across the board it is worth noting before setting out, during, and afterwards that a ‘research interview is not a natural communication exchange’ (Berg 2009: 127); the utterances of people have been granted for the purposes of a project that will then shape their words into findings and draw conclusions on the part of the researcher.
See Berg (2009: chs 4 and 9) for a discussion of narrative research techniques; Bertrand and Hughes (2004: ch. 4: 74
passim
, ch. 7: 141–51); Burnham et al. (2004: ch. 4: 80
passim
, ch. 9: 205
passim
; C. Davies (2007: ch. 7: 101–11, ch. 10; Gray (2009: ch. 14: 369–95; Silverman (2011: part IV). James and Busher (2009) look at the specifics of online interviewing.
Focus groups – a sort of group-based interview – are multilateral conversations as opposed to the bilateral format of the classic interview (see previous section); as such they are often seen as one-stop shops for researchers wanting to gather qualitative, semi-formal interview material.
Focus groups, small-group discussions led by a facilitator – or
moderator
– have been a mainstay of market research in recent years, making a comeback after a period of disuse (Gunter 2000). In the last few decades many techniques and analytical
approaches have been refined for this sort of data-gathering. The establishment of
audience research
and
reception studies
in research on popular culture, the media, and communications saw the establishment of focus groups as a popular form of data-gathering in academic research as well. Researchers in politics looking at gathering qualitative information about how groups perceive political processes, policy-making, or key issues of the hour also make use of focus groups.
This sort of data-gathering works for research projects that regard the material generated in a group, by way of people talking together, how interpersonal interactions generate more material, as the primary data. By implication this means that a variety of open-ended responses, including the psycho-emotional dimensions that govern any sort of group discussion, are relevant to the research question (Berg 2009: 158–9, Gunter 2000: 42).
This format for getting a range of nuanced information from more than one person at a time can work well on its own or together with other techniques such as surveys or one-to-one interviews; the respective importance of the focus group data will require a comparable degree of justification and working through in the research plan. The key point here is that it is a group-based response, interactively produced ideas and opinions about the topics on hand that matter. In this respect focus groups can provide an extra dimension to face-to-face interviews and more depth than anonymous surveys.
7
A key element of focus groups is that, in principle, they can gather a lot of data in a relatively short time; organizing this sort of group interview is also both cost-effective and time-effective under certain conditions. However, this assumption can be misleading. Bruce Berg provides a useful list of things to avoid when deciding about
whether a focus-group approach is suitable for your research project, echoed by others (Berg 2009: 160–2, 165–6).
Not enough care taken in these areas before and during the session can result in material that has little application for the project; the chance that participants may not show up or that discussions will either not take off, become dominated by one or two dominant members, or stray too far from the topic to generate anything worth analysing. Nonetheless, the effect of getting a group of people together to discuss a topic, consider a set of focused questions, is often a positive one; energy and ideas can emerge in ways that a formal/semi-formal interview between two parties cannot.
Like a tutor – even a media anchor person or talk show host, albeit with a clear scholarly outcome in mind – the moderator can make or break a focus group. No matter how fascinating the discussion may be or how strongly you (if you are wearing the researcher/moderator hats at once) feel engaged with the topics and opinions raised, you are not a fully participating member. As Berg notes, professional ‘moderators tend to get professional results’ (Berg 2009: 163).
The moderator’s role is a distinctive element in well-run focus groups; it is not the same as the interviewer’s role in that a moderator consciously guides and redirects the conversation. Their rapport with the group is at once more simple – more distant,
and more complex – group dynamics work between moderators and participants as well as between the latter. Your participants also need to have a clear idea about the purpose of the group, of how their input will be treated and what the research project is about.
One thing that focus groups and interviews have in common, however, is the way in which the moderator/interviewer needs to use their intuition, if not experience, in deciding when to leave one topic and move onto another.
For focus groups in which participants are asked for their ideas about an item – product or TV programme for instance – a checklist of specific questions may be substituted by a single general topic. If the group is assembled to consider a range of issues arising from a more general topic – the effects of gaming or social networking sites in people’s everyday life for instance – then a list of questions does need to be devised; more is less in this respect, hence the formulating of clear questions that also inspire discussion takes time, and trying out.
Below we encapsulate these practicalities in terms of some do’s and don’ts; these points can also be adapted for interviews, face-to-face or computer-mediated. As a rule:
The latter point is relevant for any sort of interactive research. With these practicalities in mind, here are some observations about the knock-on effect of more unseen, unknowable factors when conducting this sort of research:
A well-prepared researcher should be able to cope with these contingencies and interrogate them in light of the research question. In short, you need to think about how this sort of group (collectively)-generated qualitative data relates to your inquiry; what it can, or cannot offer with respect to other approaches.
Particularly popular in media and communications, market research, political polling, and current affairs TV programmes,
8
focus groups are an extension of interviewing; they are less time-consuming as they cluster groups by various ways and for various research ends.
A focus group is just that however: a highly selective – focused –
group
and a
focused
set of group-based interactions. They do not produce any sort of representative sample nor are members necessarily coming to the task disinterested (payment, forms of manipulation). They also need to be managed carefully in that generally speaking focus groups are not free-for-all discussions. That said, an astute researcher who takes the requisite amount of time to read and analyse the records closely will notice that sometimes the conversations that are ‘off-message’ can prove informative, rich material.
Finally, a point for the next chapter, written transcripts and recordings of focus-group sessions, as is also the case with interviews, do not speak for themselves. They generate audio-visual and written matter that is usually characterized by contradictions and a lack of standardized responses (
contra
survey work). Theses records constitute the ‘raw’ data which you, as a researcher, need to collate, filter, and make sense of in some way.