This page is older archived content from an older version of the Emerald Publishing website.

As such, it may not display exactly as originally intended.

How to... conduct interviews

Features of the interview

The interview has become one of the most prevalent genres of the second half of the twentieth century, and its popularity shows no sign of waning. Whether it is the views of politicians, the lives of celebrities, or suitability for a job (to name but a very few of its uses), it has become a favourite method to gain information, whilst in psychoanalysis a varient of the interview has come to be used as a method to help people tell their stories.

It is therefore hardly surprising that it has become a popular research method in the social and management sciences. As early as 1886, Charles Booth used interviews to triangulate his research on the social and economic conditions of the people of London, whilst it was also used in the First World War for psychological testing; George Gallup in 1935 launched its use for opinion polling with the foundation of the American Institute of Public Opinion.

An interview is a verbal interaction between two or more people where one (or occasionally more than one) person implicitly directs the flow of information (although this usage has been criticized by some particularly feminist researchers who dislike the implied inequality between interviewer and participant). It thus differs from a conversation, which is more free flowing: the person conducting the interview "stage manages" the dialogue with a view to obtaining information to meet a particular purpose, such as the collection of data to answer a research question.

Stuart Hannabus draws on the work of M.Q Patton (Qualitative Evaluation and Research Methods, Sage Publications, 1990)

"The purpose of interviewing has been defined by Patton as being "to find out what is on someone's mind…. We interview people to find out from them those things we cannot directly observe"....We want the respondents' own perspective to emerge, explore the ways in which people working together share common understandings, get insight into particular experiences, find out motives behind decisions, get a view of informal procedures, consider apparent contradictions between attitudes and behaviour, and allow respondents time to provide their answers."

"Research Interviews", New Library World, vol. 97 no. 5

Interviews as a qualitative technique

Interviews, as distinct from focus groups, can be used as either a quantitative or a qualitative technique, although here we shall be mainly concerned with its use in qualitative research. Qualitative research posits knowledge which is not "out there" and given but which is constructed and negotiated, and is heavily dependent on human experience and social interaction: it observes humans as actors and creators of knowledge. This perspective makes it particularly suitable for looking at such issues as culture, change, attitudes and consumer behaviour.

In a skillfully conducted qualitative interview, the interviewer nurtures the participant to reveal rich and varied data based on his or her understanding of the world, and is thus a partner in the creating of knowledge and data, as opposed to a mere observer.

Characteristics of interviews

The form of the interview will vary according to the research methodology used and whether it is conducted face to face, by telephone or by email, but here are some general characteristics:

Structure combined with flexibility
The researcher will have particular issues to explore which will have been determined in advance, but will also allow freedom for the interviewee to bring up issues important to him or her.

Interactivity
Data is generated by interaction between the participant and the subject.

Neutrality
The interviewer should avoid any intrusion of his or her opinions.

Iterative questioning
Initial questions will be followed by further probing further to get more in-depth answers.

Raw data capture
Tape recording is common.

Generation of new knowledge
The participant may be nudged into thinking about an issue in a new way.

Ethics
Permission from the subject should be sought and confidentiality guaranteed.

Characteristics of a good interviewer

Interviewing is a skill, as anyone who has watched a politician being interviewed on the BBC or CNN can see. Here are some characteristics of a good research interviewer:

  • warmth, ability to show interest, to put at ease and gain the trust of his or her subjects.

  • calmness and quiet confidence in their own expertise, combined with modesty and the ability not to let their experience or abilities intrude on the interview.
  • a good listener, able to show interest when the participant is talking but also to listen 'actively' and probe for what is being said only implicitly, and also decide how to respond.
  • a good memory, knowing when a point needs to be returned to.
  • curiousity, genuinely interested in what people say, and wanting to know more.
  • thorough, meticulous and capable of very careful preparation.

And perhaps above all:

  • mental agility and the ability to operate on several different mental tracks: to listen, to make sure that all the key points on the research schedule are covered, to relate comments to the research question, to be alert for contradictions, to decide how to respond with the next question, and whether to change the order of questions, to probe and get the subject to open up, to remember certain points which should not be allowed to interrupt the flow of the conversation, but which should be returned to later, and possibly also to take notes while all this is going on.

The response rate

You will have to consider how your interviews relate to your wider research perspective: is it your main technique, or are you including it with other techniques?

If it is your main technique, it is likely that your research focus is on a topic wherein psychological factors play a strong part, such as for example attitudes towards change, leadership, consumer choice etc.

In The contemplative organization, Maia Duerr looked at the extent to which organizations adopted contemplative practice, using as her main method in depth interviews with 97 people. She deliberately focussed on qualitative methods as she felt that these were more appropriate for investigating organizational culture and change.

Journal of Organizational Change Management vol. 17 no. 1

You may however also be using the interview with a number of other techniques in order to triangulate data: for example, you may be conducting a survey and want to explore the issues before formulating survey questions, or you may have a large amount of data which you want to probe further.

Use of the interviewing technique must not be done on a whim but must be part of your overall research design (see How to design your research), chosen because it is the best way of getting at your desired data, or a significant part of your desired data.

For example, you might be doing research on the extent to which companies have an ethics policy in a particular area. On the face of it, a good way of finding this information would be to interview the relevant CEOs. However, you need to think carefully about such issues as what exactly it is you want to find out, whether the CEOs are the best people to interview rather than, or in addition to, Corporate Affairs Managers etc., what other sources of data e.g. company records etc. might be useful.

There are various design issues which you will need to consider, for example:

  • You will probably have a particular issue which relates to one or all of your research questions, and this issue will form the basis of your interview questions.

  • In getting the sample for your interviews you need to consider what is your overall research population and sampling strategy, and within this what is the best method of getting a good sample.

In Shopping with consumers: reflections and innovations, Tina M. Lowrey et al. alternate in-depth interviews with observation of participants shopping to research consumer behaviour. (Qualitative Market Research: An International Journal vol. 8 no. 2).

In A comparison of two alternative interviewing techniques used within an integrated research design: a case study in outshopping using semi-structured and non-directed interviewing techniques (Marketing Intelligence & Planning vol. 14 no. 6), Denise G. Jarratt combines semi-structured and non directive interviews with two lots of quantitative questionnaires, in both cases the interview contributing to the creation of the constructs for the latter. This is represented diagrammatically as follows:

Image: Semi-structured and non directive interviews

Types of interview

The interview used for research purposes can be divided into various different types, which vary on a spectrum of structure and control from the highly structured (where questions are pre-set and the tone neutral) to unstructured, where there are no set questions and where the participant, rather than the interviewer, may even set the agenda.

Broadly speaking, the less structured the interview, the greater degree of depth of data is yielded. Again broadly speaking, the types of interview can be summed up as follows:

Image: interview types

Structured interviews

The structured interview is a quantitative technique, used in surveys, and we shall here limit our consideration to basic definitions.

Structured interviews involve asking the same questions in the same way to a large number of respondents. Questions are closed, in other words the respondent has to choose from a list of preset responses, and will probably read from a script. The main possibility of error lies in the phrasing of the questions, for example "Do you ever..." vs. "Do you never...", which is why the questions must be standardized, and phrasing, as in "Do you ever..." vs. "Do you never...", which is why the interviewer must try not to vary the tone.

The advantage of the structured questionnaire is that it is a good way of making sure that you cover a large amount of ground, because the items can be set out beforehand in a script: for example, if you are testing responses to a product and want to make sure that you cover all its features. It is also possible, because the interview is unlikely to take so long, to interview more respondents and thereby gain access to a wider sample. Data categories are likely to be predetermined and the questions pre-coded.

Structured interviews are good for explanatory research, when you are looking at the relationship between variables.

For more information, see Surveys - different methods and instruments.

Semi-structured interviews

With the semi structured interview, we move into more qualitative approaches to research, in which data collection is not to a rigid formula as with the structured interview, and is often analysed and coded after collection. As with all qualitative techniques, it is best used for issues which are less amenable to precise measurement, and which may be more complex or affective.

Stuart Hannabus has this to say about the difference between structured and unstructured interviews:

"Each approach has strengths and weaknesses, and each may be more or less suitable for particular types and areas of research: e.g. the highly-structured approach may be good for eliciting information about large numbers of people using a reference library or for finding out people’s choice of consumer product, while informality and an open structure tend to be preferable when complex, personal or sensitive issues are being probed (e.g. perceptions of discrimination at work, disagreement with company policy, fears about redundancy)."

Research Interviews, New Library World, vol. 97 no. 5, quoting Patton, M.Q.,
Qualitative Evaluation and Research Methods
(Sage, 1990)

In the semi structured interview, the researcher follows a list of questions, but these are open for the participant to respond as he or she feels appropriate. The order of the questions may also be varied if a particular response leads to one of the other questions.

Sometimes, there will not be a list of questions as such but rather an interview schedule or topic guide which lists the main issues which the researcher wishes to follow, and which will ensure consistent coverage over a number of respondents.

John Wood and Tricia Vilkinas in Characteristics associated with success: CEOs' perspectives (Leadership & Organization Development Journal, vol. 26 no. 3) use a semi structured technique to investigate the leadership self perceptions of CEOs.

William Keogh and Victoria Stewart in Identifying the skill requirements of the workforce in SMEs: findings from a European Social Fund project used two sets of semi structured interviews to analyse the strategic and human resource issues around training in SMEs. (Journal of Small Business and Enterprise Development, vol. 8 no. 2)

Maia Duerr, in The contemplative organization (Journal of Organizational Change Management, vol. 17 no. 1) used a semi structured approach to find out about the extent to which contemplative and meditative practices were used in organizations. 

Margaret Linehan, in Networking for female managers' career development, uses the semi structured technique to explore the topic of female networking "to ensure that the interviews covered the same main questions, but allowed participants to respond in a variety of ways and raise issues which were pertinent to the research" (Journal of Management Development , vol. 20 no. 10).

Non-directive in-depth interviews

In both sociological and management research, there is a move towards a type of in depth interviewing which is even less structured and directive than that where there is some sort of interviewer determined structure, and which gives more control to the interviewee. For example, the participant may be allowed to tell their own stories, or the interview may provide a framework which yields otherwise subconsciously buried information, or may redress the power balance between interviewer and interviewee so that the interviewee can raise issues which she/he considers important. (Feminist sociologists in particular have been critical of the hierarchical relationship between interviewers and their subjects.)

Stefanie C. Reissner, in Learning and innovation: a narrative analysis (Journal of Organizational Change Management, vol. 18 no. 5), researched into the issue of how organizational change affects the individual's needs to learn different skills, focusing on one organization. She used the interview technique to allow participants to tell their own stories about their life with the organization, and how change affected them personally.

When interviewing people about a subject such as organizational change to which there may be an infinite number of possible reactions, it is obviously good to use a technique that is as open ended as possible. In such circumstances, people's reactions are probably reasonable close to the surface. There are other circumstances, however, when it may be necessary to use techniques which tease out issues that are otherwise hidden and are less likely to come out in more 'standard' interviewing.

In A comparison of two alternative interviewing techniques used within an integrated research design: a case study in outshopping using semi-structured and non-directed interviewing techniques, Denise G. Jarratt uses the technique of laddering, which uses probes to uncover emotional reactions, and to probe the critical thinking behind consumer choices. (Marketing Intelligence & Planning , vol. 14 no. 6)

Ove C. Hansemark and Marie Albinsson use the technique of 'phenomenology', which

"allows people's 'life-worlds' to be discovered (Küpers, 1998), thus capturing their intersubjective experiences. Phenomenology is used to describe the structure of experiences as they present themselves to consciousness, and to find what is hidden in ordinary experiences
(Gibb, 1998)."

"Customer satisfaction and retention: the experiences of individual employees",
Managing Service Quality, vol. 14 no. 1
quoting Küpers, W. (1998), "Phenomenology of embodied productivity in services",
International Journal of Service Industry Management, vol. 9 no.4 
and Gibb, S. (1998), "Exploring career chaos: patterns of belief",
Career Development International, vol. 3 no.4

This sort of in-depth interviewing can be used to uncover very complex, affective issues as well as in the exploratory stages of research, to help uncover key issues for research questions.

Key informant interviews

These differ from 'ordinary' interviews in that the respondent is chosen because of his or her specialist knowledge, and becomes not just a provider of data but someone who helps frame the research questions. It is likely that you will here use a purposive, snowball sampling technique, whereby respondents suggest other likely interviewees. The probable determining criteria will be position in the organization, knowledge of the issue, and willingness to help.

Group interviews

A group interview occurs when there are a number of participants, and is also referred to as a focus group - see How to... conduct a focus group for a full guide. 

How does the interview take place?

Because of the emotionally intensive nature of interviewing, it is probably best if it takes place face to face, as talking face to face is more relaxing if the conversation is long and detailed. However, if the interviewees are a great distance apart, then it may be more practical to talk to them over the phone. Email interviewing is also becoming more popular, with the advantage that the medium allows for automatic recording of responses; it is also possible to allow more time for response. The disadvantage with both email and telephone, however, is that you don't see the non verbal signals, which are important in any 'in depth' conversation.

Conducting the interview

It is common to talk about "staging" an interview, because even if the interviewer is being non directive, he or she needs to set up conditions where the interviewee can be as relaxed as possible and thereby yield the best possible quality of data. There is thus an element of artificiality, which both differentiates the interview from an ordinary conversation, and ensures reliability of data, because you are setting up conditions which guarantee the best research results. The staging element needs to be borne in mind from the point where the interviewer is setting up the interview through the actual "conversation" to the point where the interview is concluded.

Practical considerations

You will need to agree with the participants the following:

  • The date and time of the interview.
  • The duration of the interview – less than an hour may be difficult to cover issues in depth, more than two hours will be difficult to sustain concentration.
  • The place of the interview – should be somewhere quiet and conducive to concentration. At the beginning of the interview, negotiate that telephones be turned off, also arrange for seating to be informal, for example facing chairs (if the interview takes place in the CEO's office, try and arrange it in the part of his or her room reserved for informal interviews, rather than being separated by a desk).
  • The recording of the interivew – if you want to tape record, you should ask permission.
  • The purpose of the interview – explain the background of the research, if necessary sending your subject a brief explanatory paper.
  • The conditions under which you will use the interview data – most particularly, you should ensure confidentiality.

The stages of the interview

Conducting an interview is a bit like diving: you start off on the surface, then go deeper, and then come up for air at the end. There are a number of stages:

Arrival and introduction

The first task of the interviewer on arriving is to put the participant, who may be nervous, at ease, and establish rapport. The conversation should be on neutral subjects, completely unconnected with the actual topic of the interview.

When you are both relaxed, introduce the purpose of the interview within the context of the research. You may well have sent the participant some information, but the chances are he or she may not have had a proper chance to look at it. This is the time to establish the 'housekeeping' procedures – is the seating arrangement suitable, can you forestall any interruptions, do you have permission to record, what will happen after the interview etc. Reiterate the fact that the discussion will be confidential.

The interview

Begin the interview with general questions that define terms and provide contextual information for what follows. By asking easier questions, you will be able to guage how the participant responds, and how you should approach the interview, before the conversation becomes too deep and complex.

Once you have the necessary background information, it is time to get into the "meat" of the interview, and move from the general to the particular. Here you will be following either pre-set questions or some sort of interview schedule, both of which will relate to the key research questions to which you are trying to find answers. You may have to ask supplementary questions in order to make sure that the topics are fully covered, and that you probe in sufficient depth to get below the surface. You may also need to "explain" certain questions by providing a link with the research.

Ending the interview

Signal the end of the interview by some such conversational device as "and moving on to the final topic". Make sure that the participant has no buring issues which have not been covered, and particularly if the interview has been sensitive, that there is no "unfinished business". Thank the participant for their time, and say how their contribution will help the research.

Facilitating rapport: the interviewer-interviewee relationship

We considered earlier (Characteristics of a good interviewer) the qualities that the interviewer needs to have. The key role of the researcher is to be a neutral observer who can make tacit assumptions explicit, and reveal what is otherwise hidden, even to the subject. The role of the participant is quite simply, to open up, and talk as freely as possible, whilst not deviating too much from the interests of the research. Here are some pointers and techniques to make that more likely to happen.

  • Show interest: maintain eye contact with the participant, nod, smile, ask follow-up questions.
  • Be non judgemental: don't indicate disapproval, emphasize that there are no right or wrong answers.
  • Be neutral, and don't be drawn into making comments or into argument: if asked for an opinion, respond with some comment such as "It's not my opinions that we are interested in here"; if they come up with something controversial, respond by throwing the issue back at them, "Why do you think that?"
  • Look for non verbal clues: for example, if the participant appears doubtful, and if so, try and probe further.
  • Make sure that you are entirely clear about what is being said: if not, ask for clarification.
  • However tempting it may be by way of encouragement, don't put words into the participant's mouth, finish sentences for them, or summarise what they have just said: ask another question.
  • Be prepared gently to steer the conversation back to the topic if it veers off into areas that are not relevant to the research, but if you interrupt, do so gently ("I wonder if I could just bring you back to"... "What I really meant was...").
  • Pace the interview, ensuring that you don't spend too long on one topic at the expense of others.

Stuart Hannabuss has a useful section on conducting the interview in his article "Research Interviews" (New Library World , vol. 97 no. 5).

Dealing with problems in interviews

Here are some of the possible problems which can occur in interviews and suggestions as to how to respond to them:

The interview gets onto sensitive areas
If the interview has to deal with sensitive issues, or if the participant has a strong emotional reaction, it is especially important to stay calm, to be empathic without getting drawn in to a counselling situation. Acknowledge the difficulty of the topic, and if the participant breaks down, invite a break and switch off the tape recorder. If the interviewee becomes angry, don't take it personally: try and get them to share their feelings.

The interviewee appears very anxious
Spend more time putting them at their ease, and on introductory and general questions.

The interviewee tries to dominate the agenda by for example answering questions of their choice, or repeatedly coming back to particular points
Acknowledge the importance of the points they are making, but try and gently bring them back to the topic.

The interviewee rambles
Gently interrupt, bringing them back to the point with another, perhaps more direct, question.

The interviewee is silent
Don't be put off by silences: the participant may need the silence to come up with a deeper response. On the other hand, it's important to keep the discussion going, and to make sure that it doesn't peter out.

The interviewee tries to direct questions back to the interviewer
Be warm, but don't disclose too much about yourself. This is not like a real conversation: the spotlight is definitely on the participant. Volunteering too much information can distract the participant from talking about themselves.

Framing the questions

Questions are the essence of the interview. However much you put the participant at their ease and strike a rapport with them, if you do not ask the questions in the right way you will not elicit the valuable data you would otherwise have obtained if you had phrased them differently. In this section we provide guidance as to how to frame questions, as well as on how to organize them in the first place. 

Developing the instrument: topic guides and question schedules

For semi-structured interviews, you will probably go in with a list of questions or topics which have been worked out beforehand. This is the instrument which you will use to capture the data. It serves not only as an aide memoire but also a means of standardizing the research between interviews and interviewers. It is also the main way of structuring the interview as well as the resulting data, so it is important to think carefully about its organization.

The starting point will be your research questions, or that subsection of them which you have decided to use this method to obtain the data.

It is advisable to group together questions which relate to the same issue or topic, to start with general topics, and then to focus in on more particular ones. Don't have too many questions; remember that you only have a limited time to ask them in.

The instrument should be easy to use during the interview, when you have a lot to occupy your mind – for example, have plenty of white space, and use colour or other typographic device to differentiate between sections and subsections, main questions and prompts etc.

How detailed the instrument should be is a matter of researcher preference – some will comprise solely the questions, others will also include questions for prompts and probes, the interview objectives, and wording for the start and close of the interview.

Below are a couple of examples from Emerald articles of interview designs/topic guides.

Interview design examples

This is an example of a very complete instrument with questions organized into topics. Note the way that the guide moves from general and contextual questions to more specific ones.

Image: interview design examples

The above example is from an article by Maia Duerr - "The contemplative organization" (Journal of Organizational Change Management, vol. 17 no. 1)

This is an example of a much simpler design:

"Characteristics associated with success: CEOs' perspectives" (John Wood and Tricia Vilkinas, Leadership & Organization Development Journal, vol. 26 no. 3)

Phrasing the questions

Here are some useful tips:

  1. Avoid closed questions. The key difference between structured and semi-structured interviews is that the questions for the former are always closed, whilst for the latter they are almost always open. Open questions invite open responses; closed questions allow the respondent just to say yes or no, which is not necessarily very helpful. (On the other hand, asking a brief closed question can be a good way of getting a rambling respondent to focus.)

  2. Avoid leading questions, or questions which entail value judgements. For example, when asking about threat of redundancy, phrase questions "How did you feel...?" rather than "Were you devastated...?"
  3. Ask questions clearly, avoiding any ambiguity. Make sure that your questions:
    • Do not contain double negatives.
    • Avoid unnecessary preambles, e.g. links with an earlier question which may obscure the question actually being asked.
    • Are free from abstractions, unnecessary theorizing (e.g. "Do you believe in the Theory X or the Theory Y school of management?" would be better phrased as "Do you think people work better when they are threatened with sanctions or given incentives and encouragement?").
    • Mirror the language used by your respondent.
  4. Base your questions on listening – not just to the words, but to the underlying meaning of what the participant is saying. Use the knowledge thus gained to decide which question to ask next, so that the interview is interactive, and also whether to return to a topic later on in the interview.
  5. For a non directed response, ask projective questions – that is, questions in a general form from which you can judge the values and beliefs etc. of the respondent, for example "What do you think are the qualities of a good manager?".1

Content mapping, content mining and probes

Robert Legard et al. (2003) describe two sorts of questions: content mapping and content mining. The former help open up the research territory and identify the issues; the latter probe deeper into the issues to achieve greater understanding.

Probes are questions that are asked as adjuncts to the main questions, to enable questioner and respondent to go deeper into issues.

Content mapping questions

Legard et al. (2003) cite three types of content mapping questions:

  1. Ground mapping: questions which open up the subject to the respondent and allow him or her to answer it in a way that is most relevant.
  2. Dimension mapping: questions which focus down on the issues raised by the participant, and define terms a little more narrowly.
  3. Perspective widening: whereas dimension mapping questions focus on issues raised by the respondent from the initial ground mapping questions, dimension mapping questions relate to ones which the interviewer specifically wants to raise.

Content mining questions

These are questions which are used as tools to dig deeper behind the issues raised by the content mapping questions. They are probes, the prudent use of which is one of the key skills of the good interviewer. According to Legard et al. (2003) they can be:

  • Amplificatory – encouraging further exploration of issues.
  • Exploratory – exploring feelings and affective issues that underlie events, for example "How did you feel when your idea was rejected?"
  • Explanatory – these probe the reasons behind views, feelings, behaviour etc. "What do you think made you angry with the Marketing Director when your idea was rejected?"
  • Carificatory – it is very important to be precise about issues and what they actually mean to the respondent. Clarificatory probes can be used to clarify language (if there is a term which you are not sure about, it is often sufficient to repeat it in an interrogatory fashion), or details of events, or inconsistencies.

It is important to stress here the difference between an ordinary conversation, where meanings often lie hidden and unexplored as one of the interlocutors goes down a new avenue, and the interview, where it is important to clarify every detail. In a well conducted interview, truth should never be skimped or sold short, and questioning should continue until all the underlying motivations, beliefs, reactions and sentiments beneath the more superficial response have been revealed.


Note

  1. See Schlackman, B. (1989), "Projective tests and enabling techniques for use in market research", in Robson, S., and Foster, A., (eds) Qualitative Research in Action, Edward Arnold, London.

Recording, analysing and presenting the data

This section looks at how you turn the data you gather in the interview into recordable findings.

Recording the data

Unless the interview is fairly short, much the best way of ensuring that you capture all points is by tape recording it. It will be virtually impossible to write notes and attend to your subject and your research agenda. Two things to remember here:

  • Always make sure that you have permission to record.
  • Never underestimate the time it takes to transcribe a recording: one estimate is seven hours for every one hour of interviewing.

Analysing the data

A very common method of data analysis which relies mainly on words is content analysis . Content analysis is used extensively as a tool in qualitative research for analysing the utterances of research subjects. Content analysis depends upon:

  • deciding what you unit of analysis is – sentences, paragraphs, other unit of meaning.
  • deciding how you will categorize your information.
  • coding the units of analysis according to categories.
  • setting out the data in tabular form, so that you can see the frequency of utterances, and emerging themes, and possibly carrying out some form of statistical and analytical tests, perhaps with software.

Whether or not you used pre-determined categories, or whether you develop these as you go along, will depend upon your research design. You may have decided to use categories that emerge from your literature review, which some writers maintain is a more reliable method, but developing categories from the evidence is also a characteristic of this sort of inductive, theory building technique. You may decide to go for a mixture of these approaches, adding categories that emerge from the findings to ones you have already decided to use.

In A comparison of two alternative interviewing techniques used within an integrated research design: a case study in outshopping using semi-structured and non-directed interviewing techniques (Market Intelligence & Planning , vol. 14 no. 6), Denise G. Jarratt categorizes the themes as follows:

Themes Interviews using a semi-structured technique Interviews using a non-directed technique
Perception of lack of choice in local shops

Disappointment with local instore service

Frustration with local shopping experience

Variety of need? outsourcing behaviour

Variation of media usage with level of shopping enjoyment

Lack of anonymity with local shopping

Time available influences shopping experience

Shopping as a recreation activity

Broad definition of instore service

Desire to see local businesses survive

Price is not the main issue when shopping locally

Poor knowledge of range of goods available locally

Need for information about new products on the market

See The contemplative organization (Maia Duerr, Journal of Organizational Change Management, vol. 17 no. 1) for a very detailed coding system based on her interview questions.

Analysing the data

For a research article, it is probably sufficient to present the main findings as they present themselves as themes.

Denise G. Jarrett, in A comparison of two alternative interviewing techniques used within an integrated research design: a case study in outshopping using semi-structured and non-directed interviewing techniques (Market Intelligence & Planning , vol. 14 no. 6), contrasts the data produced in the two techniques.

Maia Duerr, in The contemplative organization (Journal of Organizational Change Management, vol. 17 no. 1), presents findings according to major themes.

Carol D. Hansen and Mary Kay Willcox, in Cultural assumptions in career management: practice implications from Germany (Career Development International , vol. 2 no. 4), organize their findings round a pre-existing schema - Sackman's schema of cultural knowledge - and cite key utterances. 

John Wood and Tricia Vilkinas, in Characteristics associated with success: CEOs' perspectives (Leadership & Organization Development Journal , vol. 26 no. 3), outline their findings according to categories, and then discuss their findings in the light of other studies.

Other studies also look at implications for practice and future research.