Wooden signpost

How to...
Conduct a focus group

Read about focus groups, what they're good for, how to run them, and how to avoid common challenges.

What is a focus group & for what is it suitable?

Focus groups as a qualitative technique

Focus groups are a data gathering technique used in qualitative and ethnographic research, and as such they share many of the latter's assumptions and call for many of its skills.

The ethnographic approach to research is one where the researcher studies research subjects in their natural setting, observing behaviour and tapping into shared knowledge.

It is most appropriate for studying processes which have a strong social element, such as those which depend on group interaction, and where it is important to take account of diverse views and perspectives. It is used to obtain a rich and detailed view of a reasonably small area as opposed to a wide range of fairly objective data covering a large area: depth, rather than breadth.

It is based on an interpretivist theoretical perspective, which sees positivism as being inappropriate to any study which describes humans, their behaviour and attitudes.

The role of the researcher is as a participant, either active or passive, and he or she is less "objective" than with quantitative research. The researcher is an observer, and needs strong skills of communication, able to tease out the nuances of meaning and hear messages that are not being verbalised, of interpretation, with the ability to unpack messages, and discern patterns.

The most commonly used qualitative settings are participant observation, and the interview. The interview may be either one to one, or within a group.

Key features of focus groups

"The group interview is essentially a data gathering technique ... that relies upon the systematic questioning of several individuals simultaneously in a formal or informal setting."
(Fontana, A. and Frey, J.H., 2000, "The interview: from structured questions to negotiated text", in Denzin, N.K. and Lincoln, Y.S. (Eds), Handbook of Qualitative Research, 2nd ed., Sage Publications, Thousand Oaks, CA)

"Qualitative groups (Dick, 1999) are generally defined as groups of people brought together to participate in the discussion of an area of interest. In market research this would be for example, to discuss a product, brand or advertisement for the purposes of clarifying an area of concern for a client. In social research it could be to discuss an area of social concern."
(Boddy, C., 2005, "A rose by any other name may smell as sweet but 'group discussion' is not another name for a 'focus group' nor should it be", Qualitative Market Research: An International Journal, Vol. 8 No. 3, quoting Dick, B., 1999, Rigour without Numbers: The Potential of Dialectic Processes as Qualitative Research Tools, 3rd ed., Interchange, Chapel Hill, NC)

A focus group is always directed by a moderator/facilitator, with varying degrees of control depending on whether it is:

  • Structured, where the facilitator is more directive, having a set agenda and not letting the group deviate from it. The reason for having a structured focus group is that the researcher has a fairly good idea what the issues are and wants to spend time exploring them.
  • Unstructured, where the group discussion is more free flowing. This type of focus group may be used in product research, or for an exploration of general issues, such as how employees feel about a new pay scheme. A varient of the unstructured focus group is the nominal group, where participants are physically isolated, but views are gathered for comment with a possible later meeting, and the Delphi technique, where the idea is to reach a consensus.

As stated in the introduction to this piece, the terms "focus group" and "group interview" are used interchangeably. In his article "A rose by any other name may smell as sweet but 'group discussion' is not another name for a 'focus group' nor should it be" (Qualitative Market Research: An International Journal, Vol. 8 No. 3), Clive Boddy suggests the adoption of two terms, focus group interview and focus group discussion, the one structured and the other unstructured, as defined above.

The focus group can also either take place in a formal setting, designated by the researcher and where participants come in response to invitations, or in an informal setting, such as on a street corner, shopping mall etc. There is a benefit in having the group in a familiar setting, which can reduce the sense of "otherness".

The defining quality of the focus group is its collectivist nature: in concentrating on the group, it allows for expression of a variety of different perspectives (thereby aligning it to a non positivist, interpretivist approach which allows for multiple perspectives). Not only is it possible to gain access to the experiences of many different individuals, but also because individuals interact with one another, data is enriched as group members spark each other off, enabling views to be reformulated through exchange, nuances of meaning teased out, recall strengthened, and shy members given confidence. Data are enhanced as they are drawn from a social context, and interaction is horizontal – between participants – as well as vertical – interviewer/interviewee.

Another advantage of the focus group, and one that is especially relevant with a clearly defined research population, is that it is an efficient way to interview more people in a shorter amount of time than is possible with a one to one interview.

As a technique, the focus group has elements of both participant observation and the individual interview, calling for both a questioning style and the observation of group processes. Whether the researcher chooses the individual or group approach for interviews will depend upon a number of factors.

Individual interviews are good:

  • when you need to focus on an individual's personal circumstances or history
  • when the subject-matter is very complex, for example when you are talking about complex systems or processes
  • when confidentiality is required
  • when coming to a particular location may inhibit, and it is therefore better to interview on subject's home ground.

Group interviews are good:

  • when interaction between participants will illuminate the research issue
  • when issues are abstract or conceptual, or are concerned with attitudes or views
  • when interviewees are likely to be intimitated by a one to one setting.

Of course, the two methods can be combined as when a focus group is followed up by a series of one to one interviews with selected participants. Knowing how to combine them is a matter of knowing how their particular strengths and weaknesses will fit into your research design.

The following examples show uses of focus groups in both structured and unstructured settings:

History of focus groups

The focus group as a technique was first developed in the 1920s, mainly to help develop survey questionnaires, although also for other purposes of social research.

In the 1940s, Robert Merton and Paul Lazarfield used the technique for group interviewing people about their reactions to wartime radio programmes (see Merton, R., 1987, "Focused interviews and focus groups: continuities and discontinuities", Public Opinion Quarterly, No. 511, pp. 550-566). From that time until the 1970s, focus groups were mainly used in market research, for such matters as brand images, packaging and product choice.

By the late 1980s, however, the focus group had become recognised as a social science method in its own right, partly in reaction to perceived limitations of other methods: the individual interview was seen as liable to be influenced by the interviewer, and the closed questions of surveys too limited in terms of depth of data. As against this, quantitative methods have – particularly in the USA – dominated both the social sciences and management sciences, although there is currently much interest in qualitative methods on both sides of the Atlantic, so it seems likely that the popularity of the focus group will continue.

Advantages and disadvantages of focus groups

K. Denise Threlfall (in "Using focus groups as a consumer research tool", Journal of Marketing Practice: Applied Marketing Science, Vol. 5 No. 4) provides a good summary of the advantages and disadvantages of the focus group:

Advantages:

  • The researcher has less control over the subject, which means that more weight can be given to the data.
  • It allows for the observation of social interaction, which can enrich the data, correct the individualistic bias inherent in the one-to-one interview and take account of the social construction of meaning.
  • The participants can take control of the discussion process, moving the conversation to topics that are of relevance to them.
  • It is particularly suited to research amongst populations which may perceive their views to be devalued, for example socially excluded groups, or in an organisation where a particular group of workers may feel that their opinions are ignored by management.
  • It can act as a change agent as participants become aware of the limitations of their situation and produce an agenda for change.
  • As compared with a questionnaire, participants have more control in that they are able to respond to questions in greater depth, and probe awkward and sensitive issues.
  • It is relatively inexpensive to conduct, and less expensive and time-consuming than individual interviews.
  • It is good for researching attitudes, convictions and beliefs. 

Disadvantages:

  • May be subject to groupthink, where the predominant group view may dominate individual expression.
  • May be dominated by particular members while others may be shy.
  • Personal opinions may distort the result.
  • Cannot generalise the findings of one focus group.
  • A focus group is not a normal way of interacting socially, therefore the advantages drawn from observation of social interaction may be limited.
  • As compared with participant observation, the behavioural information is limited to verbal communication, body language, facial expression etc.
  • The presence of the facilitator may be restricting.
  • Because of the dynamic nature of the group process, greater skill in interviewing is called for.
  • Difficult for sensitive or intimate topics or if confidentiality is involved.
  • Difficult if for any reason participants don't feel comfortable with one another, as for example in a work situation where supervisors and supervisees are brought together.
  • Some populations may find it difficult to travel to a particular venue.
  • Difficult to control number as invitees may not turn up.

Uses of the focus group

Focus groups are used widely in the following areas:

  • social research, often to explore issues surrounding the delivery of services
  • marketing, to look at the effectiveness of an advertising campaign, or the likely reception of a particular product
  • politics, by political parties to explore voter concerns, for which they have received a certain notoriety as some would see their use as merely window dressing, while others criticise an over-readiness to generalise from the particular
  • in a range of academic disciplines such as anthropology, communication, education, marketing, political science, sociology, psychology, nursing and public health.

Examples

"Using focus groups to evaluate health promotion interventions" (Kirsten Mitchell and Patrick Branigan, Health Education, Vol. 100 No. 6) describes how to use focus groups in health promotion.

"Using focus groups as a consumer research tool" (K. Denise Threlfall, Journal of Marketing Practice: Applied Marketing Science, Vol. 5 No. 4) looks at its use in marketing research.

"Using focus groups to explore children’s perceptions of smoking: reflections on practice" (Lorna Porcellato et al.Health Education, Vol. 102 No. 6) discusses the use of focus groups with children.

Focus groups are also used widely with the following applications:

  • exploratory research, to define a research problem 
  • to generate a hypothesis
  • to carry out a needs assessment, for example if you are developing a training or educational product and you want to know what people need
  • to identify key informants who can later be interviewed
  • as a precursor to a survey, to help determine both the type of questions and their wording, the latter through listening to the type of language used by participants
  • to obtain rich accounts of particular events
  • to explore attitudes, for example towards new work practices, changes in a company
  • to triangulate with other research findings, and where you want to explore findings in greater depth.

How to facilitate a focus group

Skills required

We shall be looking further down this page at the tasks of steering a focus group and framing questions, but in general terms the group leader needs to:

  • be objective, remembering that this is a research exercise and he or she should encourage equally material which does and does not support the hypothesis
  • be flexible, allowing free flowing discussion but ensuring that there is enough structure to cover key topics, and knowing the right point to intervene if someone is too domineering, or if the group is straying off the topic
  • be empathic and a good listener, able to put people at ease, and with the right social skills to be able to draw in someone who is shy, perhaps by looking at that person in an encouraging way
  • possess a combination of tact and assertiveness, able to project him or herself onto the group without dominating
  • possess the courage to stay with a silence if that silence can engender further thoughtful contribution.

Whether or not the researcher and the facilitator are the same person will depend on the project, but if they are it is all the more important to remember the above skills, and in particular to be objective.

The stages of group process

Like any group, focus groups are subject to certain collective behaviours. Drawing on research done on small groups, Helen Finch and Jane Lewis ("Focus groups" in Ritchie, J and Lewis, J. (Eds), 2003, Qualitative Research Practice, Sage Publications, Thousand Oaks, CA) describe five stages of the group process:

  1. Forming. At this stage, everyone is relatively tense and guarded, and concerned with inclusion.
  2. Storming. Relatively early on, tensions and disagreements begin to emerge. Data from this stage of the group discussion will not be very reliable.
  3. Norming. The group begins to settle down, observe ground rules, and cooperate.
  4. Performing. Here, the group's discussion is at its best and most interactive, the researcher can relax and let the group conduct itself.
  5. Adjourning. The group works towards an ending. Sometimes, there is also mourning, when people seem reluctant to leave.

The stages of a focus group

The facilitator needs to be aware of the different stages the focus group should go through, and how to conduct each stage.

Setting the scene and laying ground rules

As participants arrive, the facilitator should welcome them and thank them for coming. Once everybody is present and seated, the facilitator should provide a brief introduction outlining the purpose of research, laying ground rules (one person to talk at a time, all views welcome, confidentiality) and to stress that there is no hidden agenda, and that all views will be treated in confidence.

Next, each participant introduces themselves, giving name and a brief bit of background. The facilitator should make a spatial note of where people are sitting, which will serve as a useful aide memoire when looking at the session notes or listening to the recording.

The opening topic

The facilitator introduces the opening topic/question, which should be fairly general, and capable of generating discussion. Attempts should be made to make everyone contribute as a way of breaking into the group. It may be necessary for the facilitator to intervene quite a bit by asking questions, and generally keeping the discussion going.

Discussion

When the discussion gets under way, the facilitator will need to tread a fine line between ensuring that all points are covered, attending to particular details and noting the odd interesting piece of data that emerges independent of the official questions, and promoting group discussion, between heeding thoughtful pauses and making sure that the discussion does not dry up. Active listening is very important, and they may need to probe at particular points.

Ending the discussion

It's very important to end the discussion on a positive note, especially if difficult material has been raised, and also to thank people for coming.

Framing questions

The facilitator will have prepared a guide with a list of topics/questions to be covered, the amount of detail and specificity of which will depend upon the degree of structure desired of the group. It is advisable however to have memorised this schedule in advance, as to read from questions will look forced and inhibit discussion.

Questions should be:

  • sufficiently stimulating and keep the discussion going
  • relevant to participants
  • capable of providing concrete data
  • open, so as to lead to in-depth responses and rich data
  • clear, avoiding ambiguity
  • couched in language that participants understand
  • sensitively worded if they explore (particularly difficult) feelings.

They should not:

  • be worded in such a way as to influence or "lead" the answer
  • cover more than one topic per question.

The number of questions and their detail will vary depending on how structured the focus group is to be, but in general five to eight questions is considered a good number, and these should move from the general to the particular.

Examples

Janice Dreachslin, in "Focus groups as a quality improvement technique: a case example from health administration education" (Quality Assurance in Education, Vol. 7 No. 4) quotes Krueger (1994) as recommending the following format for questions:

  • An opening question which is factual in nature and establishes the participants’ commonality.
  • A series of introductory questions that acquaint participants with the topic and initiate conversation.
  • Transition questions that move discussion to the focus group’s central topics
  • Two to five key questions that are central to the analysis.

An ending question to accomplish one of the following objectives:

  • encourage participants to state their final position on key topics
  • give participants the opportunity to comment on the accuracy of a facilitator-generated summary of the key input generated by the group
  • enable participants to offer any additional comments relevant to the group’s key purpose.

Dreachslin also discusses the pros and cons of focus groups.

See also Krueger, R.A. and Casey, M.A., 2000, Focus Groups: A Practical Guide for Applied Research, Sage Publications, Thousand Oaks, CA.

The Appendix of "Using focus groups to assess student needs" (Britt Anna Fagerheim and Sandra J. Weingart, Library Review, Vol. 54 No. 9) contains an example of focus group questions.

"Using focus groups to explore children’s perceptions of smoking: reflections on practice" (Lorna Porcellato et al.Health Education, Vol. 102 No. 6) also provides a copy of the interview schedule in the Appendix.

Conducting a discussion

Leading a good focus group discussion is a balancing act between on the one hand flexibility, allowing a free-flowing conversation, and on the other structure, ensuring that the conversation does not stray too far from the research objectives, that it remains relevant, inclusive of everyone and not dominated by one person. It is very important to obtain responses from the maximum number of people in the group to get the fullest coverage.

Ideally, the discussion should regulate itself and the issues be covered by the respondents without much steering. However, the following techniques will be useful:

  • Steering. Techniques for steering the discussion include reconciling and commenting on divergent views, making links.
  • Probing. The facilitator may need to probe interesting points to get a fuller response, for example by opening up the discussion to others ("How do other people feel about this?"), asking a question ("Can you say a bit more about this?"), repeating what the participant has said, or highlighting different views.
  • Listening to material which may seem divergent. A  participant may launch into an anecdote which may appear to be off the topic but which may contain a nugget of highly relevant information. At the same time it is also necessary to watch out for emergent issues, which may not be in the notes but which may lead to fertile areas for research.
  • Be alert for non verbal signals. For example, nodding or shaking the head to indicate agreement or disagreement, a frown, surprise. Such non verbal signals are best verbalised so that they can be translated into data.
  • Controlling the balance between participants. It's important to moderate between the dominant person and the shy person, and to encourage equality of contribution. This should be done tactfully, with avoidance of hurting someone or embarrassing them. A lot can be achieved by eye contact: withdraw it from the dominant person, look at the shy person encouragingly. If this fails, then try a verbal intervention: "That was very interesting, can we perhaps hear some other views?", "You haven't had a chance to say what you think".
  • Encourage expression of personal views. Participants need to be encouraged to say what they think, which they may find difficult if they feel that it is not politically or strategically acceptable (for example, complaining about a long hours culture). The facilitator should encourage this by stressing that disagreement, and challenging received attitudes, is acceptable.

Example

See "Workplace health concerns: a focus group study" (Annette M. Jinks and Ruth Daniels, Journal of Management in Medicine, Vol. 13 No. 2) for a discussion of the role of the researcher in their focus groups.

Design issues of focus groups

The focus group within the overall research design

Whether you are using just focus groups as a research method will depend on whether the technique can answer all your research questions by itself, and whether it alone is an effective way of reaching your target population. It is in any case highly likely that you will want to triangulate it with another method in order to increase reliability.

Which other method you choose will depend on the type of data you are trying to collect. You may for example wish to explore some issues in further depth through individual interviews, or you may wish to combine the focus groups with a survey (see our accompanying How to.. design a survey guide), for which the former act as a method of exploring the issues.

In "Consumer acceptance of online banking: an extension of the technology acceptance model" (Internet Research, Vol. 14 No. 3), Tero Pikkarainen et al. use focus groups to test a questionnaire, and to verify the hypothesis that affective factors were relevant to online banking acceptance.

"Probing user perceptions of service quality: using focus groups to enhance quantitative surveys" (Gwyneth H. Crowley and Charles L. Gilreath, Performance Measurement and Metric, Vol. 3 No. 2) describes the use of this technique to gain greater depth of data following a survey.

One focus group is difficult to generalise from, and you should have at least two, and more if you want to examine the views of more than one target population, say for example employed and unemployed, or people at different levels in an organisation. Generally, fewer groups are necessary for more structured, exploratory work, and more for unstructured work.

In "Application of the means-end value hierarchy model to understanding logistics service value" (International Journal of Physical Distribution & Logistics Management, Vol. 27 No. 9/10), Mentzer et al. explore the value of logistics in terms of enhancing value perceptions to customers. They held 13 different focus groups each featuring a different product but covering the same broad areas. The moderator introduced a few general topics to get the conversation started, but afterwards let the discussion roam freely as this unstructured technique was felt to be most appropriate to gather underlying themes.

Sampling considerations, group size and composition

For recruitment, convenience sampling (see our companion "How to... design a research study" guide), whereby participants are recruited based on their easy accessibility to the researcher, will be cheaper than probability sampling, although you should consider whether or not this will provide sufficient rigour for your design. Snowball sampling, whereby participants recruit others, may also be possible, and purposive sampling may also be appropriate.

See "Community policing and social capital" (Nathan W. Pino, Policing: An International Journal of Police Strategies & Management, Vol. 24 No. 2) for an account of how he recruited participants from neighbourhood groups as these were more likely to be sufficiently informed.

Care should be taken over the wording of the invitation to attend, particularly if inviting members of an organisation as the invitation can seem like an instruction.

Estimates of desirable group size vary between four to 12 and six to eight. It's a good idea to invite more people that your target number in case some don't turn up. A small group may be biassed by existing relationships among participants and silent members, whereas a large group may be difficult to control and splinter into sub-groups. Less than four people will lose the character of a group. The size will vary according to a number of factors:

  • sensitivity of the issue – for a sensitive issue a small group is better
  • the breadth/depth of data required – if you want breadth, go for large, if you want depth, go for small
  • the target population – if this is highly articulate, it will probably be better to go for small
  • the structure of the focus group – if the group is to undertake relatively structured work say in a workshop, then it's better to have a larger group.

In determining group composition, it's important to maintain a balance between homogeneity and heterogeneity. In studying disadvantaged groups, reticence may be lessened by ethnic and linguistic homogeneity, and it is generally better to avoid large differences in status and power. For example, when researching issues in an organisation it is best to have different focus groups for different levels, and to select a horizontal slice through the organisation. Homogeneity is particularly important when studying sensitive topics – for example in a study of people's attitudes towards threatened redundancy, you should not include people who don't feel threatened.

See "Focus group exploration of firm-employee relationship strength" (Carmel Herington et al.Qualitative Market Research: An International Journal, Vol. 8 No. 3) for a discussion of the use of focus groups and how the authors justified size and composition.

Group members should all have a degree of commonality in terms of how they relate to the topic, for example if researching reactions to a service, all group members should be users. On the other hand, too much homogeneity can impede discussion as agreement may make people fail to articulate views sufficiently clearly.

It is also important to avoid tokenism, for example just one male, one member of an ethnic group. If the group have a demographic feature in common, then it is better that the leader also shares the feature, e.g. is of the same ethnic origin as the group, or, in the case of a group studying issues of female managers, the leader should also be female.

It may be desirable in some instances that the group are strangers to one another, as this may make discussion easier. If you find that some people know one another, then get them to sit apart. On other occasions, for example if you are investigating issues within an organisation, members of the group will already know one another.

Time, place and other practicalities

In setting the time for the focus groups, you need to consider potential clashes with major events, for example, sporting fixtures or major holidays, as well as the needs of your target population. (If researching in an organisation, there may be particular days of the week which are allocated to meetings and it is best to avoid these.) For example, mothers with school age children are more likely to come to an event during school hours, busy managers more likely to show up if the event is organised around lunch time, especially if lunch is provided (see below). If planning more than one focus group, have these at different times to accommodate people's schedules.

The venue also needs to be appropriate to the participants. If using a pre-existing group, then there is some advantage to meeting on familiar territory, for example a church hall, or meeting room of a place of work (although in the latter case, the setting should be as neutral as possible). If there are no such places available, then you will have to give some thought to hire of a room in, say, a pub or local hotel – in which case, you need to make sure that you have the funds in your budget.

The room needs to be of adequate size and quiet, so that group members are not distracted. Pay attention to the arrangement of the seating – are the chairs arranged in such a way as to facilitate discussion? If you are planning to have the groups back to back, you need to have a second room for the next lot of participants to wait in.

It is quite common to offer some sort of incentive to attend the focus group, such as refreshments (which should be offered anyway). For example, publishers of school textbooks offer vouchers for educational material.

"Using focus groups to assess student needs" (Britt Anna Fagerheim and Sandra J. Weingart, Library Review, Vol. 54 No. 9) describes how the authors went about recruiting students to their focus group, and how they organised the practical issues.

Focus groups online

Needless to say, once you remove focus groups from the conventional physical environment there needs to be considerable differences in the design, and despite the growing sophistication of Internet technologies, focus groups in a virtual environment are still a relative novelty. As with any form of virtual communication, the loss of non verbal clues creates a challenge, with the use of probes being particularly problematic as silence has different connotations.

Advantages are that people do not have to travel to a particular location, and those who are shy may have a greater degree of confidence. There is the very real plus that data is automatically recorded.

Care needs to be taken to create a genuine "virtual space", which should have its own design, make people feel at home and be dedicated, so that it can be free of other distractions.

The following three articles, though none of them very up-to-date, discuss virtual focus groups.

"Focus groups in cyberspace: using the Internet for qualitative research" (Henrietta O'Connor and Clare Madge, Qualitative Market Research: An International Journal, Vol. 6 No. 2)

"Designing and conducting virtual focus groups" (Casey Sweet, Qualitative Market Research: An International Journal, Vol. 4 No. 3)

"On-line focus groups: conceptual issues and a research tool" (Mitzi Montoya-Weiss et al.European Journal of Marketing, Vol. 32 No. 7/8)

Recording

An important consideration is, how will you record the data? There are various ways:

  • Relying on memory.
  • Taking notes, in which case you will need a second facilitator who sits next to you, as you will be too tied up with moderating the group to take notes as well.
  • Tape recording. This is probably the best way, but you should remember the length of time it takes to transcribe the results. You should also have a good quality tape recorder with a multidirectional microphone placed on the table, and a good quality tape close to the recorder (trying to transcribe from a poor quality tape is a very frustrating experience). You should also ask permission before you record, and check that the recorder is functioning before you start.
  • Video recording. The advantage here is that you can record body language as well as verbal data, but it's very intrusive and you will definitely need permission.

Analysis

The level of sophistication of analysis may vary – you may feel that a full-scale content analysis (which we shall deal with as a separate topic) is called for, or it may be sufficient to do a more "informal" analysis of the themes.

Whichever method you choose, it is advisable to listen to the tape recording or look at your notes, and analyse the main themes and sub-themes, as well as the participants' characteristics. You need to think about how you are going to substantiate your findings: what is the real evidence? Are there particular characteristics of the participants which might affect the data? Are there any particularly strong statements? Does a consensus emerge?

Having done this, you can then summarise in a series of statements expressing the main themes.

Most of your data will be in the form of text; if you use numerical analysis it should be to give an impression, for example a certain percentage felt... It is not usual to subject the findings of focus group research to statistical tests, although recently analytic methods for qualitative data have become more sophisticated.

When you write your report, you should organise according to major themes, noting any significant differences between groups but not writing individually on each group. More detailed analysis should go in an Appendix.

In "Workplace health concerns: a focus group study" (Journal of Management in Medicine, Vol. 13 No. 2), Annette M. Jinks and Ruth Daniels describe how they analysed their data using content analysis and "meaning units" or words or phrases that characterised categories.

"Recruiting and retaining scarce information technology talent: a focus group study" (Diane Lockwood and Al Ansari, Industrial Management & Data Systems, Vol. 99 No. 6) provides an example of reporting the findings of focus groups.

"Using an ANN-approach for analyzing focus groups" (Marcus Schmidt, Qualitative Market Research: An International Journal, Vol. 4 No. 2) describes a quantitative approach to analysis of focus group data.

Resources

Books

Websites

As is to be expected, there is much information on focus groups on the Web, much of it from organisations with other main objectives, e.g. advising particular groups, marketing etc.