Aims and structure

Before working out the who, where, when and how of setting up a focus group, start with the why. Its purpose must be clear. What is it meant to achieve? What are the questions that need answers? Figure 7.1 suggests a process for preparing and running a focus group. Whether using a structured format or not, prepare a list of topics to guide the session. Less is more; covering a few key topics in depth will yield better information than squeezing lots of issues into the time available. Having said that, the group make-up can play a part here. Homogenous groups can get through a list of topics quicker, as there tends to be greater consensus, whereas more mixed groups usually have a wider range of views and more potential for debate. So take these factors into account in working out what to cover in the session.

If several focus groups are planned, consider the level of question standardisation between them. Asking the same questions in each session means the material can be aggregated, but it can reduce the scope to pursue other interesting topics. Another option is a rolling question schedule. Participants discuss a set of topics in the first session, then issues arising from that session are brought to the second session, then the third session includes new topics that arose in the second, and so on. Although it allows less comparison between sessions, the groups will be able to cover more issues and generate more material (this may not necessarily be seen as an advantage at the analysis stage however!).

Using the funnelling technique described at section 6.4, start a session with general and easily answered questions and get everyone to contribute; then address more challenging topics. It’s sensible to cover the most important issues relatively early on while participants are still fresh and there’s plenty of time. When presenting information about the proposal, remember that participants may be unused to discussing planning and design issues in depth, so explain any terminology that might be unfamiliar, use everyday concepts and vocabulary, and keep checking for understanding. Discussing controversial or emotionally charged topics can be strenuous for participants, so this should also be factored into the schedule and followed by a break or a more enjoyable activity. Allow about 90 minutes maximum for the session overall.

Aim to gather information about participants’ experiences, opinions, feelings and preferences, with a good amount of content in each category. This means asking lots of open-ended questions, starting with, ‘What’s your experience of …’, ‘What do you think about …’ or ‘How would you feel if …’, for example. Ensure that questions are relevant to the group, that everyone will understand and will be able to answer them, and remember to ask follow-up questions for more details or explanations: ‘Why do you say that?’, ‘Could you say a bit more about what happened?’ And a question that’s always worth asking: ‘Does anyone have a different view?’

Selecting participants

Having drafted a workable session strategy and time plan, the next step is to put together a good group of people to answer the questions. Send a short screening survey to prospective participants first, explaining the objectives and gathering some details to help select the people who can most usefully contribute. Also include information about recording the session, confidentiality, details of the time and venue, accessibility, length of the session and what will be covered, and check they’re happy with these arrangements.

Normally the optimum group size is eight to twelve, although homogenous groups can work with slightly fewer or more than this. The group composition depends on the research objectives. Demographically similar groups will be needed for some issues, but a general mix is suitable when a broad range of perspectives is wanted. Demographic factors play a part in group dynamics, which in turn affects the responses offered. For instance, it’s been shown that men tend to dominate discussions in mixed-gender focus groups (Stewart and Shamdasani, 2014), something for the facilitator to monitor.

Venue

Choose a venue that’s fully accessible, that everyone can easily reach and where everyone will feel comfortable. Quiet and privacy are essential, so there are no interruptions or distracting background noise. Focus groups are more focused in rooms with natural light. Room size and layout also affect interactions, as people are more relaxed and forthcoming when they have a good amount of personal space, whereas smaller spaces tend to produce more intense debate and polarised opinions (Sanoff, 2000). Discussions flow better when everyone can see each other, so a round table is ideal. The facilitator’s position also influences dynamics; if they’re positioned outside the group, participants are more likely to talk to each other.

The day will run more smoothly with one or two colleagues helping out: greeting people as they arrive, taking care of refreshments, setting up recording equipment and making notes. They will also be able to help review and analyse material afterwards. Recording the session is essential, whether audio or video, and the equipment must be tested before the session to ensure it will pick up everyone’s voices clearly; a digital voice or audio recorder, or a video microphone with an omnidirectional setting works best with groups. Try and position recording equipment unobtrusively so participants are not conscious of it. Make sure batteries are fully charged and spares are on hand, that enough memory is available and that there’s a back-up option just in case.

If it’s possible to provide refreshments, then do so; participants will feel appreciated and more relaxed, and it helps create a more congenial atmosphere for discussion. For longer sessions it’s almost always necessary, and can provide a welcome break. But don’t have refreshments on the table during the session; eating and drinking sounds can easily obliterate the discussion!

Figure 7.2a

Figure 7.2b

Figure 7.2c

Gathering data

The facilitator might start with an icebreaker exercise, depending on the group’s make-up and purpose, or move straight on to the main business. As with the funnelling technique in questionnaires, start with factual and easily answered questions before discussing more complex issues.

Once the session is under way, the facilitator has two main tasks:

1. Keeping to time. This means balancing participants wanting to pursue issues of interest with the need to cover the topic schedule within the time available. A good knowledge of the matter under discussion will enable the facilitator to judge the potential value of conversational tangents.
2. Ensuring that all participants contribute. Any focus group will include talkative and reticent participants. The facilitator has to involve the reticent members, particularly in mixed groups, and manage the talkative ones. They might ask individuals to hold back so others can have their say, and the group may also self-regulate the discussion. To reduce the influence of more vociferous members and ‘first speaker effect’ (see the Appendix on types of bias) the facilitator can ask everyone to write down their responses to a question, then pool the answers for all to read.

One more thing that requires a facilitator’s constant vigilance throughout a session is their own neutrality. The facilitator should mask their personal reactions to participants’ views, guarding against verbal and non-verbal responses that imply approval or disapproval. As well as keeping comments and body language neutral, they need to give participants equal encouragement and time to speak, whether their comments are favourable or unfavourable, and avoid placing more emphasis on positive comments when summing up. This can be particularly challenging when designers run focus groups to get feedback on their work. But all participants’ responses must be treated equally, and the facilitator should seek to build rapport, listening attentively to everyone and expressing interest through tone of voice and body language. If participants look to the facilitator for validation of their views, as often happens, they can deflect the question back to the group and ask what other people think. This is another argument in favour of the facilitator sitting a little outside the group, reducing the temptation for participants to look for their reactions. Another key aspect of neutrality is avoiding leading questions, which either imply a preferred answer or guide participants away from certain responses. Desired outcomes shouldn’t shape how questions are worded.

The facilitator will need to take written notes during the session, without trying to capture every detail or being over-selective. It makes sense to record things beyond the substance of what was said: people’s expressions, non-verbal responses, group dynamics, comments that caused a notable reaction, and so on. The facilitator can usefully spend a few minutes after participants have left reviewing their notes to check for clarity, adding in detail and explanation as required, highlighting any significant themes or connections, and noting anything unexpected that may need to be looked at further. They should hold off at this stage from forming interpretations, and focus on asking questions about the data rather than trying to find answers.

Activities and exercises

Activities and exercises in focus groups are completely optional. For some groups or topics, discussion alone will yield enough good material; other groups may be more responsive to interactive elements. As always, start with the desired end results, look at what information is required and then work backwards to find the best ways to obtain it. Given the time constraints, any activities should be quickly explainable, short, fun and/or interesting, and easy for participants to complete. (Lengthier and more complex tasks, or idea-generating exercises suit workshops better.) Activities that work in focus groups call for immediate personal responses without needing any prior knowledge, such as:

• Prioritising/sorting exercises using cards, post-it notes, etc.
• Map- or model-based activities. Ask participants to identify things like places that they like and dislike, or routes they prefer and those they avoid.
• Image- or object-based activities. Pictures or artefacts are a good way to generate discussion, and are also effective in working with more sensitive topics. ‘Mood boards’ and collage can help people consider design options and identify preferences.

A useful insight from UX focus group research is that if participants are shown images of a product concept, they can form an opinion, but if they’re given a prototype to examine, they can give much more valuable feedback based on their first-hand experience. There’s certainly scope to use focus groups to consider usability issues more in spatial design, using fly-throughs or physical and virtual models: something for designers to bear in mind.

During group activities, the facilitator should continue observing the proceedings, checking that everyone contributes. The process itself can be revealing and the insights it offers can be as useful as the outcome, so it can help to ask questions as participants work on their task to understand their thinking.

Close the session on a positive note with a simple or light-hearted question or two, especially if there were any difficult dynamics. Ask if anyone has any questions before the group breaks up, and outline what will happen to the material from the session, the next steps in the design and decision-making processes, and further opportunities for participation. Thank the group warmly for their time and input, and generally aim to make everyone feel glad they came. Send a follow-up email shortly afterwards thanking participants again, asking for feedback on the experience, and whether they would like updates on the project and future events Add them to the mailing list if so.

Managing data

There’s no standard procedure for analysing the large amounts of complex qualitative material that focus groups produce, but the ‘Quick guide to analysing qualitative data’ gives some suggested steps. Often the first step is transcribing the discussion. This can take a lot of time (it’s normal to spend an hour transcribing ten minutes of speech), so it may be necessary to select material to focus on, guided by the initial review. The time investment is worthwhile when a deeper analysis is needed, to discover the underlying reasons for people’s attitudes or behaviour, for example (and if less depth is needed, focus groups are probably aren’t the best method).

Analysing data

Qualitative data analysis always entails disassembling and reassembling material. This might involve investigating various themes, questions or types of users, but always using the same approach of taking data apart and rearranging it, and noting patterns and linkages that emerge. The disassembly stage involves pulling out everything relating to a subject, then tagging these data chunks with keywords and adding comments as necessary. Then start on the next subject, pull out all relevant chunks, add new keywords and comments, and repeat with all the remaining subjects.

A common approach to this phase in UX is affinity diagramming. This is an interactive visual method of identifying themes within the data, which starts by summarising each comment on a post-it note, and then clustering together comments relating to specific objectives or themes that emerge during analysis (see Figure 7.3). It works well for smaller datasets, short programmes or where there aren’t a wide range of issues under consideration. For large and/or complex datasets, I recommend coding, employed widely in qualitative analysis in social science. This entails creating a structure of categories and keywords to index the data, and then filtering, sorting, and cross-referencing to identify themes and connections. The ‘Quick guide to analysing qualitative data’ gives more detail. Both approaches can be usefully done as a team, using index cards or post-it notes, maps and plans of the scheme, and the results of desk research and any previous engagement work, for context.

Whatever the method, the key tasks are to compare and contrast within the material, looking for similarities and differences, patterns and relationships, consistencies and inconsistencies, and the unexpected and anomalous. Designers and visual thinkers will probably find that sketching diagrams, flow charts, mind maps and so on help in exploring and giving coherence to the data. As visually presented information is more accessible and engaging for many laypeople too, these sketches are worth keeping for future communications. It’s also important to keep questioning throughout this stage. Are we asking the right questions, and the right people? Does this data confirm or contradict findings from other research?

Drawing conclusions

If focus groups are held at the initial stages of a project to understand the general local context, it will be too early to draw conclusions, so it’s best just to summarise the range of topics covered, the main areas of interest or concern, and suggestions for next steps. For focus groups dealing with more specific issues, the interpretation process should be guided by the research questions. It’s important to bring all the relevant content into the interpretation phase, and to resist the temptation to cherry-pick the more interesting material or that which supports a desired outcome. Recalling the validity and reliability imperatives (see section 2.2), analyse data impartially and represent it truthfully. However, remember that focus groups aren’t a microcosm of the wider community, and findings aren’t generalisable; it’s impossible to know whether participants’ views are representative, or how much weight can be given to them. And studies have shown flaws in ‘groupthink’; for example, people can find other group members’ opinions more convincing than the factual information given (Weinschenk, 2011).

Figure 7.3 Affinity diagramming