Focus Groups

What is a focus group?

Focus groups are a qualitative research methodology. To recap qualitative research, focus groups are inductive (theory building), use a phemological epistemiology and the ontological position is constructionalist. They are used to seek social meanings and answer the critical question of 'why'. In other words, you can find out about things that cannot be seen or heard, such as the interviewee's inner state, the reasoning behind their actions and feelings. A qualitative researcher starts with a research question and little else. The theory develops during the data collection process and is built from or grounded within the data. Building theory is done by making comparisons - for example, observe an event and look for similarities or differences. Focus groups, unlike interviews, involve small groups of people, usually between 8 and 20 people. There is no optimum number as the size of the group depends upon the purpose of the study and the type of questions asked. Instead of being an interviewer, the researcher operates in the role of moderator or facilitator of a discussion. The participants of the focus group discuss a specific issue and that has been selected for discussion and the researcher acts to create a friendly atmosphere conducive to the discussion.

Focus groups can be defined as the explicit use of group interaction to produce data and insights that would be less accessible without the interaction found in the group. The aim is to gather detailed narrative data on participants, insights, perceptions and attitudes. Participants should share similar characteristics. They should have been brought together for their commonality rather than for their diversity. It is also good practice to not mix genders within a focus group. Presence of the other gender can intimidate participants - for example, men may be more likely to show off and women may remain quiet on an issue. Either gender may be more inclined to give a desirable answer to look good in front of the other. (This is sometimes called 'desirability bias'). Focus groups should be all female or all male. You should also ensure that participants do not know each other too well. This may result in things being left unsaid. Participants may not state relevant pieces of information as they would be disinclined to either repeat a story that their friends already know about or may not want their friends to find out.

Stages of Focus Groups

Planning

All focus groups should be planned well in advance. This includes the budgeting of salaries of moderators, costs of audio-taping, transcribing, participants' travel costs, retention fee for their time, venue costs etc. You should also plan how many participants you want for each focus groups and how many focus groups are needed before you reach saturation. The term saturation refers to the point where you find that your data is being repeated and you are not finding out any new material or themes. To get an understanding of how many groups you'll need and of what size, look at similar studies done on your topic and see how many they used. This will give you a good idea of what you will require. You will also need to work out the timing of a session (typically, a focus group lasts between 1 and 2 hours) and how you will invite your participants. Sampling is always purposive as you are aiming for inclusiveness not representativeness (and therefore generalisability to the population). However, sampling could be snowballing or you could recruit through advertisements.

Conducting

Before conducting the focus group, have an interview guide drawn up. This is very similar to the interview guide used in unstructured interviews. This guide should be based upon the existing literature. All questions should be open-ended and have some questions ready for prompting discussion. However, be very careful that the focus group does not turn into a question and answer session. Let the group carry the discussion and facilitate it, don't dominate it! Don't forget to welcome the participants at the start and explain the purpose of the focus group. At the end, thank them for their time. It is usual form that the group is observed by another researcher. This can be overtly done or the observer can be behind a one way mirror. The role of this observer is to take notes during the discussion which can be compared with the transcript. The observer should also take note of where participants are seated and prescribe pseudo-names, numbers or letters to each participant. These names should not be declared to the participants. The observer can used them during the focus group to denote who spoke when and in what order. Any body language should also be noted as should any changes to the dynamic of the group (for example, was there a dominant personality who prevented quieter members from speaking their views?)

Analysing

Analysis takes the form of thematic analysis. Basically, the data is analyzed by organizing it into categories on the basis of themes (or concepts). From this new concepts are developed, definitions formulated and relationships between concepts can be examined. Concepts are then linked together as opposites or as sets of similar categories which are then made into theoretical statements. Coding is an integral part of qualitative data analysis. It is guided by the research question and has two functions - (1) mechanical data reduction and (2) analytic categorisation of data. There are three types of coding and a researcher would use a different type on three occasions and so the raw data would be passed through three phases. Open coding is performed at the first pass. Its purpose is to locate themes and assign initial codes or labels. It is an attempt to condense the mass of data into categories. This is the time when you would note critical terms, key events or themes by writing a preliminary concept or label at the edge of the record. You can do this using brightly coloured ink (and a different colour for each theme if you like). Make sure you are open to finding different themes - do not from the onset decide to only look for an predetermined number! From this coding phase, make a list of all themes in the study. This can then be reorganised, sorted, combined, discarded or extended in further analyses. Axial Coding is used in the second pass through the data. You begin with the organised set of initial codes and focus on these codes/themes more than on the data (unlike open coding). Note any additional codes or new ideas that may emerge during this pass. Review and examine initial codes and start organising ideas or themes and identify the key axis of key concepts in analysis. You should also look for any linkages between categories. You can also drop some themes or examine others in depth. Selective Coding is the last pass through the data. Major themes of data have been identified so this phase will involve scanning the data and previous codes. You should selectively look at cases that illustrate themes and make comparsions and contrasts. Concepts are now well developed so you should seek to organise your analysis around several core generalisations or ideas. The major themes guide your search. Don't forget to look at what is missing? This can reveal a great deal! (Are people hiding information? And why?)

There are a number of computer packages that can help you with your coding such as ATLAS and NUD*IST. If you have never used one of these programmes before and you have only a small study (for example two focus groups of eight members per group), you can use the cut and paste option in MS Word. Simply, once you have identified your themes on pen and paper, type in the titles of your themes into Word and cut and paste the relevant sections from the transcripts under each theme. Make sure you are able to identify which section came from which focus group - for example, focus group 1 could be reproduced in italics, while focus group 2 remains in ordinary print. Also paste in whole quotes and dialogues so you know who said what and in what order!

Grounded theory is a widely used approach in qualitative research and can be used for the analysis of focus group data. It uses a systematic set of procedures to develop an inductively derived theory about a phenomenon. It is a method for discovering a new theory. The researcher compares unlike phenomena with the view to learn similarities and s/he seeks a theory that is comparable with the evidence. It pursues generalisation by making comparisons across social situations.

Reporting

Since focus groups are qualitative, they are theory generating. Focus groups can be used for exploratory purposes or can be phenomenological. In health research, focus groups can also be used in audits, particularly for the immediate impact of a health promotion campaign. In this instance, the analysis of the focus groups will remain thematic but the instead of developing a theory, the results will be combined to make recommendations for changes to improve a service or a campaign.

Advantages of a Focus Group

Group dynamics can be used to stimulate the discussion and the interaction of participants produces deeper insight than interviews or questionnaires.
It is an appropriate methodology when an issue is not well studied, where there is limited knowledge about people's perspectives and when exploring 'context embedded' behaviour.
It is a quick and efficient form of data collection, allowing in-depth information to be gathered in a short space of time. It can also produce richer more complex data than interviews.
It enables the spontaneous exchange of ideas within the safety of a group.
As facilitators interact with the participants, responses may be clarified and supplementary questions asked.
Participants can build upon one another's responses, producing more data and ideas than if they were interviewed separately.
Focus groups have been shown to be less inhibiting than one-to-one interviews and are suitable for capturing the dominant discourses of a social group.

Disadvantages of a Focus Group

There is the inability to generalize from small groups of non-randomly selected participants. (Note that this is inappropriate use of qualitative methodology anyway but many people have misguidedly tried to use focus groups in a quantitative manner).
Facilitators can bias results by cuing participants on the responses sought.
There can be difficulties transcribing and analysing the data.
There is the potential of non-attendance.
The facilitator may be unable to create a dynamic interaction.
There may be one or two participants dominating the discussion so shy people may be shying away from talking. People may also fear expressing an opinion if there are aggressive people in the group.
Shared experiences may deflect participants away from the topic and some topics may be too sensitive for focus group discussions.