A focus group discussion is a guided conversation among roughly six to ten participants, assembled by the researcher because they share some relevant characteristic or experience, and led by a moderator working from a topic guide. Its distinguishing feature is that the unit of data is the group interaction itself. Participants react to one another, qualify, contest, joke, concur and correct; the researcher is interested in this collective process, not merely in the sum of ten opinions. Its ancestor is the focused interview developed by Robert Merton and Patricia Kendall in wartime studies of propaganda and morale.

The method rests on a broadly interactionist premise: attitudes are not fixed possessions carried inside individuals but are formulated, justified and revised in talk with others. A focus group therefore approximates the natural setting in which opinion is actually produced — the tea shop, the self-help group meeting, the courtyard conversation — and is valued for what it reveals about shared norms and the boundaries of the sayable.

Composition, size and number of groups

Selection is purposive, never random. Groups are usually made deliberately homogeneous on the dimensions that would otherwise inhibit speech — women separate from men, young from old, agricultural labourers separate from landholders — while being kept heterogeneous enough on opinion to sustain discussion. Size is a compromise: too few and the conversation thins into serial interviewing, too many and quieter members drop out and turn-taking collapses. Participants who know each other well bring existing familiarity but also existing hierarchies and things they have agreed never to mention aloud.

A single group establishes nothing, because any one discussion may be dominated by one personality or one incident. Studies therefore run multiple groups, usually segmented by the categories of interest, and continue until saturation — the point at which additional groups produce no new themes. Sessions typically run sixty to ninety minutes, are audio-recorded where consent permits, and are supported by an assistant who records the seating plan, the sequence of speakers and the non-verbal reactions, since a transcript alone cannot show who fell silent.

Moderation

The moderator's task is to be present without being authoritative. The guide moves from a broad opening question through a small number of key questions to a closing summary that participants are invited to correct. Skill lies in the management of the group's own dynamics: dominant voices must be redirected without being publicly rebuked, silent members drawn in by name or by written exercises such as ranking cards, and group conformity actively counteracted. The pressure towards a single respectable answer is the method's chief internal threat, and experienced moderators deliberately invite dissent — asking whether anyone sees it differently, or presenting a contrary view as something heard elsewhere.

What it reveals and what it cannot

Focus groups are strong on collective phenomena. They expose shared norms and the taken-for-granted, because participants police one another and, in doing so, state rules that no individual would have thought worth mentioning. They surface contested meanings, since disagreement makes visible the fault lines within a supposedly single community. They yield the group's own vocabulary — the categories, euphemisms and jokes in which a subject is actually discussed — which is why the technique is often used to design survey instruments and information campaigns. They are also quick, relatively cheap, and can be experienced by participants as engaging rather than extractive.

The limits are equally definite. Data cannot be attributed to individuals with confidence, because a person's public statement in a group is not their private view; nor can the group be used to estimate prevalence, since it is neither random nor of a size that permits inference — counting how many participants agreed is a misuse of the technique. Sensitive disclosure about violence, illness, sexuality or debt is unlikely and may be actively unsafe, since confidentiality cannot be guaranteed by the researcher once several participants know what was said. Transcripts of overlapping speech are hard to analyse, results are sensitive to moderator skill, and the artificiality of the assembled group remains.

Indian development research

The focus group became a standard instrument of Indian development and health research through the participatory tradition associated with Robert Chambers and Participatory Rural Appraisal, where group discussion combined with mapping, ranking and seasonal calendars to elicit local knowledge, and through large programme evaluations in reproductive health, nutrition, sanitation and microfinance where self-help groups offered ready-made assemblies.

The pitfalls are structural. Where a group mixes castes, the presence of dominant-caste participants can render Dalit members silent or assenting, and the resulting transcript will read as consensus. Where women speak in the presence of men, or younger women before older ones, deference produces the same false unanimity. Where an NGO convenes the group, participants may report programme success to protect an existing benefit. The practical answers are segmentation of groups by caste, gender and age, moderators matched to the participants and fluent in the local language, convening in a neutral location, and treating focus group findings as one strand within triangulation rather than as standalone evidence.

For the UPSC answer

Define the focus group by its unit of data — interaction, not aggregated individual opinion — and credit Merton's focused interview as the origin. Give the design essentials briefly: purposive and homogeneous composition, small size, several groups run to saturation, a moderator with a topic guide. Then set the distinctive gains, shared norms and contested collective meanings, against the three things it cannot do: individual attribution, prevalence estimation and sensitive disclosure. Finish with Indian development research, where caste, gender and age hierarchies manufacture a false consensus unless groups are segmented — a point that shows the method's validity depends on social structure, not merely on technique.

References & further reading

  1. Merton, R. K. and Kendall, P. L. (1946). The Focused Interview. American Journal of Sociology, 51(6), 541–557.
  2. Krueger, R. A. and Casey, M. A. (2009). Focus Groups: A Practical Guide for Applied Research. Sage.
  3. Morgan, D. L. (1997). Focus Groups as Qualitative Research. Sage.
  4. Kitzinger, J. (1994). The Methodology of Focus Groups: The Importance of Interaction between Research Participants. Sociology of Health and Illness, 16(1), 103–121.
  5. Chambers, R. (1994). The Origins and Practice of Participatory Rural Appraisal. World Development, 22(7), 953–969.
  6. Bryman, A. (2012). Social Research Methods. Oxford University Press.