Focus Groups in Social Research

Michael Bloor, Jane Frankland, Michelle Thomas, Kate Robson

• Publisher: SAGE Publications Ltd
• Year: 2000
• Print ISBN: 9780761957430, 9780761957423
• eBook ISBN: 9781446223666
• Edition: 1
• Topic: Social Sciences
• Subtopic: Social Science Research & Methodology
• Index: Social Sciences

There is an increasing divergence of focus group practice between social researchers and commercial market researchers. This book addresses the key issues and practical requirements of the social researcher, namely: the kinds of social research issues for which focus groups are most and least suitable; optimum group size and composition; and the designing of focusing exercises, facilitation and appropriate analysis.

The authors use examples, drawn from their own focus groups research experience, and provide exercises for further study. They address the three main components of composition, conduct and analysis in focus group research and also acknowledge the increasing impact the Internet has had on social research by covering the role and conduct of `virtual focus groups?.

---

1

Trends and uses of focus groups

---

CONTENTS

Beginnings

Access to group meanings, processes and norms

Focus groups as an adjunct to other methods

Pre-pilot focus groups

Focus groups within the main study or as aid to interpretation

Focus groups and public participation

Conclusion

Exercises

---

Beginnings

A cherished didactic method in the academy is that of beginning the study of every topic with a brisk canter through the founding fathers and occasional founding mothers. It is a method that has its apogee in the notorious Oxford University English curriculum that bewildered and bored generations of students with studies of Anglo-Saxon (taught by the inventor of Bilbo Baggins and Hobbitry, Prof. J.R.R. Tolkien). Sociology departments have tried to follow the same path, and even courses that end (appropriately) with postmodernism may start by attempting an exegesis of wordy antediluvians like Herbert Spencer. The Founding Father Method has several advantages for the academic: it demonstrates an impressive breadth (if not depth) of learning; it economizes on the need for originality in one’s own thinking by lengthy recapitulation of the thoughts of others; and it elevates one’s own puny thoughts by emphasizing their continuity with the hallowed precepts of The Past.

So, to begin at The Beginning, focus groups as a research method originated in the work of the Bureau of Applied Social Research at Columbia University in the 1940s. Under the leadership of Paul Lazarsfeld, the Columbia bureau was conducting commercial market research on audience responses to radio soap operas and the like. The arrival of Robert Merton at Columbia coincided with Lazarsfeld receiving a government contract (from the delightfully named Office of Facts and Figures) to assess audience responses to the government’s own wartime radio propaganda programmes and Lazarsfeld invited Merton to work with him on the project. By Merton’s own account (Merton, 1987), he found an established experimental procedure in operation: groups of approximately 12 people at a time would be seated in the radio studio and each chair would have a red and a green button at the side; members of the group were asked to press the red button each time they responded negatively to what they heard and to press the green button when they felt positive about something. Dissatisfied with an approach which simply quantified positive and negative responses, Merton set about developing an interviewing procedure for the groups, which would help researchers to describe the subjective reactions of the group members to the programmes they heard. Over a series of audience studies (involving print and film audiences as well as radio), researchers at the bureau (not just Merton, but also Alberta Curtis, Marjorie Fiske, Patricia Kendall and others) evolved a fairly standardized set of procedures for these interviews. These procedures were summarized in Merton and Kendall’s (1946) article for the American Journal of Sociology, ‘The focused interview’. In 1956, Merton, Fiske and Kendall collaborated on a book with the same title.

The works of founding fathers in sociology are rarely ‘A good read’, witness the collected works of Vilfredo Pareto. But the Merton and Kendall paper still repays study. For example, their apt description of the mental state they sought to inculcate in their research subjects during the interview: they termed this state ‘retrospective introspection’ (Merton and Kendall, 1946: 550). Or again, their emphasis on the importance in analysing their group data of concentrating on discrepancies (`deviant cases’) in reports between different groups (Merton and Kendall, 1946: 542–544): an analytic procedure more commonly referred to as ‘analytic induction’, following Znaniecki (1968).

Good read or not, the Merton and Kendall paper passed into a degree of obscurity. The interviewing procedures developed at Columbia became part and parcel of the methodology of the individual depth or qualitative interview and academic sociologists rarely conducted group interviews. However, the Columbia bureau had started out doing audience research for commercial radio and the commercial potential of focus groups to marketing organizations and advertisers remained. Small wonder that focus groups should resurface as a commercial market research technique in the 1960s. Greenbaum of Groups Plus, who was conducting focus groups as a market researcher in American living rooms in the early 1970s, finds no continuity with the earlier Columbia studies:

Focus groups have been commonly used in market research since the late 1960s, although some packaged food marketing organizations used the technique as early as the late 1950s, and some people even trace the beginning of the focus group technique back to publication in 1941 [sic] of The Focused Interview by Robert K. Merton, Marjorie Fiske and Patricia Kendall. Most research practitioners agree, however, that the technique began to be used regularly only in the late 1960s and early 1970s and that it has grown in popularity every year since. (Greenbaum, 1998: 167)

Tom Greenbaum and market research focus groups have both come a long way since those early days of the Proctor & Gamble Charmin Toilet Tissue campaign (Greenbaum, 1998: xv). Greenbaum quotes estimates that more than a 1,000 Americans earn the bulk of their livelihoods conducting focus groups and the average full-time focus group moderator conducts over 100 groups per year. This considerable industry has its own literature – trade magazines, websites and books (Greenbaum’s among them) – and it is not our intention to contribute to that large literature with this small book. Instead, we wish to address those academic researchers who are seeking to adapt commercial focus group practice to academic research needs.

The evident success of focus groups as a marketing tool in the private sector eventually led public sector organizations to use focus groups for their own marketing purposes – to assess the impact of health education campaigns, for example. Often this public sector research was contracted to private sector marketing organizations with previous experience of focus group work for the private sector and, of course, the techniques used in these early public sector studies were the same as those in the private sector. Even where the public sector organizations chose to do the focus group work ‘in house’ in their own research departments, the techniques used were initially those of the private sector because no other models of practice were available. However, this parallelism of public sector and private sector practice has not been maintained.

The divergence of public sector and private sector focus group work has two sources. First, the requirement to keep costs competitive has led private sector researchers to adopt, as standard, practices which compromised between economy and quality. This is not to say, of course, that the private sector has failed on quality: self-evidently, this booming industry has delivered value for money – it could hardly be so successful otherwise. Additionally, where quality of data, depth of analysis and rigour of comparison were considerations for the client that outranked economy, then of course it was always an option for the client to make this clear in the contract and for the contractor to modify practice and price accordingly. Although not all clients would perhaps have appreciated that they had this option. The compromise between economy and quality can be seen most clearly in respect of the analysis of focus group data. Krueger (1994: 143–144) identifies four possible analysis strategies: transcript-based analysis, the most rigorous; tape-based analysis, based on careful listening to the tapes but not on the study of transcripts; note-based analysis; and memory-based analysis. However, standard practice for private contractors to the UK public sector has involved a mix of the last two strategies, with group facilitators/moderators using their notes as background for an oral debriefing to a report writer, who in turn will collate the efforts of several facilitators (perhaps working in different parts of the country) in his/her report to the client; audio recordings of the groups may be made, but used sparingly, and mainly as a check that the contracted-for groups have indeed been run as contracted. There is no need to dwell on the loss of understanding of group interaction involved in the double selectivity of recall of first the facilitator and then the report writer.

A second reason for the divergence of public sector focus group practice from that of the private sector has lain in the realization that focus groups can be used for more than the mere assessment of group reactions to stimuli. It is this issue of the extended uses of focus groups to which we turn now.

Access to group meanings, processes and norms

Focus groups can be used for more than, say, the generation of information on collective views on what is the optimum sized gap between the top of the soap powder packet and the level of soap powder inside, or on whether or not groups react positively to pictures of a prematurely bald UK politician wearing a baseball cap. Focus groups can yield data on the meanings that lie behind those group assessments – do groups perhaps believe that manufacturers are trying to misrepresent the amount of soap powder they are offering the consumer when a large empty space is found in the top of a newly opened packet? Similarly, focus groups can yield data on the uncertainties, ambiguities, and group processes that lead to and underlie group assessments – is the politician’s baseball cap indicative of an unseemly sensitivity towards one’s baldness and unbecoming vanity about one’s appearance? or is it a praiseworthy attempt to signal one’s solidarity with the concerns of younger people? but if the latter, isn’t it the case that any UK citizen over 25 wearing a baseball cap looks like a complete prat? and doesn’t the politician’s failure to realize this demonstrate that he is therefore completely out of touch with young people? Relatedly, focus groups can also throw light on the normative understandings that groups draw upon to reach their collective judgements – manufacturers are expected to attempt to mislead consumers and no older person can successfully pass themselves off for any length of time as ‘A young person’. It is the access that focus groups are said to provide to these group meanings, processes and norms that accounts for much of the interest currently being shown by academic researchers.

Even late-modern societies remain normatively ordered; norms of conduct remain the mainsprings of human action. Of course, late modern societies are characterized by agency and choice, where people (all but the considerable minority of the most disadvantaged) can reflexively construct their identities and motives, mixing and matching from a diverse range of materials and be influenced by diverse groups and sources (Giddens, 1991). Relatedly, many of the principal sources of normative influence of 50 years ago (such as the workgroup and the neighbourhood) have lost much of their determining force, undermined by technical and economic changes and cross-cut by new sources of influence from soccer to soap operas. Moreover, individuals choose their own affiliations, constructing their own selves in the process, perhaps literally like groups of bodybuilders, or playfully like Klingon-speaking Trekkies. Nevertheless, human behaviour is still normative, all that has changed is that the sources of normative influence are more diverse, complex and interactive; our selves are reflexive constructs, but they are very much more likely to be collective than individual constructs; choice does not equate with freedom. It is precisely because behaviour remains normative, but is more subtly and variously influenced than the past, that interest has grown in research methods like focus groups which can access the rich texture of these influences.

In everyday life, the normative order underlying behaviours and opinions is rarely articulated. It is part of our taken for granted stock of knowledge (Schutz, 1964a; Schutz and Luckmann, 1974); it only has that degree of clarity and determinacy required for the conduct of everyday activities; and it is assumed to be shared by our associates (family, friends, workmates) until experience proves otherwise. The circumstances under which we are led to examine, elaborate and assess our normative assumptions are unusual – Schutz instances the estrangement of the war veteran returning to his old job as a cigar clerk (Schutz, 1964b) – under normal circumstances and in normal interaction we only refer to those normative assumptions briefly, allusively and in passing. There is no need to spell them out in more detail: we assume their rectitude and assume others share our views. Indeed, the ability to recognize these allusions (the ‘indexical expressions’ of ethnomethodology (Garfinkel, 1967)) is the hallmark of the competent collectivity member. And the very force of these normative influences on the collectivity may lie partly in their unexamined character (Bourdieu, 1977). It follows therefore that these normative assumptions are only slowly and progressively revealed to the ethnographer immersing her/himself in a collectivity, and even then (being alluded to, instead of articulated) they are largely inductively elaborated rather than directly recorded.

The situation of the focus group, in principle and with a fair wind, can provide the occasion and the stimulus for collectivity members to articulate those normally unarticulated normative assumptions. The group is a socially legitimated occasion for participants to engage in ‘retrospective introspection’, to attempt collectively to tease out previously taken for granted assumptions. This teasing out may only be partial (with many areas of ambiguity or opacity remaining) and it may be disputatious (as limits are encountered to shared meanings), but it may yield up as much rich data on group norms as long periods of ethnographic fieldwork. There are a number of problems with focus group data (problems we will examine in detail in subsequent chapters) and there are many possible objectives of sociological data collection in addition to that of gathering data on group norms. But in respect of that one limited objective – the study of group norms – focus groups should be the sociological method of choice, providing concentrated and detailed information on an area of group life which is only occasionally, briefly and allusively available to the ethnographer over months and years of fieldwork.

The ambiguity of group norms revealed by focus group analysis is a characteristic structural feature of any normative order. As early ethno-methodological studies by Garfinkel and his followers convincingly demonstrated (Coulter, 1974; Zimmerman and Wieder, 1971), all rules are essentially contingent and defeasible in the sense that any competent collectivity member is able in principle to elaborate on the sense of the rule to justify their current behaviour, or to specify the limits of the rule’s application in order to excuse current behaviour as not breaching the rule. The same ethnomethodological studies did not, of course, destroy the warrant for seeing social behaviour as rule-governed, because norms may be essentially contingent and defeasible while remaining practically fixed and constraining under normal circumstances. This is because most collectivity members under most circumstances are not motivated to elaborate on stable formulations of rules: these stable formulations of rules become familiar, routinized and invested with reverence; and they may only be renegotiated or elaborated with superordinates or intimates at some cost to the individual (Bloor, 1980). However, the focus group process not only lays bare the rule, it also makes apparent (to facilitator and group members alike) that its stability is illusory, that the rule in question is essentially contingent and defeasible: the very process of the focus group itself (inviting group members to inspect, elaborate upon, and question rules that we normally take for granted) calls forth and demonstrates the intrinsically ambiguous character of group norms. So focus groups are simultaneously the best method for accessing group norms and also the best method for demonstrating that the group norms thus elicited cannot be unproblematically applied in organizational decision-making or public policy. One of the challenges for focus group researchers (which we revisit later in this book) is to find ways of incorporating focus group methods into participative public decision-making, rather than having the focus group findings treated simply as a resource in expert deliberations (Cunningham-Burley et al., 1999; Johnson, 1996). But even as a resource for expert deliberations, the ambiguous character of group norms derived from focus groups does not allow their unproblematic utilization by experts.

The application of any group norm in any given setting requires a prior act of interpretation, requires the attribution of meaning. Such interpretations are only rarely unique to the individual, more commonly they are shared with others in the individual’s social groups, part of a common stock of knowledge. This common stock of knowledge is the basis of social action, since recipes for action are tied to given interpretations of the situation, but the various elements of this stock of knowledge are not always clear and distinct. Interpretations have only that degree of clarity required for the person’s purpose at hand: if that purpose changes (if, say, an initial interpretation is questioned), then the interpretation may be further elaborated, different and competing interpretations may be considered, and the appropriateness of the initial recipe for action may be now thought problematic (Schutz, 1962; 1970). It therefore follows that the discussions occurring within focus groups will provide rich data on the group meanings associated with a given issue. But the very act of making such group meanings a topic of group discussion will lay bare the provisional character of such interpretations. Just as focus group data on norms may demonstrate the essential ambiguity of norms, so focus group data on meanings may demonstrate the essential ambivalence of interpretations: the rather chaotic character of the findings is not a defect of the method, it is a faithful reflection of the subject matter.

The group meanings accessed in focus group discussions are, of course, expressed in the argot and everyday language of the group, not translated into the terminology of the researcher. Since the researcher is present simply to facilitate the discussion, then ideally the group participants should be addressing each other and therefore using group terms and group categories, so-called ‘indigenous coding systems’ (Holstein and Gubrium, 1995). The focus group may therefore give the researcher privileged access to in-group conversations which contain ‘indigenous’ terms and categories in the situations of their use.

The same focus group discussions also reflect internal group processes and formal and informal group structures. Meetings of the Society of Friends (the Quakers) are formally democratic and unstructured, with no minister to lead the meeting and with all members waiting quietly for one of their number to be inspired to say something. But it is said that 100 years ago meetings in the York Meeting House found York Quakers routinely deferring to members of the Rowntree (chocolate) family, whose social and economic prominence in the town meant that they always spoke first in the meetings. Similarly, in focus groups formed from a pre-existing social group, some of the processes of the preexisting group may be captured within the focus group. So, for example, a focus group being conducted with a pre-existing workgroup may reflect the hierarchical relationships within that workgroup: a focus group conducted in a health centre team meeting may find, for example, the nurses in the team deferring to the general practitioners. Likewise, mixed sex groups of adolescents may capture within the group process broader patterns of boy–girl interaction. It may be possible to address such differences in the planning and composition of focus groups: for example, by holding separate girl and boy groups as well as mixed sex groups, or by holding a group composed of nurses from several different health centres, so that differences in group processes between differently composed groups may be highlighted.

Of course, there are difficulties of execution and (...

Table of contents

• Cover Page
• Title
• Copyright
• Contents
• Preface
• 1 Trends and uses of focus groups
• 2 Composition of groups
• 3 Preparation and conduct
• 4 Analysis
• 5 Virtual focus groups
• 6 Conclusions
• References
• Index