Editor's note: Judith L. Robinson, Ph.D., is a Cincinnati-based qualitative research consultant. Funding for the research discussed in this article was provided by the Qualitative Research Consultants Association (QRCA).

The focus group interview is a widely used technique of data collection in marketing research. Its purpose is to gain insight into consumers' thoughts, feelings and behaviors, through an unstructured interview in a group setting. While other methodologies like individual interviews, dyadic interviews and naturalistic observations can also obtain qualitative data, the popularity of the focus group has made it synonymous with the term "qualitative research" in marketing research.

There are several assumptions about the advantages of group interviewing that contribute to its wide usage. Generally, it is presumed that being part of a group lowers respondents' anxiety and provides a more comfortable setting for discussion (Greenbaum, 1988).

This assumption rests on the notion that a one-on-one situation can make respondents anxious and inhibit candid discussion, while "safety in numbers" diffuses this anxiety. However, the risk is that the overall sense of responsibility and involvement may also be diffused among group members, resulting in less information and superficial responses.

Another popular assumption is that interaction of group participants stimulates talking and makes it easier for people to open up (Goldman and McDonald, 1987). In the same sense as a therapeutic session, it would seem that the supportive presence of other participants with similar concerns makes for a more comfortable environment where members are more willing to share information they would not divulge in an individual interview.

McQuarrie and McIntyre have offered a third hypothesis on the theoretical superiority of group interviews (1988, 1990). This hypothesis focuses on differences in the type of information that group and individual interviews provide. It postulates that groups elicit and give more discussion time to group information (normative information), while individual interviews evoke and give more time to information that is
unique to the individual (idiosyncratic information). McQuarrie and McIntyre suggest that this "group difference" makes group interviews the superior setting when the objective is examination of commonly held attitudes, beliefs and behaviors.

To understand why focus groups are popular, and where they are appropriately used, we explored some of the assumptions that underlie the use of focus groups as a means of data collection. Specifically, we wanted to find out if there is a "group difference" in the type of responses a group interview generates compared with an individual interview. We also investigated some differences between group and individual interview participants, in terms of anxiety, involvement and concerns about performance evaluation.

Methodology

To explore the effects of interview setting on information obtained, we conducted focus groups and individual interviews, with the primary purpose of comparing the information discussed in the two settings.
Both focus groups (eight participants) and individual interviews collected data on the same topics. The focus groups took approximately one hour; the individual interviews were 20 to 30 minutes. Both were conducted by professional moderators according to the same discussion guide. There were 10 focus group interviews and 40 individual interviews conducted for this research.

The interviews focused on two topics: considerations in the decision to visit a fast food restaurant; and environmental concerns.

The responses generated from these two topics were listed by the moderator during the interview and they were coded from the audio tape by an independent researcher to assure agreement. The output of the individual interviews was used as the basis for identifying normative versus idiosyncratic responses: Responses that were mentioned in 15% or more of the individual sessions were classified as "normative" while those mentioned in less than 15% were considered "idiosyncratic." Additionally, order of the responses and amount of time discussed were recorded.

At the close of the interview, respondents filled out two questionnaires. The first questionnaire presented four concepts for "environmentally considerate" product/service offerings from fast food restaurants and asked for evaluations of: degree of liking, purchase intent, and amount of acceptable price increment. The evaluations of these concepts were used as an unobtrusive measure of effort and involvement in the interview process.

The second questionnaire measured self-reports of interview anxiety, effort, and concern with evaluation from the moderator and from other group members.

Results

When comparing absolute number of normative responses to idiosyncratic responses, investigation results did not support the hypothesis that group interviews produce more normative or commonly held responses. Respondents were just as likely to discuss widely shared information in a one-on-one setting as in a group setting. In absolute terms, the individual and the group interviews produced the same number of normative responses, regardless of the topic of discussion.

However, the pattern of response production did differ between the two interview settings. In group interviews, normative responses occurred earlier and were discussed longer than in individual interviews, while idiosyncratic responses occurred later and were not discussed at length. There was no such pattern in the individual interviews: Normative and idiosyncratic responses occurred throughout the discussion and were discussed for approximately the same amount of time. (See Figs. 1, 2 and 3.)

The self-reports and the unobtrusive measure also did not support the hypothesis of varying levels of anxiety in the two interview settings. Group participants were no less anxious or concerned with evaluation than were individual interviewees. Furthermore, group participants put just as much effort into evaluating concepts as did the individual interview participants.

Implications

It has long been assumed that the reasons to choose a group interview setting over other qualitative methods are: reduction of anxiety, stimulating talk and self-disclosure, and direct observation of peer influences. The results presented here confirm the third assumption: that focus group discussions are a way to observe how social forces affect a given topic.

Although the group setting does not elicit a greater number of commonly held responses, it does lead to early discussion of shared thoughts and to lengthier discussion of this information. By observing the pattern of responses in the group - which information comes first and what is discussed at length - we can separate the commonly shared thoughts from the highly personal. The same responses are present in an individual interview, but they are mixed in with highly individualistic information and are thus harder to identify. The group interview can provide an immediate sense of which opinions and behaviors are pervasive in the population. The same information can only be discerned from individual interviews after a comparative analysis of responses.

This research also indicates that there is no difference in participants' level of anxiety or involvement in the two interview settings. At least according to self-reports, group interview members were neither more nor less uncomfortable with the interview than individual interviewees; nor did they feel more withdrawn from the interview.

This finding suggests that either focus groups do not diffuse "interview anxiety" or that anxiety about individual interviews is not as high as expected. Because this research was not designed to measure absolute levels of anxiety, it is difficult to say which of these hypotheses is true. Observation suggests that the one-on-one setting was not as anxiety provoking as anticipated, at least for the topics under discussion. In this case, it appears that concerns over respondent discomfort in individual interviews are unfounded.

This research does not directly address the issue of greater talkativeness. The opportunity to talk varied with number of participants: Individual interviewees had at least 20 minutes to talk, while group participants averaged 7.5 minutes (60 minutes divided by 8 respondents). However, while each group participant may have had less time to talk, the amount of information overall was similar between groups and individual interviews. This similarity was largely due to a reduction in redundancy in the group discussion. While the redundancy in the individual interviews can be reduced during the analysis, selective discussion of repetitive information during the group session is a more immediate and less time-consuming process.

However, the advantage of greater talkativity in groups may also relate to specific types of topics. The topics under discussion weren't particularly sensitive; it may be that the impact of the group would have been more evident discussing a more personal issue, like an illness, a personal problem or a financial concern.

Overall, we set out to understand the validity of the assumptions that guide the decision to use focus groups instead of individual interviews. We found that both interview settings provide a comfortable environment in which to discuss attitudes and behaviors. We could not reach conclusions about enhanced talkativity. But we did find evidence that focus groups are superior to individual interviews in providing immediate observation of attitudes and behaviors that are common within a given population.
Ultimately, the decision to use group or individual interviews rests on a combination of practical and theoretical considerations. Issues like the sensitivity of the topic or proprietary information, type of respondents, geographic dispersion of respondents, sample size, timing and cost all have an impact. This data suggests that when the research objectives are to examine what "most consumers" think, and when peer influence is an issue, focus groups are the best means to observe and understand responses.

References

Goldman, A.E., and McDonald, S.S. (1987). The Group Depth Interview. Englewood Cliffs, N.J.: Prentice Hall.

Greenbaum, T. (1988). The Practical Handbook and Guide to Focus Group Research. Boston: DC Heath.
McQuarrie, E.F., and McIntyre, S.H.

(1988). ``Conceptual underpinnings for the use of group interviews in consumer research." Advances in Consumer Research, 1 5, 580-586.

McQuarrie, E.F., and McIntyre, S.H. (1990). "What the group interview can contribute to research on consumer phenomenology." In E.C. Hirshman (Ed.), Research in consumer behavior. Vol. 4, 165194. Greenwich, Conn.: JAI Press.