The Importance of Member Homogeneity to Focus Group Quality

Citation:

Kim P. Corfman (1995) ,"The Importance of Member Homogeneity to Focus Group Quality", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 354-359.

Advances in Consumer Research Volume 22, 1995      Pages 354-359

THE IMPORTANCE OF MEMBER HOMOGENEITY TO FOCUS GROUP QUALITY

Kim P. Corfman, New York University

An issue that is widely debated among those who provide and use focus group services is the importance of group member homogeneity to the quality of a group's output. The underlying thesis of this research is that the relative importance of homogeneity on each of two dimensionsCexogenous homogeneity and issue homogeneityCdepends on how the dimension relates to the focus group topic. Theory from social psychology on small group processes is used to support hypotheses concerning the impact of homogeneity on self-disclosure in focus groups. They are tested using data from 99 participants in 11 focus groups on a sensitive topic. Results suggest that homogeneity may not be as important as is commonly believed when the topic is of sufficient interest to the participantsCeven when the topic is potentially sensitive.

The last decade has produced a resurgence in the popularity of focus groups as a marketing tool. Despite their widespread use and an abundance of "how to" books and articles on the subject (e.g., Krueger 1988; Morgan 1988; Templeton 1987), little systematic research has been conducted on the theoretical foundations of the approach and little formal testing has been performed of the quality of output produced by various methods (Stewart and Shamdasani 1990).

An issue that is widely debated among those who provide and use focus group services is the importance of group member homogeneity to the quality of a group's output. As the following quotations illustrate, recommendations range from complete homogeneity to homogeneity on a subset (sometimes specified) of characteristics to an analysis of the trade-offs involved to a careful balancing of similarities and differences:

"Homogeneous groups....are generally more comfortable and open with each other, whereas mixed sex, ethnic, or socioeconomic groups make it more difficult to achieve a high degree of group interaction" (Keown 1983, p. 66)

"Mixing participants from distinct market segments into a single group is not recommended because each person's segment has different requirements" (Welch 1985, p. 247).

"The goal is homogeneity in background, not homogeneity in attitudes..." (Morgan 1988, p. 46).

"A number of individuals may be very different in national origin, religious beliefs, political persuasion, and the like; but if they share a common identity relevant to the discussion...., a group can form." (Goldman 1962, p. 62)

"Sometimes a varied group is wanted, for the interplay of diverse views on a topic that all can discuss.... However, sharp diversity or division in the group is hazardous." (Levy 1979, p. 30)

"....it is usually helpful to provide for both homogeneity and contrast within specific groups." (Wells 1974, p. 4)

Group member homogeneity can be defined on a number of dimensions. The underlying thesis of this research is that the importance of homogeneity on these dimensions depends on how the dimension relates to the focus group topic. Two forms of homogeneity are defined, exogenous homogeneity and issue homogeneity, and their influence on individual self-disclosure is examined in an exploratory study of focus groups on a sensitive topic.

BACKGROUND

Although experience with focus groups has led to a large body of popular wisdom on the subject, very little formal research has been conducted on factors affecting the productivity of focus groups, even less on the influence of group composition, and none on the effects of member homogeneity. Factors that have been investigated include acquaintanceship (Fern 1982; Nelson and Frontczak 1988), group size (Fern 1982), the importance of the moderator (Fern 1982), and moderator philosophy (McDonald 1992, 1993).

There is, however, a large literature in social psychology on small group processes and the influence of a variety of group member characteristics on interaction and group performance. The research that relates to our interests considers the effects of homogeneity in ability (Cartwright 1968; Goldman 1965; Laughlin, Branch and Johnson 1969), gender (Aries 1976; Reitan and Shaw 1964; Wyer and Malinowski 1972), age and education level (Cartwright 1968), race (Fenelon and Megargee 1971; Ruhe and Allen 1977), religion (Fiedler and Meuwese 1963), culture (Fiedler 1966), attitudes, opinions, and values (Cartwright 1968; Fisher 1980; Terborg, Castore and DeNinno 1976), and personality (Hoffman 1959; Triandis, Hall and Ewen 1965). In these studies the kinds of tasks employed are predominately more structured and clearly defined than the typical focus group "task," objective performance quality is much more easily determined, and the groups studied are far smaller (usually two to three members) than those normally assembled for focus groups. However, some of this work provides a useful foundation for theory building in the focus group context.

THEORY AND HYPOTHESES

Focus groups are used for a wide variety of purposes including generating hypotheses about behavior, brainstorming or testing ideas for new products, packaging, advertising, etc., gaining insight into how consumers view a brand or category and its users, what they want from it, and what associations they make among product attributes, learning about purchase and usage behavior, testing questionnaire language, observing the process of opinion formation, and explaining puzzling survey results. Calder (1977) proposes a taxonomy of focus group approaches with which the many uses of focus groups may be classified into three distinct groups: exploratory, phenomenological, and clinical. Although the demands placed on participants with these uses and approaches differ, in order to be successful, all rely on the willingness of participants to be open with their feelings, beliefs, ideas, behavior, etc. and to discuss them candidly with other group members. The factors that may influence the quality of a participant's involvement and willingness to self-disclose in a focus group include characteristics of the individual, the nature and sensitivity of the discussion topic, the composition of the group, moderator traits and style, and the physical environment. The focus of this study is the composition of the group, specifically, the degree of homogeneity among its members and how homogeneity affects self-disclosure in focus groups. As noted by Shaw (1981), "groups are homogeneous or heterogeneous with respect to specific characteristics, not all of which are relevant to the group's activities" (p. 238). Two types of homogeneity are considered here: exogenous homogeneity and issue homogeneity.

Exogenous Homogeneity

Exogenous homogeneity implies similarity in such characteristics as gender, ethnicity, social class, religion, personality, attitudes, values, and age when the factor is not highly correlated with response to the issue under investigation. Many have observed that exogenous homogeneity is important because consumers who differ greatly in social class and stage in the family life cycle have such different resources, problems, experiences, and perceptions that they may have difficulty communicating with each other (Krueger 1988; Merton, Fiske and Kendall 1990; Wells 1974). Researchers have explored the effects of homogeneity on such traits as ability (Goldberg et al. 1966; Goldman 1965; Laughlin, Branch and Johnson 1969; Shaw 1960), personality (Triandis, Hall and Ewen 1965; Hoffman 1959) and race (Fenelon and Megargee 1971; Ruhe and Allen 1977; Ruhe and Eatman 1977). Most of these studies indicate that heterogeneous groups outperform homogeneous groups on clearly defined intellective problem-solving tasks that are better accomplished by groups whose members have differing and complementary knowledge bases and experiences upon which to draw. These results have limited relevance to focus groups because most focus groups are designed to acquire information possessed by the group members about themselves and not to determine how well they solve artificial problems created by the researcher. (Exceptions may be the use of focus groups for more creative tasks such as brainstorming new product ideas.)

Research that is more closely related to the concerns of focus groups examines the effects of homogeneity on cohesiveness, which has been shown to enhance several aspects of group interaction. The evidence indicates that group members will be more attracted to each other and, thus, become more cohesive when they agree in their attitudes, have similar values, and have similar abilities and opinions (Cartwright 1968; Shaw 1981; Terborg et al. 1976). (A small amount of contradictory evidence comes from the context of work groups whose goals are such that some complementary differences in abilities might be valued and some similarities be judged irrelevant to the task, Gross 1956; Seashore 1954.) Cartwright (1968) and Shaw (1981) examine the effects of cohesiveness on group interaction and conclude that individuals in cohesive groups are likely to participate and communicate more (Back 1951; Lott and Lott 1961), feel greater self-esteem and less anxiety (Julian, Bishop and Fiedler 1966; Myers 1962; Seashore 1954), be more cooperative, accepting, and trusting of other members (Back 1951; Shaw and Shaw 1962), feel more secure (Pepitone and Reichling 1955), and be more effective in achieving the group goals (Goodacre 1951; Schacter et al. 1951; Shaw and Shaw 1962; Van Zelst 1952). The observation is also made that members of cohesive groups are more likely to conform to the norms of the group (Berkowitz 1954; Lott and Lott 1961; Schacter et al. 1951; Wyer 1966). This effect is expected to be weaker than the positive effects of cohesiveness; however, it can have negative consequences for focus groups if it leads to the repression of differing opinions and perspectives. In general, it seems reasonable to conclude that greater exogenous homogeneity will generally result in improved quality of interaction and increased self-disclosure in focus groups.

Issue Homogeneity

Issue homogeneity implies similarity in response to the focal issue, e.g., product usage, preference, attitude, motivation, etc. Lower levels of issue homogeneity are often better for provoking introspection through exposure to contrasting perspectives, revealing reasons behind differences in behavior and attitudes, and exposing the researcher to a broader range of consumer response (Krueger 1988; Wells 1974). However, very low levels may make productive discussion difficult because participants may be unable to understand each other's needs and behavior (Krueger 1988; Levy 1979). Due to the benefits of greater cohesiveness that result from homogeneity in attitudes, opinions, and values it seems reasonable to conclude that moderate levels of issue homogeneity will improve the quality of member interaction and encourage self-disclosure while allowing sufficient variation among members to stimulate insightful discussion.

It is important to note that the above arguments hold only for benign issue-related factors. When the topic is sensitive, it is important to have high levels of issue homogeneity. Sensitive issue-related factors affect willingness to discuss a particular issue openly (e.g., alcoholism when the topic is alcohol use, gender when the topic is contraception). When the use of a product or the need for it is a sensitive issue, only those who have similar needs and usage patterns may feel comfortable discussing it together (Levy 1979; Morgan 1988). Heterogeneity in ability to purchase a product or the impact of the purchase on an individual's budget can also create discomfort. Thus, income may be a sensitive factor for high ticket items. The presence of respected authority or someone who is superior in rank may also inhibit the open discussion (Cunninghis 1992).

Even when the topic is sensitive or there are sensitive issue-related factors, low levels of exogenous homogeneity may be less damaging to group interaction (and self-disclosure) when issue homogeneity is high. Morgan (1988) reports on work with groups of widows that were very mixed with regard to social class (exogenous heterogeneity), in which this factor had little impact because their bereavement created an important and fundamental similarity.

Hypotheses

In this study the focus is on the individual and how group composition influences his or her contributions to the discussion. At the individual level, heterogeneity determines a participant's feelings of uniqueness. From the discussion above, the following basic hypothesis is derived:

H1: Feelings of uniqueness on the part of focus group participants result in decreased self-disclosure.

Both exogenous and issue homogeneity are expected to encourage self-disclosure. However, when the topic is sensitive, homogeneity in sensitive issue-related factors is expected to be more important:

H2: When the discussion topic is sensitive, feeling unique in issue-related factors has a larger impact on self-disclosure than feeling unique in exogenous factors.

Although the focus of this study is the effects of uniqueness, other individual, relationship, and environmental factors that have been proposed as determinants of focus group quality are also examined.

METHOD

Topic and Subjects

The topic and subjects were selected to ensure that discussion would be both relevant and sensitive. A corporate sponsor was located who was interested in the results of focus groups conducted with college students on the subject of diet and weight loss products. An undergraduate subject pool provided 99 participants. Pre-screening determined that none had participated in another focus group within the preceding six months or ever taken part in one on weight control, and that they were interested in the subject. Eleven groups were formed on the basis of the students' schedules, each containing from 7 to 12 members, with approximately equal representation of males and females. All sessions were conducted within a 22 day period. Subjects were asked not to discuss their experiences with anyone who might participate at a future date.

Data Collection

The sessions were held in a conference room equipped with a tape recorder, a video camera, and a table of refreshments. As they arrived, participants were given tents cards with their first names, offered refreshments, and invited to sit where they chose around an oblong conference table. Then they were asked to complete a three page prequestionnaire which requested information on the following exogenous factorsCincome, ethnicity, traditionalism, religiousness, social values, and their acquaintanceship with other membersCand a variety of issue-related factorsCheight, weight, bone structure, desire to lose weight, diet habits, and use and harmfulness of exercise, weight loss programs, and diet pills.

The moderator for all 11 groups was a clinical psychologist, experienced in group therapy and a trained focus group moderator. He was thoroughly briefed on the topic and had input into the design of the discussion guide, but was not aware of the nature of the study. At the beginning of the session, he gave a standard focus group introduction and explained that the purpose of the group was to talk about "your feelings about eating and dieting, whether you are concerned about weight loss and why, and what you do about it if you are." After a warm up task, the discussion covered eating and exercise habits and use of weight loss products and programs. Of the 1.5 hour discussion, approximately a half hour at the end was devoted to discussion of two diet aids manufactured by the sponsorCone currently on the market and the other in testing.

At the conclusion of the discussion, participants were asked to complete a postquestionnaire which asked for self-reports on items relating to the focus group session: self-disclosure, group process (Open Group Process, Work Group Functioning Scale, Seashore et al. 1982), cohesiveness (Gross 1957), learning about themselves (insight), homogeneity (Group Homogeneity, Work Group Functioning Scale), interest and involvement in the discussion, and attitude toward the moderator. When they had completed this form they were given a final questionnaire and told that it was for an independent study on personality. (Subjects in this pool are accustomed to completing multiple independent experiments and surveys in a single session.) The personality questionnaire asked for self reports on items relating to adolescent self-esteem (Rosenberg 1965), interpersonal orientation (Swap & Rubin 1983), traditionalism (Corfman & Lehmann 1987), religiousness (Corfman & Lehmann 1987), and social values (Kahle's 1983 List of Values). The entire procedure lasted less than two hours.

ANALYSIS AND RESULTS

Variables and Indices

Items from the first postquestionnaire which relate to the focus group session itself were factor analyzed with varimax rotation. Seven factors were retained (using the eigenvalues greater than one criterion). Items loading at greater than .5 on a factor are examined. The factors are easily interpreted and where loadings are not consistent with expectations, they make sense in the context of the index. Indices were created by summing the relevant items. The following are the resulting indices and their Cronbach alphas: Self-disclosure (a=.83), Interest/involvement (a=.83), Cohesiveness (a=.73), Insight (a=.77), Homogeneity (a=.64), and Comfort (a=.65). (These alpha levels are generally considered reasonable for exploratory research, Nunnally 1978.) The Comfort index is composed of a moderator item, a self-disclosure item, and a group process item, all of which relate to how comfortable the participant felt in the group. The eighth factor had two items relating to attitude toward the moderator, but their correlation was only .49, so the items were used separately.

Items from the second postquestionnaire relating to three personality traits were also factor analyzed. Four factors were retained, the first three of which are clearly the three traits: Self-esteem (a=.86), Traditionalism (a=.86), and Religiousness (r=.74). The fourth factor has only two items and is not easily interpreted.

The nine items in Kahle's List of Values (LOV) scale were factor analyzed and loaded on three factors: Self-Oriented (a=.74), Social (a=.75), and Stimulation (a=.45). Two dummy variables indicated to which of these categories the subject's most important value belonged.

Other variables were created as follows. The Interpersonal Orientation scale is the sum of 29 items and has a Cronbach alpha of .78. A Weight vs. Ideal variable was created by subtracting the subject's objective ideal weight (Bender 1973) from his or her actual weight. Negative scores were corrected to zero. (No one in the study was seriously underweight.) The uniqueness variables were created by calculating the proportion of the group that was the same as the subject on each factor. For example, for Diet Pill Harm, a subject who indicated in the prequestionnaire that he or she believed diet pills were harmful would have a uniqueness rating corresponding to the proportion of the group that believed they were harmful.

Regression

A regression was run using Self-disclosure as the dependent variable and four categories of independent variables: uniqueness dimensions, individual characteristics, relationship factors, and characteristics of the physical environment. The regression would have been run with a dummy variable for group membership as a control for the repeated measures design. However, group size served the same purpose as it was a linear combination of the group membership variables. Table 1 contains the variables and the results of the analysis. Comfort, Cohesiveness, and Insight were highly correlated. Thus, only one, Comfort, was included in the analysis. Gender uniqueness was omitted because there was virtually no variance; the numbers of men and women were close to equal in all groups. The R² for the regression is .68 (adjusted R²=.42) and only six variables are significant.

The only uniqueness variable that is significant is exercise harm, indicating that those who felt they were more unique in their beliefs about the dangers associated with exercise were more reluctant to talk about themselves. Because few subjects felt exercise was harmful, the unique participants were those who felt it could be harmful. As the significance level is only p<.10 and the number of independent variables is large, the significance of this estimate may have been due to chance. (The single significant individual characteristic, the dummy variable associated with being from the West Indies, may safely be ignored, as there was only one individual in that category.) To check whether subjects' perceptions of homogeneity were related to the uniqueness variables examined here, correlations were examined. While Homogeneity is significantly correlated with four of the uniqueness variables (Traditionalism, Ethnicity, Weight vs. Ideal, and Diet Pill Use), again Homogeneity and Self-Disclosure are not significantly correlated.

TABLE 1

REGRESSION: SELF-DISCLOSURE

Three relationship factors were significant. Those who felt more comfortable in the group and those who were interested and involved in the discussion were more like to self-disclose. The negative sign associated with Liking of Moderator is misleading and can be better understood through an examination of the relationship between the two moderator variables and Self-disclosure. While the moderator variables have a .49 correlation (p<.001), Effectiveness has a .29 (p<.05) correlation with Self-disclosure while Liking of Moderator is not significantly correlated with Self-disclosure. From this it appears that the regression signs are a result of estimate instability (due to collinearity) and perceptions of Moderator Effectiveness result in greater self-disclosure. One aspect of the physical environment has a significant effect. Participants who sat closer to the moderator disclosed more about themselves. Subjects chose where to sit and, it seems, those who were more comfortable sitting near the moderator were also those who were willing to talk about themselves.

CONCLUSIONS

In summary, the results of this exploratory study found that participants who were more unique in their beliefs that exercise was harmful may have disclosed less about themselves, although this finding was weak. Those who were comfortable, interested, felt the moderator was good at getting the group to talk, and sat closer to the moderator disclosed more. The weak findings for the importance of homogeneity suggest that homogeneity may not be as important as is commonly believed when the topic is of sufficient interest to the participantsCeven when the topic is potentially sensitive.

Some limitations of this study suggest stronger tests of the hypotheses. First, in this study the dependent variable, self-disclosure, is a self-rating. While one would hope that there is a relationship between perceived self-disclosure and objective self-disclosure, it is not clear how close that relationship is and the latter could be much more revealing. Transcripts of focus groups session could be content analyzed and varieties of self-disclosure identified and tallied. For example, it would be useful to distinguish among disclosure of past behavior, opinions, and reports of others' behavior and opinions, the latter being less personal and revealing. Another problem is the issue of accuracy of self-disclosure. Further, while subjects' disclosures were entirely plausible, this study could not discriminate true self-disclosure from inaccurate reports of behavior, feelings, etc. and it was not possible to tell who was withholding relevant information. It would be very interesting, although somewhat difficult, to design a study involving subjects about whom the researchers had confidential information. Second, there may have been insufficient variation in many of both the exogenous and issue-related factors to pick up the effects of heterogeneity on self-disclosure. Future studies could attempt to create heterogeneity on specific dimensions through selection of their members. Finally, it would also be useful to distinguish between kinds of uniqueness in a heterogeneous group. "Good" uniqueness, such as being the sole marathon runner in the group, may have a different effect on self-disclosure than "bad" uniqueness, such as being the sole diet pill taker.

Given the widespread use of the focus group technique in marketing practice, the almost complete absence of formal research on the subject is disturbing. A large number of generally accepted principles and practices have never been subjected to formal examination in an environment that controls for the many factors that interact to create a particular experience, nor have they been evaluated on the basis of more than intuitive assessments of session "quality." This study addresses commonly made untested recommendations which concern focus group composition. The results of this investigation provide some preliminary insight into the factors that influence self-disclosure in focus groups and should help users and providers of focus group services begin to design more effective studies.

REFERENCES

Aries, E. (1976), "Interaction Patterns and Themes of Male, Female, and Mixed Groups," Small Group Behavior, 7 (February), 7-18.

Back, K. W. (1951), "Influence through Social Communication," Journal of Abnormal and Social Psychology, 46, 9-23.

Bender, Arnold E. (1973), Nutrition and Dietetic Foods.

Berkowitz, Leonard (1954), "Group Standards, Cohesiveness, and Productivity," Human Relations, 4 (February), 509-519.

Calder, Bobby J. (1977), "Focus Groups and the Nature of Qualitative Marketing Research," Journal of Marketing Research, 14 (August), 353-364.

Cartwright, Dorwin (1968), "The Nature of Group Cohesiveness," in Dorwin Cartwright and Alvin Zander, eds., Group Dynamics, New York: Harper and Row, 91-109.

Corfman, Kim P. and Donald R. Lehmann (1987), "Models of Group Decision-Making and Relative Influence: An Experimental Investigation of Family Purchase Decisions," Journal of Consumer Research, 14 (June), 1-13.

Cunninghis, Burt (1992), lecture, Focus Group Moderators' Workshop, Consumer Sciences Inc., (November).

Fenelon, J. R. and E. I. Megargee (1971), "Influence of Race on the Manifestation of Leadership," Journal of Applied Psychology, 55, 353-358.

Fern, Edward F. (1982), "The Use of Focus Groups for Idea Generation: The Effects of Group Size, Acquaintanceship, and Moderator on Response Quantity and Quality," Journal of Marketing Research, 19 (February), 1-13.

Fiedler, F. E. (1966), "The Effect of Leadership and Cultural Heterogeneity on Group Performance: A Test of the Contingency Model," Journal of Experimental Social Psychology, 2, 237-264.

Fiedler, F. E. and W. A. T. Meuwese (1963), "Leader's Contribution to Task Performance in Cohesive and Uncohesive Groups," Journal of Abnormal and Social Psychology, 67, 83-87.

Fisher, B. Aubrey (1980), Small Group Decision Making, New York: McGraw Hill.

Goldberg, M. L., A. H. Passow and J. Justman (1966), The Effects of Ability Grouping, New York: Teachers College Press.

Goldman, Alfred E. (1962), "The Group Depth Interview," Journal of Marketing, July, 61-68.

Goldman, M. (1965), "A Comparison of Individual and Group Performance for Varying Combinations of Initial Ability," Journal of Personality and Social Psychology, 1, 210-216.

Goodacre, D. M. (1951), "The Use of a Sociometric Test as a Predictor of Combat Unit Effectiveness," Sociometry, 14, 148-152.

Gross, E. (1956) , "Symbiosis and Consensus as Integrative Factors in Small Groups," American Sociological Review, 21, 174-179.

Gross, E. F. (1957), "An Empirical Study of the Concepts of Cohesiveness and Compatibility," unpublished honors thesis, Harvard University, Department of Human Relations.

Hoffman, L. Richard (1959), "Homogeneity of Member Personality and its Effect on Group Problem-Solving," Journal of Abnormal and Social Psychology, 58, 27-32.

Julian, J. W., D. W. Bishop and F. E. Fiedler (1966), "Quasi-therapeutic Effects of Intergroup Competition," Journal of Personality and Social Psychology, 3, 321-327.

Kahle, Lynn R. (1983), Social Values and Social Change, New York: Praeger.

Keown, Charles (1983), "Focus Group Research: Tool for the Retailer," in Thomas J. Hayes and Carol B. Tathum, eds., Focus Group Interviews: A Reader, Chicago: American Marketing Association, 64-70.

Krueger, Richard A. (1988), Focus Groups: A Practical Guide for Applied Research, Beverly Hills, CA: Sage.

Laughlin, Patrick R., Laurence G. Branch and Homer H. Johnson (1969), "Individual Versus Triadic Performance on a Unidimensional Complementary Task as a Function of Initial Ability Level," Journal of Personality and Social Psychology, 12 (2), 144-150.

Levy, Sidney J. (1979), "Focus Group Interviewing," in James B. Higginbotham and Keith K. Cox, eds., Focus Group Interviews: A Reader, Chicago: American Marketing Association, 29-37.

Lott, A. J. and B. E. Lott (1961), "Group Cohesiveness, Communication Level, and Conformity," Journal of Abnormal and Social Psychology, 62, 408-412.

McDonald, William J. (1992), "The Influence of Moderator Philosophy on the Content of Focus Group Sessions: A Multivariate Analysis of Group Session Content," in Robert P. Leone and V. Kumar, eds., Enhancing Knowledge Development in Marketing, vol. 3, Summer, Chicago: American Marketing Association, 540-545.

McDonald, William J. (1993), "Focus Group Research Dynamics and Reporting: An Examination of Research Objectives and Moderator Influences," Journal of the Academy of Marketing Science, forthcoming.

Merton, Robert K., Marjorie Fiske, and Patricia L. Kendall (1990), The Focused Interview, 2nd edition, New York: Free Press.

Morgan, David L. (1988), Focus Groups as Qualitative Research, Beverly Hills, CA: Sage.

Myers, A. E. (1962) , "Team Competition, Success, and Adjustment of Group Members, Journal of Abnormal and Social Psychology, 65, 325-332.

Nelson, James E. and Nancy T. Frontczak (1988), "How Acquaintanceship and Analyst Can Influence Focus Group Results," Journal of Advertising, 17 (1), 41-48.

Nunnally, Jum C. (1978), Psychometric Theory, New York: McGraw-Hill.

Pepitone, A. and G. Reichling (1955), "Group Cohesiveness and the Expression of Hostility," Human Relations, 8, 327-337.

Reitan, Harold T. and Marvin E. Shaw (1964), "Group Membership, Sex-Composition of the Group, and Conformity Behavior," Journal of Social Psychology, 64 (October), 45-51.

Rosenberg, Morris (1965), Society and the Adolescent Self-Image, Princeton, NJ: Princeton University Press.

Ruhe, J. A. and W. R. Allen (1977), "Differences and Similarities between Black and White Leaders," in Proceedings of the American Institute of Decision Sciences, Northeast Division, April, 30-35.

Ruhe, J. A. and J. Eatman (1977), "Effects of Racial Composition on Small Work Groups," Small Group Behavior, 8, 479-486.

Schacter, Stanley, Norris Ellertson, Dorothy McBride and Doris Gregory (1951), "An Experimental Study of Cohesiveness and Productivity," Human Relations, 4, 229-238.

Seashore, S. E. (1954), Group Cohesiveness in the Industrial Work Group, Ann Arbor, MI: University of Michigan Press.

Seashore, S. E., E. E. Lawler, P. Mirvis and C. Cammann (1982), eds., Observing and Measuring Organizational Change: A guide to Field Practice, NY: Wiley.

Shaw, Marvin E. (1960), "A Note Concerning Homogeneity of Membership and Group Problem Solving," Journal of Abnormal and Social Psychology, 60, 448-450.

Shaw, Marvin E. (1981), Group Dynamics: The Psychology of Small Group Behavior, New York: McGraw Hill.

Shaw, Marvin E. and Lilly May Shaw (1962), "Some Effects of Sociometric Grouping upon Learning in a Second Grade Classroom," Journal of Social Psychology, 57 (August), 453-458.

Stewart, David W. and Prem N. Shamdasani (1990), Focus Groups: Theory and Practice, Beverly Hills, CA: Sage.

Swap, Walter C. and Jeffrey Z. Rubin (1983), "Measurement of Interpersonal Orientation," Journal of Personality and Social Psychology, 44 (1), 208-219.

Templeton, J. F. (1987), Focus Groups: A Guide for Marketing and Advertising Professionals, Chicago: Probus.

Terborg, James R., Carl Castore and John A. DeNinno (1976), "A Longitudinal Field Investigation of the Impact of Group Composition on Group Performance and Cohesion," Journal of Personality and Social Psychology, 34 (November), 782-790.

Triandis, H. E., E. R. Hall and R. B. Ewen (1965), "Member Heterogeneity and Dyadic Creativity," Human Relations, 18, 33-55.

Welch, Joe L. (1985), "Researching Marketing Problems and Opportunities with Focus Groups," Industrial Marketing Management, 14 (November), 245-253.

Wells, William D. (1974), "Group Interviewing," in James B. Higginbotham and Keith K. Cox, eds., Focus Group Interviews: A Reader, Chicago: American Marketing Association, 2-12.

Wyer, R. S., Jr. (1966), "Effects of Incentive to Perform Well, Group Attraction, and Group Acceptance on Conformity in a Judgmental Task," Journal of Personality and Social Psychology, 4, 21-26.

Wyer, R. S., Jr. and C. Malinowski (1972), "Effects of Sex and Achievement Level upon Individualism and Competitiveness in Social Interaction," Journal of Experimental Social Psychology, 6, 255-263.

Van Zelst, Raymond H. (1952), "Validation of a Sociometric Regrouping Procedure," Journal of Abnormal and Social Psychology, 47 (April), 299-246.

----------------------------------------