Looking for Groupthink: Detecting Problems in Small Group Decision Making. by Scott Stone Thesis Sp«Com c. 2 Stone, Scott. 1994 S8801 Looking for Detecting problems m 1994 Submitted in Partial Fulfillment of the Requirements for the Master of Arts in Communications Studies Degree Approved by: Dr. Kathleen Golden Chairperson, Thesis Committee Edinboro University of Pennsylvania 7/'r/w Date D/ Mary Alice Dye Committee Member O Dr. Timothy Thompson Committee Member 7f Date J\6C-&!28 C-"2- Looking for groupthink: Detecting problems in small group decision-making. Scott Stone Dr. Kathleen M. Golden April 22, 1994 ACKNOWKlanewaiNTS Sincere appreciation is extended to the many people who participated or assisted in the preparation and completion of this study. Special thanks to Thomas Egloff for his time and effort in coupleting this study. Special appreciation is paid to Dr. Kathleen M. Golden, thesis director, for her excellent guidance, encouragement, continuous wise counsel, and friendship. Gratitude for their interest and advice is expressed to the members of the thesis committee: Dr. Mary Alice Dye and Dr. Timothy Thompson. This study was influenced greatly by Dr. Randy Hirokawa, whose advice, information, and time are greatly appreciated. To Rose, my father, and my mother, the writer acknowledges his indebtedness for the many sacrifices and never-ending encouragement which brought a mutually-shared goal to realization. This study would not have been possible to complete without their help and support. ABSTRACT Hie purpose of this investigation was to examine the reasons why small groups may make effective or ineffective decisions. This research combined two related studies to examine both the effects of group cohesion on small group decision quality and the groupthihk phenomenon effect on small groups. A qualitative approach was used in examining the data. The attempt to create cohesion among group members proved highly successful, and as a result many of the problems that are often associated with cohesive groups appeared. The results of this study lend support to Hirokawa and Pace's theory that group decision-making quality is dependant upon key aspects of interaction between group members. Furthermore, the propositions that proved true continue the research into Janis' theory of groupthink. Overall this study offers further insight into the reasons why small groups make effective or ineffective decisions. TABLE OF CONTENTS Page I. INTRODUCTION 1 A. Research Problem 2 B. Terms 2 II. REVIEW OF THE LITERATURE . 5 A. Irving Janis ....................... 5 B. Data to Theory Researchers 5 C. Theory to Data Researchers 7 D. Summary............................... 11 E. Table of Researchers . . . 13 III. METHOD 14 A. Judging 16 B. Propositions 21 C. Verification 22 22 IV. RESULTS A. Propositions ....................................... 22 B. Promotive and Disruptive Influences 23 C. Demographics ................................... 25 ....................................... 26 D. Limitations V. DISCUSSION ........................................... VI. RECOMMENDATIONS FOR FUTURE RESEARCH 28 29 Introduction The concept of groupthink originated from the research of Irving Janis, who was trying to determine why groups comprised of highly intelligent and effective decision­ making individuals often combined to formulate some of the most ill-conceived decisions of our time. Janis states "Groupthink refers to a deterioration of mental efficiency, reality testing, and moral judgment that results from in­ group pressures." (Janis, 1972, p.28) Janis has spent a great deal of time examining some of recent history's worst decision-making events, trying to summarize how the groups found themselves making such terrible decisions. Hirokawa is one of many people who has examined some of Janis's ideas. Hirokawa and Pace (1983) researched what elements go into making an effective or ineffective decision. This team did an extensive study on small group decision-making. The major emphasis of this study comes from an experiment Hirokawa and Pace conducted in which they examined effective and ineffective groups to find out what aspects of decision making separated them. The current study continues the experiments of Hirokawa and Pace with a closer emphasis on Janis's symptoms of groupthink. These symptoms are described in the terms section of this paper. By replicating the experiment Hirokawa and Pace directed, changing only the four propositions to compare to groupthink symptoms, experimental results lend additional support to Janis's theory. Looking for groupthink 2 Research Problem By replicating Hirokawa and Pace (1983) and by changing the propositions to include some of the symptoms of groupthink, the hope is to add a greater degree of substance to the premise that Janis's symptoms of groupthink do indeed lead to low quality decisions. The research question asks: Do the symptoms of groupthink really exist in small group decision-making and if so, are they contributors to ineffective decision-making? Some of the problems examined were: Why do some small groups make ineffective, low quality decisions? Since this problem is very general, the scope of the research was narrowed to examine these problems: Are there any key elements that ineffective groups share that are not found within effective groups? Are there any key elements that effective groups share that are not found in ineffective groups? Terms Below are the key terms which guide this study. Groupthink - "Groupthink refers to a deterioration of mental efficiency, reality testing, and moral judgment that results from in-group pressures." (Janis, 1972, p.28) Effective group - a group whose decision received a rating of at least a "6," on a scale from 1 to 7, on all four criteria by all three judges (Hirokawa and Pace 1983) . Looking for groupthink 3 Ineffective or-mip - a group whose decision received a rating of no higher than "2," on a scale from 1 to 7, on all four criteria by all three judges (Hirokawa and Pace 1983). Symptoms of groupthink 1. An illusion of invulnerability, shared by most or all members, which creates excessive optimism and encourages taking extreme risks; 2. Collective efforts to rationalize in order to discount warnings which might lead members to reconsider their assumptions before they recommit themselves to past policy decisions; 3. An unquestioned belief in the group's inherent morality, inclining the members to ethical and moral consequences of the decisions; 4. Stereotyped views of enemy leaders as too evil to warrant genuine attempts to negotiate, or as too weak and stupid to counter whatever risky attempts are made to defeat their purpose; 5. Direct pressure on a member who expresses strong arguments against any of the group's stereotypes, illusions, or commitments, making clear that this type of dissent is contrary to what is expected of all loyal members; 6. Self-censorship of deviations from the apparent group consensus, reflecting each member's inclination Looking for groupthink 4 to minimize to himself/herself the importance of his/her doubts and counter-arguments; 7. Shared illusion of unanimity concerning judgements conforming to the majority view (partly resulting selfcensorship of deviations, augmented by the false assumption that silence means consent); 8. The emergence of self-appointed mindguards - members who protect the group from adverse information that might shatter their shared complacency about the effectiveness and morality of their decisions. (Janis 1972, Montanari & Moorhead 1989) Looking for groupthink 5 Review of literature Tiie study of group decision-making has been around as long as people have joined in an attempt to create a higher quality decision. Much of the literature reviewed is dated, but necessary to understand the concepts that came about from this early work. Janis (1972, 1982) has been a pioneer in the study of group decision-making. Janis believes there are distinct elements which hinder a group's ability to make high-quality decisions. In the books Groupthink and Victims of Groupthink, Janis illustrates his theory by applying it to some of the worst decisions ever made by high-powered groups. He examines the Kennedy administration's errors that led to the Bay of Pigs fiasco and Nixon's Watergate coverup. In a counterpoint, he shows us how Kennedy' s administration learned from its mistakes and avoided groupthink during the Cuban Missile Crisis. These books were the catalysts from which many other studies and theories of group decision-making arose. Many researchers, like Janis, took the data-to-theory approach and applied groupthink to obvious decision-making fiascoes creating case studies. These data-to-theory researchers include Raven (1974), who examined the Nixon administration and the Watergate coverup focusing on the administration's inability to accept that they had been caught; Huseman & Drive (1979), who applied Janis' theory to Looking for groupthirik 6 both small and large business groups to show both were susceptible to groupthink; and Manz & Sims (1982), who attempted to elaborate on Janis' theory to include autonomous work groups. These early studies were marked by a constant emphasis on the need for further research into group decision-making. Continuing to examine flawed decisions, Smith (1984) looked at the possibility of groupthink playing a large part in the failure to liberate hostages from Teheran, and Hensley & Griffin (1986) studied the Kent State University Board of Trustees' decisions during the 1977 gymnasium controversy. This study was continued through many articles and has still not reached a complete resolution. The Challenger Shuttle disaster left N.A.S.A. under the microscope of many researchers. Among the first to examine this catastrophe were Esser & Lindoerfer (1989), who questioned the decision of N.A.S.A. to launch the shuttle, using recent studies conducted by others in the field as support for their research. This study thoroughly examined verbal interactions recorded just before the shuttle launch and tried to tie in the symptoms of groupthink with the interaction of the people involved. The most recent case study reviewed was by Moorehead, Ference, & Neck (1991), and continued the analysis of decision fiascoes by reexamining the Space Shuttle Challenger decision and revising the groupthink framework. Looking for groupthink 7 Much of Moorehead, Ference, & Neck's research was based on McCauley's (1989) research when he reviewed seven independent cases for possible groupthink trying to establish a clearer definition of the groupthink phenomenon. Whyte (1989) proposed a new look at groupthink by combining the pressures for uniformity with the aspects of framing, risk, and group polarization. This unique view provides a wider, but often more confusing, way of examining fiascoes. In 1990 Hart (1990) applied his revised model of groupthink to Irangate and the failure to prevent the 1940 Nazi invasion of Holland. These resources give the reader a good basis of knowledge into the possible reasons why these fiascoes occur. However, this only helps us to understand why these poor decisions were made. Very little information is supplied on how to keep from falling victim to the symptoms of groupthink. We need to better understand the ways in which groups go about making high-quality decisions, and what they are doing differently from groups making predominantly low-quality decisions. This is the approach taken by the theory-to-data analysts. The theory-to-data approach was led by Flowers (1977) who used 120 undergraduates to examine how group cohesiveness and leadership style affect a group's decision­ making process. He found that group cohesiveness was not a significant factor in the quality of the decision, but that Looking for groupthink leadership style was. 8 The combination of leadership style and group cohesiveness was also not found to be a significant enough factor to influence the decision-making process. Since this is an aspect of particular importance to this study, his findings were examined very closely. Close attention was also paid to Courtright (1978), who used 96 undergraduates to study the effect of group cohesion and leadership style on the quality of decisions reached by groups. He could not prove any significance between group cohesion or leadership style when associated with the quality of decisions but used many interesting techniques of gathering the large amount of data. For example, Courtright told twelve groups that they were highly cohesive, and told another twelve groups that they would probably not get along. Using these two different sets of data, he was able to compare and contrast the interpersonal reactions of the groups. Since Courtright could not prove any significance of group cohesion or leadership style on the quality of the decisions, this experiment attempted to create cohesion among the members of the groups. Janis believed cohesion is a key element in the groupthink phenomenon and therefore this experiment will attempt to introduce that element. Using content analysis, Tetlock (1979) researched American foreign policy decisions looking for possible groupthink situations. By evaluating the complexity of the speaker, the political group with which the speaker Looking for groupthink 9 identifies, and the opponents of the speaker, Tetlock observed the effects of these elements on the quality of the decisions made. This study has come under scrutiny because of the limited scope of information examined by Tetlock. In the wake of Tetlok's experiment Fodor and Smith (1982) reexamined Flowers study regarding group cohesion and leadership style using 80 undergraduates. Again group cohesion showed little significance to the quality of the decisions reached by the group, while leadership style proved to be a significant factor. In another large experiment Callaway & Esser (1984), using 128 undergraduates, examined the problem-solving procedure undertaken by these groups, along with the effects of group cohesion. The students were randomly assigned to 32 four- person groups. Half the groups were given a decision-making problem concerning the amount of money a horse trader made during a series of trades, while the other half were given a problem regarding ranking 15 items by importance they would want if lost at sea. Only the lost at sea problem showed any significance in group cohesion, while the problem solving procedures showed no significance at all. However, when combined, the group cohesion and the problem solving procedure showed significance in both the lost at sea and horse trader problems. This experiment proved very informative in both its results and the way in which the experiment was carried out. Many of the minor changes made Looking for groupthink 10 from Hirokawa and Pace's study can be linked to aspects of Callaway and Esser's experiments, Some of these changes include the use of the personality questionnaire to create cohesion and the post experiment questionnaire to measure cohesiveness. Continuing their research Callaway, Marriott, and Esser (1985) used 112 undergraduates to examine the effects of dominant members and procedural guidelines in decision­ making groups. The results showed that dominant members made higher quality decisions, exhibited less anxiety, took more time to reach a decision, verbally agreed and disagreed more, and reported more group influence on members. The decision-making procedures affected only the amount of time it took to reach a decision. In the same year, Leana (1985) used 52, four-person groups comprised of undergraduates and recreated Flowers' experiment with a few more dependant variables. Group cohesiveness showed significance on one- fourth of the decision process while leadership style showed significance on three-fourths of the decision process. The combination of leadership style and group cohesion again proved insignificant but was judged to be worthy of further study. Moorhead and Montanari (1986) had 45 teams of three to five college seniors, and used factor and path analysis to study group cohesion, insulation, and leadership style. By attempting to introduce symptoms of groupthink into the group, they studied its effects on the quality of the Looking for groupthirik 11 decisions produced. The results showed that the effects of groupthink on the quality of decisions were not as substantial as Janis had implied. Since it is such a difficult task to introduce any factor consistently into a group's decision-making, especially factors dependant on intergroup dynamics, this experiment left considerable room for further research into groupthink. Hirokawa (1980, 1983, 1987, 1989) began examining Janis's theory of groupthink and eventually branched off to examine other possible problems with small group decision­ making. Hirokawa has published numerous articles on the topic of small group decision-making, looking more toward why groups make effective or ineffective decisions. In his first publications, Hirokawa examined the elements used by both effective and ineffective groups while coming to a decision. He found a strong correlation between group interaction and the quality of the decision reached. These articles cover his reasoning, procedures for gathering data, findings, and reactions. Hirokawa's later work focuses more on how effective groups go about making high-quality decisions and what people can do to avoid the problems associated with low quality decision groups. A substantial amount of research still must be done before any concrete answers will develop as to why some groups make high-quality decisions while others cannot. Hopefully, by examining as many aspects of group decision- Looking for groupthink 12 making as we can, we will be able to focus in on those characteristics that keep some groups from achieving highquality decisions. The concepts of cohesion, groupthink, leadership style, group interaction, and problem solving procedures have been analyzed from many different points of view. This experiment adds another piece to the puzzle of group decision-making by combining ideas to examine group interaction from yet another angle, adding to the general pool of information from which the next researcher will gather his or her ideas. The table on the following page presents the focus and the approach of the research reports that were studied during the planning of this experiment. Looking for groupthink 13 Authors by Focus and Approach Data Theory to to Theory Data Groupthink Group Cohesive -ness Janis (1972, 1982) Raven (1974) Flowers (1977) Courtright (1978) Huseman & Drive (1979) Tetlock (1979) Hirokawa (1980, 1983, 1987, 1989) Fodor (1982) Manz & Sims (1982) Callaway & Esser (1984) Smith (1984) Callaway, Marriott & Esser (1985) Leana (1985) Hensley & Griffin ■ (1986) Moorhead & Montanari (1986) Esser & Lindoerfer (1989) :kkk kkk kkk kkk McCualey (1989) kkk kkk kkk kkk kkk kkk kkk kkk Whyte (1989) Hart (1990) Moorehead, Ference & Neck (1991) kkk kirk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk kkk Looking for groupthink 14 Method This study began by attempting to obtain samples of "effective" and "ineffective" groups. To do this, undergraduate communication classes at a North-eastern University were used. The volunteers were told that they would be participating in a study of group decision-making including a 30-minute, group discussion that would be videotaped. A portion of the 16PF (Schuerger, 1992) personality factor instrument was administered to three classes of approximately 25 students. A copy of this personality questionnaire is included in Appendix E. It was explained to the students that the personality factor questionnaire would be the basis for assigning each of them to groups. They were also told that by using these questionnaires to examine each of their individual personalities, we would be able to create groups who should work well together and be able to reach high-quality, effective decisions. A written copy of what the students were told can be found in Appendix A. Janis believed that groupthink is a result of the cohesiveness of the group. To create cohesion among the group members, a technique used by Callaway and Esser (1984) was used. The classes participating in the study were administered the personality questionnaire and the subjects were led to believe that the questionnaire was the basis for assigning them to groups with similar personalities. In Looking for groupthink 15 reality, the questionnaires were no more than a tool to allow the students the freedom to believe in their group's ability to come up with a high-quality decision. The only piece of information collected from the questionnaire was the student's name. Each of the questionnaires collected was numbered one through the number of people in the class. A random number generating computer program was used to randomly assign each of these numbers to one of four groups for each class. Each of these numbers was used to find the corresponding student in each class and set up the groups. Students who were not assigned to one of the four groups per class were listed as alternates for absent class members. After the groups had been formed, half of the groups were given the task of deciding on a recommendation for dealing with a journalism student who was caught plagiarizing on a class assignment. The other half of the groups were asked to produce a recommendation for dealing with a dormitory resident who maliciously damaged her roommate's stereo system. Each group was given written background information pertaining to their assigned case, and each had 30 minutes to reach the best decision they could. All the groups were videotaped while making their decision. A post-test was administered to all group members to check the cohesiveness manipulation. can be found in Appendix D. A copy of this test The subjects responded to a Looking for groupthink 16 seven-point Likert-type (Likert, 1932) scale to questions concerning different aspects of their interpersonal communication within the group. These data provided a better understanding as to the subjects' confidence in the group's decision quality, agreement with the group's decisions, extent of attempted influence by the group, and extent of actual influence by the group. A small demographic questionnaire was placed on the back of the post-test. Although this information was not used in this experiment, it may be helpful for future reference and further study. After completing the questionnaire, the students were debriefed as to the actual purpose of the study. copy of the debriefing can be found in Appendix B. A written After the debriefing the students were allowed to ask questions pertaining to the study and were asked not to discuss any of the experiment with individuals outside of their class until the next week. The next step was to evaluate the decisions reached by the groups in order to identify both high and low-quality decisions for further study. of judges were utilized. To do this, two separate sets Each set of judges consisted of three individuals who have the knowledge and expertise to properly evaluate the decision reached by the groups. The judges for the plagiarism case consisted of two journalism professors and an English professor while three senior Looking for groupthink 17 resident assistants judged the vandalism case, The judges were asked to evaluate the group's decisions in four categories: appropriateness, warrantedness, reasonableness, and fairness. The judges used the scaling system below. Appropriate _1_:_2 Warranted _1_:_2' Reasonable _1_:_2' Fair 1 : 2 3_:_4_:_5_:_6_:_7_ '3_:_4_:_5_:_6_:_7_ 3_:_4_:_5_:_6_:_7_ 3 : 4 : 5 : 6 : 7 Inappropriate Unwarranted Unreasonable Unfair The next step was to identify "effective" and "ineffective" groups. Hirokawa and Pace believed that the four evaluation criteria were equal in importance, and therefore the researcher will not try to weight one category more than another. Using the judges' evaluations as a basis, and by defining an "effective" group as a group whose recoirmendation received at least a "6" on all four criteria by all three judges, and an "ineffective" group as a group whose recommendation received no higher than a "2" on all four criteria by all three judges, "effective" and "ineffective" groups were selected. By these criteria one effective group and one ineffective group were chosen. These group's written decisions as well as judges I evaluations can be found in Appendix C. Analysis of the group discussions created a small problem for Hirokawa and Pace. The decision to use a structured or unstructured approach for analyzing the data took much consideration. For this study, an unstructured Looking for groupthink 18 approach was used to analyze the discussions in a more free­ form manner, without the aid of any a priori observational categories. The reason this approach was taken was that an unstructured format was deemed more appropriate for this kind of study; since the discussions were to look at the problem as broadly and as deeply as possible, Hirokawa and Pace did not want to restrict the judges' opinions, In this experiment the judges were given checklists of observational categories to the judges for two main reasons, First, since the experiment was looking for specific aspects of the decision-making process, it was deemed unnecessary for the judge's to have to labor over each communication utterance. Second, since we have the benefit of Hirokawa and Pace's findings, we already have an extensive list of promotive and disruptive influences to create checklists for use by the judges. Procedures for analyzing the group discussions were set up by creating two separate research teams. Each team consisted of two principal observers, who were carefully instructed as to the objective of the study and the procedures that would be errployed for analyzing the discussions. One team began by examining the "effective" groups independently of each other. The goal of the team was to "identify all observable aspects of the discussions that might have accounted for the groups' high-quality decisions"(Hirokawa & Pace, 1983). Each judge was Looking for groupthink 19 instructed to carefully read through the list of promotive influences and independently view the videotape looking for an instance of promotive influence - that is, "communication utterances which function to help the group arrive at a high-quality decision" (Gouran & Hirokawa, 1980) . Once the judges had looked at the video tape, they met as a team and compared their respective observational lists, reexamining any discrepancies. By comparing these lists, the team came up with one general list of communication characteristics which appear to account for the decision-making success of the "effective" groups. To formulate this team list the judges had to agree that the promotive influence actually occurred during the discussion and the promotive influence had to be common to both judges. The judges' next job was to observe and analyze the "ineffective" groups. The judges watched the video tape independently of one another and attempted to "identify all observable aspects of the discussions that might have accounted for the groups' low-quality decisions"(Hirokawa & Pace, 1983). Each judge was instructed to carefully read through the list of disruptive influences and independently view the videotape, looking for any instance of disruptive influence - that is, "communication utterances which function to prevent the group from arriving at a highquality decision" (Gouran & Hirokawa, 1980). Once the judges had looked at the video tape they met again as a team Looking for groupthink 20 and compared their respective observational lists, reexamining any discrepancies. By comparing these lists, the team came up with one general list of communication characteristics which appear to account for the decision­ making failure of the "ineffective" groups. To formulate this team list the judges had to agree that the disruptive influence actually occurred during the discussion and the disruptive influence had to be common to both judges. The next step was to repeat the judging in reverse with the second team of judges first examining the "ineffective" groups for disruptive influences and then studying the "effective" groups for promotive influences. Then the two teams met and compared their respective team lists for the purpose of reducing results to two final lists: a list of disruptive influences not found in effective groups, and a list of promotive influences not found in ineffective groups. Looking for groupthink 21 Propositions Hirokawa and Pace developed four propositions that they believed would be proven by their study, The researcher changed these to more closely represent symptoms of groupthink and tested these propositions to add substance to the idea that Janis's theory of groupthink does affect the quality of decision-making. Proposition 1: The quality of a group's decision may be dependent upon the manner in which group members examine and evaluate the risks involved with each decision (illusion of invulnerability) . Proposition 2: The quality of a group' s decision may be dependent upon the manner in which group members allow each other to present and evaluate alternatives (direct pressure on dissenters). Proposition 3: The quality of a group's decision may be dependent upon the group's ability to assess the ethical and moral consequences of their decisions (unquestioned belief in the group's inherent morality). Proposition 4: The quality of a group's decision may be dependent upon the group's willingness to rationalize decisions and discount warnings that might lead the members to reconsider their assumptions (collective rationalization). Looking for groupthink 22 Verification The research was verified by turning the four propositions into four seven-point rating scales, and then having other observers use those scales to rate the quality of the group discussions. Hirokawa and Pace reasoned that if the observations were valid, the raters should rate the discussions in a manner consistent with the predictions of the four propositions. The researcher believes this to be a valid conclusion, so he too used this method of verification. The results of the verification showed that there are large differences in "effective" and "ineffective" groups. Results The goal of this experiment was to add substance to Janis' theory that highly cohesive groups sometimes make low-quality decisions, not because of any flaw in the individuals themselves, but in their interaction with each other and their method of decision-making. The researcher set about this by establishing four propositions, directly related to groupthink symptoms, that he felt would separate • effective decision-making groups from ineffective ones. Of these four propositions, three were found to exist in ineffective groups but not in effective groups. The three propositions found to be valid were: Proposition 1: The quality of a group's decision may be dependent upon the manner in which group members examine and Looking for groupthink 23 evaluate the risks involved with each decision (illusion of invulnerability). Proposition 3: The quality of a group's decision may be dependent upon the group's ability to assess the ethical and moral consequences of their decisions (unquestioned belief in the group's inherent morality). Proposition 4: The quality of a group's decision may be dependent upon the group's willingness to rationalize decisions and discount warnings that might lead the members to reconsider their assumptions (collective rationalization). The second proposition was rejected because it was found in both the effective and ineffective groups. This proposition stated: Proposition 2: The quality of a group's decision may be dependent upon the manner in which group members allow each other to present and evaluate alternatives (direct pressure on dissenters). After the judging process, lists of promotive influences common only to effective groups and disruptive influences common only to ineffective groups were formed. These lists are supplementary to the propositions but proved to be very enlightening. These list are: Pramotive influences not found in ineffective groups. 1. The group determined the difference between facts, evidence, and speculation. Looking for groupthink 24 2. The group had a member who controlled communication channels and promoted evenness of participation. 3. The group asked for clarification on vague ideas presented by other group members. 4. The group avoided rationalizing their decisions with irrelevant information, sticking with the facts to back up their conclusions. 5. The group was able to assess the ethical and moral consequences of their decision. 6. The group was able to examine and evaluate the risks involved with each decision. Disruptive influences not found in effective groups. 1. The group had a member(s) who withdrew from group discussion by refusing to participate and/or engaged in irrelevant conversations. 2. The group made no attempt to determine the difference between facts, evidence, and speculation while making their decision. 3. The group had a member who monopolized group time with long, drawn-out monologues. 4. The group had a member or members who went along with one member1s ideas without examining other possibilities. 5. The group went along with one member's vague ideas without asking for clarification. 6. The group had a member who goes along with other members' ideas without questioning them or offering any ideas of their own. 7. The group rationalized their decisions with information other than facts. 8. The group failed to examine the ethical and moral consequences of their decision. 9. The group failed to examine and evaluate the risks involved with each group. A key variable in this experiment was the group members' perceived cohesiveness with one another. The post- Looking for groupthink 25 experiment evaluation was administered to measure how cohesive the group members themselves felt within their group. With the copy Post Experiment Evaluation IVfean Scores of the post experiment evaluation that can be found in Appendix D and the chart here, we can see that the is Q1 Q2 Q3 Q4 □ Overall ■ Ineffective □ Effective Q5 overall cohesion level among group members was quite high. Questions three and four are deceiving due to the nature of the scaling, but are showing highly-cohesive groups. Both the effective and ineffective groups show a higher than average cohesion level in four of the five questions. The groups' mean cohesion levels were well above the acceptable mean of four on all five questions. Ages of Students Demographic information is presented in charts as seen here. The average age of the students and the average class rank were surprising with □ ■ □ ■ □ ■ □ ED 18 19 20 21 22 23 24 25+ eight non-traditional students and more juniors than any other class rank. The more advanced age of the students could have contributed to the overall high-quality of the decisions reached. With the older students bringing more group decision-making experience into the resolution process, the overall conclusions reached were Looking for groupthink 26 enhanced. Groups containing at least one member of the over 24 age group scored an average of 1.66 points better in the overall decisions rated by the judges. This calculates out to a 5.9% increase in their overall score. Further research into the role of age and experience in the group decision­ making process might provide more insight into another aspect of quality decision-making. Since the average age of undergraduate students at large Universities is under age 21, the more mature average age of students in this study could have produced higher-quality decisions than might occur in similar studies done at other large Universities. Women outnumbered men in this study as shown below reflecting the demographics of the University where female students outnumber male students. Women were not the majority in every group due to factors involving random placement into groups. Demographic information may prove useful in further research on this subjectveer inSchool Number of Females to Males 0 2 4 6 8 1012141618 20 22 24 26 28 30 32 34 □ Freshman ■ Sophmore □ Junior ■ Senior Limitations With this experiment, the attempt was to test Janis's theory of groupthink by replicating a previous study with minor changes, The limitations that exist are that it took at least fifty volunteer subjects and eleven judges to replicate Hirokawa and Pace's study. A study using this Looking for groupthink 27 many people takes considerable planning, resources, and time to accomplish. Time proved to be a considerable limitation in the collection of data. The experiment hinged on misleading the students into believing the groups were formed to allow the highest quality decisions. In the post­ experiment debriefing the students were told about the random group assignments and were asked not to discuss this with others outside of class until the following week. To reduce the chance of the students discussing the experiment with other individuals, the entire data collection phase of the experiment was conpleted in just three days. Another limitation was that Janis's theory of groupthink is very broad, and it was difficult to define the symptoms precisely. To define these symptoms as influences of communication was also not an easy task and took much consideration. Looking for groupthink 28 Discussion This experiment was originally conceived to examine the effects of groupthink on small group decision making. As the research into current literature progressed, the idea of what aspects allow some groups to create higher quality decisions than other groups came as a natural extension of the groupthink problem. When it came time to begin collecting data, the overriding goal was to try to find the aspects of decision making used by effective groups that were not found in ineffective groups. The propositions relating to groupthink came as extensions to the superordinate goal of the decision making aspects. The experiment confirmed three propositions and rejected one while the attempt to create cohesion among group members proved effective. The attempt to create cohesion among group members was a key element in establishing a successful environment for groupthink to occur. Cohesion usually takes place over time and thus time is often thought of as a necessary ingredient of groupthink. Since time was limited for this experiment another option had to be found in the way of creating cohesion. After carefully examining the literature it was found that another research group (Callaway & Esser, 1984) had administered, with outstanding success, a bogus personality questionnaire in an attempt to create cohesion. This method was chosen since it allowed the study to remain Looking for groupthink 29 in a limited time frame. Callaway & Esser's experiment also addressed the aspect of zero-history groups sunmarizing that with the cohesive manipulation, the lack of prior group interaction was insignificant. A wide variety of decisions was formulated by the groups and can be found in Appendix C. Although many of the decisions were similar in the individual ideas the group presented, the final judgment often differed substantially. The fact that the judges had to rate the decisions on the overall conclusion reached, and not the elements within each decision, frustrated some of the judges. The expert judges who examined the written decisions expressed a feeling that many of the conclusions that were rated as poor or mediocre included excellent elements and ideas, but were not altogether fair when brought together as a package. This eliminated some decisions that may have been excellent except for one or two elements. An example of this can be found on the second page of Appendix C. Group 4 had many good ideas but two of the judges felt that asking the young woman's parents to attend counseling was beyond their right. Unfortunately, there is no easy way to avoid this problem since the total decision must be examined. Recommendations for Further Research When this study was originally conceived it was not intended to be such a vast undertaking for the researcher. Many unexpected and interesting problems presented Looking for groupthink 30 themselves along the way. This study could have been improved by increasing the number of groups and judges used. This increased amount would allow a more narrow scope of research while lending more credibility to the data. Allowing the students to meet a few times in their groups and make decisions similar to the type used in this experiment, prior to videotaping groups, could prove useful in future research. The extended time would allow the groups to become more comfortable with one another, and possibly could improve group cohesion. In turn, this could increase the interval between high-quality and low-quality decisions by eliminating many of the interpersonal problems associated with first-time meetings of group members. The experimenter gained a great deal of knowledge into both the subject matter reviewed for formulating this experiment and the planning, effort, and time it takes to conduct such an investigation. The data generated by this study should allow better insight into small group decision­ making and the problems associated with it. The researcher will gain much more information than he expected concerning the effort and time constraints of such an experiment, the generosity with time and information of many University Professors across the country, the political process that researchers must wade through, and the patience of friends and family. Bibliography Barge, J.K. & Hirokawa, R.Y. (1989). Toward a communication competency model of group leadership. Small Group Behavior., 20 (2) ,167-189. Callaway, M.R. & Esser, J.K. (1984) . Groupthink: Effects of cohesiveness and problem-solving procedures on group decision-making. Social Behavdor and Personality. 12 (2), 157-164. Callaway, W.R., Marriott, R.G. & Esser, J.K. (1985). Effects of dominance on group decision-making: Toward a stress reduction explanation of groupthink. Journal of Personality and Social Psychology. 49.949-952. Cline, R.J.W. (1990). Detecting groupthink: Methods for observing the illusion of unanimity. Comrnum cation Quarterly. 28. (2),112-126. Courtright, J.A. (1978) . A laboratory investigation of groupthink. Communication Monographs. 45,229-246. Esser, J.K. & Lindoerfer, J.S. (1989) . Groupthink and the space shuttle challenger accident: Toward a quantitative case analysis. Journal of Behavioral Decision-making, 2,167-177. Flowers, M.L. (1977). A laboratory test of some implications of Janis's groupthink hypothesis. Journal of Personality and Sonia1 Psychology. 22,888-896. Fodor, E.M. & Smith, T. (1982) . The power motive as an influence on group decision-making. Journal of Personality and Social Psychology. 42,178-185. Hart, P. (1990) . Groupthink In government-: A study of small .groups and policy failurer Published doctoral dissertation, Rijksuniversiteit, Leiden, Netherlands. Hensley, T.R. & Griffin, G.W. (1986) . Victims of groupthink: The Kent State University Board of Trustees and the 1977 gymnasium controversy. Journal of Conflict Resolution. 30. 497-531. Hirokawa, R.Y. (1987). Why informed groups make faulty decisions: an investigation of possible interaction-based explanations. Small Group Behavior,. 18 (l),3-29. Hirokawa, R.Y. (1980). A comparative analysis of communication patterns within effective and ineffective decision-making groups. Communication Monographs. 47,314-321. Hirokawa, R.Y. (1988). Group communication and decision-making performance. Human Communication Research. 14 (4),487-515. Hirokawa, R.Y. & Johnston, D. (1989). Toward a general theory of group decision-making: Development of an integrated model. .Small Group Behavior. 20 (4),500-523. Hirokawa, R.Y. & Pace, R. (1983) . A descriptive investigation of the possible communication-based reasons for effective and ineffective group decision-making. Communication Monographs. 50.363-379. Huseman, R.C. & Driver, R.W. (1979). Groupthink: Implications for small group decision-making in business. Readings in organizational, behavior. Boston: Allyn and Bacon. Janis, I.L. (1982) . Groupthink. Boston: Houghton Mifflin Company. Janis, I.L. (1972) . Victims of groupthink. Boston: Houghton Mifflin Company. Leana, C.R. (1985) . A partial test of Janis' groupthink model: Effects of group cohesiveness and leader behavior on defective decision-making. Journal on Management, 11,5-17. Likert, R. (1932) . A technique for the measurement of attitudes. Archives of Psychology, No. 140. Manz, C.C. & Sims, H.P. (1982) . The potential for groupthink in autonomous work groups. Human Relations. 15.,773-784. McCauley, C. (1989) . The nature of social influence in groupthink: Compliance and internalization. Journal of Personality and Social Psychology, 57 , 250-260. Montanari, J.R. & Moorhead G. (1989) . Development of the groupthink assessment inventory. Educational and Psychological Measurementf 49,209-219. Moorhead, G., Ference, R. & Neck C.P. (1991) . Group decision fiascoes continue: Space shuttle Challenger and a revised groupthink framework. Human Relations, M (6),539-550. Moorhead, G. & Montanari, J.R. (1986) . An empirical investigation of the groupthink phenomenon. Human Relations, 38 (5), 399-410. Park, W. (1990) . A review of research on groupthink. Journal of Behavioral Decision-makincr, a (4),229-245. Posner-Weber, C. (1987) . Update on groupthink. Small Group Behavior, 13. (1),H8-125. Raven, B.H. (1974) . The Nixon group. Journal of Social Issues. aa, 117-126. Smith, S. (1984) . Groupthink and hostage rescue mission. British Journal of Political Science, 15 f117-126. Schuerger, J.M. (1992) . The Sixteen Personality Factor Questionnaire and its junior versions. Journal of Counseling; & Developement, 21,231-244. Tetlock, P.E. (1979) . Identifying victims of groupthink from public statements of decision makers. Journal of Personality and Social Psychology, 37r1314-1324. Whyte, G. (1989) . Groupthink reconsidered. Academy of Management! Review. 14 (1),40-56. Looking for groupthink Appendix A Appendix A Pre-experiment: Congratulations, you have been selected as members of the Honors and Ethics Committee at Edinboro University. Each committee will consist of five members and be given a different case on which to render a decision. These groups have been carefully selected to ensure the highest quality decisions based on the individual personality traits of the committee members. Please read over the case thoroughly, paying close attention to all the relevant information. You will be given approximately 30 minutes in which to deliberate the case among yourselves. At the end of that time you will be asked to render a written decision regarding the action to be taken. The time remaining will be announced at the 15 and 25 minute marks. Take a few minutes to carefully read over the case. Only the written information in the case is available. Any questions about other elements of the case will not be answered. Please do not write on the cases and good luck. (Hand out cases and start recording) Looking for groupthink Appendix B Appendix B Post-experiment: (Collect written decisions by the group. Turn off the video recorders. Administer the post experiment evaluation and collect it . ) _ Now for the truth. The personality questionnaire was not the basis _for assigning people to the groups. Though the guestionnaire is a legitimate instrument in communications study, its only purpose here was an attempt to create the illusion of cohesiveness within the group. In other words to allow the group members to believe in their committee's ability to get along and make good decisions. The quality of the decisions reached will determine which video tapes are to be studied further. The highest quality and lowest quality decisions will be examined to find out what elements of decision making were used by high quality groups that were not used by the low quality groups. The post experiment evaluation will be used to determine how well the false personality questionnaire created cohesion in the group. Any questions about any aspect of this experiment can be directed to your professor or Scott Stone. All information gathered during this experiment will be held in strict confidentiality. We ask that you please do not discuss this experiment with anyone outside of class until Friday since this experiment also involves other classes. Thank you for your time. Looking for groupthink Appendix C Appendix C Vandal ism Case 2/1/94 Group 2: Ms. Lewis and Ms. Belovorak should be written up for excessive noise. Ms. Lewis should also receive a one year probation and pay Ms. Belovorak $170 in damages for the stereo. - Ms. Belovorak should be moved to an available room while Ms. Lewis remains in her present room since she was there first. Appropriate Warranted Reasonable Fair 1..2..3..★..5..6..7 1..2..3..4.*5..6..7 1..2..3+.4..5..6..7 1..2..3..4*.5..6..7 Inappropriate Unwarranted Unreasonable Unfair Group 4: (1) Ms. Lewis needs to pay for the stereo system. (2) Ms. Lewis will be put on probation and several weeks of counseling. (3) The two girls will be separated. They will both be given the choice of who will stay or leave. In the event that neither would like to leave, Ms. Lewis will remain in the room due to seniority. Proper placement will be given to Ms. Belovorak. Appropriate Warranted Reasonable Fair 1..2.★3..4..5..6..7 1..2..3..4.*5..6..7 1.*2..3..4..5..6..7 1..2.*3..4..5..6..7 Inappropriate Unwarranted Unreasonable Unfair 2/2/94 Group 2: - Ms. Lewis should have mandatory counseling if available. - Ms. Lewis should give restitution for the stereo by getting employment through the University. - Ms. Lewis should no longer be eligible for on carrpus housing beginning next semester. - For the Remainder of the semester Ms. Lewis should room alone. Appropriate Warranted Reasonable Fair 1..2..4..5..6..7 1..2. .3. .4..*..6..7 1..2..4..5..6..7 1..2..3..4.*5..6..7 Inappropriate Unwarranted Unreasonable Unfair Group (4: ^ ^13 wm reimburse the $170 plus tax by the end of the semester. _ . _ (2) Ms. Lewis will be on probation for one year. Looking for groupthink Appendix C (3) If Ms. Lewis receives one more write up she is out of the dorms. (4) A written apology to Ms. Belovorak. (5) A letter to Ms. Lewis' parents. (6) Ms. Lewis has 24 hours to find a new roommate or one will be appointed. Appropriate Warranted Reasonable Fair 1..2..3..4.^5..6..7 1..2..3*.4..5 6..7 1..2..3..4*.5 6..7 1..2..3..4.*5..6..7 Inappropriate Unwarranted Unreasonable Unfair 2/3/94 Group 2: Ms. Belovorak _ should receive a semester of probation and has the option of staying in the room or moving to a new one. She will get a private room. Ms. Lewis should receive a semester of probation, 30 hours of community service and the choice to pay for a private room or to get another roommate. Appropriate Warranted Reasonable Fair l..*..3..4..5..6..7 1.+2..3..4 5..6..7 l.*2..3..4 5..6..7 1.*2..3..4..5..6..7 Inappropriate Unwarranted Unreasonable Unfair Group 4: Our recommendation is that Ms. Lewis have mandatory counseling concerning her behavior. We also request that her parents be notified as well as be present at the first counseling session. (This may be stemming from some past family problems or events.) Also we suggest that she participate in community service activities such as working with the handicapped or young children in order to instill the value of patience. As a final request, Ms. Lewis shall be on career probation in order that she be reprimanded for the belligerent behavior eg. arguing, fighting, disrupting others, etc. She shall be expelled from school permanently. We also suggest that she pay for the damage to the stereo, regardless of whether Ms. Belovorak presses charges. We also suggest that the R.A. be spoken to by her superiors for her inability to take charge in the past situations involving the two roommates (their fighting). Also, she should not have immediately accused Ms. Lewis for having broken the stereo without concrete evidence. If she already has an explosive personality this will immediately put her on the offense. Appropriate Warranted Reasonable Fair 1..2..*..4..5..6..7 1..2..3*.4..5..6..7 1..2..*..4..5..6..7 1..2..★..4..5..6..7 Inappropriate Unwarranted Unreasonable Unfair Looking for groupthink Appendix C Plagiarism Case 2/1/94 Group 1: _ Our conclusion is based on admitting Mr. Jordan plagiarized his assigranent, and he is a senior in the journalism department. We feel that these two factors play a large role in our decision. We wish we could have viewed the students and William F. Buckley's paper. Our recommendation is failure of this class, although plagiarism will not appear on his final transcript. Appropriate Warranted Reasonable Fair l..2..3..★..5..6..7 1..2..3..★..5..6..7 1..2..3..4*.5..6..7 1..2..3..4..5*.6..7 Inappropriate Unwarranted Unreasonable Unfair Group 3: We, the members of the Ethics Committee, find that since Mr. Jordan did have some knowledge that what he was doing was plagiarism, we feel that he should receive a zero for this assignment. Also, his future professors should be made aware. If this were to occur again, further and more severe action will be taken against Mr. Jordan. Appropriate Warranted Reasonable Fair 1..2..3..★..5..6..7 1..2..3.★4..5..6..7 1..2..3.★4..5..6..7 1..2..★..4..5..6..7 Inappropriate Unwarranted Unreasonable Unfair 2/2/94 Group 1: James Jordan receives an "F" for the class and must retake the course with a different professor. Professor Adamson should be lectured by the department head on how to respect students and the proper way to handle this type of situation. Appropriate Warranted Reasonable Fair 1..2..3..4..5..6*.7 1..2..3..4..5..6*.7 1..2..3..4..5..6*.7 1..2..3..4..5..6*.7 Inappropriate Unwarranted Unreasonable Unfair Looking for groupthink Appendix C Group 3: We decided to give James Jordan a failing grade on the paper with no make-up. James Jordan should give an oral apology to his professor and classmates and also a written apology to the University. The reason for this conclusion is because of his academic background and he has not had any previous incidents. If this same event occurs again he will be immediately expelled from the University. Appropriate Warranted Reasonable Fair 1..2..3..4..5i.6..7 1..2..3..4..5i.6..7 1..2..3..4.*5..6..7 1..2..3..4.*5..6..7 Inappropriate Unwarranted Unreasonable Unfair 2/3/94 Group 1: We felt that James Jordan deserves an "F" for the course. He has the option to retake the course. If he continues to plagiarize, he will be expelled from the University. Also, there will be no trace of this circumstance on his record. Appropriate Warranted Reasonable Fair 1..2..3..★..5..6..7 1..2..3..★..5..6..7 1..2..3..★..5..6..7 1..2..3..*..5..6..7 Inappropriate Unwarranted Unreasonable Unfair Group 3: The course of action to be taken against the student will result in a failing grade for the paper and lowering of the final grade by one letter. Remark will be made on transcripts to read that he was brought up on plagiarism charges. Appropriate Warranted Reasonable Fair 1..2..3. .4..5..★..7 1..2..3. . 4..5.*6..7 1..2 3..4..5..★..7 1..2 3..4..5..★..7 Inappropriate Unwarranted Unreasonable Unfair Appendix D Looking for groupthink Appendix D Post Experiment Evaluation 1. How responsive were other group members to differences of opinions and criticism from you? Resentful |_1_|_2_|_3_|_4_|_5_|_6_|_7_| Encouraging 2. How attractive, interesting, fun to be with, and enjoyable was the group? Not Attractive |-1-|-2-|-3-|-4-|-5-|-6-|-7-| Very Attractive 3. How would you rate the ability and competence of the group? Highly Competent |-1-|-2-|-3-|-4-|-5-|-6-|-7-| Incompetent 4. How successful do you feel the experimenter was in matching individuals who worked well together? Very Successful [-1-|-2-|-3-|-4-|-5-|-6-|-7-| Unsuccessful 5. How willing would you be to participate in future research as a member of the same group? Not Interested |-1-|-2-|-3-|-4-|-5-|-6-|-7- Delighted Background Information Name: __________________ Home Town or City: Major:____________ Age:____ Sex: State: Year in School: Appendix E Looking for groupthink Appendix E Personality Questionnaire Date: Name: Please circle the number that you feel is closest to your own personality trait. Answer truthfully because this will be the basis for assigning groups that work well together. Reserved 123456789 10 Warm Concrete Thinking 12345678910 Bright, Abstract Emotional 123456789 10 Calm Adaptable 123456789 10 Assertive Serious 123456789 10 Enthusiastic Expedient 123456789 10 Conscientious Shy 123456789 10 Bold Tough-Minded 12345678 9 10 Tender-Minded Trusting 12345678 9 10 Suspicious Practical 12345678 9 10 Imaginative Forthright 12345678 9 10 Socially Proper Composed 12345678 9 10 Worrisome Conservative 12345678 9 10 Progressive Group-Oriented 12345678 9 10 Self-Sufficient Spontaneous 123456789 10 Self-Disciplined Relaxed 123456789 10 Tense