Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Groupthink in decision making : testing for its existence, effects and prevention Kyle, Neil John 1980

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1980_A1 K94.pdf [ 7.51MB ]
Metadata
JSON: 831-1.0095476.json
JSON-LD: 831-1.0095476-ld.json
RDF/XML (Pretty): 831-1.0095476-rdf.xml
RDF/JSON: 831-1.0095476-rdf.json
Turtle: 831-1.0095476-turtle.txt
N-Triples: 831-1.0095476-rdf-ntriples.txt
Original Record: 831-1.0095476-source.json
Full Text
831-1.0095476-fulltext.txt
Citation
831-1.0095476.ris

Full Text

GROUPTHINK IN DECISION MAKING: TESTING FOR ITS EXISTENCE, EFFECTS AND PREVENTION by NEIL JOHN KYLE B.A., Simon Fraser University, 1971 M.A., University of B r i t i s h Columbia, 1976 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n THE FACULTY OF GRADUATE STUDIES Department of Psychology We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1980 © N e i l John Kyle, 1980 In presenting th is thes is in p a r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make it f ree ly ava i l ab le for reference and study. I fur ther agree that permission for extensive copying of this thes is for scho la r ly purposes may be granted by the Head of my Department or by h is representat ives . It is understood that copying or publication of th is thes is fo r f i n a n c i a l gain s h a l l not be allowed without my writ ten pe rm i ss ion . Department of P s y c h o l o g y The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T.1W5 Date August 20, 1980 i i Abstract Groupthink i s a theory concerning decision making developed by Janis (1972) on a case study basis. He uses the theory to explain several i n t e r -n a t i o n a l fiascoes such as the Bay of Pigs invasion of Cuba. Highly cohesive, i s o l a t e d groups operating i n s t r e s s f u l circumstances, under highly a s s e r t i v e leaders may support the leader's p o s i t i o n i n the attempt to maintain group consensus. Experimental research reviewed by the present author i s re-interpreted as not supporting the theory. C r i t i c i s m s concerning the o r i g i n a l development of the theory and the p o s s i b i l i t y of obtaining disconfirming evidence are outlined. The present experiment obtained for subjects 192 s t a f f members from government and corporate organizations. Leadership s t y l e , cohesiveness and stress were manipulated to simulate the groupthink conditions. The four-person groups attempted to develop solutions to two current s o c i a l problems: (1) Canadian immigration, and (2) c a p i t a l punishment. The sessions were taperecorded and subjects were given post-experimental questionnaires. The questionnaires provided information f o r manipulation checks and the attempt to observe s p e c i f i c symptoms of groupthink. The audio-tapes were rated by two independent observers for the q u a l i t y of the decision making discussions. The proposed solutions ( i n t r a n s c r i p t form) were rated for t h e i r q u a l i t y by experts from Immigration and law. Groupthink theory was not supported by the analyses at the various l e v e l s . The independent v a r i a b l e manipulations were moderately successful. The r e s u l t s indicated that leadership s t y l e played the dominant r o l e i n a f f e c t i n g both the group atmosphere and the q u a l i t y of the decision making. These findings are interpreted as being consistent with e a r l i e r research i i i on groupthink theory. I t was also found that the time l i m i t a t i o n s employed i n the study influenced the task oriented dimensions of the group processes It was suggested that the ro l e of leadership may be more c r u c i a l to group-think theory than i s the r o l e of group cohesiveness. However, before reaching a decision on the v a l i d i t y of the groupthink theory based upon the current laboratory research, i t i s recommended that groupthink be tested i n a fashion more appropriate to the l e v e l of analysis of the theory. i v Table of Contents Page Abstract i i L i s t of Tables v i i i L i s t of Figures i x Acknowledgement x Introduction 1 General Decision Making Models 1 A l l i s o n ' s Decision Making Model 2 Janis and Mann's Decision Making Model 5 Decision Making Stress 6 Optimal Decision Making 7 Stages i n the Decision Stress Model 8 Group Cohesiveness 12 Groupthink Theory 15 Examples of Groupthink 16 Groupthink Preconditions 17 Groupthink Symptoms 19 Groupthink Decision Process 20 Groupthink Decision Outcome 20 Avoiding Groupthink 21 Problems i n Groupthink Theory 25 Case Studies and Experimental Research on Groupthink 30 Groupthink i n the Watergate C r i s i s 30 Flowers' Research on Groupthink 33 Courtright's Research on Groupthink 37 Tetlock's Research on Groupthink 41 V Page Proposal for the Present Study 45 Method 47 Subjects 47 Procedure 52 P i l o t Studies 52 Main Study: A c q u i s i t i o n of Subjects 52 Selection of Groups f or Antecedent Conditions 52 Selection of Group Leaders 53 Procedures During the Research Sessions 54 Procedures Following the Research Sessions 56 Procedure f o r S t a t i s t i c a l Analysis 57 Manipulation Check 57 Dependent Variables 58 Post Hoc M u l t i p l e Regression Analysis 6Q Results 62 Manipulation Checks on the Independent Variables 62 Manipulation of the Antecedent Conditions 62 Manipulation of Leadership Style 65 The Production of Concurrence Seeking 67 Analysis of the Groupthink Symptoms 67 Analysis of the Group Process Data 68 Assessments of In t e r - r a t e r R e l i a b i l i t y 7Q Manova and Anova Analyses 7Q Analysis of the Decision Outcome Data 75 C a p i t a l Punishment: Assessments of Inter- r a t e r R e l i a b i l i t y 77 Ca p i t a l Punishment: Manova and Anova Analyses 77 Immigration: Assessments of Inter- r a t e r R e l i a b i l i t y 77 v i page Immigration: Manova and Anova Analyses 80 C a p i t a l Punishment and Immigration: Assessments of 80 the R e l i a b i l i t y of Ratings Across the Two Problems C a p i t a l Punishment and Immigration: Manova and Anova's 80 on the Combined Data Mu l t i p l e Regression P r e d i c t i o n of Quality of Decision Outcome 81 P r e d i c t i o n Based Upon the C a p i t a l Punishment Data 81 P r e d i c t i o n Based Upon the Immigration Data 81 P r e d i c t i o n Across Both Problems: Canonical Correlation 85 Analysis Discussion 86 Concurrence Seeking and Groupthink Symptoms 86 Quality of the Decision Making Process 87 Quality of the Decision Outcome 92 References 1Q0 Appendix A: Pre-Experimental Session, Group Leadership Training 105 Procedures Appendix B: Leadership Training Questionnaire 110 Appendix C: General Information About the Study Given to the 113 Subject Groups Appendix D: Instructions to the Antecedent Conditions Present 115 Groups Appendix E: Instructions to the Antecedent Conditions Absent 118 Groups Appendix F: Consent Form 121 Appendix G: C a p i t a l Punishment and Immigration Problems 123 v i i Page Appendix H: Instructions to Begin the Session and Instructions 127 to Time Limited Groups During the Session Appendix I: Post-Experimental Questionnaire f o r the Group 129 Leader Appendix J : Post-Experimental Questionnaire for the Group 136 Members and Notes on the Post-Experimental Questionnaire Appendix K: Debriefing Information 145 Appendix L: Example of the Feedback Given to a Group 148 Appendix M: Rating Scales for Assessing the Quality of the 157 Decision Making Process Appendix N: Scales f or Expert Ratings of the Quality of the 163 Decision Outcome Appendix 0: A d d i t i o n a l Analyses of the Quality of the Decision 166 Outcome Ratings v i i i L i s t of Tables Table 1: Group Means for the Antecedent Conditions Questions Reaching S i g n i f i c a n c e Table 2: Group Means f o r Groupthink Symptom Questions Reaching S i g n i f i c a n c e Table 3: Group Process Data: Percentage of Agreement, Inter-Rater R e l i a b i l i t y Check Table 4: Group Process Data: Pearson Correlations, Inter-Rater R e l i a b i l i t y Check Table 5: Group Means for the Decision Process Ratings by the Two Independent Audio-Tape Reviewers Table 6: Group Means for the Decision Process Ratings by the Experimenter Table 7: Decision Outcome Data: Percentages of Agreement, Inter-Rater R e l i a b i l i t y Check Table 8: Decision Outcome Data: Pearson Correlations, Inter-Rater R e l i a b i l i t y Check Table 9: M u l t i p l e Regression P r e d i c t i o n of Decision f o r C a p i t a l Punishment Table 10: Correlations Between Predictor Variables and C r i t e r i o n Variables Table 11: Step-Wise Mul t i p l e Regression P r e d i c t i o n of Decision Outcome for Immigration Page 64 69 71 72 74 76 78 79 82 83 84 i x L i s t of Figures Page Figure 1. A l l i s o n ' s C l a s s i f i c a t i o n of the Components 3 Within Decision Making Theories Figure 2. Decision Stress Model of Janis and Mann (1977) 9. Figure 3. Schematic Outline of Components Within Groupthink 22 Theory Figure 4. Outline of P o t e n t i a l Influences on the Public 44 Statements used i n Tetlock's (1979) Research Figure 5. Design f o r Testing the Groupthink Hypothesis 46 Figure 6. Dependent Variables: Groupthink Symptoms 4& Hypothesized Interaction Figure 7. Dependent Variables: Decision Processes 49 Hypothesized Interaction Figure 8. Dependent Variables: Quality of Decision Outcome 50 Hypothesized Interaction Figure 9. Interaction E f f e c t s : Comprehensiveness Measure 169 Figure 10. Interaction E f f e c t s : F e a s i b i l i t y Measure 170 X Acknowledgement In a project that has spanned two years and involved the cooperation of numerous organizations, plus the help received from s t a f f and friends at U.B.C, i t i s impossible to name or adequately thank everyone. To the many people without whose help t h i s research could never have been accomplished, my sincerest thanks. In p a r t i c u l a r , I would l i k e to express my deepest gratitude to my wife who has provided her support and love, and stood by me through the many years of u n i v e r s i t y . In addition, I would l i k e to thank the members of my committee, Dr. P. Suedfeld, Dr. R. Knox, Dr. R. Hakstian, Dr. S. Butt, and Dr. R. Taylor, for the advice, support, and generous help that they have provided. I have greatly appreciated t h e i r time and e f f o r t on my behalf. Many thanks also to Mr. B. B o u t i l i e r and Mr. M. Boyes for the extensive hours that they spent r a t i n g the audio-tapes. F i n a l l y , I would l i k e to extend my thanks to the senior executives and the several hundred s t a f f members from the following organizations, who through t h e i r p a r t i c i p a t i o n made th i s research p o s s i b l e : B r i t i s h Columbia (B.C.) Attorney General's Department; B.C. Department of Health, Mental Health Branch; B.C. Department of Human Resources; B.C. Telephone Company; Department of National Defense; and, MacMillan Bloedel Company. - 1 -Groupthink i n Decision Making: . Testing for Its Existence, E f f e c t s and Prevention Decision making i s a process engaged i n by both i n d i v i d u a l s and groups. There are numerous ways of conceptualizing the decision making process and there i s a large number of problems that may hinder the task of reaching a decision. In considering group de c i s i o n making, one p o t e n t i a l d i f f i c u l t y i s that the members may become caught up i n a process that Janis (1972; Janis and Mann, 1977) has l a b e l l e d "groupthink". This i s a tendency for cohesive groups under strongly opinionated and assertive leaders to react to a s t r e s s -f u l decision making task by supporting the leader's view. In i t ' s attempt to maintain consensus on the decision, the group may dramatically decrease i t ' s c r i t i c a l analysis of the problem and thus impair the q u a l i t y of the subse-quent decision. Before dealing with the s p e c i f i c s of groupthink theory we w i l l consider the r e l a t i o n s h i p between i t and general theories on decision making. Of p a r t i c u l a r i n t e r e s t i s the r e l a t i o n s h i p between groupthink theory and both the e a r l i e r decision making theory developed by A l l i s o n (1971) as well as the decision making theory developed subsequently to groupthink theory by Janis and Mann (1977). General Decision Making Models Writers i n many areas of organizational behaviour and psychology have outlined t h e i r perceptions of the stages associated with the decision making process. Theorists i n systems analysis ( B e l l , Keeney & R a i f f a , 1977; Dickerson & Robertshaw, 1975; Katz & Kahn, 1978; King, 1978) have discussed the decision making aspects of problem assessment, generating a l t e r n a t i v e s , evaluating a l t e r n a t i v e s , s e l e c t i n g an a l t e r n a t i v e , implementation and evaluation. The l i t e r a t u r e on operations research (Ackoff & S a s i e n i , 1968), - 2 -decision analysis (La V a l l e , 1978), program evaluation (Ross, 1980) and s o c i a l psychology (Hoffman, 1978) also makes reference to problem i d e n t i f i -cation and formulation, idea generation, s o l u t i o n evaluation or an a l y s i s , s e l e c t i o n , and implementation. S i m i l a r l y , there are many references within the area of organizational behaviour which discuss the basic components of decision making outlined above (Alexander, 1979; Bonczek, Holsapple, & Whinston, 1979; MacCrimmon & Taylor, 1976; Miner, 1979; Thompson & Tuden, 1971). In p a r t i c u l a r , basic texts on organizational behaviour (e.g., DuBrin, 1978; Gibson, Ivancevich & Donnelly, 1979) usually contain very e x p l i c i t outlines and discussions of the de c i s i o n making process. To si m p l i f y and extract the basic, agreed upon components, the decision making process i s generally viewed as occurring i n the following sequence. Decision making s t a r t s with problem i d e n t i f i c a t i o n and t h i s leads to the generation of p o t e n t i a l s olutions. The a l t e r n a t i v e s are then evaluated and a choice i s made. The l a s t step i s the implementation of the decision and the evaluation of the outcomes. A l l i s o n ' s Decision Making Model The theory outlined by A l l i s o n (1971) suggests that the above view of the d e c i s i o n making process represents only one dimension within possible conceptualizations of decision making. A l l i s o n (1971) adds a second dimen-sion concerned with the type of decision making model that one employs as an explanation of the decision making process. He proposed that theories of decision making processes can be viewed i n terms of the following three models: (1) the r a t i o n a l actor model; (2) the organizational model; and (3) the p o l i t i c a l process model. Within each of these models A l l i s o n (1971) v i s u a l i z e d three d i s t i n c t components: (1) information input; (2) the d e c i -sion making component; and (3) the implementation aspect (see Figure 1). - 3 -Figure 1 A l l i s o n ' s C l a s s i f i c a t i o n of the Components Within Decision Making Theories \ Dimension 2: \ Decision Model 1 Model 2 Model 3 \ Making Rational Actor Or ganizat i o n a l P o l i t i c a l Dimension l : \ M o d e l Decision \ Making Process \ Model Process Model Process Model Information Input Decision Making Implementation - 4 -These l a t t e r three components encompass the dimension concerned with the decision making process. The r a t i o n a l actor model i s the c l a s s i c a l model i n t h i s area. The basic components of the r a t i o n a l actor model are: (1) Goals and Objectives — T h e actor has s p e c i f i c goals and objectives that he attempts to meet when dealing with the problem. (2) A l t e r n a t i v e s — T h e r a t i o n a l actor must choose among a set of a l t e r n a t i v e p o t e n t i a l solutions. (3) Consequences—There are s p e c i f i c consequences associated with each p a r t i c u l a r decision choice and these consequences can be ranked i n terms of preferences. (4) Choice—The r a t i o n a l actor simply chooses the most preferable or most optimum a l t e r n a -t i v e . This means that, by and large, the r a t i o n a l actor behaves i n order to maximize h i s payoff. In t h i s model, organizations are conceived of as unitary agents with the information input, decision making and implementation functions being handled by that one agent. Thus, we speak of "the government's d e c i s i o n " or "the corporation implementing cost cutting measures". The organizational process model extends the r a t i o n a l actor model by describing the organizational influences that a f f e c t both the information input component and the implementation component of the decision model. The organization i s no longer conceived of as a unitary agent, nor are the orga-n i z a t i o n a l processes " r a t i o n a l " i n the sense outlined previously under the r a t i o n a l actor model. The basic concepts of the organizational process model are: (1) There i s often no i n t e r n a l consensus concerning the orga-nizatio n ' s goals and objectives. Instead there are multiple goals, and c o n f l i c t s among the goals of subunits. (2) Organizations attempt to avoid uncertainty by: (a) solving immediate problems rather than dealing with long range strategies and (b) regulating environmental v a r i a b i l i t y through the use of plans, standard operating procedures, industry t r a d i t i o n s , etc. - 5 -(3) They engage i n " s a t i s f i c i n g " behaviour regarding the achieving of goals and objectives, that i s , involvement i n a course of action that i s adequate, that produces s a t i s f a c t o r y outcomes rather than maximal outcomes. (4) Orga-n i z a t i o n a l behaviour i s generally stable but does change as the r e s u l t of experience. The p o l i t i c a l process model considers the influence imposed on the decision making by the a c t i v i t y that can be described as bargaining among the members or c o a l i t i o n s within the organization. The actions of the organiza-t i o n are seen as the r e s u l t of the i n t r i c a t e and subtle maneuveurs made by both the c e n t r a l figures and the formal and informal subgroups within the organization. The process i s analyzed i n terms of which i n d i v i d u a l s and groups are involved; what determines t h e i r r e l a t i v e degrees of power or influence; and how a l l of the groups, posi t i o n s and influences combine to y i e l d the organizational decisions and actions. The organization i s not seen as a unitary agent and the aspects of information input, decision making, and implementation are seen as being quite d i s t i n c t . Goals, i n t e r e s t s and power bases may clash and the moves are generally played out according to esta-blished rules of the game. The s i g n i f i c a n t outcome i s that the events at a l l three stages, from information input to implementation, are influenced by the p o l i t i c a l forces within the organization. Janis and Mann's Decision Making Model In terms of the taxonomy of decision theories developed by A l l i s o n (1971), most decision making theories, including the theory of dec i s i o n making proposed by Janis and Mann (1977), are located within the r a t i o n a l actor model. In addition, the theory of Janis and Mann i s concerned p r i -marily with the decision making component, although i t does take note of some of the informational input aspects of the r a t i o n a l actor model. I t i s - 6 -i n t e r e s t i n g that i n the 1972 book on groupthink Janis b r i e f l y proposes what he c a l l s a group dynamics model for the analysis of decision making. This model would be i n addition to the r a t i o n a l actor, organizational process, and p o l i t i c a l process models outlined by A l l i s o n (1971). Janis' model i s concerned with how the group dynamics associated with a small group of d e c i -sion makers may influence the r a t i o n a l i t y of the decision reached. As f a r as the present author knows there i s no s p e c i f i c research comparing the d i f f e r -ent models. The u t i l i t y of the d i f f e r e n t models seems to l i e p r i m a r i l y i n the extent to which the reader i s convinced that one model or combination of models rather than another constitutes- the most accurate or u s e f u l t h e o r e t i c a l representation of r e a l i t y . Janis and Mann (1977) point out that two major l i m i t a t i o n s on the q u a l i t y of decision making stem from the l i m i t e d nature of man's capacity to process information and from the need to make decisions which s a t i s f y the demands of bureaucratic t r a d i t i o n s , procedures and p o l i t i c s . Other l i m i -tations stem from the personality c h a r a c t e r i s t i c s of the decision makers such as the degree to which they tend to mentally s i m p l i f y complex material (conceptual complexity) , t h e i r stereotyping behaviour,, tolerance for ambi-guity, and self-esteem. Decision making st r e s s . The c e n t r a l thesis of Janis and Mann's theory i s that apart from factors such as those l i s t e d above, there are also severe l i m i t a t i o n s placed upon the q u a l i t y of decision making by the f a c t that engaging i n or preparing to make decisions produces psychological s t r e s s . This stress varies with the psychological make-up of the i n d i v i d u a l and with the perceived magnitude of the d e c i s i o n to be made. They outline a series of f i v e basic assumptions about stress and reactions to s t r e s s : (1) The degree of stress i s proportional to the expected number of u n s a t i s f i e d goals and t h e i r importance; (2) When - 7 -considering possible new courses of a c t i o n , d e c i s i o n a l stress i s a function of the extent of commitment to the present course of action: (3) Defensive avoidance of threat cues w i l l occur as a r e s u l t of high l e v e l s of d e c i s i o n a l stress i n s i t u a t i o n s where a l l perceived a l t e r n a t i v e s are associated with serious r i s k s and there i s l i t t l e hope of f i n d i n g a s o l u t i o n better than the least objectionable choice; (4) Under high threat conditions where there i s an a n t i c i p a t i o n of i n s u f f i c i e n t time to choose a course that w i l l avoid serious losses, there w i l l e x i s t a state of severe decision c o n f l i c t and stress that w i l l increase the l i k e l i h o o d of hypervigilance (panic) i n the decision maker; (5) Threat producing a moderate l e v e l of stress w i l l r e s u l t i n a v i g i l a n t search for a l t e r n a t i v e s which may y i e l d an adequate s o l u t i o n , i f the decision maker f e e l s that i t w i l l be possible to f i n d such a s o l u t i o n . Optimal d e c i s i o n making. Janis and Mann (1977) also specify seven i d e a l procedures that, i f followed, would i n d i c a t e that the decision maker was operating under a condition they characterize as " v i g i l a n t information pro-cessing". They suggest that as best p o s s i b l e , within the l i m i t a t i o n s of h i s cognitive a b i l i t i e s , the decision maker should: (1) survey a wide range of a l t e r n a t i v e p o l i c i e s , (2) consider the f u l l scope of the objectives and values that need to be met, (3) c a r e f u l l y consider the costs and benefits associated with each a l t e r n a t i v e , (4) i n t e n s i v e l y seek out a d d i t i o n a l i n f o r -mation on the a l t e r n a t i v e s , (5) a s s i m i l a t e without bias a l l the information both supporting and not supporting the preferred choices, (6) reexamine a l l the information on each known a l t e r n a t i v e , including those considered unac-ceptable, before making a commitment to a f i n a l choice, (7) o u t l i n e d e t a i l e d plans f o r implementing the decision and for dealing with contingent r i s k s . To the extent that these conditions for v i g i l a n t information processing are met, i t i s suggested that there i s an increased p r o b a b i l i t y f o r success-f u l d e c i s i o n making. Given t h i s underlying set of premises concerning stress - 8 -and v i g i l a n t information processing, the major focus of the theory i s con-cerned with the causes and consequences of dec i s i o n making when the conditions for v i g i l a n t information processing are not met. The theory may perhaps be best described by considering a representational drawing (see Figure 2) which amalgamates the features from three aspects of the model outlined by Janis and Mann (1977, see Figures 3, 4 and 8). Stages i n the dec i s i o n stress model. The decision stress model of Janis and Mann (1977) i s concerned with the antecedent conditions, mediating pro-cesses, and consequences that e x i s t at the f i v e d i f f e r e n t stages i n the decision making process. I n i t i a l l y , there must be a challenge to the e x i s t i n g course of action that threatens the decision maker with the p o s s i b i l i t y of serious losses or the p o s s i b i l i t y of a f a i l u r e to gain valuable outcomes. This i n i t i a l challenge motivates the i n d i v i d u a l (or group) to enter the f i r s t stage of the decision making process, that i s , the process of appraising the seriousness of the challenge. In t h i s f i r s t stage, information i s reviewed concerning possible losses i f the present course of action i s not al t e r e d . I f the r i s k s of maintaining the current p o l i c y or behaviour are not serious then the challenge i s rejected and the i n d i v i d u a l adheres to h i s or her present p o s i t i o n with l i t t l e d e c i s i o n a l c o n f l i c t . However, i f i t appears that there may be some r i s k from not changing, the i n d i v i d u a l proceeds to the second stage of dec i s i o n making and prepares to survey the possible a l t e r -native courses of action. In t h i s second stage, a l t e r n a t i v e s are considered and i f unacceptable are discarded; i f acceptable, then the r i s k s associated with the a l t e r n a t i v e are considered. I f the associated r i s k s are not serious then there may occur a change, with l i t t l e d e c i s i o n a l c o n f l i c t , to accepting the a l t e r n a t i v e as a solut i o n . Janis and Mann (1977) note that decision makers attempting to solve problems by using a s a t i s f i c i n g strategy, often repeat t h i s process - 9 -Antecedent Conditions START: Challenging Negative Feedback or Opportunity Figure 2 Decision Stress Model of Janis and Mann ( 1977 ) Mediating Processes Consequence s Additional Infor-mation About Losses From Continuing Unchanged iTAGE 1: Appraising the Challenge J Risks serious if I don't change^-Kaybe or Yes Discard of Unacceptable A l t e r n a t i v e J n forma t ion About Losses From Changing j(—-jls this alternative acceptable ?J Signs Of No More Information or Other Resources A v a i l a b l e Information About Deadline and Time. Pres sure s Information About Whether Others Can Be Involved Unconflicted END. Stage 5 of Adherence Prior Decision TAGE 2 j Surveying Alternatives JSearch for Another Alternative" J^Risks serious if change to this?[ " ,-** 1 -•—* aybe or Yes |ls it ilistic to [hope to find better olution? Are the risksserious if I postpone the deciion? Maybei^or Yes Can I turn the decision over to someone else? Is there sufficient time to bearch for and evaluate a better alternative? Maybe or Yes' Have I sufficiently surveyed the a11ernatives ? Unconflicted [change Defensive Avoidance Type A: Procrastinating-Lack of Interest Issue. Mo Search, Appraisal or Contingency Planning END. Weak Stage 5. Vulnerabilitl Unanticipated Challenges Defensive Avoidance Type B: Shifting Responsibility-Commitment to Someone Else's Choice, No Search Appraisal or Contingency Plans Defensive Avoidance Type C: iBolstering-Commitment to Least Objection-Eble Alternative, ith Biased Search] ppraisal and STAGE 3: Weighing of Alternat Might a modifi ed alternative be better? (Further Search For and Evaluation of Consequence s) Which alternative is best? Can I relax the requirements? Could the best alternative meet the (essential requirements; Attenuated Stages 3 £ 4: Superficial Weighing of Alternatives and Little Deliberation Abou t Commitment STAGE 4: Deliberating About Commitment Shall 1 adopt the best alternative and allow others to know? Strong Stage 5. Decision Based Upon Thorough Search, Appraisal and Contingency Planning. Adherence Despite Negative Feedback. Low Vulnerability to Changes Anticipated in Stages 1-4. - 10 -over and over again each time the a l t e r n a t i v e runs i n t o d i f f i c u l t y . This leads to a pattern of incremental changes i n the p o l i c y or action with no serious consideration of the f u l l range of p o s s i b i l i t i e s . I f there are serious r i s k s associated with an a l t e r n a t i v e i t i s then decided whether or not there i s reason to hope that a better s o l u t i o n may be found. I f the answer to t h i s question i s negative then the decision maker i s l i k e l y to engage i n what Janis and Mann (1977) term defensive avoidance. Defensive avoidance i s the evasion of threatening cues which arouse awareness of the p o t e n t i a l losses. According to Janis and Mann, defensive avoidance i s probably the most pervasive reaction that occurs i n response to important and s t r e s s f u l decision making and i s also the most d i f f i c u l t response to prevent or correct. The dec i s i o n maker w i l l postpone the decision i f possible or s h i f t the r e s p o n s i b i l i t y to someone else; or i f a l l else f a i l s , he w i l l accept the least objectionable s o l u t i o n as the best a l t e r n a t i v e possible and bol s t e r the decision by emphasizing the benefits and down-playing the anticipated losses. I t i s t h i s l a s t case of defensive avoidance through b o l s t e r i n g that i s of p r i n c i p a l i n t e r e s t i n t h i s paper, f o r i t i s t h i s process that, when i t occurs i n decision making groups, i s l a b e l l e d groupthink by Janis and Mann (1977). Janis and Mann (1977) o u t l i n e s i x main b o l s t e r i n g t a c t i c s that they have observed i n t h e i r studies on de c i s i o n making: (1) Exaggeration of the favourable outcomes, (2) Minimization of the unfavourable outcomes, (3) Denial of the aversive nature of fe e l i n g s associated with unfavourable out-comes, (4) Once a decision i s made, exaggerating the length of time that w i l l occur before action on the decision need be taken, (5) Minimizing the degree of s o c i a l s u r v e i l l a n c e of the decision commitment, (6) Minimizing personal r e s p o n s i b i l i t y for the decision. Groupthink i s defined by Janis and Mann (1977) as a c o l l e c t i v e pattern - 11 -of defensive avoidance through b o l s t e r i n g . The major di f f e r e n c e between the bo l s t e r i n g of i n d i v i d u a l decision makers and dec i s i o n makers i n a moderately or highly cohesive group i s that the group supplies a context of s o c i a l sup-port or disapproval f o r the actions. Which of the above t a c t i c s w i l l be employed by a dec i s i o n maker i s considered to be a product of the type of dec i s i o n , the s i t u a t i o n a l context, and the i n d i v i d u a l ' s p r e d i s p o s i t i o n s . I t should be noted that Janis and Mann view the occurrence of b o l s t e r i n g under some conditions as u s e f u l . When a thorough search and appraisal of the alt e r n a t i v e s has been made and the i n d i v i d u a l has selected one which i s c l e a r l y the best choice, b o l s t e r i n g the decision may be valuable i n producing a confident commitment by avoiding l a s t minute demoralization over f e e l i n g s of uncertainty about the choice. However, to complete our discussion of the second stage of the theory, i f the group (or ind i v i d u a l ) f e e l s that i t i s reasonable to hope to f i n d a better s o l u t i o n they must then decide whether or not they have s u f f i c i e n t time to consider the other a l t e r n a t i v e s . In a s i t u a t i o n where the group mem-bers f e e l that they do not have s u f f i c i e n t time because of r a p i d l y approach-ing deadlines or the apparent c l o s i n g o f f of a l t e r n a t i v e s , they may be subject to high l e v e l s of psychological stress. This w i l l lead to a state of rapid o s c i l l a t i o n between choice, s u p e r f i c i a l information intake, and obsession with the apparently i n e v i t a b l e losses, and w i l l often r e s u l t i n the hasty s e l e c t i o n of the f i r s t p o t e n t i a l means of escaping the perceived danger. Janis and Mann (1977) l a b e l t h i s state of panic as hypervigilance. On the other hand, i f the decision makers f e e l that i t i s r e a l i s t i c to hope for better solutions and there i s s u f f i c i e n t time to search for these solutions, then the degree of stress w i l l be lowered to a moderate l e v e l and they w i l l enter into v i g i l a n t d e c i s i o n making. At t h i s point, the decision maker w i l l proceed through the t h i r d stage of decision making by engaging i n - 12 -a v i g i l a n t evaluation of the a l t e r n a t i v e s . When ah a l t e r n a t i v e has been found that w i l l meet the e s s e n t i a l requirements of a good s o l u t i o n , the fourth stage i s reached, which involves making a decision to p u b l i c l y commit oneself to the choice. When t h i s l a s t phase of commitment i s completed the i n d i v i d u a l or group has reached a strong d e c i s i o n (stage 5) on the basis of a thorough research and evaluation of the p o t e n t i a l r i s k s , b e n e f i t s , and losses. Janis and Mann (1977) suggest that i n d i v i d u a l s or groups completing the f i f t h stage w i l l also have developed contingency plans to deal with po-t e n t i a l d i f f i c u l t i e s and the decision w i l l be less vulnerable to the a n t i c i -pated p o s s i b i l i t i e s of negative feedback or required a l t e r a t i o n s . Janis and Mann (1977) consider t h e i r theory to be applicable both f o r decision making on an i n d i v i d u a l l e v e l and decision making accomplished through a group process. The model i s also meant to be applicable with r e s -pect to the making of v i t a l l y important decisions as w e l l as routine, common-place decisions. Janis and Mann (1977) stress the fact that the theory i s designed f o r the understanding of decision making where the outcomes of the decision are of some consequence to the i n d i v i d u a l or group. They suggest that t y p i c a l laboratory research on hypothetical decisions i s us e f u l f o r looking at basic cognitive processes but has l i m i t e d a p p l i c a b i l i t y to r e a l -l i f e decisions. The theory i s based upon the importance of c o n s e q u e n t i a l l y i n d e c i s i o n making, but i n terms of normal decisions the consequences need not be l i f e or death; "we also expect the model to apply to everyday personal decisions concerning such matters as how to handle work assignments, whether or not to carry out s o c i a l o b l i g a t i o n s , and the l i k e — p r o v i d e d that they evoke at least a mild degree of worry about the outcome" (Janis and Mann, 1977, p. 75). Group Cohesiveness In a d d i t i o n to understanding the r e l a t i o n s h i p between groupthink and - 13 -Janis and Mann's (.1977) general theory on decision making, i t i s also important to consider the r e l a t i o n s h i p between groupthink and group cohesive-ness. Group cohesiveness i s a v a r i a b l e that i s of importance to the group decision making process and has received extensive psychological study. Early research by Schachter (1951) considered the e f f e c t s of group cohesiveness and task relevance upon the group decision making process. He manipulated cohesiveness by forming four bogus campus clubs where the high cohesiveness clubs were composed of members a l l of whom had indicated a strong preference for j o i n i n g that p a r t i c u l a r club. In contrast, the low cohesiveness groups contained members who had indicated a strong preference for j o i n i n g a d i f f e r e n t club. Task relevance was manipulated by choosing a discussion topic of i n t e r e s t to some clubs but i r r e l e v a n t to the i n t e r e s t s of the other clubs. He found that under conditions of high group cohesive-ness and task relevance, an experimental confederate who disagreed consis-t e n t l y with the group's viewpoint became the focus of e f f o r t s by the others to change his a t t i t u d e . In addition, continued r e f u s a l to change h i s p o s i t i o n threatened the s o l i d a r i t y of the group. The outcome was that the deviant was rejected, d i s l i k e d and i n a s o c i a l sense, excluded from the group. A number of authors have described the cohesiveness of a group i n terms of the sum of the forces that act to cause members to remain i n the group (Festinger, Schachter, & Back, 1950; Sherif & S h e r i f , 1964, 1969). Three major factors contributing to group cohesiveness were f e l t to be the a t t r a c -tiveness of the group, the duration of the past a s s o c i a t i o n , and the extent to which the group could mediate the attainment of important goals. Some in d i v i d u a l s j o i n groups to s o c i a l i z e or learn new s k i l l s , and others j o i n to gain prestige, but regardless of the reason, members highly a t t r a c t e d to the group have been found to frequently attempt to re c o n c i l e differences of - 14 -opinion among other members (Back, 1951). In addition, groups composed of i n d i v i d u a l s who l i k e d the other members tended to communicate more, reached quicker consensus on the group's p o s i t i o n , and then applied group pressure on r e c a l c i t r a n t p a r t i c i p a n t s to comply (Lott & Lott , 1965). This i s not a s u r p r i s i n g f i n d i n g since a number of authors have noted that highly a t t r a c -t i v e groups exert influence over t h e i r members' opinions and behaviours (Berkowitz, 1954; Festinger, 1950; Festinger, Schachter, & Back, 1950). In terms of general factors a f f e c t i n g group cohesion, i t has been found that competition, threat, and stress which are exerted upon the group by an external force t y p i c a l l y produce increases i n i n t e r n a l group cohesion (Coser, 1956; Ferguson & K e l l y , 1964; J u l i a n , Bishop, & F i e d l e r , 1966; LeVine & Campbell, 1972). In t h i s case, group cohesiveness i s often defined i n terms of the lack of i n t e r n a l c o n f l i c t . This r a i s e s an important issue i n terms of defining group cohesiveness since, as mentioned e a r l i e r , some authors have defined i t as a t t r a c t i o n to the group. The research of Weiss and h i s co-workers (Lombardo, Weiss, & S t i c h , 1973; Weiss, Lombardo, Warren, & Kelley, 1971) has demonstrated how apparent i n t e r n a l c o n f l i c t i n the form of d i s -agreement may, under the r i g h t conditions, increase the cohesiveness of the group. Their studies i n d i c a t e that while a t t r a c t i o n and s o c i a l cohesion may be due, i n part, to a t t i t u d e s i m i l a r i t y , these outcomes may also be due to the r e i n f o r c i n g e f f e c t s of being able to reply to another's comments i n cases where at t i t u d e c o n f l i c t i s demonstrated through verbal disagreement. The a b i l i t y to reply reduces the noxious state caused by the disagreement and thus, based upon learning p r i n c i p l e s , may ei t h e r increase a t t r a c t i o n to the other person or decrease the amount of d i s l i k e f o r the person. In a group s i t u a t i o n , t h i s would therefore tend to enhance cohesiveness, p a r t i -c u l a r l y i n comparison to s i t u a t i o n s where group norms r e s t r i c t e d a f u l l "give and take" discussion of disagreements. - 15 -I t i s of value to note that the groupthink theory of Janis (19.72) appears to r e l y most heavily on the d e f i n i t i o n of group cohesiveness as a t t r a c t i o n to the group. The cohesiveness defined as a t t r a c t i o n to the group i s seen to e x i s t as a necessary precursor to groupthink, before the decision stress on the group produces the "second" kind of cohesiveness, which i s a lessening of i n t e r n a l c o n f l i c t and the seeking of consensus. Groupthink Theory As mentioned e a r l i e r , an important aspect of Janis and Mann's (19.77)_ theory on decision making i s i t s d e s c r i p t i o n of the circumstances leading to the defensive avoidance technique of b o l s t e r i n g . Groupthink has been ex-plained by Janis and Mann as simply being the b o l s t e r i n g of an apparent group decision before i t has been completely accepted. In s i t u a t i o n s where there are serious r i s k s associated with the a l t e r n a t i v e s , where there are apparent-l y no better solutions to be found, and the decision cannot be postponed or s h i f t e d to someone el s e , a group w i l l be subject to severe stresses. In a moderately or highly cohesive group i t i s postulated that these stresses w i l l lead to strong fee l i n g s of a f f i l i a t i o n which i n turn produce conformity pressures and a s t r i v i n g for unanimity. The least objectionable choice i s made and the group provides shared support, r a t i o n a l i z a t i o n s , etc., to b o l -ster the decision. In an a r t i c l e written i n 1971, and i n a subsequent book published i n 1972, Janis discusses the e n t i r e groupthink process at length. To f a c i l i t a t e a det a i l e d discussion of h i s theory i t w i l l be u s e f u l to out-l i n e the steps that w i l l be taken i n examining h i s work. F i r s t , we w i l l mention the examples which Janis (1972) used as the basis of support for the derivation of the groupthink p r i n c i p l e s . Second, we w i l l consider the many group processes and personality factors that may act as the necessary conditions for producing groupthink. The nature of the "groupthink - 16 -symptoms" w i l l then be discussed. Other aspects of the decision making pro-cess apart from the groupthink symptoms w i l l be considered, as w e l l as the decision outcome which i s the r e s u l t of groupthink. Then, several examples and a summary of ways to avoid groupthink w i l l be outlined. From there we w i l l move to a b r i e f mention of the problems associated with non-cohesive group decision making approaches. We w i l l consider some of the problems associated with the way Janis (1972) has presented the theory of groupthink and then discuss his concepts concerning the type of studies which are needed i n future groupthink research. Last, we w i l l examine the case studies and experimental research that have attempted to test groupthink theory. Examples of groupthink. Janis (1972) provides f i v e examples of the c i r -cumstances surrounding major i n t e r n a t i o n a l decisions that resulted i n what he f e e l s to be fiascoes. He uses these examples as the basis for o u t l i n i n g the types of processes and outcomes involved when decision making groups engage i n groupthink. The f i r s t example concerns the Bay of Pigs invasion of Cuba i n 1961 by Cuban e x i l e s , supported, developed and financed by the United States govern-ment under the Eisenhower and Kennedy administrations. The second example i s found i n the unwillingness on the part of the naval commanders i n Hawaii to accept the p o s s i b i l i t y that Pearl Harbor might be attached by the Japanese as a prelude to war with the United States. The decision by the Truman administration to allow General MacArthur to invade and attempt to conquer a l l of North Korea, that r e s u l t e d i n the Chinese intervention and subsequent major American m i l i t a r y losses, provided the t h i r d case of groupthink pro-cesses at work. The fourth i l l u s t r a t i o n was the decision on the part of President Johnson's p o l i c y making group to escalate the war i n Vietnam during the mid-1960's. The l a s t example i s r e l a t e d to the events surrounding the decisions made by N e v i l l e Chamberlain's group i n the attempt to deal with - 17 -Nazi Germany under H i t l e r j u s t p r i o r to World War I I . Groupthink preconditions. As a r e s u l t of examining these cases i n d e t a i l , Janis (1972; Janis & Mann, 1977) proposes that a number of conditions underlie the production of groupthink i n group decision making s i t u a t i o n s . The f i r s t , and one of the most e s s e n t i a l conditions, i s the existence of a moderate or high degree of group cohesiveness. High group cohesion r e s u l t s from strong commitment to the group, a highly valued membership i n the group, and i s associated with fee l i n g s of camaraderie. Mutual friendships and l o y a l t i e s are valuable bonds e s t a b l i s h i n g cohesiveness preconditions. Simi-l a r l y , acceptance by the others i s an important feature of group cohesive^ ness that leads to the freedom to speak one's views without fear of r e j e c t i o n or l o s i n g one's standing i n the group, e s p e c i a l l y when those views deviate from the group norm. Janis (1972) suggests that group cohesiveness may be increased through the f e e l i n g s of competence produced by s u c c e s s f u l l y dealing with problems as a group and through s o c i a l rewards such as friendships within the group and the prestige of belonging to the group. Janis (1972) views group cohesiveness as a major condition promoting groupthink, "the more a m i a b i l i t y and e s p r i t de corps among the members of an in-group of p o l i c y makers, the greater i s the danger that independent c r i t i c a l thinking w i l l be replaced by groupthink..." (p. 198). Apart from cohesiveness, there are several a d d i t i o n a l variables that are seen as antecedent conditions for the production of groupthink. F i r s t , i s o -l a t i n g the decision making group from outside expert opinion r e s t r i c t s the input of v i t a l information and c r i t i c a l opinion that may r e c t i f y errors that are due to the concurrence seeking of the group's members. Second, i n con-junction with the group's i s o l a t i o n , the group w i l l often f a i l to system-a t i c a l l y search out and appraise the a l t e r n a t i v e courses of a c t i o n . This w i l l allow cursory and biased judgements to be made on p o l i c y issues. A - 18 -t h i r d condition i s the s e t t i n g of a strong norm by the leader for a p a r t i -cular type of decision or s o l u t i o n . This d i r e c t i v e leadership coupled with the leader's power w i l l help to induce conformity to the proposed s o l u t i o n on the part of the group members. This state leads to the f i n a l p a i r of antecedent conditions. Under s t r e s s f u l conditions, the leader's favouring of a p a r t i c u l a r a l t e r n a t i v e may tend to produce i n the group a f e e l i n g that there i s l i t t l e hope of f i n d i n g a better or more acceptable s o l u t i o n . Group cohesiveness, i s o l a t i o n , poor search and appraisal procedures, d i r e c t i v e leadership, d e c i s i o n a l s t r e s s , and a lack of hope for a better s o l u t i o n a l l combine to produce concurrence seeking on the part of the group members. This pressure to maintain consensus on important issues w i l l r e s u l t i n the leader's decision being bolstered by the group members. It i s suggested that the s t r i v i n g f o r unanimity found i n groupthink overrides the group member's a b i l i t y and motivation for independent or counter-normative thinking. This leads the group to r e j e c t the views of mem-bers who deviate, producing compliance even among members who may have doubts concerning the group's decisions. In combination with group process factors that; influence groupthink, Janis (1972) suggests that c e r t a i n personality c h a r a c t e r i s t i c s such as fear of s o c i a l disapproval and r e j e c t i o n may make i n d i v i d u a l s more susceptible to conformity pressures and subsequent groupthink. He notes that personality predispositions for groupthink may include s o c i a l support seeking by people with: (1) low self-esteem, (2) high responsiveness to s o c i a l disapproval, and possibly (3) low l e v e l s of moral development (footnote 6, pp. 238-239). What are the dynamics behind groupthink? Groupthink f i l l s some very powerful psychological needs. The s t r i v i n g for concurrence helps to main-t a i n f e e l i n g s of harmony and togetherness within the group. In a groupthink s i t u a t i o n i t appears to the members that there i s consensus on the decision, - 19 -that everyone i s b a s i c a l l y i n agreement. Groupthink therefore functions to help reduce one's anxiety and possible g u i l t about having made the correct choice. The groupthink process of s t r e s s i n g the p o s i t i v e features and down-playing the negative features of the proposed s o l u t i o n helps avoid damage to one's self-esteem, and i t also helps to avoid the p a i n f u l f e e l i n g s associated with severe c r i t i c i s m ofione's pet ideas by respected associates. Groupthink symptoms. As noted e a r l i e r , the actual existence of the phenomenon of groupthink i s established by the existence of a s p e c i f i c set of group processes. These processes may occur when the conditions outlined above p r e v a i l during the attempt to reach a s i g n i f i c a n t group deci s i o n . Janis (1972) outlines these processes as the eight major symptoms of group-think: (1) i l l u s o r y f e e l i n g s of i n v u l n e r a b i l i t y and sec u r i t y that lead to excessive optimism and r i s k y decisions; (2) the use of r a t i o n a l i z a t i o n s to avoid warnings of danger that might otherwise lead to a reconsideration of assumptions underlying past p o l i c y decisions; (3) unquestioned f e e l i n g s of moral correctness leading to decisions that are assumed to be moral; (4) the development of shared negative stereotypes of the enemy leaders as too e v i l to be genuinely negotiated with, or too stupid or weak to counter the actions being considered against them; (5) the a p p l i c a t i o n of d i r e c t pressure upon members who are not seen as being l o y a l to the group when they strongly question the i l l u s i o n s , stereotypes or commitments of the group; (6) the use of self-censorship to minimize the importance of one's doubts and counter-arguments and thus to i n h i b i t deviation from apparent group consensus; (7) a shared sense of unanimity concerning the majority view that i s i l l u s o r y due to the existence of self-censorship and the f a l s e assumption that s i l e n c e means agreement; (8) the phenomenon of self-appointed "mind guards" who function to protect the leader and group from strongly opposing views and disturbing information concerning the effectiveness and morality of the - 20 -decisions being made. According to Janis (1972) the "symptoms of groupthink w i l l be found most often when a decision poses a moral dilemma, e s p e c i a l l y i f the most advan-tageous course of action requires the p o l i c y makers to v i o l a t e t h e i r own standards of humanitarian behavior. Under these conditions, each member i s l i k e l y to become more dependent than ever on the in-group f o r maintaining h i s self-image as a decent human being and accordingly w i l l be more strongly motivated than ever to maintain a sense of group unity by s t r i v i n g f o r concurrence" (p. 206). Groupthink decision process. The symptoms of groupthink are i n d i c a t i v e of a poor decision making process. Assuming that the c r i t i c a l evaluation of as many a l t e r n a t i v e s as possible within time and processing constraints i s the method most l i k e l y to se l e c t the best d e c i s i o n , decision making groups that do not meet t h i s c r i t e r i o n are operating at a suboptimal l e v e l . Sub-optimal d e c i s i o n processes would apply to s i t u a t i o n s where the previously l i s t e d symptoms of groupthink were apparent. U n r e a l i s t i c optimism, the f a i l u r e to heed warnings, and misperceptions of the opponents a l l i n d i c a t e a poor information processing and decision making procedure. S i m i l a r l y , i s o -l a t i o n from outside expert opinion, the s t a t i n g by the leader of a preferred s o l u t i o n at the beginning of dicussion, allowing excessive group pressure to be exerted against those with deviant views, and the existence of "mind guards" r e s t r i c t i n g the input of important information to the group, a l l suggest a f a u l t y decision process. Janis (1972) reveals very early i n h i s writing that he considers t h i s i n f e r i o r decision process to be one of the most, i f not the most, detrimental e f f e c t of groupthink. Groupthink decision outcome. When the decision making process i s poor, more often than not the product w i l l be poor also. Janis (1972) comments that decision making groups which e x h i b i t a l l or most of the groupthink - 21 -symptoms w i l l perform i n e f f e c t i v e l y and w i l l be more l i k e l y to f a i l to meet t h e i r objectives. He f e e l s that the more frequently a group displays the symptoms, the worse w i l l be the q u a l i t y of t h e i r decisions. This may occur even when some symptoms are absent, i f other symptoms are p a r t i c u l a r l y pronounced. The display i n Figure 3 schematically o u t l i n e s , i n a s i m p l i f i e d form, the various components of the groupthink theory of Janis (1972). Avoiding groupthink. At t h i s point, Janis (1972) examined two cases of i n t e r n a t i o n a l s i t u a t i o n s where the decision making groups appeared to f i t most of the preconditions necessary for producing groupthink. However, rather than f a l l v i c t i m to groupthink they managed to avoid i t . The f i r s t example involved the Kennedy administration during the 1962 Cuban m i s s i l e c r i s i s . Some of the steps taken by President Kennedy to avoid bad decision making were as follows: (1) Individuals were not allowed to be only spe-c i a l t y experts but were required to think about and c r i t i c a l l y discuss a l l aspects of the issues. Two i n d i v i d u a l s were asked to take on the s p e c i a l r o l e of considering i n depth a l l the possible flaws i n various plans; they b a s i c a l l y became d e v i l ' s advocates. (2) Outside experts were brought i n to present fresh views and discussion was not l i m i t e d by an imposed agenda. (3) The major group was divided into sub-groups to discuss p o l i c y indepen-dently and then reassembled for general debate. (4) The leader (President Kennedy) d e l i b e r a t e l y absented himself from some meetings ( p a r t i c u l a r l y preliminary ones) to avoid r e s t r i c t i n g a free ranging discussion by having others f a l l i n l i n e with hi s views on a t o p i c . The second groupthink-avoidance case study concerned the development of the Marshall plan f o r the post-WWII reconstruction of Europe. The major p o l i c y development group was headed by Mr. G. Kennan. Some of the processes which were used i n t h i s s i t u a t i o n to minimize poor decision making were as - 22 -Figure 3 Schematic Outline of Components Within Groupthink Theory Preconditions 4-Groupthink Symptoms 4-Decision Process 4-Decision Outcome (high cohesiveness, concurrence seeking, s t r e s s , leadership style) (e.g., f a l s e unanimity, s e l f -censorship, mindguards, i l l u s i o n s of i n v u l n e r a b i l i t y ) (e.g., decreased information input, decreased number of suggested solutions) (poorer q u a l i t y , more often wrong) - 23 -follows: (1) Instead of e s t a b l i s h i n g a norm supporting group unity, from the beginning i t was attempted to develop a norm centered on the c r i t i c a l ques-tioning of ideas. (2) The group leader did not impose h i s own views. Rather, he presented them for c r i t i c a l evaluation by the others. The leader a t -tempted to maintain i m p a r t i a l i t y and generate genuine open debate. (3) There was a d e l i b e r a t e attempt by leaders at a l l l e v e l s , up to and including the President, to avoid placing pressure on subordinates to comply with the leader's i n i t i a l preferences. (4) Several subgroups were used to i n i t i a l l y grapple with the problem; these groups then came together again to combine t h e i r ideas and work out t h e i r d i f f e r e n c e s . S i m i l a r l y , several major groups were formed to deal with separate important components of the p o l i c y ; as before, these e f f o r t s were l a t e r reviewed and integrated, i n t h i s case by an evaluation review committee. On the basis of these two case studies, Janis (1972) provides a summary of techniques for avoiding groupthink: (1) "The leader of a policy-forming group should assign the r o l e of c r i t i c a l evaluator to each member, encour-aging the group to give high p r i o r i t y to a i r i n g objections and doubts. This p r a c t i c e needs to be reinforced by the leader's acceptance of c r i t i c i s m of hi s own judgments i n order to discourage the members from soft-pedaling t h e i r disagreements" (p. 209). This for Janis i s the most important recom-mendation, one which i s necessary before other moves are taken to avoid groupthink; i t e n t a i l s the development of a norm f o r the c r i t i c a l assessment of ideas. (2) The leader should be i m p a r t i a l rather than s t a t i n g preferences, expectations or s p e c i f i c solutions that he would l i k e to see accepted. This approach fosters the group's consideration of a wider range of a l t e r n a t i v e s . (3) Several planning and evaluation groups under d i f f e r e n t leaders should be used to work on the same problem. (4) In conjunction with the previous suggestion, a policy-making group may be s p l i t on occasion into several - 24 -subgroups that meet separately and then recombine to work out the differences i n t h e i r p o s i t i o n s . (5) Each group member should consult with trusted experts i n h i s own department or area to gain a d d i t i o n a l perspective on the issue and report the views back to the group at large. (6) At i n t e r v a l s , outside ex-perts should be i n v i t e d to present t h e i r views to the core group and to c r i -t i c a l l y question the views of the members. (7) A d e v i l ' s advocate should be appointed as a c r i t i c a l scrutineer of ideas at a l l meetings where a l t e r n a -t i v e s are being considered. (8) When the decision involves dealings with another organization or nation, s p e c i f i c time should be spent considering the various types of communications and signals from the opposing group. This would be done i n the attempt to understand t h e i r intentions and in.order to consider the various possible r i s k s associated with d i f f e r e n t i n t e n t i o n s . The group should then attempt to meet these r i s k s by developing contingency plans to deal with the outcomes of d i f f e r e n t types of intentions. Janis (1272) sug-gests possibly using a Cassandra's advocate, a term describing the Greek mythological prophetess of doom who was not believed. The r o l e of the Cassandra's advocate would be to present a worst-case analysis of the a v a i l -able information concerning the intentions of the opposing side. (9) When a preliminary consensus has been reached, a meeting should be held where mem-bers are s p e c i f i c a l l y expected to expose and discuss any remaining personal doubts that they have on the issue before making a f i n a l d e c i s i o n . When considering the fiascoes c i t e d as evidence f o r groupthink, i n con-t r a s t to the successful decisions c i t e d as evidence f o r avoiding groupthink, there may be an i m p l i c a t i o n that we should u t i l i z e a l l possible techniques for avoiding groupthink. This may be considered appropriate since these techniques are good for decision making, while groupthink i s bad for decision making. It should be noted however, that over-zealous use of these b e n e f i c i a l techniques raises the p o t e n t i a l for bogging the group down i n what may be unnecessary - 25 -complexity or time consuming e f f o r t s . In comparison to the costs of delaying a decis i o n , these e f f o r t s may have diminishing returns on increasing the qua-l i t y of the group's decision making. In discussing the r o l e of groupthink i n decision making, i t i s important to note that groupthink i s not the only a f f l i c t i o n that reduces the q u a l i t y and outcome of the decision making process. In the view of Janis (1972), low cohesive groups lacking mutual acceptance, a m i a b i l i t y or e s p r i t de corps may not display the groupthink symptoms, but may display worse decision making and produce even greater fiascoes than moderately or highly cohesive groups. There are two major problems that often beset non-cohesive groups. In some cases the group may create acquiescence i n members out of a fear of r e j e c t i o n or recrimination (whereas i n groupthink the acquiescence i s out of a more p o s i t i v e motivation to maintain unanimity or harmony among f r i e n d s ) . In other cases, a d i f f e r e n t set of motivations may dominate. For example, i f the group i s composed of members of incompatible views or members working only . as representatives of other fac t i o n s or i n t e r e s t s , rather than as members committed to the larger group, they may become entangled i n power struggles and win-lose s t r a t e g i e s . Friendship and e s p r i t de corps are necessary to lessen competitiveness, develop t r u s t , and t o l e r a t e disagreements. Janis (1972) f e e l s that what i s needed for optimal decision making are cohesive groups (defined i n terms of a t t r a c t i o n to the group rather than a lack of i n t e r n a l c o n f l i c t ) where groupthink tendencies are c o n t r o l l e d . Problems i n groupthink theory. There are two problems that may e x i s t i n groupthink theory. F i r s t , did the generation of the groupthink hypo-thesis involve c i r c u l a r reasoning? Second, under what conditions does a case study provide e i t h e r disconfirming or supporting evidence f o r the theory of groupthink? In h i s book on groupthink Janis notes that h e . i n i t i a l l y had several - 26 -dozen cases where, i n h i s opinion, the outcomes were fiascoes a t t r i b u t a b l e i n part to the group processes involved i n the decision making. He reduced the l i s t by looking f or cases where defective decisions had been made by small cohesive groups. "In other words, the fiascoes that I selected f or analysis deserved (sic) to be fiascoes because of the grossly inadequate way the policy-makers c a r r i e d out t h e i r decision making tasks" (Janis, 1972, p. 10). Now, although i t would seem that Janis was using case study as a means of hypothesis generation i t appears that he had already conceptualized what constituted good group d e c i s i o n making features (see the s i x major decision making defects Janis, 1972, p. 10). He then selected f o r case study impor-tant fiascoes that lack these f e a t u r e s — " E a c h (case):clearly meets two impor-tant c r i t e r i a . . . (1) the decision making group was cohesive and that (2) decision making was extremely d e f e c t i v e " (p. 11). Following t h i s i t appears as though Janis then looked for other consistent group processes, found eight, and c a l l e d them groupthink symptoms. At the same time he looked for consistent preconditions and found several, among them being group cohesive-ness. Then i n developing the theory of groupthink Janis s t a t e s that these preconditions cause the groupthink. In addition, the preconditions cause other group processes that are also l i k e l y to be p o o r — f o r example, the s i x defects previously mentioned above ("I assume that these s i x defects and some related features of inadequate d e c i s i o n making r e s u l t from groupthink". Janis, 1972, p. 10), F i n a l l y , i t i s f e l t that a l l of these poor d e c i s i o n making processes "increase the l i k e l i h o o d of poor outcome" (Janis, 1972, p. 12). To summarize, i t appears that the generation of the groupthink .hypo-thesis perhaps proceeded through the following steps. (1) The i d e n t i f i c a t i o n of poor d e c i s i o n making processes. '(2) The s e l e c t i o n of cases for analysis r e s t r i c t e d to cohesive groups where the decision making processes are poor. (3) The i d e n t i f i c a t i o n of consistent group processes among the cases and the - 27 -l a b e l l i n g of these processes as groupthink symptoms. (4) The i d e n t i f i c a t i o n of consistent preconditions among the cases, noting that group cohesiveness (among others) i s a major precondition. (5) The statement that these pre-conditions cause the groupthink symptoms, the other poor decision making pro-cedures and the poor decision -outcomes. To the present author i t would seem that t h i s type of hypothesis generation involves c i r c u l a r reasoning to some extent. The preconditions necessary for groupthink form an important part of Janis' theory. A major precondition i s group cohesiveness, but he had i n i -t i a l l y r e s t r i c t e d h i s s e l e c t i o n of cases for intensive study to those i n -volving small cohesive groups. Having done that, the extent to which the groupthink symptoms might also be found i n non-cohesive groups i s not c l e a r . In t h i s sense, the importance of group cohesiveness as a necessary precon-d i t i o n for the groupthink symptoms i s not f i r m l y established. When one i s faced with assessing the q u a l i t y of a si n g l e d e c i s i o n , then the assessment of the d e c i s i o n process, including perhaps the judgment of experts regarding the choice made i n comparison to the a l t e r n a t i v e s , would seem to be the most reasonable approach. Due to the " i n t e r v e n t i o n " of c i r -cumstances between the time that a d e c i s i o n i s made and the time that i t i s implemented, p a r t i c u l a r l y i n cases of organizational decisions, i t would be hazardous to judge a si n g l e decision by i t s outcome. The intervening c i r -cumstances may allow a poor decision to succeed and may cause a wise d e c i s i o n to f a i l . Contingency planning i n d e c i s i o n making helps reduce the p o s s i b i -l i t y of circumstances causing good decisions to go awry but does not eliminate i t . At the same time, i f one assumes that there i s a random d i s t r i b u t i o n across decisions of intervening circumstances having negative e f f e c t s upon decision outcome, and i f one has a large number of decisions that can be - 28 -studied, then good de c i s i o n making processes should on the average produce better outcomes than poor decision making processes. Thus, an a l t e r n a t i v e method for generating hypotheses concerning group decision making processes would be to select " f i a s c o e s " or other major decisions made by groups, based on c r i t e r i a such as t h e i r seriousness, importance, or s o c i a l s i g n i f i c a n c e . Then one could look to see what type of group processes distinguished the good decisions (as defined by the outcomes) from the bad decisions. Using t h i s type of "discriminant a n a l y s i s " approach i n studying group processes may y i e l d a firmer basis for making judgments concerning the correctness of the i n i t i a l assumptions about good de c i s i o n making features. Janis himself i n d i r e c t l y comments on the p o s s i b i l i t y of using an a l t e r -native method to generate hypotheses. In discussing the question of "how widespread i s groupthink?", Janis (1972) states, "at present we do not know what percentage of a l l natio n a l fiascoes are a t t r i b u t a b l e to groupthink... A l l that can be said from the h i s t o r i c a l case studies I have analyzed so far i s that groupthink tendencies sometimes play a major r o l e i n producing large-scale fiascoes. In order to estimate how large the percentage might be for various types of decision-making groups, we need in v e s t i g a t i o n s of a v a r i e t y of p o l i c y decisions made by groups of executives who have grossly m iscal-culated the unfavourable consequences of t h e i r chosen course of a c t i o n " (pp. 192-193). In a statement of f a i t h , he suggests that probably " c l e a r symptoms of groupthink are present i n at least a s u b s t a n t i a l minority of a l l miscal-culated executive decisions—governmental and nongovernmental, American and forei g n " (p. 196). Furthermore, he expects that the analysis of any one group's decisions over a length of time w i l l show that "a s i z a b l e percentage of that group's decision errors probably w i l l prove to be a t t r i b u t a b l e to groupthink tendencies, i f the group i s moderately or highly cohesive" (p.197) .. Perhaps t h i s type of research should have been done while the groupthink - 29 -hypothesis was being formulated. We can now proceed to the problem concerning' the p o s s i b i l i t y .of. d i s con-firming the groupthink hypothesis. Janis (1972) states that not a l l fiascoes are produced by.groupthink; they.may be the r e s u l t of ignorance, erroneous i n t e l l i g e n c e information, fatigue, etc. S i m i l a r l y , as noted e a r l i e r with respect to p r a c t i c a l outcomes, groupthink i s f e l t to increase the l i k e l i h o o d of decision making er r o r s , but at the same time w i l l not always lead to a f i a s c o . Sometimes, fort u i t o u s events such as the other side's greater stu-. p i d i t y may prevent d i s a s t e r . In h i s book, Janis (.1972) considered that the poor decision making processes underlying fiascoes such as the Bay of Pigs provided supporting evidence for h i s groupthink theory. However, Janis (1972, p. 12) claimed that even i f an event such as the Bay of Pigs decision had turned out to be successful, i t would s t i l l support the groupthink hypo-thesi s . In t h i s case, the groupthink hypothesis would be supported whether the Bay of Pigs was a success or a f i a s c o , simply because the group de c i s i o n making process was poor. If one accepts the argument raised e a r l i e r that across a seri e s of decisions good de c i s i o n making processes should, on the average, lead to better outcomes than poor decision making processes then i t would seem tenuous to claim support for the groupthink theory independent of the decision outcome. The groupthink hypothesis would be more us e f u l i f i t s ultimate v a l i d i t y were measured by an external c r i t e r i o n such as, " i f one considered a large number of decisions made under groupthink conditions, there would be an increased p r o b a b i l i t y of the outcome of any one decision being a f a i l u r e . " I f one accepts the concept of an external v a l i d i t y c r i -t e r i o n such, as "these processes w i l l tend to lead to wrong dec i s i o n s " then c l e a r l y i f the Bay of Pigs decision had been successful i t could not c o n s t i -tute support for the groupthink hypothesis. In terms of future research, Janis (1972, p. 197; footnote 2, p. 237) - 30 -proposes that the appropriate technique for t e s t i n g the groupthink hypothesis i s through the comparative f i e l d study of decisions and decision making groups. An a d d i t i o n a l p o s s i b i l i t y e x i s t s i n simulations inv o l v i n g high ranking government executives, even though he f e e l s that these simulations do not create the severe stress and need for s o c i a l support that i s generated by r e a l l i f e c r i s e s . Most research on conformity i n decision making i s based upon groups of strangers who meet once and do not expect to meet together again. In contrast, "to understand the predispositions conducive to group-think, we need studies of groups that meet together for many weeks and work on decisions to which each member w i l l be committed" (Janis, 1972, p. 191). In 1977, f i v e years l a t e r , Janis s t i l l advocated the use of f i e l d studies to test the p r i n c i p l e s of groupthink. However, as noted i n an e a r l i e r quota-t i o n , he now saw groupthink as being a possible part of most normal groups i n everyday s i t u a t i o n s , "... we propose to apply the model, as w e l l , to the more commonplace decisions made by executives i n routine meetings ..." (Janis and Mann, 1977, p. 75). This should make groupthink a much more e a s i l y r e -searched topic than i f one could only study groups that had worked together for weeks on a decision. Case Studies and Experimental Research on Groupthink Since the time that Janis f i r s t published h i s theory on groupthink i n decision making i n 1972, there have been two case studies and three experi-mental studies that have attempted to test h i s ideas. These studies have provided some support f o r h i s theory but they have been l i m i t e d by a number of serious problems i n t h e i r design. I w i l l s t a r t with the two case studies and then proceed to the experimental studies. Groupthink i n the Watergate c r i s i s . The two case studies were concerned with the Nixon e l i t e responsible for the Watergate coverup (Green and - 31 -Conolley, 1974, c i t e d i n Janis and Mann, 1977; Raven, 1974). The paper by Green and Conolley was presented at a conference but was never published. The paper by Raven was given as the P r e s i d e n t i a l Address to the Society f o r the Psychological Study of S o c i a l Issues i n August of 1974. In the opinion of Janis and Mann (1977, footnote 7, p. 423) the evidence presented i n these two studies as supporting groupthink i s not e n t i r e l y consistent, e s p e c i a l l y with respect to the cohesiveness of Nixon's group. Since the unpublished paper by Green and Conolley covers the same case as Raven's (1974) study, and both were negatively commented upon by Janis and Mann (1977), the following section w i l l r e f e r only to the published a r t i c l e by Raven. Raven's (1974) study was concerned with explaining the i n i t i a l d e cision to commit the Watergate break-in and the l a t e r s e r i e s of decisions involved i n covering up the Watergate episode. Raven attempted to demonstrate the existence of the necessary preconditions for groupthink, e.g. , group cohe-siveness, a s s e r t i v e leadership and i s o l a t i o n . The i s o l a t i o n of the Nixon group does seem to be f a i r l y w e ll established. Nixon, as the leader of the o v e r a l l group, was c e r t a i n l y a s s e r t i v e i n s t r e s s i n g h i s "attacking" s t y l e of dealing with the opposition, but from Raven's account i t i s not clear exactly how as s e r t i v e he was concerning h i s views on s p e c i f i c plans. The extent to which the group was cohesive i s also not c l e a r . Raven notes that there were two major antagonistic factions within the general group ( M i t c h e l l vs. Haldeman), but he f e e l s that there was s t i l l a strong sense of cohesiveness since a l l the members very strongly wanted to belong to the general Nixon-group. This a t t r i b u t e of a t t r a c t i o n to the group i s only one measure of group cohesiveness, and must be considered i n the context of other measures such as the extent of interpersonal friendship and the degree of i n t e r n a l c o n f l i c t . The lack of friendships across the two factions and the i n t e r -personal c o n f l i c t suggests that the o v e r a l l group was not highly cohesive. - 32 -In addition, i t i s predicted by Janis' theory that a group that meets a l l the groupthink preconditions s t r i v e s f o r concurrence i n the decision making; however, i t appears from Raven's account that the s t r i v i n g i n Nixon's group was usually for one f a c t i o n to win out at the expense of the other f a c t i o n , rather than attempting to achieve agreement on a decision. Thus, i t appears that the existence of the necessary preconditions i s only weakly established. An a d d i t i o n a l point i s that groupthink theory deals with group de c i s i o n making processes; i t i s not clear to what extent the i n i t i a l burglary de-c i s i o n and subsequent cover-up decisions were the product of what we would normally consider to be a group decision making process. Raven (1974) next looks at the evidence concerning the existence of s i x groupthink symptoms. There i s reasonable evidence to support the existence of an i l l u s i o n of morality, conformity pressures, the suppression of personal doubts and the existence of self-appointed mindguards. The evidence i s much more tenuous concerning the existence of i l l u s i o n s of i n v u l n e r a b i l i t y and unanimity. Given the fact that evidence concerning the existence of the necessary preconditions was weak and the f a c t that i t should be p o s s i b l e f o r the appropriate non-groupthink preconditions to produce some of the same group processes as those produced by groupthink, f i n d i n g several groupthink symptoms associated with a f i a s c o cannot be taken as strong proof supporting the theory of groupthink i n decision making. Overal l , the evidence i s weak that the theory of groupthink explains the behaviour of the Nixon e l i t e during the Watergate c r i s i s . Raven (1974) notes that there are many other explanations f o r the Watergate behaviour, such as the strong norm to be ha r d - h i t t i n g and merciless i n dealing with„opponents, or the processes of re c e i v i n g authorization from superiors, r o u t i n i z a t i o n of the work, and dehumanization which can lead to immoral decisions. Groupthink may form part of the explanation, in.conjunction with other f a c t o r s , but the - 33 -evidence i s not strong that i t forms the s i n g l e explanation or even a major explanation of the behaviour of the Nixon group during Watergate. (A s i m i l a r c r i t i c i s m has been made by Katz and Kahn (1978, pp. 514-515) concerning the rol e of groupthink as the sole explanation of decision errors such as the Bay of Pigs invasion of Cuba.) Flowers' research on groupthink. In 1977, Flowers performed the f i r s t experimental test of the groupthink hypothesis. Her subjects were under-graduate students i n a r o l e playing s i t u a t i o n where they were school adminis-t r a t o r s attempting to deal with the case of an e l d e r l y woman teacher who could no longer handle the d i s c i p l i n a r y problems i n her c l a s s . Flowers mani-pulated two independent v a r i a b l e s , leadership s t y l e (open versus closed) and group cohesiveness (high versus low). An open s t y l e of leadership stressed the importance of reaching a wise decision through discussing a l l the sug-gested solutions, whereas the closed leadership s t y l e stressed unanimity and the leader openly favoured h i s own preferred s o l u t i o n . Group cohesiveness was defined on the basis of the i n d i v i d u a l ' s a t t r a c t i o n to the group, where high cohesiveness was expected to be a property of groups of acquaintances and low cohesiveness was expected to be found among groups of strangers. One i n d i v i d u a l out of each group of four subjects was trained to be eith e r an open or closed s t y l e leader. A l l of the subjects were given sheets o u t l i n i n g the school administrator r o l e that they were to play during the session. Also contained on these sheets were s i x or seven fa c t s about the case that were not given to any other member. Each group was asked to a r r i v e at a so l u t i o n within approximately 30 minutes. Flowers (1977) f e l t that t h i s design incorporated those aspects of a decision making s i t u a t i o n outlined by Janis (1972) as precursors f o r groupthink: c o n t r o v e r s i a l points of view, a moral dilemma concerning what to do about the teacher, a sense of c r i s i s since an immediate decision was needed, c o n f l i c t with an outside group (a - 34 -m i l i t a n t teachers' union), and the f a c t that no s i n g l e member knew a l l of the relevant information. The sessions were tape-recorded. The actual time taken to reach a decision was between 15 and 45 minutes, following which the subjects f i l l e d i n a post-session questionnaire and were debriefed. The questionnaire con-tained f i v e 8-point bipolar scales that considered: (a) the i n d i v i d u a l ' s pre- and post-session agreement with the team de c i s i o n , (b) the i n d i v i d u a l ' s perceived freedom to express personal opinions, (c) the person's willingness to role-play another s i t u a t i o n with the same team, and (d) the degree to which the team was a t t r a c t i v e to the i n d i v i d u a l . The subjects were also asked to rank order members with respect to t h e i r influence on the d e c i s i o n , and to i n d i c a t e the number of team members that they had known before the experiment. Flowers (1977) predicted that groups engaged i n groupthink would: (1) suggest a smaller number of solutions to,the problem, (2) consult fewer out-side sources of information and (3) consider l e s s information before making the decision. O v e r a l l , Flowers has attempted to test J a nis' groupthink theory by designing a 2 x 2 between groups study where the conditions for groupthink were manipulated to be e i t h e r present or absent. Unfortunately, while she tested for the q u a l i t y of the decision making process, she did not test f or the presence or absence of the groupthink symptoms nor the q u a l i t y of the d e c i s i o n outcome. A manipulation check indicated that the leaders had role-played cor-r e c t l y the open and closed s t y l e s of leadership. S i m i l a r l y , a check on group cohesiveness indicated a s i g n i f i c a n t d i f f e r e n c e between high and low cohesive groups i n terms of the number of members i n the group that knew each other previous to the experiment and also i n terms of the attractiveness of the group. - 35 -The r e s u l t s indicated that the open leadership s t y l e teams proposed a s i g n i f i c a n t l y greater number of solutions than the closed leadership s t y l e groups. In contrast, the number of solutions did not vary s i g n i f i c a n t l y be-tween groups d i f f e r i n g i n cohesiveness, nor were there cohesiveness-leadership s t y l e i n t e r a c t i o n s . The o v e r a l l number of facts considered from the r o l e sheets was s i g n i f i c a n t l y r e l a t e d to leadership s t y l e but not to group cohe-siveness, nor were there any i n t e r a c t i o n s . Open s t y l e leadership groups considered s i g n i f i c a n t l y more of these outside sources of information before reaching t h e i r decision. Last, i t was found that open s t y l e leadership groups considered s i g n i f i c a n t l y more of the r o l e sheet fa c t s p r i o r to a r r i v i n g at a consensus on the s o l u t i o n than did closed groups. In contrast, the closed groups considered s i g n i f i c a n t l y more of the r o l e sheet fa c t s i n the discussion period a f t e r they had a r r i v e d at a consensus but before they had indicated to the experimenter that they were f i n i s h e d . Again, i n terms of when the r o l e sheet fa c t s were discussed there was no s i g n i f i c a n t d i f f e r e n c e between high and low cohesive groups, nor were there i n t e r a c t i o n e f f e c t s . Flowers (1977) concluded that with respect to group cohesiveness and leadership s t y l e , the l a t t e r had the more important e f f e c t upon groupthink. To account for the lack of findings with respect to cohesiveness, she notes that there may be several explanations. F i r s t the l e v e l of cohesiveness i n her groups was considerably lower than the cohesiveness of the groups studied by Janis (1972); second, the closed leadership s t y l e involved the leader's advocacy of a preferred s o l u t i o n and t h i s may have overshadowed the e f f e c t s of group cohesiveness; and t h i r d , the theory of Janis (1972) may need to be revised, downplaying the r o l e of cohesiveness. She points out that i n the groups that Janis (1972) studied, the leaders had a great deal of s o c i a l power over the other members ( i . e . , i n terms of a b i l i t y to reward and punish, and a b i l i t y to exert legitimate, expert, and possibly even referent power). - 36 -She f e e l s that t h i s a d d i t i o n a l v a r i a b l e of leadership power might be very important to consider i n r e v i s i n g Janis' theory and might be a prime f a c t o r i n accounting for the increased groupthink i n highly cohesive groups. Flowers (1977) also notes that the role-played c r i s i s s i t u a t i o n was far from the magnitude of the c r i s e s studied by Janis (1972). Despite these d i f f e r -ences, she f e l t that groupthink did occur since her r e s u l t s supported some of the predictions made by Janis' (1972) groupthink theory. In commenting upon t h i s research by Flowers (1977), several questions need to be considered. F i r s t of a l l , did the experiment meet the conditions outlined by Janis (1972) as necessary to produce groupthink? Second, was groupthink produced? And f i n a l l y , did the open leadership s t y l e reduce groupthink? With respect to the f i r s t question, Janis (1972; Janis & Mann, 1977) contends that the prime conditions for groupthink e x i s t when there i s a cohe-sive group whose members s t r i v e to achieve concurrence concerning a s t r e s s f u l or d i f f i c u l t d e c i s i o n . The experiment did appear to meet the necessary pre-conditions i n terms of leadership s t y l e and cohesiveness. However, i t was a r o l e - p l a y i n g s i t u a t i o n , where a time l i m i t was not enforced and the issue was of l i t t l e personal or s o c i a l s i g n i f i c a n c e . It would have been valuable to check on the subjects' stress l e v e l or motivation to do the task. Perhaps most important of a l l from the perspective of groupthink theory, there was no d i r e c t evidence from the experiment that cohesiveness or leadership s t y l e led to e x p l i c i t concurrence seeking on the part of the group members. Con-currence seeking was only proposed to e x i s t as a part of the manipulated leadership s t y l e wherein closed leaders indicated t h e i r preferred s o l u t i o n and stressed agreement on a s o l u t i o n . While Flowers (1977) should be applauded for taking a necessary f i r s t step i n attempting to bring J a n i s ' (1972) proposals i n t o the laboratory for t e s t i n g , the existence of a l l the - 37 -necessary groupthink preconditions was not c l e a r l y demonstrated i n her study. Was groupthink produced, and d i d the open s t y l e of leadership reduce groupthink? There are two ways of answering t h i s question. As indicated by Flowers (1977, p. 890) i n her p r e d i c t i o n s , the existence of groupthink would be demonstrated by a s i g n i f i c a n t i n t e r a c t i o n between cohesiveness and leader-ship s t y l e . However, i n her r e s u l t s there were no s i g n i f i c a n t i n t e r a c t i o n e f f e c t s on any of the dependent measures. In addition, Janis (1972) outlines eight major symptoms of groupthink. In Flowers' study there was no t e s t made for the existence of any of these symptoms. I t appears, therefore, that to the best of our knowledge groupthink was not produced by the experimental conditions. I f groupthink did not e x i s t , then we surely cannot know i f the open leadership s t y l e would have reduced groupthink. What then i s the i n t e r p r e t a t i o n to be placed upon Flowers' results? The f i n a l outcome of Flowers' (1977) research demonstrated that leadership s t y l e , but not her manipulation of group cohesiveness, nor the i n t e r a c t i o n between these two f a c t o r s , affected the process of a r r i v i n g a decision. Contrary to the i n t e r p r e t a t i o n placed upon these r e s u l t s by Flowers, i t seems to the pre-sent author that there has been a f a i l u r e to demonstrate the existence of groupthink. Instead, i t appears as though Flowers (1977) has demonstrated the influence of leadership factors upon the q u a l i t y of the decision making process. Courtrlght's research on groupthink. In addition to the work of Flowers, John Courtrlght (1978) has also attempted a laboratory i n v e s t i g a t i o n of the groupthink phenomenon. Unfortunately, Courtright did not get o f f to a good s t a r t since i n h i s introduction he..confused things somewhat by t r y i n g to convince the reader that several of the groupthink symptoms (e.g., the as-sumption of morality, the i l l u s i o n of i n v u l n e r a b i l i t y , and the existence of mindguards) were instead antecedent conditions for groupthink. - 38 -Courtright had 96 male and female undergraduates as subjects i n four-person experimental groups. Subjects sat around a table facing a videotape camera and were given a 10 minute warm-up discussion time. Following the warm-up period the students were asked to produce a sing l e recommendation on the problem of "what i s the best method for r e c r u i t i n g new students to the University of Iowa?" Courtright was interested i n manipulating two v a r i a b l e s . One v a r i a b l e was the degree of cohesiveness of the group and the second v a r i a b l e was the extent to which the group f e l t that there were l i m i t s upon the type of acceptable solutions. A leader a s s e r t i v e about h i s or her point of view i s the usual source of such l i m i t s i n groupthink theory. However, Courtright attempted to impose the l i m i t s by manipulating the problem solving d i r e c t i o n s . There were three d i f f e r e n t types of l i m i t s i n hi s study: (1) a "freed" con-dition—where the i n s t r u c t i o n s implied ample time and stated that the.best solutions are the r e s u l t of competition among many incompatible ideas; (2) a " l i m i t e d " condition—where the i n s t r u c t i o n s implied i n s u f f i c i e n t time and stressed cooperation i n deriving a s o l u t i o n from a small number of i n i t i a l ideas; (3) a "no i n s t r u c t i o n s " c o n d i t i o n — a control group that was t o l d of the time a v a i l a b l e and asked to proceed. The second independent v a r i a b l e was group cohesiveness, which was mani-pulated through a t t i t u d e s i m i l a r i t y and propinquity. Following the warmup session, subjects i n the low cohesiveness condition were s p l i t - u p and r e -formed i n t o new groups, whereas the high cohesiveness groups remained i n t a c t . The high cohesiveness groups were t o l d that the r e s u l t s of an e a r l i e r questionnaire that the members completed showed that they were very s i m i l a r , that they should be very compatible and should do w e l l on the task. Low cohesiveness groups were t o l d that t h e i r questionnaire r e s u l t s indicated the opposite. - 39 -It was predicted that groups i n the high cohesiveness, l i m i t e d condition (groupthink condition) would develop fewer a l t e r n a t i v e solutions and would generate more statements of agreement and fewer statements of disagreement than would groups i n the other conditions. Second, i t was predicted that subjects i n the groupthink condition would produce decisions of i n f e r i o r q u a l i t y . Courtrlght (1978) found that on the measures of cohesiveness given to the groups before they attempted to solve the problems, there were s i g n i f i -cant differences i n the expected d i r e c t i o n between the high cohesiveness and low cohesiveness groups. However, h i s measures indicated that these s i g n i -f i c a n t differences between the groups i n cohesiveness had disappeared by the end of the experimental session. Courtright used a MANOVA (multivariate analysis of variance) to test h i s predictions on the three process measures; he found no s i g n i f i c a n t main or in t e r a c t i o n e f f e c t s . However, since the i n t e r a c t i o n e f f e c t approached s i g n i f i -cance (p <.07), he further analyzed i t by conducting a univariate ANOVA on each of the three dependent measures. He found that the major source of variance was associated with statements of disagreement, while the va r i a b l e s concerning the number of solutions proposed and the statements of agreement contributed n o n - s i g n i f i c a n t l y to the i n t e r a c t i o n e f f e c t . A simple e f f e c t s analysis of the "statements of disagreement" ANOVA i n t e r a c t i o n term indicated that the only s i g n i f i c a n t simple e f f e c t was that high cohesive, l i m i t e d groups produced s i g n i f i c a n t l y less disagreement than did the low cohesive, l i m i t e d groups. The q u a l i t y of the decisions was independently rated by f i v e graduate students each using the same f i v e scales (effectiveness, f e a s i b i l i t y , creat-i v i t y , s i g n i f i c a n c e , and competence). A MANOVA was performed and orthogonal contrasts applied to the f i v e scales. Neither the main e f f e c t s nor the - 40 -i n t e r a c t i o n between the cohesiveness and experimental i n s t r u c t i o n s v a r i a b l e s achieved s i g n i f i c a n c e . The l i n e a r component of the i n t e r a c t i o n approached s i g n i f i c a n c e (p <.12). (It should be noted that the use of orthogonal con-t r a s t s i s questionable since i t i s d i f f i c u l t to consider the f i v e scales to be orthogonal questions. Non-orthogonal comparisons would y i e l d a more con-servative test of t h i s " q u a l i t y - o f - t h e - d e c i s i o n " question that, i n turn, would reduce the " s i g n i f i c a n c e " of the l i n e a r component of the i n t e r a c t i o n term.) In h i s discussion of the r e s u l t s , Courtright suggests that he has pro-vided a t e n t a t i v e a f f i r m a t i o n of the existence of groupthink and that i t can be studied i n a laboratory situation... In addition, he suggests that the absence of disagreement among group members may be the most important i n d i -c a tion of the groupthink process. This absence of disagreement would i n d i c a t e a lack of c r i t i c a l analysis of the factors discriminating poor q u a l i t y solutions from good q u a l i t y solutions. There are a number of problems associated with the Courtright study. F i r s t , i n terms of e s t a b l i s h i n g the existence of the necessary preconditions for groupthink, i t would be expected that the group would display a high l e v e l of motivation or s t r e s s . However, there was no check on stress l e v e l or motivational l e v e l i n the experiment. This may be a c r u c i a l issue, since the problem that the students were dealing with was l i k e l y to be of low personal importance and i n t e r e s t . In a d d i t i o n , by the end of the experimental session the groups no longer d i f f e r e d i n terms of cohesiveness. Since i t was pre-dicted that the i n t e r a c t i o n of cohesiveness with the " l i m i t i n g " i n s t r u c t i o n s would create groupthink, f i n d i n g that the e f f e c t of one of the independent v a r i a b l e manipulations had disappeared during the experiment could only reduce, i f not eliminate, the p o s s i b i l i t y of generating groupthink. Second, as mentioned i n the discussion of the Flowers' (1977) research, there are two ways to test for the existence of groupthink. The f i r s t way of - 41 -t e s t i n g f or groupthink i s to manipulate the factors outlined by!Janis (1972) as preconditions for the production of groupthink. The i n t e r a c t i o n of these factors should produce the d e f i c i e n t processes and outcomes associated with groupthink. The second way i s to t e s t for the existence of a s p e c i f i c set of group processes c a l l e d groupthink symptoms. In Courtrlght's (1978) study :the i n t e r a c t i o n e f f e c t s approached s i g n i f i c a n c e but were no n - s i g n i f i c a n t , and there was no test made for the existence of the groupthink symptoms. Third, the o v e r a l l lack of s i g n i f i c a n t findings (the two MANOVA's were non- s i g n i f i c a n t , only one out of three process measures was s i g n i f i c a n t , and none of the f i v e q u a l i t y of d e c i s i o n outcome measures were s i g n i f i c a n t ) pro-vides, at best, weak support for groupthink theory. In summation, the p o s i t i v e assessment by Courtrlght of the r e s u l t s as supporting the theory of groupthink, and h i s i n t e r p r e t a t i o n that the number of disagreements may be the best discriminator between groupthink and non-groupthink groups, i s premature. Tetlock's research on groupthink. A d i f f e r e n t type of study was under-taken by Tetlock (1979). He applied a content analysis to the p u b l i c s t a t e -ments of the key decision makers i n the groupthink and non-groupthink c r i s e s studied by Janis (1972). He predicted that the p u b l i c statements issued by groupthink decision makers would: (1) be more s i m p l i s t i c and (2) show more p o s i t i v e attitudes towards t h e i r own group and more negative attitudes towards outgrpups ( i . e . , domestic and p o l i t i c a l opponents). A technique for assessing i n t e g r a t i v e complexity (the extent to which information i s d i f f e r e n t i a t e d and integrated) was used to test the f i r s t hypothesis (see Suedfeld, 1978). Osgood's (1959) evaluative assertion analysis was used to test the second p r e d i c t i o n . The content analysis data were obtained by s e l e c t i n g 12 paragraphs from a r c h i v a l records of statements made by the President of the U.S.A. and the - 42 -Secretary of State concerning each of the following f i v e c r i s e s : the Mar-s h a l l plan (also analyzed were 12 paragraphs from the Under-Secretary of State, George Marshall); the Cuban m i s s i l e c r i s i s ; the invasion of North Korea; the Bay of Pigs invasion; and the Vietnam war e s c a l a t i o n . The Marshall plan and the Cuban m i s s i l e c r i s i s were selected as examples of c r i s e s where groupthink was avoided, while the remaining cases were selected as examples of groupthink i n decision making. The case of Prime Minis t e r Chamberlain's decision making and the Pearl Harbour c r i s i s studied by Janis were not used by Tetlock. The r e s u l t s indicated that, as predicted, the decision makers i n the non-groupthink groups made statements that were more i n t e g r a t i v e l y complex than did the decision makers under the groupthink conditions. S i m i l a r l y , as predicted, the groupthink de c i s i o n makers evaluated t h e i r own groups much more p o s i t i v e l y than d i d the non-groupthink decision makers. However, the t h i r d p r e d i c t i o n was not supported. There was no diffe r e n c e between the groupthink and non-groupthink decision makers i n t h e i r evaluation of domestic and foreign opponents (the out-groups). A discriminant analysis indicated that i n t e g r a t i v e complexity was the major factor which could be used to d i s -tinguish public statements under groupthink c r i s e s from public statements under non-groupthink c r i s e s . An i n t e r e s t i n g a d d i t i o n a l f i n d i n g concerned the consistency of the decision makers across groupthink and non-groupthink s i t u a t i o n s . Truman, Acheson and Kennedy showed changes i n t h e i r statements that indicated s h i f t s i n t h e i r thinking processes depending upon whether they were i n groupthink or non-groupthink s i t u a t i o n s . President Kennedy appeared to change the most dramatically. In contrast, the statements of Rusk during three d i f f e r e n t c r i s e s showed no s i g n i f i c a n t s h i f t s i n cognitive s t y l e as a r e s u l t of being engaged i n non-groupthink and groupthink s i t u a t i o n s . Thus, there i s preliminary evidence that some decision makers eit h e r react more - 43 -strongly to the pressures of the groupthink s i t u a t i o n or are able to learn from being a groupthink s i t u a t i o n and modify t h e i r behaviour to avoid such de c i s i o n making problems i n future c r i s e s . In summary, the evidence found by Tetlock provides reasonable support f o r the theory of groupthink. However, there are two l i m i t a t i o n s which should be borne i n mind. These are the basic problems associated with c o r r e l a t i o n a l studies. F i r s t , Tetlock i s taking the f i n a l r e s u l t of the p u b l i c statements as i n d i c a t i v e of a s p e c i f i c type of group process c a l l e d groupthink. However, other factors such as personality t r a i t s , or s i t u a t i o n a l v a r i a b l e s l i k e im-pression management are also l i k e l y to have contributed to the nature of the public statement. Since these v a r i a b l e s are not c o n t r o l l e d f o r , i t i s impossible to make statements causally l i n k i n g the groupthink process to the pu b l i c statement. (Tetlock has chosen one p o t e n t i a l t h i r d v a r i a b l e expla-nation, propaganda s t r a t e g i e s , and dealt with i t as the a l t e r n a t i v e and competing explanation; i n f a c t , there may be numerous other t h i r d v a r i a b l e explanations possible.) Second, with a c o r r e l a t i v e study c a u s a l i t y i s not demonstrated to be u n i d i r e c t i o n a l . I t i s t h e o r e t i c a l l y possible that t h i s i s an i n t e r a c t i v e system where groupthink leads to p u b l i c statements showing c h a r a c t e r i s t i c s of poor decision making and early p u b l i c statements help com-mit the leader to preliminary posi t i o n s and decisions that are poor; thus, the process of public commitment, as w e l l as groupthink, may lead to poor decision making. Figure 4 provides an o u t l i n e of the s i t u a t i o n that may have existed. In conclusion, i n the context of the research presently a v a i l a b l e , the work of Tetlock provides the strongest supporting evidence for the groupthink theory. At the same time though, given the d i f f i c u l t i e s noted above, the support i s of a l i m i t e d nature and experimental work needs to be done to provide a stronger test of theory. - 44 -Figure 4 Outline of P o t e n t i a l Influences on the Public Statements used i n Tetlock's (1979) Research A. Character t r a i t s (eg., conceptual complexity) S i t u a t i o n a l -determinants of group processes Groupthink or the lack of group-think S i t u a t i o n a l f a c tors (eg., impression management) -> Private decision <-making and state-ments i n the group A -> Public statements - 45 -To summarize the r e s u l t s of a l l the studies bearing on the groupthink hypothesis, with the exception of Tetlock's (1979) reanalysis of the o r i g i n a l groupthink cases, there i s l i t t l e evidence to support the existence of group-think as proposed by Janis (1972; Janis and Mann, 1977). Proposal for the Present Study There has been no experimental study that has used decision makers from a subject population perhaps more i n keeping with Janis' (1972) cases (for example, personnel from government or corporate organizations); nor has there been a study that has tested f o r the existence of the groupthink symptoms that are postulated by groupthink theory. The present study w i l l therefore attempt to correct or avoid the problems described i n the e a r l i e r research and at the same time look f o r the existence of groupthink symptoms within a sample of government and corporate decision makers. See Figure 5 for an outline of the design of-the study. An attempt w i l l be made to manipulate the antecedent conditions of cohe-siveness, and stress or motivation, as w e l l as leadership s t y l e . S p e c i f i -c a l l y , i t i s predicted that the antecedent conditions present groups w i l l be more cohesive, more motivated and under greater stress than the antecedent conditions absent groups. I t i s also predicted that group leaders i n the assertiveness conditions w i l l place more emphasis upon t h e i r own views and state these views e a r l i e r i n the discussion than w i l l group leaders i n the non-assertiveness conditions. It i s hypothesized that the i n t e r a c t i o n between leadership s t y l e arid the cohesiveness/stress antecedent conditions w i l l produce groupthink. Group-think i n turn i s predicted to r e s u l t i n the existence of groupthink symptoms, a decrease i n the q u a l i t y of the dec i s i o n making processes, and a decrease i n the q u a l i t y of the actual decisions reached. S p e c i f i c a l l y , i t i s predicted - 46 -Figure 5 Design for Testing the Groupthink Hypothesis Independent Variable Antecedent Conditions (Conditions Present) High Cohesiveness/High Stress/High Motivation High Time Pressure (Conditions Absent) Low Cohesiveness/Low Stress/Low Motivation/Low Time Pressure Independent Variable Leadership Style Group 1 non-assertive; leadership Group 2 as s e r t i v e leadership Group 3 non-as s er t ive leadership Group 4 assertive leadership Dependent Variable Groupthink Symptoms Dependent Variable Decision Process Dependent Variable Decision Outcome - 47 -that i n the MANOVA tests of the groupthink symptoms, the group process v a r i -ables, and the decision outcome measures, the presence of groupthink w i l l be demonstrated by the presence of s i g n i f i c a n t i n t e r a c t i o n terms. F i r s t , i t i s predicted that there w i l l be an i n t e r a c t i o n between leadership s t y l e and the antecedent conditions such that when the antecedent conditions are present, an assertive s t y l e of leadership w i l l produce groupthink symptoms, whereas a non-assertive s t y l e of leadership w i l l produce fewer or no groupthink symp-toms. When the antecedent conditions are absent, i t i s predicted that there w i l l be no differences i n the production of groupthink symptoms under the two leadership s t y l e s (see Figure 6). In terms of the group processes, i t i s predicted that when the antecedent conditions are present groups under asser-t i v e leaders w i l l be rated as performing poorer on the group process measures than groups under non-assertive leaders. However, when the antecedent con-d i t i o n s are absent i t i s predicted that there w i l l be no differences on these measures between groups under as s e r t i v e or non-assertive leaders. This i n t e r -action e f f e c t i s outlined i n Figure 7. F i n a l l y , i n terms of the decision outcome data, i t i s predicted that an i n t e r a c t i o n w i l l e x i s t between leader-ship s t y l e and the antecedent conditions. When the antecedent conditions are present, groups under a s s e r t i v e leaders w i l l produce decisions that w i l l be rated by experts as i n f e r i o r to the decisions produced by groups under non-assertive leaders. In contrast, when the antecedent conditions are absent, there should be no differences i n the ratings of decision q u a l i t y under the assertive or non-assertive leaders (see Figure 8). Method Subjects The subjects p a r t i c i p a t i n g i n the study were s t a f f members of the - 48 -Figure 6 Dependent Variables: Groupthink Symptoms Hypothesized Interaction High Degree of Symptoms Present Low Degree of Symptoms | Present Antecedent Conditions Present Antecedent Conditions Absent Non-Assertive Leadership Style Assertive Leadership Style - 49 -Figure 7 Dependent Variables: Quality of the Decision Processes Hypothesized Interaction Processes High i n Quality Processes Low i n Quality + Antecedent Conditions Present Antecedent Conditions Absent Npn-Assertive Leadership Style Assertive Leadership Style - 50 -Figure 8 Dependent Variables: Quality of Decision Outcome Hypothesized Interaction Outcomes High i n Quality Outcomes Low i n Quality Antecedent Conditions Present Antecedent Conditions Absent Non-Assertive Leadership Style Assertive Leadership Style - 51 -following organizations (the number of p a r t i c i p a n t s i s indicated i n brackets): B r i t i s h Columbia (B.C.) Attorney General's Department: Corrections, S t a f f Development (11), B.C. P o l i c e Academy (8); B.C. Department of Health, Mental Health Branch: Port Coquitlam, Surrey and Whalley Mental Health Teams (12), Riverview Mental Hospital (67); B.C. Department of Human Resources: Woodlands (20); B.C. Telephone Company (20); Department of National Defense: Armed Forces Base, Chilliwack (52); MacMillan Bloedel Company (13). To minimize the d i f f i c u l t i e s associated with obtaining large groups and to make the group s i z e comparable with previous research i n t h i s area ( c f . , Courtright, 1978; Flowers, 1977) each group i n the study consisted of four individuals."'" There were 48 four-person groups f or a t o t a l of 192 subjects. In addition to these 48 research groups, there was one group that was not included and the f i r s t two s t a f f groups were used as p i l o t s . The non-included group was eliminated from the condition that had one extra group and was selected by taking the group where the leadership manipulation appeared to be least e f f e c t i v e . Before venturing into the community to work with the s t a f f groups, the experimenter p i l o t tested the questionnaires and procedures on 43 four-person groups of University of B r i t i s h Columbia undergraduate students. In the 48 groups i n the study there were 103 males and 89 females, with 28 groups led by males and 20 groups led by females. The s t a f f were generally i n t h e i r mid-twenties or older. In terms of organizational h i e r a r c h i e s , the s t a f f occupied pos i t i o n s which ranged from "front l i n e " workers to senior management. An attempt was made to r e c r u i t as many management s t a f f as pos-s i b l e and many groups were composed e i t h e r p a r t i a l l y or t o t a l l y or management s t a f f . I f management l e v e l personnel were part of a group, one of them was asked to take on the leadership function. One group had only three members and one group had f i v e members. - 52 -Procedure P i l o t studies. A series of p i l o t studies was conducted to assess the e f -fectiveness of the manipulations of the independent v a r i a b l e s as well as the effectiveness of the questionnaires. When i t appeared that there would be no major d i f f i c u l t i e s i n carrying out the experimental procedures, arrangements were made to conduct the study with the aforementioned organizations. Main study: A c q u i s i t i o n of subjects. Gaining access to employees i n government and business organizations was a d i f f i c u l t p u b l i c r e l a t i o n s task that extended over the f u l l nine months of data c o l l e c t i o n . In most cases, a f t e r a series of contacts with various senior s t a f f , an arrangement would be made for the experimenter to work through a " l i a i s o n " person within the organization. The l i a i s o n person would ei t h e r r e c r u i t or a s s i s t i n r e c r u i t i n g volunteers for the study. The groups were assigned to the conditions i n a s t r a t i f i e d random procedure. That i s , within each p a r t i c i p a t i n g organization an attempt was made to have approximately equal numbers of groups f a l l within each of the conditions. This was done to prevent p o t e n t i a l a r t i f a c t u a l r e -su l t s which might occur i f one organization was randomly placed into only one or two conditions. Selection of groups for antecedent conditions. The groups were selected to meet the c r i t e r i a for the two independent v a r i a b l e s : antecedent condi- .... tions present or absent, and leadership s t y l e a s sertive or non-assertive. The antecedent conditions present groups were selected to be cohesive and were subjected to time and motivational pressures during the experimental sessions. To choose a cohesive group the l i a i s o n person was asked to recom-mend groups that had worked together before i n a department or on a committee. The antecedent conditions absent groups were selected to be very low i n cohesiveness and an attempt was made to downplay time and motivational pressures during the experimental sessions. The low cohesiveness groups were - 53 -to be composed of i n d i v i d u a l s who had not worked with each other before, and preferably did not even know each other. Selection of group leaders. The l i a i s o n person usually a s s i s t e d i n sele c t i n g the group group leader, since i n most cases he or she would know a l l or most of the group members. In cohesive groups, an i n d i v i d u a l would be selected as the group leader i f she normally f i l l e d that r o l e i n the group ( i . e . , as a supervisor), or i f she were the most senior i n rank, or, i f n e i -ther of these c r i t e r i a f i t , the l i a i s o n person would be asked to sel e c t the one i n d i v i d u a l out of the four that would be l i k e l y to u n o f f i c i a l l y take on the leader r o l e by being the most dominant person i n the group. In non-cohesive groups, the same basic strategy was followed—choosing the most senior i n rank, the most dominant, v e r b a l , or ente r p r i s i n g i n d i v i d u a l to be the group leader. However, i n some cases since the i n d i v i d u a l s were of near-l y equal ranks and none of them had worked together before, s e l e c t i n g a leader who might n a t u r a l l y have taken on that r o l e i n the group was a d i f f i -c u l t judgment to make. Once the group leader was chosen, arrangements were made for him or her to meet with the experimenter p r i o r to the experimental session. Usually, t h i s "pre-meeting" with the leader occurred immediately before the experi-mental session. The half-hour pre-meeting was used as a t r a i n i n g session to i n s t r u c t the leader i n the ass e r t i v e or non-assertive r o l e that he or she was asked to undertake during the group session (see Appendix A, Pre-experimental session, group leadership t r a i n i n g procedures). This time was also used to c o l l e c t a measure of h i s or her t y p i c a l leadership s t y l e , through the use of F i e d l e r ' s (1966, 1967, 1971) least preferred co-worker test (see Appendix B, Leadership t r a i n i n g questionnaire). When the t r a i n i n g procedures and the leadership questionnaire had been completed, the leader was given a copy of the two problems to read and think about before the - 54 -experimental session. The two problems were concerned with c a p i t a l punish-ment and immigration. Approximately one page of background information was provided on each of these issues and the reader was requested to develop what he or she thought would be the best possible s o l u t i o n to each problem. The leader usually had about ten minutes to study the problems before the rest of the group arrived. This was an e s s e n t i a l aspect of the t r a i n i n g session since i t was found i n the p i l o t studies that i t was extremely d i f f i c u l t f o r many people i n the leader r o l e to be strongly assertive about t h e i r point of view at the s t a r t of the session unless they had had some time to think about the problems and prepare t h e i r p o s i t i o n . Procedures during the experimental sessions. The experimental sessions were conducted i n a room with a table, around which the four subjects sat. If the subjects belonged to a non-cohesive group, they were introduced to each other. The general purpose of the study and the way i n which the session would proceed was explained to the group (see Appendix C, General Information about the study). Then the antecedent conditions present groups (the cohe-sive groups) were t o l d that they would have only 30 minutes within which to produce solutions to the two problems. They were to work under the d i r e c t i o n of the group leader, who would be asked at the end of the session to f u l l y describe the nature of the decisions that they had reached. To attempt to increase t h e i r motivation, they were t o l d that i t was important to reach the best decisions possible. I t was said that the tape of the session would be rated f or the q u a l i t y of the solutions proposed and that some tapes would be anonymously shown as examples or cases i n decision making to student classes or p r o f e s s i o n a l seminars (see Appendix D, Instructions to the antecedent con-d i t i o n s present groups). In contrast, the antecedent conditions absent groups were t o l d that they could have as long as they l i k e d to a r r i v e at t h e i r decisions. No mention of - 55 -time was made to these groups by the experimenter while they were engaged i n the research session. In f a c t , some groups spent up to two hours i n discussion of the problems,. No mention was made of r a t i n g the q u a l i t y of the solutions that the group proposed, nor of using the tapes as examples of group decision making. An attempt was made to put the group at ease through some small t a l k or joking. This was done to lessen the seriousness of the occasion, i n contrast to the other groups where motivation was stressed through attempting to make t h e i r e f f o r t s seem important (see Appendix E, Instructions to the antecedent conditions absent groups). A l l experimental sessions were tape-recorded. The experimenter sat at the side of the room with a small black switch-box. I t was explained to each group that previous experience had shown that when l i s t e n i n g to the tapes i t was very d i f f i c u l t to t e l l the di f f e r e n c e between voices. Therefore, the experimenter would use the switch-box, that was wired into the tape-recorder, to put a tone on the tape for each person. Thus, each time the same person spoke, a sequence of one, two, three or four "beeps" would be o v e r l a i d on the tape by the experimenter. By doing t h i s , i t was possible to l i s t e n to the tapes l a t e r and i d e n t i f y how often each person spoke and what he s a i d . (This was a necessary procedure i n order to be able to l a t e r make some of the group process ratings.) The experimenter did not p a r t i c i p a t e i n the discussions, and avoided eye contact with the group during the session. Once the i n s t r u c t i o n s were given to each group, the members were asked to sign a consent form f o r p a r t i c i p a t i n g i n the research (see Appendix F, Consent form). Following that, the group was t o l d that they would be given three minutes to read the background material on the immigration and c a p i t a l punishment problems (see Appendix G, C a p i t a l punishment and immigration problems). At the end of the three minutes, the experimenter instru c t e d them to s t a r t the discussion and started the timing of the half-hour f o r the time-- 56 -l i m i t e d groups. At the end of 15 minutes, the time-limited groups were i n -terrupted and t o l d that they had only 15 minutes l e f t . Near the end of the 30 minutes the group was interrupted again and t o l d that they had two to three minutes l e f t i n which to complete t h e i r proposals (see Appendix H, _ . Instructions to begin the session and i n s t r u c t i o n s to time l i m i t e d groups during the session). At the end of 30 minutes they were t o l d that the session was f i n i s h e d . Subjects i n the antecedent conditions absent groups continued t h e i r discussions without i n t e r r u p t i o n from the experimenter u n t i l they informed the experimenter that they were f i n i s h e d . Procedures following the experimental sessions. At the end of the d i s -cussion period the experimenter asked the group leader to o u t l i n e the nature of the two decisions reached by the group. The leader's response was tape-recorded. When the leader f i n i s h e d speaking, a post-experimental question-naire was given to each group member. The leader received the same ques-tionnaire as the other members, with the exception of two questions dealing with leadership s t y l e (numbers 7 & 8), that were worded i n second person singular rather than t h i r d person singular. In addition, the leader's ques-tionnaire contained F i e d l e r ' s (1967) group atmosphere scale, which provided the information necessary to make observations based upon F i e d l e r ' s theory of leadership (see F i e d l e r and Chemers, 1974). The group leader's questionnaire i s contained i n Appendix I and the group members' questionnaire i s contained i n Appendix J . The post-experimental questionnaire was designed to perform a manipu-l a t i o n check on the independent v a r i a b l e s , i . e . , to determine whether the appropriate preconditions for groupthink were met i n the study. In addition the questionnaire contained measures designed to i n d i c a t e the presence of f i v e of the eight groupthink symptoms postulated by Janis (1972). The f i v e symptoms tested were the i l l u s i o n of unanimity, the degree of self-censorship, - 57 -the degree of conformity pressure, the i l l u s i o n of i n v u l n e r a b i l i t y , and the degree of morality perceived i n the decisions. Given the nature of the ex-perimental sessions and the type of problems that the groups were dealing with, i t was f e l t that i t would not make sense to test for the remaining three groupthink symptoms. The missing symptoms consisted of r a t i o n a l i z i n g ' o u t s i d e warnings, holding stereotyped views of the enemy (there was no w e l l defined enemy), and developing mindguards i n the group (mindguarding requires at-tempted outside input). The r e l a t i o n s h i p between the manipulation checks, groupthink symptoms and the various questions on the questionnaire i s ex-plained i n more d e t a i l i n the section of Appendix J t i t l e d , "Notes on the post-experimental questionnaire". When the subjects had completed the post-experimental questionnaire they were debriefed (see Appendix K, Debriefing information). At that point, for groups who were inter e s t e d , an arrangement was made to meet again i n about one week's time to have the experimenter provide feedback to the group on the session. In the feedback discussion the experimenter would ou t l i n e the c h a r a c t e r i s t i c s of the group, such as i t s l e v e l of cohesiveness, and mention what the average f e e l i n g s of the group members were concerning the s e s s i o n — for example, t h e i r f e e l i n g s of conformity pressure. The experimenter would then note a few of the c h a r a c t e r i s t i c s of the decision making process that had occurred during the session, such as the number of a l t e r n a t i v e solutions considered by the group. An example of the feedback provided to a group i s contained i n Appendix L. Procedure for S t a t i s t i c a l Analysis Manipulation check. Based upon questions 1 - 12 i n the post-experimental questionnaire, a MANOVA was used to assess the successfulness of the mani-pulations of the independent v a r i a b l e s . When the o v e r a l l MANOVA F-test was - 58 -s i g n i f i c a n t , i n d i v i d u a l ANOVA tests were run on each question. This helped to determine which questions were responsible f o r the s i g n i f i c a n t differences found i n the MANOVA. Dependent v a r i a b l e s . Based upon questions 13 - 20 i n the post-experimental questionnaire, a MANOVA was used to determine i f there were any groupthink symptoms appearing as the r e s u l t of the main or i n t e r a c t i o n e f f e c t s of the independent v a r i a b l e s . When the MANOVA F-test for a main or i n t e r -action e f f e c t was s i g n i f i c a n t , i n d i v i d u a l ANOVA tests were run on each ques-t i o n to determine which questions were causing the MANOVA e f f e c t s . A s i m i l a r procedure was undertaken with the group process ratings based on the tape-recorded sessions and the quality-of-the-decision-outcome ratings made on the tr a n s c r i p t s of the group's decisions. MANOVA's were the i n i t i a l test f o r o v e r a l l s i g n i f i c a n c e and i f such s i g n i f i c a n c e was found then ANOVA's were run on the i n d i v i d u a l questions. The aforementioned group process r a t i n g s were made by two graduate stu-dents (not including the experimenter) following t h e i r independent review of the audiotape of each session. The raters were also b l i n d to the experi-mental condition of the research groups. Each group was rated on the basis of the following eight quality-of-the-decision-process v a r i a b l e s , taken l a r g e l y from the suggestions of Janis and Mann (1977, p. 11): (1) the extent of the discussion of the group's objectives; (2) the number of a l t e r n a t i v e solutions generated; (3) the extent of the consideration of the p o s i t i v e and negative aspects of the a l t e r n a t i v e s ; (4) the extent to which general i n f o r -mation r e l a t e d to the problem was discussed; (5) the degree of p a r t i c i p a t i o n of a l l the group members; (6) the extent to which a l t e r n a t i v e s and i n f o r -mation were reviewed before a f i n a l d e cision was made; (7) the extent to which the group leader stressed h i s or her viewpoint; (8) the extent to which the group leader proposed h i s or her solutions f i r s t . (See Appendix M, - 59 -Rating scales for assessing the q u a l i t y of the decision making process.) The two group process raters were trained to use the r a t i n g scales p r i o r to assessing the research groups. In addition, a f t e r each seri e s of f i v e or ten groups was rated the experimenter would hold a session to discuss any consistent differences i n the ratings which appeared between the two r a t e r s . I f the two r a t e r s ' assessments d i f f e r e d by two or more categories (on the seven point s c a l e s ) , the differences were discussed u n t i l a c l o s e r agreement was achieved. Rating differences of only one category were averaged. Two assessments of i n t e r - r a t e r r e l i a b i l i t y were made. The f i r s t r e l i a b i l i t y assessment consisted of tabulating the percentage of responses of the two raters that were 100% consistent, then c a l c u l a t i n g the percentage of responses that were only one category d i f f e r e n t on the seven point scale, or two cate-gories d i f f e r e n t , etc. The second assessment of r e l i a b i l i t y was a Pearson c o r r e l a t i o n of the ratings across the two r a t e r s . The MANOVA and ANOVA analyses were based upon the ratings that had been adjusted a f t e r the raters had discussed major dif f e r e n c e s . The quality-of-the-decision-outcome ratings were made by four indepen-dent experts. Two experts rated the immigration solutions proposed by each group and two d i f f e r e n t experts rated the c a p i t a l punishment solutions. Transcripts of the solutions were rated on the following seven c r i t e r i a , based l a r g e l y upon Leathers' (1972) p r o d u c t i v i t y r a t i n g scales (see Appendix N, Scales for expert ratings of the q u a l i t y of the decision outcome): (1) s o l u t i o n effectiveness; (2) f e a s i b i l i t y of the s o l u t i o n ; (3) c r e a t i v i t y of the s o l u t i o n ; (4) basis i n s i g n i f i c a n t information; (5) s o l u t i o n comprehen-siveness; (6) r i s k associated with implementing the s o l u t i o n ; (7) o v e r a l l q u a l i t y of the s o l u t i o n . The two experts on immigration were s t a f f members provided by the Federal Department of Immigration. One expert was the manager f o r - 60 -recruitment and s e l e c t i o n f o r the B.C./Yukon region, while the other expert was an Immigration counsellor. The two experts on c a p i t a l punishment were Professors of Law within the U.B.C. f a c u l t y who taught i n the area of c r i -minal law. One of the experts was a member of the B.C. Bar Association who has worked f o r the Law Reform Commission of Canada and has published works i n the area of cri m i n a l law. The second expert has a doctorate i n crimino-logy, has served as the Chairman of the B.C. P o l i c e Commission, has worked for the Le Dain Royal Commission and the Law Reform Commission of Canada and i s presently a Senior P o l i c y Advisor to the B.C. Attorney General's Department. Each of the two sets of experts was trained i n using the r a t i n g scales before assessing the t r a n s c r i p t s of the decisions. Due to the considerable imposition upon t h e i r time, an extensive p r e - t r a i n i n g process was not pos-s i b l e . During the time that the experts were making t h e i r r a t i n g s , i t was attempted to have the experimenter review the material following each seri e s of 10 or 15 groups. Discussion of any major differences i n ratings would then be i n i t i a t e d by the experimenter. This turned out to be f e a s i b l e f o r the immigration experts but impossible to carry out with the c a p i t a l punish-ment experts. A si n g l e r a t i n g for each of the seven scales was obtained on the c a p i t a l punishment problem and on the immigration problem, for each group, by averaging the ratings made by the experts. Two assessments of i n t e r - r a t e r r e l i a b i l i t y were made. One assessment was the percentage of responses with 10Q% correspondence, d i f f e r i n g by one category, d i f f e r i n g by two categories, etc. The second assessment was the c a l c u l a t i o n of a Pearson c o r r e l a t i o n across the ratings of the two experts. Post hoc multiple regression and canonical c o r r e l a t i o n analyses. Two step-wise multiple regression analyses were performed on the qual i t y - o f - t h e decision data. For the pr e d i c t i o n c r i t e r i o n one set of analyses used the - 61 -averaged ratings of o v e r a l l decision q u a l i t y that were provided by the two immigration experts, while the second set of analyses used the averaged ratings of o v e r a l l decision q u a l i t y from the two experts on c a p i t a l punish-ment. The reason for making the separation between the two problems was that the c o r r e l a t i o n between the ratings of two experts within one problem was moderate, but when the averaged ratings of the two experts within one problem were correlated to the averaged ratings of the two experts within the second problem, the c o r r e l a t i o n was very low (see Tables 7 and 8). The following fourteen v a r i a b l e s were used as the pr e d i c t o r s : (1) assignment to antecedent conditions present or absent; (2) leadership s t y l e , a s s e r t i v e or non-assertive; (3) i n t e r a c t i o n between leadership s t y l e and antecedent conditions; (4) lea s t - p r e f e r r e d co-worker (L.P.C.) score; (5) group atmosphere score (G.A.S.); (6) i n t e r a c t i o n between L.P.C. score and G.A. score (#'s 4, 5, and 6 stem from F i e d l e r ' s (1967) research); (7) the "average-degree-of-friendship" r a t i n g f o r each group (as a measure of cohe-sivenss); (8) the average degree of concurrence seeking f o r each group; (9) the leadership s t y l e , which was taken as the average across questions seven and eight on the post-experimental questionnaire; (10) the average degree of motivation i n the group; (11) the average degree of perceived time pressure i n the group (the above pr e d i c t o r s , #'s 7 - 11, were obtained from the post-experimental questionnaire); (12) the sex of the group leader; (13) the type of organization that the subjects came from, mental health or (14) m i l i t a r y (which included the subjects from the Attorney General's Department). The coding was designed so that i f the subjects came from neither the mental health organizations nor the m i l i t a r y then they came from the t h i r d category, business organizations. The step-wise multiple regression analysis was expected to provide in-" formation on which of the predictors made s i g n i f i c a n t and independent - 62 -contributions to the p r e d i c t i o n of the c r i t e r i o n . Since there was some ques-t i o n concerning predictors that operated both within the c a p i t a l punishment problem and within the immigration problem, a canonical c o r r e l a t i o n analysis was performed. This analysis used the c r i t e r i o n measures of c a p i t a l punish-ment decision q u a l i t y and immigration decision q u a l i t y as one set of v a r i -ables. The second set of variables consisted of the fourteen pr e d i c t o r items. I t was expected that the canonical constructions would provide i n f o r -mation about the consistency of predictors across the two problems. Results In discussing the r e s u l t s we w i l l f i r s t consider the analyses concerning the successfulness of the independent v a r i a b l e manipulations. The production of concurrence seeking w i l l then be discussed, followed by the analyses of the groupthink symptoms. The group process data and the re l a t e d i n t e r - r a t e r r e l i a b i l i t i e s w i l l next be considered. Then, the q u a l i t y - o f - t h e - d e c i s i o n -outcome analyses plus the i n t e r - r a t e r r e l i a b i l i t i e s of the expert judges w i l l be reviewed. Last, the post hoc multiple regression predictions of decision q u a l i t y w i l l be discussed. Manipulation Checks bri the Independent Variables Manipulation of the antecedent conditions. It was predicted that the antecedent conditions present groups would be s i g n i f i c a n t l y more cohesive, more motivated, and under greater stress than the antecedent conditions absent groups. The MANOVA of the antecedent conditions questions (questions 1 - 5 , and 9 - 12 on the post-experimental questionnaire) showed a s i g n i f i -cant main e f f e c t for antecedent conditions, (F_ „, = 9.49, p <.0001), The —y ,5b *-main e f f e c t for leadership s t y l e and the i n t e r a c t i o n e f f e c t s were both - 63 -n o n - s i g n i f i c a n t . The ANOVA's indicated that the following questions had s i g n i f i c a n t main e f f e c t s for the antecedent conditions: (1) the degree of friendship i n the group, (F^ 44 = 55.96, p_ <.001); (2) the number of group members previously worked with, (F^ ^  = 78.82, p_ <.0001); (3) the i n t e r a c t i o n of the number of members previously worked with by the degree of i n t e r e s t i n p a r t i c i p a t i n g with those members, (F q = 17.11, p_ <.0005); (4) the l e v e l of perceived time pressure (J_^ ^  = 8.37, £ <.01). The group means for each of these questions i s shown i n Table 1. The r e s u l t s i n d i c a t e that there was a s i g n i -f i c a n t l y greater degree of frie n d s h i p , more f a m i l i a r i t y with each other from previous non-experimental groups, and a greater degree of perceived time pressure among the antecedent conditions present groups than among the ante-cedent conditions absent groups. At the same time though, the following questions relevant to the ante-cedent conditions manipulations had no s i g n i f i c a n t main or i n t e r a c t i o n ef-fect s on the ANOVA's: (1) the degree of past i n t e r e s t i n p a r t i c i p a t i n g i n groups that current members were a part of; (2) the degree of i n t e r e s t i n p a r t i c i p a t i n g with the current group members; (3) the interpersonal a t t r a c -tiveness of the current group members as work-mates: (4) the l e v e l of group motivation; (5) the l e v e l of perceived importance of the problems, and (6) the l e v e l of o v e r a l l stress. The antecedent conditions manipulations were e f f e c t i v e , but with l i m i -t a t i o n s . The antecedent conditions present groups were s i g n i f i c a n t l y more cohesive on three of the s i x cohesiveness measures. In addition, i n r e t r o -spect i t does not seem reasonable to have thought that the assignment of a subject to the antecedent conditions groups would a f f e c t h i s or her i n t e r e s t i n having p a r t i c i p a t e d with some of the members i n previous groups. This i n t e r e s t i n past p a r t i c i p a t i o n with members was one of the three non-- 64 -Table 1 Si Group Means for the Antecedent Conditions Questions Reaching Significance Antecedent Antecedent Conditions Conditions F_-Ratio^ Present Absent Degree of Friendship 3.28 4.91 55.96 Members Worked with Previously 1.58 3.30 78.33 Previous Members by Interest 3.24 7.04 17.11 c Perceived Time Pressure 2.86 3.74 8.37 The group means are within a range of p o t e n t i a l values from 1 to 7. The lower the score the greater the degree of the c h a r a c t e r i s t i c being measured. k d.f. = 1, 44; except for "previous members by i n t e r e s t " where d.f. = 1, 39. Scoring reversed from that found on the question i n Appendices I and J . ft p <.01 ftft p_ <.00Q5 p_ <.0001 - 65 -s i g n i f i c a n t measures of differences i n group cohesiveness. The o v e r a l l l e v e l of motivation across a l l of the groups was rated as being between "strongly motivated" and "more than some motivation" (M = 2.83, question #9, post-experimental questionnaire). Thus, while there was no s i g n i f i c a n t difference i n motivation between the antecedent conditions pre-sent (M = 2.87) and absent (M = 2.79) groups there was a moderately high degree of motivation i n a l l groups. From the experimenter's observations i t would seem that the f a i l u r e to produce s i g n i f i c a n t differences i n motivation between groups was due to the f a c t that h i s attempts to increase motivation had no marked e f f e c t with subjects that already displayed a r e l a t i v e l y high l e v e l of commitment. I t was expected that the problems would be seen as high i n importance and there was no p r e d i c t i o n of differences i n these ratings between groups. The r e s u l t s showed that the problems were rated on the average as being between "somewhat important" and "important" (M = 4.57, question #10, post-experimental questionnaire). However, i t was found that even though there was a s i g n i f i c a n t d i f f e r e n c e i n perceived time pressure, with o v e r a l l motivation and perceived problem importance being moderate, these e f f e c t s were not translated i n t o s i g n i f i c a n t differences between the groups with respect to t h e i r f e e l i n g s about .the o v e r a l l l e v e l of s t r e s s . In summary, concerning the manipulations of the antecedent conditions, the groupthink preconditions of cohesiveness and perceived time stress were s i g n i f i c a n t l y stronger i n the antecedent conditions present groups than they were i n the antecedent conditions absent groups. In a d d i t i o n , a l l groups demonstrated a moderate l e v e l of involvement i n the task as indicated through t h e i r ratings of motivation and perceived problem importance. The major negative outcome was the lack of s i g n i f i c a n t differences between the groups i n t h e i r o v e r a l l f e e l i n g s of s t r e s s . Manipulation of leadership s t y l e . I t was predicted that group leaders - 66 -requested to perform a s s e r t i v e l y would place s i g n i f i c a n t l y more emphasis upon t h e i r own views and state these views s i g n i f i c a n t l y e a r l i e r i n the d i s -cussion than would group leaders requested to perform non-assertively. The MANOVA of the leadership s t y l e questions (questions 7 & 8 on the post-experimental questionnaire) showed a s i g n i f i c a n t main e f f e c t for leadership s t y l e (F_2 ^ = 64.11, p_ <.0001). The main e f f e c t f o r antecedent conditions and the i n t e r a c t i o n e f f e c t s were both no n - s i g n i f i c a n t . The ANOVA's indicated that the questions concerning the leader's emphasis upon h i s own point of view and the timing of the presentation of h i s views both had s i g n i f i c a n t main e f f e c t s for leadership s t y l e (emphasis on view: F_ = 44.14, £ <.0001; timing of presentation: F_ = 64.84, £ <.0001). A consideration of the means for both of those questions indicates that a s s e r t i v e leaders stressed t h e i r views more strongly than did non-assertive leaders (M's = 2.47, 4.39 respectively) and the ass e r t i v e leaders presented t h e i r views e a r l i e r i n the discussion than did the non-assertive leaders (M's = 3.07, 5.39 r e s p e c t i v e l y ) . These r e s u l t s concerning the effectiveness of the leadership s t y l e mani-pulation are borne out in.the analysis of the data provided by the two i n d i -v iduals independently r a t i n g the audio-tapes of the sessions. As part of the i r ratings of the q u a l i t y of the decision making process, the two raters assessed the leadership s t y l e exhibited i n each group (see Appendix M, questions 7 & 8). The findings show that the o v e r a l l MANOVA for the leader-ship questions had a s i g n i f i c a n t main e f f e c t for leadership s t y l e ( F D = '—o ,51 22.13, _p <.0001). The ANOVA's of the two leadership questions y i e l d e d the same r e s u l t s as outlined above. There were s i g n i f i c a n t main e f f e c t s for leadership s t y l e concerning the leader's emphasis on h i s view (F^ ^ = 72.52, £ <.0001) and the timing of the leader's presentation of h i s view (F_^ ^ = 143.52, £ <.0001). Leaders under the assertive condition put more stress upon t h e i r own point of view i n comparison to non-assertive leaders - 67 -(M's = 3.17, 5.63 respectively) and presented t h e i r views e a r l i e r than did non-assertive leaders (M's = 2.69, 6.15 r e s p e c t i v e l y ) . In summary, the manipulation of leadership s t y l e was e f f e c t i v e . The Production of Concurrence Seeking The ANOVA of the question concerning group concurrence seeking (question 6, post-experimental questionnaire) indicated that there were no s i g n i f i c a n t main e f f e c t s f or leadership s t y l e or antecedent conditions nor a s i g n i f i c a n t i n t e r a c t i o n e f f e c t . The mean r a t i n g f or concurrence seeking across a l l groups was 4.38, while the mean r a t i n g f or the antecedent conditions present, leader-ship s t y l e a s s e r t i v e , c e l l was v i r t u a l l y the same at 4.26. On the post-experimental questionnaire these ratings i n d i c a t e a moderate degree of concur-rence seeking across a l l groups. The p r e d i c t i o n of groupthink theory that s i g n i f i c a n t l y more concurrence seeking would take place as a r e s u l t of the i n t e r a c t i o n of an assertive leadership s t y l e with the existence of antecedent conditions such as cohesiveness and stress was not borne out i n t h i s study. Analysis of the Groupthink Symptoms It was predicted that there would be a s i g n i f i c a n t i n t e r a c t i o n between leadership s t y l e and the antecedent conditions such that when the antecedent conditions were present, an assertive s t y l e of leadership would produce group-think symptoms, whereas a non-assertive s t y l e of leadership would produce fewer or no groupthink symptoms. When the antecedent conditions were absent, i t was predicted that there would be no s i g n i f i c a n t differences i n the pro-duction of groupthink symptoms between the two leadership s t y l e s . Contrary to the groupthink theory p r e d i c t i o n of a s i g n i f i c a n t i n t e r a c t i o n e f f e c t , the MANOVA of the groupthink symptoms questions (questions 13 - 20 i n the post-experimental questionnaire) revealed a non-significant i n t e r a c t i o n term. - 68 -The main e f f e c t s f or the antecedent conditions were also found to be non-s i g n i f i c a n t . However, there was a s i g n i f i c a n t main e f f e c t f or leadership s t y l e (Fg 3 ? = 2.47, p_ <.05). The ANOVA's of the groupthink symptoms questions indicated that the leadership s t y l e main e f f e c t was rel a t e d to the areas of conformity pressure and the perceived r i s k i n e s s of the decision. There were leadership s t y l e main e f f e c t s found for the following questions: (1) the degree of conformity pressure perceived to be placed on others i n the group (F = 11.03, p_ < .005); (2) the degree of conformity pressure f e l t by the subject (F_ = X , HQ 5.17, p_ <.05); and (3) the degree of r i s k seen i n the decisions (F = 4.11, p_ <.05). A consideration of the c e l l means displayed i n Table 2 shows that group members led by ass e r t i v e leaders f e l t more conformity pressure ;from the group both upon themselves and upon other members than was f e l t by group members under non-assertive leaders. Group members led by assertive leaders also perceived the decisions that were made to be more r i s k y than did group members led by non-assertive leaders. In addi t i o n , the ANOVA r e s u l t s and the c e l l means indicated that the groups led by assertive leaders tended to be less accurate i n t h e i r perception of the amount of agreement among group mem-bers over the proposed decisions than were the groups led by non-assertive leaders ( t h i s did not reach s i g n i f i c a n c e , F_ = 2.74, p - .10). There were no s i g n i f i c a n t leadership s t y l e ANOVA main e f f e c t s f or the groupthink symptom questions dealing with the i l l u s i o n of agreement, s e l f -censorship, the i l l u s i o n of morality, nor the degree of optimism by the mem-bers over the proposed solutions (under groupthink theory high optimism and r i s k taking are i n d i c a t o r s of an i l l u s i o n of i n v u l n e r a b i l i t y ) . Analysis of the Group Process Data I t was predicted that there would be an i n t e r a c t i o n between leadership - 69 -Table 2 Group Means for Groupthink Symptom Questions Reaching Si g n i f i c a n c e Assertive Non-Assertive cl Leadership Leadership F-Ratio Style Style Conformity Pressure on Others 2.50 1.90 11.03"' b * Conformity Pressure on Subject 1.20 0.82 5.17 Riskiness of the Decision 2.76 2.35 4.11 a d.f. = 1,44 k Scoring reversed from that found on the question i n Appendices I and J . £ <-05 £ <.005 - 70 -s t y l e and the antecedent conditions of cohesiveness and s t r e s s . When the antecedent conditions were present i t was hypothesized that groups under ass e r t i v e leaders would be rated as performing poorer on the group process measures than groups under non-assertive leaders. However, when the ante-cedent conditions were absent i t was hypothesized that there would be no s i g n i f i c a n t difference between groups under assertive or non-assertive leaders. Assessments of i n t e r - r a t e r r e l i a b i l i t y . The information contained i n Table 3 shows that .90% of the i n i t i a l group process ratings were within one category of each other on a seven point scale. The Pearson c o r r e l a t i o n s shown i n Table 4 have been adjusted by the Spearman-Brown formula f o r com-bining judgments from two ra t e r s (see G u i l f o r d , 1965, p. 466). The corre-l a t i o n s are s t a t i s t i c a l l y s i g n i f i c a n t and ind i c a t e a mean i n t e r - r a t e r c o r r e l a t i o n of 0.76. Following the assessment of i n t e r - r a t e r r e l i a b i l i t y , an arithmetic average was calculated to a r r i v e at a si n g l e r a t i n g f o r each group on each question i n cases where the ratings d i f f e r e d by one category. It was decided that i n cases where the two independent ratings d i f f e r e d by more than one category, rather than taking the average i t would be more appropriate to have the two raters discuss and adjust t h e i r ratings to within one category of each other. Once they were within one category of each other the average of the two scores was taken. MANOVA and ANOVA analyses. 1 The MANOVA of the group process data demon-strated a s i g n i f i c a n t main e f f e c t f o r the antecedent conditions (F^ = 2.65; P_ <.05). There was no s i g n i f i c a n t main e f f e c t f o r leadership s t y l e , and i n These analyses did not contain the two ratings concerned with leadership s t y l e , to avoid confounding the dependent measures with the manipulations of the independent v a r i a b l e s . - 71 -Table 3 Group Process Data: Percentage of Agreement, Inter-Rater R e l i a b i l i t y Check Complete Agreement One Category Difference Two Categories Difference Three Categories Difference Discussion of Obj e c t i v e s a 33% 46% 17% 4% Scales for Number of S o l -utions Proposed 3 58% 33% 8% Assessing the Consideration of Consequences 3 38% 54% 8% Quality of General Infor-mation Used 3 27% 50% 23% the Decision Member P a r t i c i p a t i o n 3 46% 48% 6% Making Process Review of Information 3 44% 48% 8% (See Appendix M) Leadership: Assertiveness 3 31% 67% 2% Leadership: Timing of View 3 52% 42% 2% 4% Average*5 41% 49% 9% 1% 3 N = 48 b N = 384 - 72 -Table 4 Group Process Data: Pearson Correlations^ Inter--Rater R e l i a b i l i t y Check Pearson Correlations Discussion of h .54* Obj ectives Scales for Number of Solutions Proposed^* * .44 Assessing the Consideration of Consequences^ ** .73 Quality of General Information Used b ** .75 the Decision Member P a r t i c i p a t i o n ^ * .57 Making Process Review of Information^ 3 ** .77 (See Appendix M) Leadership Style: b Assertiveness ** .91 Leadership Style: ** .94 Timing of View Mean 0 .76 a R e l i a b i l i t y c o e f f i c i e n t s adjusted by the Spearman-Brown formula. b N = 48 c Calculated through the use of Fisher transformations. * £ <.005 £ <.001 - 73 -contrast to the p r e d i c t i o n made by groupthink theory, there was no s i g n i f i c a n t i n t e r a c t i o n between leadership s t y l e and antecedent conditions. The ANOVA's of the si x dependent v a r i a b l e group process measures i n d i -cated that there were s i g n i f i c a n t main e f f e c t s f o r antecedent conditions on two of the si x measures. The two questions showing these s i g n i f i c a n t main e f f e c t s concerned the degree to which the members considered the consequences of the decision (F_^ ^  = 12.61, p_ <.001) and the amount of general information about the topic that was brought i n t o the discussion (J_^ ^  = 11.43, p_<.005). A consideration of the c e l l means on these two questions (see Table 5) shows that the antecedent conditions absent groups were more e f f e c t i v e i n terms of bringing i n more information r e l a t e d to the problems and made a stronger con-s i d e r a t i o n of the consequences of the various a l t e r n a t i v e s proposed than did the antecedent conditions present groups. I t should be noted that there was also a tendency for the antecedent conditions absent groups to produce more solutions than the antecedent condi-tions present groups, although t h i s d i f f e r e n c e did not reach s i g n i f i c a n c e (Fij. 44 = 2.59, p_ >.10). S i m i l a r l y , on two of the three remaining group pro-cesses, those of the consideration of the group's objectives and the degree of reviewing during the discussion, the antecedent conditions absent groups received more p o s i t i v e ratings than did the antecedent conditions present groups, although the differences between them were not s i g n i f i c a n t . I t i s of i n t e r e s t that the r e s u l t s stemming from the ratings of the two independent reviewers were v i r t u a l l y i d e n t i c a l with the r e s u l t s from the MANOVA and ANOVA's of the ratings made by the experimenter. During the nine months that the experimental data were being c o l l e c t e d , the experimenter provided feedback about the sessions to the p a r t i c i p a t i n g groups. In con-junction with preparing these feedback sessions the experimenter made ratings of the group processes on nearly a l l of the groups. The MANOVA of these - 74 -Table 5 Group Means for the Decision Process Ratings by the Two Independent Audio-Tape Reviewers Antecedent Antecedent Conditions Conditions F-Ratio b Absent Present Discussion of 4.54 4.92 1.64 Scales f o r Obj ectives Assessing Number of Solu-tions Proposed 0 3.85 4.08 2.59 the Quality Consideration 3.85 4.67 it -k 12.61 of the of Consequences General Infor- 3.50 4.54 11.43* Decision mation Used Making Member P a r t i c i -c pation 3.03 3.35 1.84 Process Review of Information 3.85 4.21 1.56 a The group means are within a range of p o t e n t i a l values from 1 to 7. The lower the score the greater the dej; »ree of the c h a r a c t e r i s t i c being measured. b d.f. = 1,44 Scoring reversed from that found on the question i n Appendices I and J . * p_ <.005 £ <.Q01 - 75 -ratings y i e l d e d a s i g n i f i c a n t main e f f e c t for antecedent conditions (F^ ^ = 3.02, _p_ <.05) and non-significant e f f e c t s f o r leadership s t y l e and the i n t e r a c t i o n term. The ANOVA's of the experimenter's ratings indicated that there were s i g n i f i c a n t main e f f e c t s f o r the antecedent conditions on the following questions: (1) the degree to which the members considered the consequences of the decision (F^ ^y = 12.65, p_ <.005); (2) the amount of general i n f o r -mation about the topic brought into the discussion (F_^ = 7.35, p_ <.01); (3) the number of solutions proposed (F_^ ^y = 8.65, p_ <.01). A review of the group means i n Table 6 shows that the antecedent conditions absent groups brought i n more information, made a stronger consideration of the conse-quences associated with the a l t e r n a t i v e s and produced more a l t e r n a t i v e .1 solutions than did the antecedent conditions present groups. S i m i l a r l y , the antecedent conditions absent groups had ratings that were more p o s i t i v e on the three remaining group p r o c e s s measures than did the antecedent con-d i t i o n s present groups, although the differences were not s i g n i f i c a n t . Thus, the o v e r a l l r e s u l t s show a very consistent pattern across both the independent raters and the experimenter's r a t i n g s , and across s i g n i f i c a n t and non-significant outcomes. In terms of the group process measures, the ante-cedent conditions absent groups appears to be more e f f e c t i v e than the antece-dent conditions present groups. Analysis of the Decision Outcome Data It was predicted that there would be an i n t e r a c t i o n between leadership s t y l e and the antecedent conditions of cohesiveness and s t r e s s . This i n t e r -action would be such that when the antecedent conditions were present, groups under a s s e r t i v e leaders would produce decisions that would be rated by ex-perts as i n f e r i o r i n q u a l i t y to the decisions produced by groups under non-- 76 -Table 6 Group Means for the Decision Process Ratings by the Experimenter Antecedent Antecedent Conditions Conditions F-Ratio b Absent Present Discussion of 5.05 5.50 1.06 Scales for Obj ectives Assessing Number of Solu-tions Proposed 0 3.52 5.20 * 8.65 the Quality Consideration of Consequences 2.91 3.80 ** 12.65 of the General Infor- 3.14 4.15 7.35 Decision mation Used Making Member P a r t i c i p a t i o n 2.81 3.00 0.55 Process Review of Information 3.38 3.45 0.04 The group means are within a range of p o t e n t i a l values from 1 to 7. The lower the score the greater the degree of the c h a r a c t e r i s t i c being measured. b d . f. = 1,37 c Scoring reversed from that found on the question i n Appendices I and J . * p_ <.01 p_ <.005 - 77 -asse r t i v e leaders. In contrast, when the antecedent conditions were absent, i t was predicted that there would be no s i g n i f i c a n t differences i n the ratings of the q u a l i t y of the decisions produced by groups under assertive or non-ass e r t i v e leaders. 2 C a p i t a l punishment: Assessments of i n t e r - r a t e r r e l i a b i l i t y . The i n f o r -mation contained i n Table 7 shows that a mean of 71% of the decision outcome ratings by the two c a p i t a l punishment experts were within one category of each other on a seven point scale. The Pearson c o r r e l a t i o n s shown i n Table 8 however, indicate a wide range of co r r e l a t i o n s with an average c o r r e l a t i o n of 0.66. A l l of the cor r e l a t i o n s are s t a t i s t i c a l l y s i g n i f i c a n t . C a p i t a l punishment: MANOVA and ANOVA analyses. The MANOVA of the ratings of the f i r s t c a p i t a l punishment expert indicated that the leadership s t y l e , antecedent conditions, and the i n t e r a c t i o n factors were non-s i g n i f i c a n t . The MANOVA of the second expert's ratings yielded a s i g n i f i -cant main e f f e c t for leadership s t y l e (F_^  ^y = 2.52, p_ <.05). The antece-dent conditions and i n t e r a c t i o n terms were no n - s i g n i f i c a n t . The ANOVA's did not y i e l d s i g n i f i c a n t r e s u l t s for leadership on any of the seven measures. A review of the c e l l means on the s i g n i f i c a n t MANOVA r e s u l t s indicated that o v e r a l l , the groups under the non-assertive leaders produced decisions superior i n q u a l i t y to the decisions achieved by groups under the assertive leaders. (M's = 0.04, -0.65 re s p e c t i v e l y on a scale from -3 to +3). The MANOVA of the combined data from the two c a p i t a l punishment experts yielded no s i g n i f i c a n t r e s u l t s . Immigration: Assessments of i n t e r - r a t e r r e l i a b i l i t y . The information The r e l i a b i l i t y c o e f f i c i e n t s for the Immigration, C a p i t a l Punishment, and Combined co r r e l a t i o n s shown i n Table 8 are. adjusted by the Spearman-Brown formula f o r combining judgments from two raters (see G u i l f o r d , 1965, p.466). - 78 -Table 7 Decis ion Outcome Data: Percentages of Agreement, Inter -Rater R e l i a b i l i t y Check Scales for Expert Ratings of the i Qual i ty of the Decis ion Mean E f f e c -t iveness F e a s i b i -l i t y Creat iv-i t y - S i g n i f icance - Compre-hensive R i s k i -ness Overa l l Qua l i ty Complete „ a Cap Pun 15% 25% 25% 33% 29% 25% 40% 27%b Agreement I m m i g r a t ' n ° 26% 21% 30% 15% 26% 36% 19% 25%d Combined 9% 13% 7% 15% 22% 4% 7% l l % f -One Cap Pun 3 58% 35% 56% 42% 31% 48% 38% 44%b Category Immigrat 'n C 38% 40% 34% 28% 32% 40% 47% 37%d Dif ference Combined 39% 39% 28% 28% 33% 26% 37% 33%f Two Cap Pun 3 6% 19% 17% 17% 25% 19% 17% . 17%b Categories I m m i g r a t ' n ° 13% 9% 4% 4% 17% 2% 11% 9%d Dif ference e Combined 22% 22% 30% 37% 30% 33% 17% 27%f Three Cap Pun 3 19% 19% 2% 8% 10% 8% 6% l l % b Categories I m m i g r a t ' n ° 4% 4% 4% 26% 6% 6% 6% 8%d Di f ference Combined 15% 22% 15% 13% 11% 30% 22% 18%f Four Cap Pun 3 2% 2% 4% l% b Categories I m m i g r a t ' n ° 13% 11% 17% 15% 9% 4% 13% 12%d Dif ference Combined 11% 4% 11% 4% 2% 4% 15% 7%f Five Cap Pun 3 Categories I m m i g r a t ' n ° 4% 15% 11% 11% 11% 9% 4% 9%d D i f f erence e Combined 2% 9% 2% 2% 2% 3%f Six Cap Pun 3 Categories I m m i g r a t ' n ° 2% 2% 2% l% d Dif ference Combined 2% 2% l% f C a p i t a l Punishment: 15 = 48 T o ta l N = 336 Immigration: N = 47 Tota l N = 329 Both C a p i t a l Punishment and Immigration Combined: N = 46 T o ta l N = 322 b c d e f - 79 -Table 8 Decis ion Outcome Data: Pearson C o r r e l a t i o n s 3 Inter-Rater R e l i a b i l i t y Check Pearson Cor re la t ion Pearson C o r r e l a t i o n Pearson Cor re la t ion C a p i t a l Punishment Immigration Experts Experts Between C a p i t a l Pun-ishment & Immigration Experts^ E f fec t iveness .67 .70 .00 Scales for F e a s i b i l i t y .29 .51 .11 Expert Ratings C r e a t i v i t y .82 .70 .18 of the S i g n i f i c a n c e .75 .52 .18 Qual i ty of Comprehensiveness .61 .68 .A3 the Dec is ion Risk iness .41 .77 .21 Out come Overa l l Qua l i ty .82 .69 .15 Mean .66 .66 .18 R e l i a b i l i t y c o e f f i c i e n t s adjusted by the Spearman-Brown formula. N = 48 N = 47 N = 46 Calcu la ted through the use of F isher t ransformat ions. £ <.05 * £ <.005 _p_ <.001 - 80 -shown i n Table 7 indicates that 62% of the decision outcome ratings by the two immigration experts were within one category of each other on a seven point scale. Unfortunately, there were a s u b s t a n t i a l number of responses .: that were separated by as much as four and f i v e categories. The Pearson cor-r e l a t i o n s between the two r a t e r s , shown i n Table 8, are a l l s i g n i f i c a n t , with an average c o r r e l a t i o n of 0.66. Immigration: MANOVA and ANOVA analyses. The MANOVA's of the ratings of the two Immigration experts did not y i e l d any s i g n i f i c a n t r e s u l t s . C a p i t a l punishment and immigration: Assessments of the r e l i a b i l i t y of  ratings across the two problems. The information contained i n Table 7 shows that an average of 44% of the ratings across the two problems were within one category of each other and 71% of the ratings were within two categories of each other on a seven point scale. However, a second measure of r e l i a b i -l i t y , the Pearson c o r r e l a t i o n s t a t i s t i c , indicates v i r t u a l l y no s i g n i f i c a n t c o r r e l a t i o n s between the ratings made of groups on the two d i f f e r e n t problems (see Table 8). C a p i t a l punishment and immigration: MANOVA and ANOVA's on the combined  data. The MANOVA r e s u l t s showed no s i g n i f i c a n t main or i n t e r a c t i o n e f f e c t s . (It should be noted that a d d i t i o n a l analyses were performed on the data that may be of i n t e r e s t to the r e a d e r — s e e Appendix 0.) In summary, the analysis of the decision outcome data shows expert ratings that c o r r e l a t e moderately within each of the two problems but have a low c o r r e l a t i o n across the two problems. The MANOVA r e s u l t s from.the one c a p i t a l punishment expert provide tentative support for the leadership main e f f e c t r e s u l t s found when t e s t i n g the data on the groupthink symptoms. The MANOVA r e s u l t s from the remaining experts ind i c a t e a lack of s i g n i f i c a n t findings. - 81 -Mul t i p l e Regression P r e d i c t i o n of Quality of Decision Outcome Since these analyses were performed on a post hoc ba s i s , there were no a p r i o r i p r e d i c t i o n s . P r e d i c t i o n based upon the c a p i t a l punishment data. The step-wise r e -gression analysis indicated that the following f i v e v a r i a b l e s made s i g n i -f i c a n t contributions to the p r e d i c t i o n of the q u a l i t y of the decision: (1) motivational l e v e l , (2) leadership s t y l e ( i . e . , assertive vs. non-assertive), (3) the i n t e r a c t i o n of L.P.C. with G.A.S., (4) the degree of perceived time pressure, and (5) the sex of the group leader (see Table 9). Increased moti-vation, non-assertive leaders, increased L.P.C./G.A.S. scores, decreased time pressure and male group leaders were a l l p o s i t i v e l y correlated with increased ratings i n the q u a l i t y of the decisions (see Table 10). The information contained i n Table 9 shows that these factors only ac-count f o r 23% of the variance associated with p r e d i c t i n g the q u a l i t y of the decision. When considering the contributions that the va r i a b l e s make to 2 increasing the amount of variance accounted f or (R ), i t appears that moti-v a t i o n a l l e v e l i s the only n o n - t r i v i a l predictor of decision q u a l i t y f o r the c a p i t a l punishment problem. P r e d i c t i o n based upon the immigration data. The step-wise regression analysis indicated that the following s i x va r i a b l e s made s i g n i f i c a n t c o n t r i -butions to the p r e d i c t i o n of the q u a l i t y of the decis i o n : (1) not belonging to a m i l i t a r y organization, (2) the l e v e l of concurrence seeking, (3) the i n t e r a c t i o n of the antecedent conditions with leadership s t y l e , (4) the degree of perceived time pressure, (5) L.P.C, score, and (6) belonging to a mental health organization (see Table 11). Not being i n the m i l i t a r y , increases i n concurrence seeking, decreases i n the groupthink condition (antecedent conditions present/leadership s t y l e a s s e r t i v e ) , increases i n time pressure, increases i n L.P.C. scores, and belonging to a mental health - 82 -Table 9 Step-Wise Mul t i p l e Regression P r e d i c t i o n of Decision Outcome for C a p i t a l Punishment Variable M u l t i p l e Increase i n Name R R R Anova F Ratio Motivational Level 0.3482 0.1212 0.1212 6.207b Leadership Style 0.3971 0.1577 0.0364 4.118° Interaction L.P.C./G.A.S. 0.4394 0.1931 0.0354 3.430d Time Pressure 0.4660 0.2172 0.0240 2.913 e Leader's Sex 0.4803 0.2307 0.0135 2.459 f " A l l F_ r a t i o s s i g n i f i c a n t at .05 l e v e l b d.f. = 1,45 ° d.f. = 2,44 d d.f. = 3,43 6 d.f. = 4,42 f d.f. = 5,41 - 83 -Table 10 Correlations Between Predictor Variables and C r i t e r i o n Variables Predictor Variables' Overall Quality of Ov e r a l l Quality of Immigration Decision C a p i t a l Punishment Decision Antecedent Conditions -0.092 -0.012 Leaderstyle (assigned) -0.133 -0.077 Interaction Antec./Leader -0.278 -0.036 L.P.C. 0.224 0.039 G.A.S. 0.175 0.165 Interaction L.P.C./G.A.S. 0.222 0.207 Cohesiveness 0.047 -0.032 Concurrence Seeking 0.259 0.078 Leader's Sex -0.194 0.120 Mental Health Organiz. 0.370 0.025 M i l i t a r y Organization -0.374 -0.115 Leaderstyle (as rated) 0.173 0.150 Motivational Level -0.264 -0.348 Time Pressure 0.328 -0.061 - 84 -Table 11 Step-Wise Mul t i p l e Regression P r e d i c t i o n of Decision Outcome for Immigration Variable Multiple Increase i n Name R R R Anova F Ratio M i l i t a r y Organization 0.3737 0.1396 0.1396 7.303 Concurrence Seeking 0.4599 0.2115 0.0719 5.901 Time Pressure 0.5120 0.2621 0.0506 5.092 Interaction Antecedent Conds./Leadership Style 0.5775 0.3335 0.0714 5.254 L.P.C. 0.6117 0.3741 0.0406 4.902 Mental Health Organization 0.6295 0.3962 0.0221 4.375 g a l l F_ r a t i o s s i g n i f i c a n t at .01 l e v e l d.f. = 1,45 d.f. = 2,44 d.f. = 3,43 d.f. = 4,42 d.f. = 5,41 d.f. =6,40 - 85 -organization were a l l p o s i t i v e l y correlated with increased ratings i n the q u a l i t y of the decisions (see Table 10). The information contained i n Table 11 shows that these factors account f o r 40% of the variance i n p r e d i c t i n g the q u a l i t y of the decision. However, i t appears that not belonging to a m i l i t a r y organization i s the only v a r i a b l e that accounts f or a n o n - t r i v i a l 2 amount of variance (R ) i n the p r e d i c t i o n of decision q u a l i t y . Of i n t e r e s t i s the fact that the negative c o r r e l a t i o n between the presence of the group-think conditions and decision q u a l i t y supports the predictions of groupthink theory; at the same time, the p o s i t i v e c o r r e l a t i o n between concurrence seeking and decision q u a l i t y i s contrary to the predictions of groupthink theory. (Both of these factors of concurrence seeking and the presence of groupthink conditions contribute only t r i v i a l amounts to the variance asso-ciated with p r e d i c t i n g decision q u a l i t y — s e e Table 11.) Pr e d i c t i o n across both problems: Canonical c o r r e l a t i o n a n a l y s i s . A canonical c o r r e l a t i o n analysis was performed using the fourteen p r e d i c t o r items (see Table 10) as one set of v a r i a b l e s i n the a n a l y s i s . The second set of variables i n the analysis was the c a p i t a l punishment and immigration ratings on the o v e r a l l q u a l i t y of the decisions. The r e s u l t s indicated that neither of the two canonical c o r r e l a t i o n s achieved s i g n i f i c a n c e . It was of i n t e r e s t to note that not belonging to a m i l i t a r y organization had the largest regression weight i n the canonical construction where the predictor variables correlated most highly with the immigration v a r i a b l e , This sup-ports the r e s u l t s found under the step-wise multiple regression a n a l y s i s . S i m i l a r l y of i n t e r e s t was the f i n d i n g that the fa c t o r of motivation had the largest consistent multiple regression weights across both canonical con- • st r u c t i o n s . This also supports the findings of the step-wise multiple r e -gression analysis. In summary, the multiple regression and canonical c o r r e l a t i o n analyses - 86 -i n d i c a t e that two d i f f e r e n t v a r i a b l e s y i e l d the best predictions of decision q u a l i t y for the two d i f f e r e n t problems. On the c a p i t a l punishment problem increases i n motivational l e v e l were associated with increased ratings i n d e c i s i o n q u a l i t y . On the immigration problem not belonging to a m i l i t a r y organization was associated with increased ratings i n decision q u a l i t y . The canonical c o r r e l a t i o n analysis suggested that of the two dominant pre d i c t o r v a r i a b l e s , motivational l e v e l was l i k e l y to provide the strongest help i n p r e d i c t i n g decision q u a l i t y across both problems. Discussion Concurrence Seeking and Groupthink Symptoms The o v e r a l l r e s u l t s do not provide support for the theory of groupthink as proposed by Janis (1972; Janis and Mann, 1977). The analyses indicated that i n contrast to the predictions of groupthink theory, the combination of an a s s e r t i v e leadership s t y l e with cohesiveness and perceived time stress did not r e s u l t i n s i g n i f i c a n t l y higher than normal l e v e l s of concurrence seeking among the members of these groups. S i m i l a r l y , i t was found that i n contrast to groupthink p r e d i c t i o n s , there was no s i g n i f i c a n t i n t e r a c t i o n between leadership s t y l e and the antecedent conditions r e s u l t i n g i n the production of groupthink symptoms. However, the analyses did demonstrate that leadership s t y l e by i t s e l f had a s i g n i f i c a n t e f f e c t upon several of the group processes. In comparison to groups led by non-assertive leaders, groups under assertive leaders f e l t a s i g n i f i c a n t l y greater degree of conformity pressure and viewed the d e c i -sions that were reached as being s i g n i f i c a n t l y more r i s k y . Conformity pressure i s one of the eight groupthink symptoms outlined by Janis (1972), and r i s k taking i s combined with high optimism to also y i e l d a groupthink - 87 -symptom concerning the i l l u s i o n of i n v u l n e r a b i l i t y . While the r e s u l t s i n d i -cate that leadership s t y l e had a s i g n i f i c a n t e f f e c t upon the groupthink symp-tom of conformity pressure, there i s a reservation concerning the e f f e c t of leadership s t y l e upon the i l l u s i o n of i n v u l n e r a b i l i t y . Members under asser-t i v e leaders perceived the decisions to be more r i s k y than did members under non-assertive leaders, but ratings by experts external to the group, did not y i e l d the same s i g n i f i c a n t d i f f e r e n c e between the two leadership s t y l e s . In addition, the second h a l f of the i l l u s i o n of i n v u l n e r a b i l i t y i s constituted by a high degree of optimism. Members under assertive leaders were not s i g n i -f i c a n t l y more o p t i m i s t i c concerning the decisions that they reached than were members under non-assertive leaders. Therefore, while one can consider leadership s t y l e to have had a s i g n i f i c a n t e f f e c t upon the r i s k i n e s s of the decisions as perceived by the members, one cannot consider an i l l u s i o n of i n v u l n e r a b i l i t y to have existed as a r e s u l t ^ o f leadership s t y l e i n these groups. In summary, groups under groupthink conditions were not found to display more concurrence seeking and were not found to exhibit more groupthink symp-toms than groups under non-groupthink conditions. At the same time, leader-ship s t y l e was found to have a s i g n i f i c a n t e f f e c t i n causing groups under ass e r t i v e leaders, i n comparison to non-assertive leaders, to f e e l greater conformity pressure and to view the decisions reached as being more r i s k y . These r e s u l t s do not support the predictions of groupthink theory, but they do i n d i c a t e some p o t e n t i a l problem areas within discussion groups where the leader i s strongly assertive about h i s or her point of view and strongly expresses that view at the beginning of the decision making session. Quality of the Decision Making Process In a d d i t i o n to the group processes measured by the groupthink symptoms, - 88 -t h i s study also tested s i x a d d i t i o n a l measures of the q u a l i t y of the decision making process. Again, i n contrast to the predictions made by groupthink theory, there was no s i g n i f i c a n t i n t e r a c t i o n between leadership s t y l e and the antecedent conditions of cohesiveness and time s t r e s s . That i s , there was no s i g n i f i c a n t d i f f e r e n c e i n the q u a l i t y of the decision processes found under .: the groupthink and non-groupthink conditions. However, there was a s i g n i -f i c a n t e f f e c t on two of the s i x measures when high-cohesive, high-time-stress groups (antecedent conditions present) were compared to low-cohesive, low-time-stress groups (antecedent conditions absent). These r e s u l t s suggested that groups under conditions of high cohesiveness and time stress were les s e f f e c t i v e i n terms of the amount of information brought into the discussion and l e s s e f f e c t i v e i n t h e i r consideration of p o t e n t i a l negative and p o s i t i v e consequences associated with a l t e r n a t i v e solutions. These r e s u l t s were consistent across the ratings and analyses of three d i f f e r e n t independent observers. In addition, there was a very strong i n d i c a t i o n that the cohesive, time stressed groups produced fewer solutions on the average than did the non-cohesive, non-time stressed groups. On these measures of group process the o v e r a l l p i c t u r e i s consistent i n showing the non-cohesive, non-time stressed groups to be more e f f e c t i v e than the cohesive, time stressed groups. Since most research on cohesiveness indicates that cohesive groups are, i n general, more e f f e c t i v e than non-cohesive groups (e.g., Berkowitz, 1954; Fisher, 1974; Katz & Kahn, 1978) i t i s highly l i k e l y that t h i s somewhat sur-p r i s i n g f i n d i n g i s a r e s u l t of the time manipulation rather than the cohe-siveness manipulation i n the antecedent conditions. The multiple regression analyses support the i n t e r p r e t a t i o n that the decreased effectiveness of the groups was caused by the time l i m i t a t i o n rather than the existence of cohe-siveness. Thus, while cohesive groups may normally function more e f f e c t i v e l y than non-cohesive groups, even the effectiveness of cohesive groups may be - 89 -s e r i o u s l y undermined by strong time pressures. In f a c t , these r e s u l t s i n d i c a t e that i f the time pressures upon cohesive groups are strong enough those groups may be less e f f e c t i v e than non-cohesive groups not operating under such time r e s t r a i n t s . The e f f e c t s of time l i m i t a t i o n s on decision making may operate through intervening variables such as stress or arousal l e v e l . However, i n the pre-sent study such an i n t e r p r e t a t i o n i s not consistent with the e a r l i e r r e s u l t s i n d i c a t i n g that :the s i g n i f i c a n t differences i n perceived time pressure found between groups did not manifest themselves as s i g n i f i c a n t differences i n o v e r a l l l e v e l s of stress. I t i s more l i k e l y that the strong time pressures diminished the group's eff e c t i v e n e s s , at least i n the present case, by simply putting p h y s i c a l r e s t r i c t i o n s upon the amount of information that could be processed by the group. For example, time pressures may l i m i t the f e a s i b i l i t y and ease of bringing outside relevant information into the discussion, or of generating and evaluating a large number of a l t e r n a t i v e p o t e n t i a l solutions. It i s valuable to consider the r e s u l t s concerning the groupthink symp-toms and the s i x a d d i t i o n a l group processes within the context of broader theories on group dynamics. Research and theory stemming from Bales (1950) i n the early f i f t i e s through to more current l i t e r a t u r e on groups and leader functioning (e.g., Bales, 1970; Blake and Mouton, 1969; F i e d l e r , 1967; Katz and Kahn, 1978) are consistent i n i d e n t i f y i n g at least two main aspects of group functioning: task r e l a t e d features and s o c i a l or inter-personal r e l a t i o n s features. Task related aspects of group functioning are concerned with such a c t i v i t i e s as s e t t i n g objectives, d e f i n i n g goals and engaging i n goal-oriented behaviour. Inter-personal r e l a t i o n s within groups are usually broadly defined i n terms of how the group members work together, or what the group atmosphere i s l i k e ; for example, i n d i v i d u a l s o f f e r i n g encouragement or expressing f e e l i n g s of concern over what they f e e l are r i s k y decisions being - 90 -made by the group. These o f f e r s of encouragement and f e e l i n g s of concern influence what could be c a l l e d the a f f e c t i v e or socio-emotional aspects of the group's functioning, i n contrast to the task oriented features of the group's functioning. Within t h i s context, i t appears as though the group-think symptoms that play a dominant r o l e i n the theory of Janis (1972; Janis and Mann, 1977) are b a s i c a l l y concerned with the nature of the group's a f f e c t i v e atmosphere. Feelings of optimism, moral correctness, unanimity and conformity pressure are descriptions of the psychological mood of the group. The s i x a d d i t i o n a l group processes analyzed i n t h i s study l a r g e l y f a l l within the task-oriented measures of group processes. Measures of how many solutions were produced, how much information was considered and to what extent the consequences of various a l t e r n a t i v e s were considered, are i n d i c a t o r s of how well the group was functioning i n i t s task of producing the best possible s o l u t i o n to the problem. The r e s u l t s of the present experiment i n d i c a t e that leadership s t y l e had a s i g n i f i c a n t e f f e c t on the a f f e c t i v e measures of the group processes whereas the antecedent conditions, i n p a r t i c u l a r time pressure, had a s i g n i f i c a n t e f f e c t on the task oriented measures of the group processes. Conformity pressure and perceived r i s k i n e s s of the decision were influenced by leader-ship s t y l e , while the amount of information discussed, the extensiveness of reviewing the decision consequences and the number of solutions proposed were influenced by the cohesiveness and time constraints e x i s t i n g within the groups. I t i s i n t e r e s t i n g that leadership s t y l e and the antecedent conditions had e f f e c t s upon d i f f e r e n t aspects of the group's functioning. The i n i t i a l e f f e c t s of strong time r e s t r i c t i o n s may show up f i r s t i n measures of the group's task r e l a t e d effectiveness. I t may be only under conditions of higher motivation or stress than existed i n t h i s study where strong time r e s t r i c t i o n s produce a second e f f e c t , that i s , where they influence the - 91 -group's a f f e c t i v e functioning. With respect to leadership, i t may be the case that i n t h i s s i t u a t i o n the a s s e r t i v e leaders were seen as being some-what obnoxious or "pushy" members of the group, thus exerting a negative influence on the measures of the group atmosphere. However, member moti-vati o n or involvement may not have been strong enough, or as Flowers (1977) pointed out concerning her study, the power of the group leader may not have been strong enough to compel the members to "toe the l i n e " with the leader's views, and thus also demonstrate the negative influence of leadership s t y l e upon the measures of task effectiveness. In comparison to the present study, Flowers' (1977) research demon-strated the e f f e c t s of leadership s t y l e upon task functioning (e.g., the number of solutions proposed, and the number of " r o l e sheet" fa c t s consi-dered) but showed no leadership or cohesiveness e f f e c t upon the a f f e c t i v e functioning of the group (e.g., conformity pressure, measured by Flowers as the "perceived freedom to express opinions"). There may be several reasons for the differences between Flowers' r e s u l t s and the r e s u l t s of t h i s study. Flowers found a "bottoming e f f e c t " on her measure of conformity pressure: no one seemed to f e e l any pressure at a l l . Perhaps the assertive leaders i n the present study were more i n s i s t e n t i n t h e i r views and thus exerted more conformity pressure upon the members. I f t h i s i s the case, i t i s d i f f i c u l t to explain why the assertiveness by leaders i n t h i s study did not a f f e c t the task oriented dimensions of the group processes. As mentioned above, i t may simply be due to the lack of a strong enough involvement on the part of the group members or the lack of enough power i n the r o l e of the group leader. The o r i g i n a l cases studied by Janis (1972) were analyzed predominantly with respect to the e f f e c t of leadership s t y l e , cohesiveness and stress i n producing dysfunctions i n the a f f e c t i v e aspects of the group processes - 92 -( i . e . , i n producing the groupthink symptoms). It would be i n t e r e s t i n g i f the cases were reanalyzed more s p e c i f i c a l l y i n terms of the e f f e c t of the group-think preconditions upon the task r e l a t e d aspects of the group's functioning. Quality of the Decision Outcome The r e s u l t s stemming from the analyses of the q u a l i t y of the decisions made by the groups c l e a r l y indicated one thing, that i t i s d i f f i c u l t to obtain agreement between experts i n t h e i r judgments on decision q u a l i t y . This i s not a s u r p r i s i n g f i n d i n g given the experiences of previous researchers with the ratings of experts (cf. Carr, Green, and Hinckle, 1976; Courtrlght, 1978). The experiences of the present experimenter also c l e a r l y indicated that the majority of';the group leaders i n the experiment were s e r i o u s l y lacking i n some leadership s k i l l s . For example, the a b i l i t y of the majority of the group leaders was so poor i n adequately conveying to the experimenter the decisions reached by the group, p a r t i c u l a r l y with respect to conveying the t o t a l extent of what was decided, that the q u a l i t y of the d e c i s i o n reached by a group was often highly confounded by the s k i l l of the leader i n being able to properly state the group's p o s i t i o n . This, i n turn, increased the d i f f i c u l t y of f i n d i n g s i g n i f i c a n t differences between groups i n the q u a l i t y of the decisions produced. In s p i t e of these l i m i t a t i o n s , the analysis of the ratings of one of the c a p i t a l punishment experts demonstrated a main e f f e c t for leadership s t y l e , with groups led by a s s e r t i v e leaders producing decisions that were i n f e r i o r i n o v e r a l l q u a l i t y to the decisions produced by groups under non-assertive leaders. This f i n d i n g i s s i m i l a r to the analyses of the groupthink symptoms, where a main e f f e c t f o r leadership s t y l e was also found. In both cases the non-assertive leadership s t y l e functioned better than the assertive leader- . ship s t y l e . As mentioned previously, these r e s u l t s are also consistent with - 93 -Flowers' (1977) research i n that she also found a main e f f e c t f o r leadership s t y l e , with a s s e r t i v e leaders f a r i n g poorer. The consistency of r e s u l t s within and across studies suggests that leadership s t y l e may have been the predominant factor operating i n Janis' (1972) cases of groupthink i n action. As noted i n the introduction to t h i s study and as pointed out by Flowers (1977), there may be reason to question the c e n t r a l r o l e assigned to group cohesiveness by Janis (1972). I t w i l l be valuable to see i f future experi-mental research on groupthink continues to f i n d leadership to be a major factor and continues to question the r o l e of cohesiveness i n the theory. Under those circumstances groupthink theory may simply become a s p e c i a l case of one of the more general theories on leadership. The f i n a l set of r e s u l t s concerns the post hoc analysis of s i g n i f i c a n t predictors of d e c i s i o n q u a l i t y . The r e s u l t s across the multiple regression and canonical c o r r e l a t i o n analyses were not consistent i n demonstrating one v a r i a b l e or set of variables to be the major predictors of decision q u a l i t y . Motivational l e v e l appeared to be the best p r e d i c t o r for the c a p i t a l punish-ment problem and the type of organization belonged to (or not belonged to) appeared to be the best predictor for the immigration problem. Motivational l e v e l may be the best common predictor but the evidence i s very weak for such a statement. The lack of cross-problem v a r i a b l e s for the p r e d i c t i o n of decision q u a l i t y may stem from the f a c t that the i n t e r - r a t e r c o r r e l a t i o n s of decision q u a l i t y were moderate within problems but very low between problems. In t h i s study i t seems that both the judgments of decision q u a l i t y and the predictors of that q u a l i t y were s p e c i f i c to the type of problem encountered by a group. In other words, a group faced with the two problems of immi-gration and c a p i t a l punishment may do quite well on one problem and quite poorly on the second problem. In a d d i t i o n , i f we t r i e d to p r edict decision q u a l i t y for a s p e c i f i c group, we may be best off by matching the type of - 94 -predictor v a r i a b l e to the type of problem. Such f i n d i n g s , i f consistent, may r a i s e serious problems i n e s t a b l i s h i n g the type of group conditions most conducive to high q u a l i t y decision making across d i f f e r e n t problems. It i s i n t e r e s t i n g that at least with some s o c i a l problems, being i n an organization other than the m i l i t a r y increases the p o s s i b i l i t y of obtaining higher q u a l i t y solutions. This may be a comment upon the p o t e n t i a l disad-vantages of having the m i l i t a r y i n a p o s i t i o n where they make decisions about such s o c i a l problems as immigration. There are several implications i f motivation and the organization be-longed to are major predictors of decision q u a l i t y . F i r s t , research on decision making, and groupthink i n p a r t i c u l a r , should e x p l i c i t l y take these va r i a b l e s into consideration (cf. Courtright, 1978; Flowers, 1977). Second, research by Gouran, Brown and Henry (1978) suggests that i n a decision making discussion, task oriented contributions by group members influenced the per-ception of the q u a l i t y of the discussion more than inter-personal r e l a t i o n s contributions. Thus, it.may be only the task oriented aspects of motivation and the organization belonged to that w i l l be perceived to a f f e c t the q u a l i t y of the decisions reached. The multiple regression and canonical c o r r e l a t i o n analyses are also valuable i n o u t l i n i n g the kinds of factors that do not seem to be important explanations of decision q u a l i t y . For example, i n t h i s study i t was more valuable to speak of motivational l e v e l than groupthink when attempting to explain the decision making. The use of such analyses i n future studies on decision making may also help to i n d i c a t e where i t i s and i s not valuable to use explanatory concepts such as groupthink. A number of reservations concerning t h i s research and the i n t e r p r e t a -tions of the r e s u l t s need to be considered. F i r s t , while there may be some question.concerning the way i n which Janis (1972) selected cohesiveness as a - 95 -major factor i n groupthink, i t i s also c l e a r that there may be some question concerning the effectiveness of the manipulations of group cohesiveness within the Courtright (1978) study and the present study. Second, there are l i m i t a t i o n s within both Courtright's study and t h i s study due to the mod-erate r e l i a b i l i t i e s across the ratings of experts. These r e l i a b i l i t i e s may i n d i c a t e enough error variance e x i s t i n g i n the data c o l l e c t e d that the e f f e c t s of any factor such as groupthink would have to be quite powerful i n order to be observed. The t h i r d reservation concerns the t h e o r e t i c a l problems associated with judging the q u a l i t y of a decision. As indicated e a r l i e r , with respect to unstructured and complex problems the judgment of decision or s o l u t i o n qua-l i t y may be best decided with reference to the q u a l i t y of the decision making process. This i s p a r t i c u l a r l y true when considering a s i n g l e d e c i s i o n rather than a se r i e s of decisions made by the same group. At the same time though, there has been some development i n the assessment of decision q u a l i t y based upon the decision i t s e l f . The work of F i e d l e r (1967) indicated that solutions for structured problems, that have correct or optimal outcomes, were judged on the basis of "goodness of f i t " to the required s o l u t i o n . Solutions or decisions concerning unstructured problems where there was no " c o r r e c t " ans-wer were rated by experts as to t h e i r q u a l i t y . Leathers (1972) has c a r r i e d t h i s process a step further by breaking the general r a t i n g of decision q u a l i -ty i n t o a s e r i e s of components, such as r a t i n g the c r e a t i v i t y and comprehen-siveness of the decisions. At t h i s stage we have moved beyond r a t i n g decision q u a l i t y by considering decision process to a consideration of the content of the decision. This decision content i s open to ratings on q u a l i t y that should be independent of the nature.of the decision;- for example,.the. ratings should be independent of whether the decision made on c a p i t a l punish-ment i s for or against the death penalty. More research i s required to l i n k - 96 -ratings of the q u a l i t y of the decision process to ratings on the q u a l i t y of the decision content. In add i t i o n , i t may also prove very valuable to use a large sample of decisions, and assess the r e l a t i o n s h i p that e x i s t s between the ratings of the q u a l i t y of decision content and some measure of the successfulness of the decision outcome. In the present study we have measured decision q u a l i t y by considering expert ratings on s i x components of decision q u a l i t y for each of the deci-^:'.. sions that were made by the groups. However, there may be questions con-cerning the basis upon which the experts made these ratings. For example, with respect to the immigration experts, were the judgments of what c o n s t i -tuted c r e a t i v i t y i n the solutions, themselves a product of some group dynamic within the organization? This question and the p o t e n t i a l r a t i n g bias could be answered by c o r r e l a t i n g the ratings of experts on immigration within one organization to the ratings of experts within a d i f f e r e n t organization. Within the context of questioning decisions that may be the product of group dynamics, i t could be asked whether the decisions associated with the carrying out of present research were the r e s u l t of groupthink within the supervisory group. However, t h i s would not seem to be the case. While there was a cohesiveness within the group, there was no assertive s t r e s s i n g by one i n d i v i d u a l that the research should be c a r r i e d out i n one p a r t i c u l a r way, nor was there a lack of input from experts external to ;the committee. A fourth reservation concerning the study i s the fac t that the multiple regression analyses were done post hoc and were performed on a very small sample s i z e f o r that type of an a l y s i s . The r e s u l t s may be quite unstable and need to be v e r i f i e d as a" p r i o r i predictions on new samples of groups engaged in..decision making. Last, there must be serious reservations about a l l three experimental studies (Courtright, 1978; Flowers, 1977; and the present research) with - 97 -respect to the external v a l i d i t y of the experimental s i t u a t i o n s . One major difference between these studies and the cases analyzed by Janis (1972) con-cerns the nature of the consequences involved f o r the group members i n ac-cepting any given s o l u t i o n as t h e i r preferred choice. The solutions being proposed by subjects within the various experiments had very l i t t l e impact i n terms of personal consequences for the p a r t i c i p a n t s . In addition, as Flowers (1972) noted, the leaders i n the o r i g i n a l groupthink cases wielded immensely more power over t h e i r group members than did the leaders under the experimental conditions. The o v e r a l l e f f e c t of these reservations concerning the experimental studies i s that they may well represent inadequate simulations of the o r i g i n a l groupthink conditions. In f a c t , given the lack of demonstration of the existence of groupthink within these three experimental studies i t may be somewhat f r u i t l e s s continuing to study groupthink within the laboratory. A more r e a l i s t i c t e s t i n g of groupthink theory may have to wait u n t i l i t can be conducted using normal groups functioning within t h e i r t y p i c a l working envi-ronments. By doing so we may determine whether groupthink i s involved i n the functioning of groups under everyday conditions. A negative answer to that question would s t i l l mean that we would need to determine whether or not groupthink acted as a s p e c i a l f a c t o r operating only under more c r i s i s - l i k e conditions involving highly cohesive groups, highly a s s e r t i v e leaders, high stress (and perhaps, conditions high i n personal consequences for the group members). At present,' what we know i s that there i s l i t t l e evidence that groupthink operates within groups functioning under experimental laboratory conditions. In conclusion, I would l i k e to make two general remarks concerning groupthink theory and the way i n which i t has been studied. The theory of groupthink has been a provocative entry into the f i e l d of understanding major - 98 -p o l i t i c a l events. I t has stimulated a great deal of i n t e r e s t ; that i n turn has led to research on the theory. I t may be unfortunate that the theory was soon subjected to two problems that seem commonly generated.by researchers i n psychology. F i r s t , psychology i s strongly dominated by an experimental, laboratory oriented approach to research. At no time i s t h i s code of behavior more vigorously enforced than when the neophyte i n the f i e l d attempts to demon-st r a t e h i s or her research p r o f i c i e n c y i n order to complete a Ph.D. and gain acceptance into the profession. I t may not be e n t i r e l y c o i n c i d e n t a l that the three tests of the theory that were b a s i c a l l y negative i n outcome were highly experimental i n o r i e n t a t i o n and done as doctoral d i s s e r t a t i o n s (Courtright, 1978; Flowers, 1977; and the present study), whereas the strongest support for the theory came from a non-experimental study not done under the requirements of a doctoral d i s s e r t a t i o n (Tetlock, 1979). The point i s that groupthink theory i s psychological i n nature; t h i s resulted i n i t being tested through the experimental/laboratory approaches to research t r a d i t i o n a l to t h i s d i s c i p l i n e . Such a fate may not have been the most appropriate way to judge the v e r a c i t y of the theory. Given the nature of the events studied by Janis (1972) and the contrast between groups under i n t e r n a t i o n a l c r i s i s conditions and groups i n an experiment, i t was perhaps misleading to rush Janis' theory into the laboratory for study. Thorngate (1980), i n reviewing a recent experimentally oriented text on the s o c i a l psychology of decision making, strongly suggests that t h i s area i s stag-nating because of our lack of research i n more appropriate " r e a l l i f e " s e ttings. Second, and more minor, i s the problem that i n our haste to experimen-t a l l y test theories we sometimes leap into t e s t i n g the implications of the theory before t e s t i n g the theory i t s e l f . For example, both. Flowers (1977) - 99 -and Courtright (1978) concentrated on implications such as "groups s u f f e r i n g from groupthink w i l l display fewer statements of disagreement" or " w i l l d i s -cuss less a v a i l a b l e information before the decision than w i l l non-groupthink groups". At the same time both Flowers and Courtright missed t e s t i n g two major components of the theory. One component stated that high l e v e l s of concurrence seeking would be the r e s u l t of s p e c i f i c preconditions and would lead to groupthink. The second component stated that groupthink i s defined p r i m a r i l y through the existence of the groupthink symptoms. It seems to the present author that the theory of groupthink deserves a better hearing through research more c l o s e l y tuned to Janis'(1972) l e v e l of analysis and more c l o s e l y tuned to the s p e c i f i c s of the predictions made by groupthink theory. - 100 -References Ackoff, R. L., & Sasieni, M. W. Fundamentals of operations research. New York: Wiley, 1968. Alexander, E. R. The design of a l t e r n a t i v e s i n organizational contexts: A p i l o t study. Administrative Science Quarterly, 1979, 7A_, 382-404. A l l i s o n , G. T. Essence of d e c i s i o n : Explaining the Cuban m i s s i l e c r i s i s . Boston: Little-Brown, 1971. Back, K. The exertion of influence through s o c i a l communication. Journal of Abnormal and S o c i a l Psychology, 1951, ^6_, 9-24. Bales, R. Personality and interpersonal behavior. New York: Holt, Rinehart, & Winston, 1970. Bales, R. Interaction process a n a l y s i s . Reading, Mass.: Addison-Wesley, 1950. B e l l , D. E., Keeney, R. L., & R a i f f a , H. (Eds.), C o n f l i c t i n g objectives i n decisions. New York: Wiley, 1977. Berkowitz, L. Group standards, cohesiveness, and p r o d u c t i v i t y . Human Relations, 1954, ]_, 509-519. Blake, R., & Mouton, J. S. Building a dynamic corporation through g r i d organization development. Reading, Mass.: Addison-Wesley, 1969. Bonczek, R. H., Holsapple, C. W., & Whinston, A. B. Computer-based support of organizational decision making. Decision Sciences, 1979, 10_, 268-291. Carr, D. F., Green, T. B., & Hinckle, T. W. Exploring nominal grouping from a s o c i a l f a c i l i t a t i o n context. In D. F. Ray and T. B. Green (Eds.), Expanding dimensions of management thought and action. M i s s i s s i p p i State University: Southern Management Association, 1976. Coser, L. A. The functions of s o c i a l c o n f l i c t . Glencoe, 111.: Free Press, 1956. - 101 -Courtright, J . A. A laboratory i n v e s t i g a t i o n of groupthink. Communication  Monographs, 1978, 45, 229-245. Dickerson, S. L., & Robertshaw, J . E. Planning and design. Toronto: Lexington Books, 1975. DuBrin, A. J . Fundamentals of organizational behavior: An applied  perspective. (2nd ed.) New York: Pergamon Press, 1978. Ferguson, C. K., & K e l l y , H. H. S i g n i f i c a n t factors i n overevaluation of own-group's product. Journal of Abnormal and S o c i a l Psychology, 1964, 69, 223-228. Festinger, L. Informal s o c i a l communication. Psychological Review, 1950, 57, 271-282. Festinger, L., Schachter, S., & Back, K. S o c i a l pressures i n informal  groups: A study of human factors i n housing. New York: Harper & Brothers, 1950. F i e d l e r , F. E. The e f f e c t of leadership and c u l t u r a l heterogeneity on group performance: A test of the Contingency Model. Journal of Experimental  S o c i a l Psychology, 1966, 2, 237-264. F i e d l e r , F. E. A theory of leadership effectiveness. New York: McGraw H i l l , 1967. F i e d l e r , F. E. V a l i d a t i o n and extension of the Contingency Model of leader-ship effectiveness: A review of empirical f i n d i n g s . Psychological  B u l l e t i n , 1971, 76, 128-148. F i e d l e r , F. E., & Chemers, M. M. Leadership and e f f e c t i v e management. Glenview, 111.: Scott-Foresman, 1974. Fisher, B. A. Small group decision making: Communication and the group  process. New York: McGraw-Hill, 1974. Flowers, M. L. A laboratory t e s t of some implications of Janis's groupthink hypothesis. Journal of Personality arid S o c i a l Psychology, 1977, 35, 888-- 102 -896. Gibson, J . L., Ivancevich, J . M., & Donnelly, J . H., J r . Organizations:  Behavior, structure, processes. (3rd. ed.) Dallas Texas: Business Publi c a t i o n s , 1979. Gouran, D. S., Brown, C , & Henry, D. R. Behavioral correlates of perceptions of q u a l i t y i n decision making discussions. Communication Monographs, 1978, 45, 51-63. G u i l f o r d , J . P. Fundamental s t a t i s t i c s i n psychology and education. (4th ed.) New York: McGraw-Hill, 1965. Hoffman, R. L. Group problem solving. In L. Berkowitz (Ed.), Group processes. New York: Academic Press, 1978. Janis, I. L. Groupthink among p o l i c y makers. In N. Sanford and C. Comstock (Eds.), Sanction for e v i l . San Francisco: Jossey-Bass, 1971. Janis, I. L. Victims of groupthink: A psychological study of f o r e i g n - p o l i c y decisions and fiascoes. New York: Houghton M i f f l i n , 1972. Janis, I. L., & Mann, L. Decision making: A psychological analysis of c o n f l i c t , choice, and commitment. New York: Free Press, 1977. J u l i a n , J . W., Bishop, D. W. , & F i e d l e r , F. E. Quasi-therapeutic e f f e c t s on intergroup competition. Journal of Personality and S o c i a l Psychology, 1966, 3, 321-327. Katz, D., & Kahn, R. L. The s o c i a l psychology of organizations. New York: Wiley, 1978. King, W. R. Developing u s e f u l management decision support systems. Manage- ment Decision, 1978, 16, 262-273. La V a l l e , I. H. Fundamentals of d e c i s i o n a n a l y s i s . New York: Holt, Rine-hart & Winston, 1978. Leathers, D. G. Quality of group communication as a determinant of group product. Speech Monographs, 1972, _39, 166-1973. - 103 -Le Vine, R. A., & Campbell, D. T. Ethnocentrism. New York: Wiley, 1972. Lombardo, J. P., Weiss, R. F., & S t i c h , M. H. Effectance reduction through speaking i n reply and i t s r e l a t i o n to a t t r a c t i o n . Journal of Personality and S o c i a l Psychology, 1973, 28, 325-332. Lott, A. J . , & L o t t , B. E. Group cohesiveness as interpersonal a t t r a c t i o n : A review of r e l a t i o n s h i p s with antecedent and consequent v a r i a b l e s . Psychological B u l l e t i n , 1965, 6_4, 259-309. MacCrimmon, K. R., & Taylor, R. N. Decision making and problem solving. In M. D. Dunnette (Ed.), Handbook of i n d u s t r i a l and organizational psychology. Chicago: Rand McNally, 1976. Miner, F. C , J r . , A comparative analysis of three diverse group decision making approaches, Academy of Management Journal, 1979, 22^ , 81-93. Osgood, C. E. The representational model and relevant research methods. In I. de S. Pool (Ed.), Trends i n content a n a l y s i s . Urbana: University of I l l i n o i s Press, 1959. Raven, B. H. The Nixon group. Journal of S o c i a l Issues, 1974, 30, 297-320. Robinson, J . S., & Snyder, R. C. Decision making i n i n t e r n a t i o n a l p o l i t i c s . In H. C. Kelman (Ed.), International behavior: A s o c i a l - p s y c h o l o g i c a l a n a l y s i s . New York: Holt, Rinehart & Winston, 1965. Ross, J. A. Decision rules i n program evaluation. Evaluation Review, 1980, 4, 59-74. Schachter, S. Deviation, r e j e c t i o n , and communication. Journal of Abnormal and S o c i a l Psychology, 1951, 46, 190-207. Schachter, S., E l l e r t s o n , N., McBride, D., & Gregory, D. An experimental study of cohesiveness and p r o d u c t i v i t y . Human Relations, 1951, 4_, 229-238. Sheri f , M., & Sherif, C. Reference groups: Exploration into the conformity and deviation of adolescents. New York: Harper & Row, 1964. Sherif, M., & Sherif, C. S o c i a l psychology. New York: Harper & Row, 1969. - 104 -Suedfeld, P. Chapter t i t l e , i n German: Die messung i n t e g r a t i v e r komplexitHt i n archivmaterialien. (Measuring i n t e g r a t i v e complexity i n a r c h i v a l materials). In H. Mandl, & C. L. Huber (Eds.), Kognitive komplexitHt. GHttingen: Hogrefe Verlag, 1978, pp. 179-192. Tetlock, P. E. Identifying victims of groupthink from p u b l i c statements of decision makers. Journal of Personality and S o c i a l Psychology, 1979, 37, 1314-1324. Thompson, J . D., & Tuden, A. Strategies, structures, and processes of organizational decision. In D. A. Kolb, I. M. Rubin,& J . M. Mclntyre (Eds.), Organizational psychology: A book of readings. Englewood C l i f f s , New Jersey: P r e n t i c e - H a l l , 1971. Thorngate, W. W i l l group dynamics research ever change? Contemporary  Psychology, 1980, 25, 457-458. Weiss, R. F., Lombardo, J . P., Warren, D. R., & Kelley, K. A. Reinforcing e f f e c t s of speaking i n reply. Journal of Personality and S o c i a l Psycho1:. . logy, 1971, 20, 186-199. Wright, P. H. The de l i n e a t i o n and measurement of some key va r i a b l e s i n the study of fri e n d s h i p . Representative Research i n S o c i a l Psychology, 1974, 5_, 93-96. - 105 -Appendix A Pre-Experimental Session, Group Leadership Training Procedures - 106 -Group Leadership Training Procedures Instructions to Leaders: This i s meant to be a very b r i e f i n s t r u c t i o n session before you get together with your group. B a s i c a l l y , we're interested i n the way i n which your group operates under an (assertive; non-assertive) s t y l e of leadership. Now, we know that each of you has a personal s t y l e of leading groups which may lean towards being e i t h e r a s s e r t i v e or non-assertive concerning your ideas, and that i t probably varies somewhat according to the s i t u a t i o n that you:'.re i n . However, for t h i s present study, we would l i k e you to be s t r i c t l y (assertive; non-assertive) i n your s t y l e . We are interest e d i n the contrast between groups under non-assertive leaders who d e l i b e r a t e l y db not stress t h e i r own ideas and leaders who stress agreement with t h e i r s olutions. Ok, l e t ' s t a l k about some s p e c i f i c c h a r a c t e r i s t i c s that are associated with operating i n a (assertive; non-assertive) s t y l e . A. The non-assertive s t y l e . 1. The p r i n c i p a l i n t e r e s t i s i n receiving input from a l l members so that the most information possible i s av a i l a b l e before making a deci s i o n . 2. The mode of operation i s l i k e that of a discussion group leader mixed with a c o n c i l i a t o r . At f i r s t there i s the attempt to s o l i c i t a l l the relevant information without coming to a premature dec i s i o n . Then once i t appears that most of the opinion i s a v a i l a b l e the next step becomes one of r e c o n c i l i n g the various views to come up with one s o l u t i o n that appears most l i k e l y to be correct. 3. A major t a c t i c i s to be able to develop i n the group an atmosphere of trus t within which there can be a c r i t i c a l assessment of ideas, without the fear of having f e e l i n g s hurt. 4. There are two prime d i f f i c u l t i e s that non-assertive leaders often - 107 -run into and must be avoided to set up the group atmosphere that you want. a. Group leaders often come on strong with t h e i r own ideas at the very s t a r t of a group discussion. In most groups where the leader has a p o s i t i o n of power due to status, higher q u a l i f i c a t i o n s than the others, or the a b i l i t y to strongly influence the careers of the other members, the mem- .. bers w i l l react to the leader's i n i t i a l views by becoming more conservative or hesitant i n v o i c i n g t h e i r own views. They often w i l l avoid r a i s i n g questions which may c o n f l i c t with that of the authority f i g u r e — y o u , the leader. I would l i k e you to avoid s t a t i n g your own point of view u n t i l i t i s c l e a r that a l l the other members have stated t h e i r s . I would l i k e your primary i n t e r e s t to be i n the other points of view. Try also to avoid putting your opinion i n t o black and white terms; s t a t i n g things as being c a t e g o r i c a l l y r i g h t or wrong leaves l i t t l e room for compromise without l o s i n g face, even i f you decide l a t e r that you want to change your view somewhat. This brings us to the second major d i f f i c u l t y to be overcome i n getting the group into an atmosphere of c r i t i c a l l y assessing ideas. b. To show that a l l ideas are open for c r i t i c a l assessment i t must be clear that the group leader does not react adversely when h i s own ideas are examined or flaws pointed out i n them. This i s very important because at t h i s stage the leader can make or break the desired group atmosphere. In actual f a c t , you become the model f o r the rest of the group. To summarize: The primary thing that we are interested i n i s that you  do not state your View u n t i l the others have done so and that you take the  r o l e of being d e l i b e r a t e l y non-assertive concerning your views on the  problems. (Present the two problems and the Leadership Training Questionnaire.) - 108 -B. The ass e r t i v e s t y l e . 1. The p r i n c i p a l i n t e r e s t i s to consider the problems, form some ideas about what should be done and e f f e c t i v e l y s e l l your ideas to the group. 2. The primary approach i s through s t r e s s i n g your own opinion. Work out your positions and push for them as the group leader. 3. One of the main techniques used i n operating i n t h i s s t y l e i s to es t a b l i s h control over the^ d i r e c t i o n i n which the group i s heading. Often-times, leaders, wishing to be ass e r t i v e but proceeding i n e f f e c t i v e l y , end up "competing" with one or two other group members for the actual leadership of the group. A group attempting to go i n several d i f f e r e n t d i r e c t i o n s at the same time w i l l not be an e f f e c t i v e group. One valuable and useful way for esta b l i s h i n g the group d i r e c t i o n and your leadership early on i s to ou t l i n e your s p e c i f i c ideas about the s o l u t i o n to the problem. Very often the leader w i l l s t a r t the meeting o f f by summarizing the s i t u a t i o n and the problem. Then once having presented h i s views on the problems he d i r e c t s the meeting by asking for questions and other ideas. 4. A second problem that i s often met when meetings are not direc t e d e f f e c t i v e l y i s that there tends to be a great deal of time consuming redun-dancy i n ideas and discussion. Often the same arguments and same issues w i l l be rehashed i n f i f t y d i f f e r e n t ways. One advantage of problem solving with an ass e r t i v e leader i s that they tend to minimize t h i s redundancy by asking people to be s p e c i f i c , by i n d i c a t i n g when discussions have stopped being productive and started being dogmatic statements of p o s i t i o n s . 5. The l a s t aspect of the problem solving session i s a c t u a l l y coming up with a so l u t i o n . Again, t h i s i s where an assertive leader must display considerable s k i l l . Oftentimes, i t i s v i r t u a l l y impossible to reco n c i l e a l l viewpoints i n a group and i n order to decide upon some so l u t i o n i t i s necessary f o r the leader to select what appears to be the best proposal. At - 109 -t h i s point you are b a s i c a l l y acting i n a s i t u a t i o n s i m i l a r to that of a President of Prime M i n i s t e r , when they receive input concerning a problem which may be quite diverse depending upon the group of advisors. At the end, i t i s up to the group leader (or President) to select what appears to the best s o l u t i o n i n the face of divergent opinions. 6. In general, I would l i k e you to stress your ideas and the need f o r agreement. This does not mean being t o t a l l y i n f l e x i b l e , because a l l leaders t y p i c a l l y adjust t h e i r positions on the basis of new information but i t does mean being a leader who i s ass e r t i v e about h i s point of view. To summarize: For each of the two problems s t a r t o f f by presenting your  point of view and then remain a s s e r t i v e about your view during the session. (Present the two problems and the Leadership Training Questionnaire.) - 110 -Appendix Leadership Training (Fiedler's Least Preferred Co-Worker B Questionnaire Test — F i e d l e r , 1967, p. 268) - I l l -Leadership Training Questionnaire Instructions: I would l i k e you to think back to everyone that you have ever worked with. Pick out the. one i n d i v i d u a l with whom you could work least w e l l . He or she may be someone you knew i n the past or s/he may be someone that you work with now. He or she does not have to be the person that you l i k e l e ast w e l l , but should be the person with whom you had the. most d i f f i c u l t y i n getting a job done. On the set of scales below, please describe t h i s person as s/he appears to you. C i r c l e the appropriate number for each scale. Pleasant Frie n d l y Rej ecting Helpful Unenthusiastic Tense Distant Cold Cooperative Supportive Boring 8 7 6 5 4 3 2 1 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 8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 Unpleasant Unfriendly Accepting F r u s t r a t i n g Enthusiastic Relaxed Close Warm Uncooperative H o s t i l e Interesting - 112 -Quarrelsome Self-assured E f f i c i e n t 1 2 3 4 5 6 7 8 8 7 6 5 4 3 2 1 Gloomy Open 7 6 5 4 3 2 1 1 2 3 4 5 6 7 7 6 5 4 3 2 1 Harmonious Hesitant I n e f f i c i e n t Cheerful Guarded - 113 -Appendix C General Information About the Study Given to the Subject Groups - 114 -General Information About the Study (for the subject groups) This i s b a s i c a l l y a study concerned with group problem solving or group decision making. We've a l l been i n groups or committees attempting to deal with issues or solve problems. However, I have found that when looking at how c e r t a i n groups operate under s p e c i f i c types of leaders, dealing with s p e c i f i c types of problems, we know very l i t t l e about the group processes or outcomes that occur. That i s the purpose of the present research. We w i l l be using a p a r t i c u l a r type of people to form the groups ( c i v i l servants; businessmen; m i l i t a r y personnel). The groups w i l l operate under leaders who have s p e c i f i c s t y l e s of leadership. F i n a l l y , the groups w i l l be asked to solve a p a r t i c u l a r type of p o l i c y problem. Through a l l t h i s , we w i l l be interested i n how the group deals with the problem, the s o l u t i o n that they produce and the members' reaction to the process. The way we proceed i s by meeting with the group leaders to discuss the various leadership s t y l e s sometime j u s t p r i o r to the actual group session. Then, the whole group gets together under the d i r e c t i o n of the leader and we present the group with two p o l i c y problems to solve. At the end of the session you w i l l be given a questionnaire to f i l l out so that we may get your feedback about the en t i r e process. Because of the fact that I cannot remember the proceedings of e n t i r e sessions, the sessions w i l l be tape-recorded. The group sessions usually take about one hour to complete. - 115 -Appendix D Instructions to the Antecedent Conditions Present Groups - 116 -Instructions to the Antecedent Conditions Present Groups Explain the taperecorder and my presence. You a l l know that we're here to study group de c i s i o n making. Today the problems that y o u ' l l be asked to deal with concern your views on Canadian immigration and c a p i t a l punishment. You w i l l have a t o t a l time of only 30 minutes to produce solutions to these two problems. I ' l l l e t you know when h a l f the time i s gone. Most people f i n d they have to work fast to get through i n t h i s time. A c t u a l l y , i t ' s good you're a (mixed) ( a l l male) ( a l l women) group they seem to work f a s t e r . Now, what I would l i k e you to do i s to work under the leadership of during t h i s session. When your 30 minutes i s up, I ' l l l e t you know. I ' l l ask the group leader to make a statement o u t l i n i n g as s p e c i f i c a l l y as possible the proposed solutions to the problems and the reasons f o r s e l e c t i n g those solutions. (To the leader) There's some spare paper there, I usually ask the leader j u s t to j o t down some points as you go along, not sentences, but j u s t things to jog your memory so that at the end when I ask you what decisions you've reached you can t e l l me. What you t e l l me as your statement at the end of the session w i l l be taken as the de s c r i p t i o n of the decisions reached by the group and the reasons behind the decisions. Following that, I ' l l get you a l l to f i l l out a questionnaire and the session w i l l be f i n i s h e d . Then we can t a l k about the research and I ' l l answer any of the questions that may have come to your mind. Emphasis! We would l i k e you to do your best on these problems. I t i s important that you reach the best solutions possible. We w i l l be looking at these discussions i n terms of the q u a l i t y of the solutions you propose. We w i l l also be showing these tapes anonymously to classes of students and to some pro f e s s i o n a l seminars for managers as examples or cases i n group decision making. So do the best that you can. - 117 -I t i s the U n i v e r s i t y p o l i c y concerning research that people be f u l l y informed about the study so that they know that there's no deception i n -volved and everything i s above board. These are the consent forms that we ask each person to f i l l i n . - 118 -Appendix E Instructions to the Antecedent Conditions; Absent Groups - 119 -Instructions to the Antecedent Conditions Absent Groups Start o f f by making introductions. Low Decision S.: Explain the taperecorder and my presence. You a l l know that we're here to study group d e c i s i o n making. Today the problems that y o u ' l l be asked to deal with concern your views on Canadian immigration and c a p i t a l punishment. You can have pretty well as long as you l i k e to work on the problems, u n t i l you f e e l comfortable with the solutions. There i s no set time l i m i t , but there i s usually no problem i n f i n i s h i n g the discussions and coming up with solutions within about an hour. Now, what I would l i k e you to do i s to work under the leadership of during t h i s session. When you are f i n i s h e d , you can l e t me know. I ' l l ask the group leader to make a statement o u t l i n i n g as s p e c i f i c a l l y as possible the proposed solutions to the problems and the reasons for s e l e c t i n g those solutions. There's some spare paper there, I usually ask the leader j u s t to j o t down some points as you go along, not sentences, but ju s t things to jog you memory so that at the end when I ask you what decisions you've reached you can t e l l me. What you t e l l me as your statement at the end of the session w i l l be taken as the des c r i p t i o n of the decisions reached by the group and the reasons behind the decisions. Following that I ' l l get you a l l to f i l l out a questionnaire and the session w i l l be f i n i s h e d . Then I can answer questions and t e l l you why I'm doing t h i s anyway! (A l i g h t comment to make them f e e l at ease.) I t i s the University p o l i c y concerning research that people be f u l l y informed about the study so that people know that there's no deception involved i n the study, everything I've stated i s above board, and we haven't hot wired the chairs or something. These are the consent forms that we ask each person to f i l l i n . - 120 -OK, go ahead. - 1 2 1 -Appendix F Consent Form - 122 -THE UNIVERSITY OF BRITISH COLUMBIA 2075 Wesbrook Mall Vancouver, B.C., Canada V6T 1W5 Department of Psychology CONSENT FORM The ethics review committee of the Univ e r s i t y of B r i t i s h Columbia requests c that a l l i n d i v i d u a l s p a r t i c i p a t i n g i n research associated with the Univ e r s i t y sign a consent form. By signing t h i s form you in d i c a t e that the nature of the study has been explained to you; that you are p a r t i c i p a t i n g v o l u n t a r i l y and are free to withdraw without penalty from the study at any time; that there are no r i s k s associated with t h i s study; and that a l l information w i l l be maintained with s t r i c t e s t c o n f i d e n t i a l i t y . Date: Signature: - 123 -Appendix G Ca p i t a l Punishment and Immigration Problems - 124 -Ca p i t a l Punishment The l a s t hanging i n Canada was i n 1962. From 1963 to 1976, the Federal cabinet has commuted a l l death sentences to l i f e imprisonment. At the pre-sent time there i s no "death sentence" for any crime committed i n Canada. The present law, which was passed i n 1976, c a l l s for a mandatory 25 year imprisonment sentence for 1 s t degree murder, and 10-25 years for 2 n^ degree murder. In both cases, a f t e r serving a minimum of 15 years, e l i g i b i l i t y for parole may be considered by a j u d i c i a l review panel i n i t i a l l y , and then by the National Parole Board. Over the past 15-20 years, the general approach taken by the government has been to reduce the number of crimes foriwhich the death sentence could be applied and increase the minimum length of time that must be served before being e l i g i b l e for parole. The government has taken the p o s i t i o n that they f i n d the death penalty both morally unacceptable and i n e f f e c t i v e as a deterrent. Many people have supported t h i s view, including many churches who view c a p i t a l punishment as being contrary to the s p i r i t and teachings of C h r i s t i a n i t y . In contrast to th i s p o s i t i o n , p u b l i c opinion p o l l s have shown that i n recent years the majority of Canadians have favoured the retention of the death penalty. This support, according to the Gallup P o l l s , h i t a low point of 42% i n 1966 and since that time has increased to i t s present standing of approximately 60% of Canadians being i n favour of c a p i t a l punishment for some crimes. People who want i t returned often dispute the evidence concerning whether or not c a p i t a l punishment i s an e f f e c t i v e deterrent. They often comment that even i f i t did not deter others, i t would c e r t a i n l y stop that i n d i v i d u a l from ever k i l l i n g again. Some view the death penalty as a form of punishment or r e t r i b u t i o n , i n b i b l i c a l terms "an eye f o r an eye". Others contend that i t i s required to protect the p o l i c e and prison guards by - 125 -decreasing the r i s k they face i n doing t h e i r jobs. In your view, what would be the best law or laws that Canada could have regarding c a p i t a l punishment? Immigration In 1976 a new immigration act became law. It allowed Canada to set t o t a l admission l i m i t s on immigration and disallowed d i s c r i m i n a t i o n on the basis of race, ethnic o r i g i n , c olor, r e l i g i o n , or sex. There are three classes of immigrants: (1) Immediate family members—who are not assessed under the point system. (2) Refugees—who are assessed on the basis of assistance v a v a i l a b l e to them and t h e i r a b i l i t y to adapt to Canadian l i f e , they are not given a r a t i n g under the point system. People who would not normally be admitted.as refugees may gain admittance under relaxed s e l e c t i o n c r i t e r i a i n times of c r i s i s . (3) Other immigrants—who are assessed under the point system. The point system i s based on the factors of education, vocational preparation, experience, occupational demand, previously arranged employment, lo c a t i o n i n Canada, age, language a b i l i t y , personal s u i t a b i l i t y , and . r e l a t i v e s . In general, immigrants must be awarded a minimum of 50 out of a 100 a v a i l a b l e points to be able to immigrate. Employment rel a t e d factors account f o r almost ha l f of the possible r a t i n g points. During the period of time from 1946-1973, approximately 3.3 m i l l i o n people immigrated from Europe, A u s t r a l i a , New Zealand, and the U.S.A. In comparison, approximately 1/2 m i l l i o n people immigrated from A f r i c a , North, Central and South America (excluding the U.S.A.) and A s i a (approximately 1/2 came from A s i a ) . During the s i x year period from 1968-1973, approximately 600,000 people immigrated from Europe, etc., and approximately 295,000 people - 126 -immigrated from A f r i c a , etc., (approximately 1/2 of these came from A s i a ) . The trend appears to be that i n recent years an increasing proportion of the Canadian immigration i s coming from A f r i c a , North, Central and South America (excluding the U.S.A.) and As i a . The l e v e l of Canadian immigration has always been determined by the r e s u l t of many opposing forces. On the one hand, pressures to increase immi-gration have resulted from factors l i k e the demands of t h i r d world countries and the s p e c i a l refugee problems that continually a r i s e . S i m i l a r l y , Canada has f e l t moral pressure because of i t s p o s i t i o n as one of the wealthy, developed nations. Canada has many times turned to foreign immigration to provide the people needed f o r our continued i n d u s t r i a l growth and development, a matter of p a r t i c u l a r concern when our natural population growth i s ap-proaching zero. Foreign immigration also provides Canada with a r i c h d i v e r s i t y of new views and cultures. On the other hand, immigration has been c i t e d by some as r e s u l t i n g i n increased unemployment, increased s t r a i n s on the housing market, unwarranted demands on the school system, a d d i t i o n a l costs for s o c i a l s e r v i c e s , and as the underlying cause behind the increasing number of r a c i a l i n c i d e n t s . At the present time, to achieve the best balance between the opposing points of view, what would you consider to be the best immigration p o l i c i e s that Canada could have with respect to foreign countries? - 127 -Appendix H Instructions to Begin the Session and Instructions to Time Limited Groups During the Session - 128 -Instructions to Begin the Session OK, here are the problems. You w i l l have three minutes to read the problems and then I w i l l t e l l you to go ahead. The session w i l l s t a r t and you w i l l have (30 minutes to produce your solutions; as long as you f e e l necessary to be comfortable i n the solutions that you produce). (TURN ON THE TAPERECORDER) Instructions to Time Limited Groups During the Session (Antecedent conditions present) Excuse me, l e t me interrupt f o r a minute. You have reached half-time, you have only 15 minutes l e f t to come up with proposals for the two . problems! Go ahead! Excuse me, l e t me interrupt t h i s l a s t time. You have only 2-3 minutes l e f t to f i n a l i z e your proposals for the two problems! Go ahead! (Handout questionnaires properly!) - 129 -Appendix I Post-Experimental Questionnaire for the Group Leader - 130 -Post-Experimental Questionnaire Sex: M FM ( C i r c l e one) 1. We would appreciate the following information concerning the people that you worked with today i n t h i s session. Please consider each member of the group i n turn ( i t does not matter which person you c a l l member 1, 2, e t c . ) . Look at the "friendship/acquaintanceship" scale below and choose the scale item that most c l o s e l y defines your degree of friendship with that member. Then, place the number for t h i s scale item i n the r a t i n g blank f o r that member. A l l information contained on t h i s questionnaire w i l l remain s t r i c t l y c o n f i d e n t i a l . Degree of Friendship/Acquaintanceship 1 2 3 4 5 6 among my a good a moderately a s l i g h t an acquain- never met best f r i e n d good f r i e n d f r i e n d tance before friends today Example Rating: Member 5. 6 Your ratings of the other group members: Member 1. Member 2. Member 3. Based on your experiences within the group during t h i s past session, we would appreciate your feedback on the following questions. At the end of the questionnaire we would also appreciate any comments that you might have con-cerning t h i s research. In questions where there are ra t i n g scales, please c i r c l e the appropriate number above your answer. 2. How many of the members of today's group had you previously worked with - 131 -i n a group-meeting s i t u a t i o n or a committee, i n the attempt to discuss or deal with problems concerning your organization(s)? Number: If you had previously worked with some of the other members i n a group-s i t u a t i o n or a committee, how would you rate your degree of i n t e r e s t i n p a r t i c i p a t i n g with that previous group? 1 2 3 4 5 6 7 very interested somewhat somewhat uninter- very no previous interested interested uninter- ested uninter- group or ested ested committee contact with any members How would you rate your degree of i n t e r e s t i n p a r t i c i p a t i n g with the present group members today? 1 2 3 4 5 6 very interested somewhat somewhat uninter- very un-interested interested uninterested ested interested O v e r a l l , how would you rate the group members that you were with today? 1 2 3 4 5 6 very unenjoyable somewhat somewhat enjoyable & very unenjoyable & uninter- unenjoyable enjoyable i n t e r e s t i n g enjoyable & & uninter- esting to & uninter- & i n t e r - to work with i n t e r e s t i n g esting to work with esting to esting to to work with work with work with work with To what extent was the attempt to reach group agreement on the solutions a consideration i n your own e f f o r t s to help the group solve the problems? 1 2 3 4 5 6 the least an unimpor- a somewhat a somewhat an important the prime consider- tant consid- unimportant important consider- consider-ation I had eration 1 . consider- consider- a t i o n I had ation I • i n solving had i n ation I had ation I' i n sol v i n g had i n the s o l v i n g the i n solving had i n the problems solving problems problems the solving the problems the problems problems - 132 -7. To what extent did you stress your own point of view concerning the best solutions? 1 2 3 4 5 6 very strongly more than some l i t t l e very l i t t l e strongly asserted some assertion., assertion assertion assertion asserted own view of own view of own of own of own view own view view view 8. After o u t l i n i n g the problems, to what extent did you propose your own solutions before the other group members had much opportunity to discuss the problems? 1 2 3 4 5 6 to a very to a to more than to some to a small to a very large degree large some degree degree degree small degree degree 9. What was your l e v e l of motivation to have the group produce solutions to the problems? 1 2 3 4 5 6 very strongly strongly more than some weakly very weakly motivated motivated some motivation motivated motivated motivation 10. How would you rate the importance to you of the problems that your group was attempting to solve? 1 2 3 4 5 6 very unimpor- unimpor- somewhat somewhat important very tant tant unimpor- important important tant 11. Did you f e e l pressure because of the time period allowed f o r solving the problems? 1 2 3 4 5 6 very l i t t l e l i t t l e some more than much very much pressure pressure pressure some pressure pressure pressure 12. How would you rate your own o v e r a l l l e v e l of stress during t h i s problem solving session? 1 2 3 4 5 6 very relaxed somewhat somewhat s t r e s s f u l very relaxed relaxed s t r e s s f u l s t r e s s f u l - 133 -13. At the end of the session, to what extent did you f e e l the other group members were i n agreement concerning the proposed solutions to the problems? 1 2 3 4 5 6 everyone l a r g e l y more agreement more d i s - l a r g e l y everyone agreed agreed; than disagree- agreement disagreed; disagreed; completely; few or ment; some than agree- many or very much no doubts weak doubts ment; more strong or very doubts than some doubts strong . doubts doubts 14. At the end of the session, to what extent did you agree with the proposed solutions to the problems? 1 2 3 4 5 6 disagreed l a r g e l y disagreed agreed more l a r g e l y completely completely; disagreed; more than than d i s - agreed; agreed; very many many or agreed; agreed; some few or no doubts or very strong more than doubts weak strong doubts some doubts doubts doubts 15. After the session was f i n i s h e d , to what extent did you have questions or doubts about the proposed solutions that you eit h e r didn't t a l k about or didn't f e e l comfortable i n str e s s i n g as much as you would have l i k e d during the session? 1 2 3 4 5 6 very many many doubts more than some doubts, few doubts very few doubts or or questions some doubts or questions or ques- doubts or questions not rai s e d or questions not rai s e d tions not questions not raised or not not raised or not raised or not raised or not stressed or not stressed not or not stressed stressed stressed stressed 16. Did you f e e l that the group exerted pressure upon some members when they raised opposing ideas? 1 2 3 4 5 6 very l i t t l e l i t t l e some more than much very much pressure pressure pressure some pressure pressure pressure - 134 -17. Did you f e e l that the group exerted pressure upon you when you rai s e d opposing ideas? 1 2 3 4 5 6 very much much more than some l i t t l e very l i t t l e pressure pressure some pressure: pressure pressure pressure 18. To what extent do you see the solutions reached by the group as being the most e t h i c a l l y or morally correct ones possible? 1 2 3 4 5 6 to a very to a large to more . to some to a small to a very large degree than some degree degree small degree degree degree 19. How o p t i m i s t i c are you about having produced the best possible solutions to the problems, within the time that you had available? 1 2 3 4 5 6 very o p t i m i s t i c somewhat somewhat p e s s i m i s t i c very o p t i m i s t i c o p t i m i s t i c p e s s i m i s t i c p e s s i m i s t i c 20. We are interested i n the r i s k associated with the proposed solutions to these problems. I f you were i n a s i t u a t i o n where you were to use these solutions as a basis f o r making some personal d e c i s i o n , how r i s k y do you think i t would be? 1 2 3 4 5 6 very l i t t l e a l i t t l e somewhat more than much very r i s k r i s k r i s k y somewhat r i s k r i s k y r i s k y As the group leader, we are interested i n how you would describe the group with which you were working today. For each question please c i r c l e the number which describes your f e e l i n g s about the group. 1. Pleasant : : : : : : : : Unpleasant 8 7 6 5 4 3 2 1 2. Frien d l y : : : : : : : : Unfriendly 8 7 6 5 4 3 2 1 3. Bad : : : : : : : : Good 1 2 3 4 5 6 7 8 Comments: - 135 -4. Worthless 1 2 3 4 5 6 7 Valuable 5. Distant 6. Cold 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 Close Warm 7. Quarrelsome 1 2 3 4 5 6 7 8 Harmonious 8. Self-assured 7 6 5 4 3 2 1 Hesitant 9. E f f i c i e n t 7 6 5 4 3 2 1 I n e f f i c i e n t 10. Gloomy 1 2 3 4 5 6 7 Cheerful - 136 -Appendix J Post-Experimental Questionnaire f or the Group Members and Notes on the Post-Experimental Questionnaire - 137 -Post-Experimental Questionnaire Sex: F FM ( C i r c l e one) 1. We would appreciate the following information concerning the people that you worked with today i n t h i s session. Please consider each member of the group i n turn ( i t does not matter which person you c a l l member 1, 2, e t c . ) . Look at the "friendship/acquaintanceship" scale below and choose the scale item that most c l o s e l y defines your degree of f r i e n d -ship with that member. Then, place the number for t h i s scale item i n the r a t i n g blank for that member. A l l information contained on t h i s questionnaire w i l l remain s t r i c t l y c o n f i d e n t i a l . Degree of Friendship/Acquaintanceship 1 2 3 4 5 6 among my a good a moderately a s l i g h t an acquain- never met best f r i e n d good f r i e n d f r i e n d tance before friends today Example Rating: Member 5. 6 Your ratings of the other group members: Member 1. Member 2. Member 3. Based on your experiences within the group during t h i s past session, we would appreciate your feedback on the following questions. At the end of the questionnaire we would also appreciate any comments that you might have concerning t h i s research. In questions where there are r a t i n g scales, please c i r c l e the appropriate number above your answer. - 138 -2. How many of the members of today's group had you previously worked with i n a group-meeting s i t u a t i o n or a committee, i n the attempt to discuss or deal with problems concerning your organization(s)? Number: 3. I f you had previously worked with some of the other members i n a group-s i t u a t i o n or a committee, how would you rate your degree of i n t e r e s t i n p a r t i c i p a t i n g with that previous group? 1 2 3 4 5 6 7 very interested somewhat somewhat uninter- very no previous interested i n t e r - uninter- ested uninter- group or ested ested ested committee contact with any members 4. How would you rate your degree of i n t e r e s t i n p a r t i c i p a t i n g with the present group members today? 1 2 3 4 5 6 very interested somewhat somewhat uninter- very interested interested uninterested ested uninterested 5. Over a l l , how would you rate the group members that you were with today? 1 2 3 4 5 6 very unenjoyable somewhat somewhat enjoyable & very unenjoyable & uninter- unenjoyable enjoyable i n t e r e s t i n g enjoyable & & uninter- esting to & uninter- & i n t e r - to work with i n t e r e s t i n g esting to work with esting to esting to to work work with work with work with with 6. To what extent was the attempt to reach group agreement on the solutions a consideration i n your own e f f o r t s to help the group solve the problems? 1 2 3 4 5 6 the l e a s t an unimpor- a somewhat a somewhat an important the prime consider- tant consid- unimportant important consider- • consider-ation I had eratipn I : consideration consider- ation I had ation I i n solving had i n I had i n ation I i n solving had i n the solving the solving the had i n the solving problems problems problems solving problems the the problems problems - 139 -7. To what extent did the group leader stress t h e i r own point of view concerning the best solutions? 1 2 3 4 5 6 very strongly more than some l i t t l e very l i t t l e strongly asserted some as s e r t i o n assertion assertion assertion asserted own view of own view of own of own of own view own view view view 8. Aft e r o u t l i n i n g the problems, to what extent did the leader propose his or her own solutions before the other group members had much opportunity to discuss the problems? 1 2 3 4 5 6 to a very to a large to more than to some to a small to a very large degree some degree degree degree small degree degree 9. What was your l e v e l of motivation to have the group produce solutions to the problems? 1 2 3 4 5 6 very strongly strongly more than some weakly very weakly motivated motivated some motivation motivated motivated motivation 10. How would you rate the importance to you of the problems that your group was attempting to solve? 1 2 3 4 5 6 very unimpor- unimpor- somewhat somewhat important very tant tant unimpor- important important tant 11. Did you f e e l pressure because of the time period allowed for solving the problems? 1 2 3 4 5 6 very l i t t l e l i t t l e some more than much very much pressure pressure pressure some pressure pressure pressure 12. How would you rate your own o v e r a l l l e v e l of stress during t h i s problem solving session? 1 2 3 4 5 6 very relaxed somewhat somewhat s t r e s s f u l very relaxed relaxed s t r e s s f u l s t r e s s f u l . - 140 -13. At the end of the session, to what extent did you f e e l the other group members were i n agreement concerning the proposed solutions to the problems? 1 2 3 4 5 6 everyone l a r g e l y more agreement more d i s - l a r g e l y everyone agreed agreed; than disagree- agreement disagreed; disagreed; completely; few or ment; some than agree- many or very many no doubts weak doubts ment; more strong or very doubts than some doubts strong doubts doubts 14. At the end of the session, to what extent did you agree with the proposed solutions to the problems? 1 2 3 4 5 6 disagreed l a r g e l y disagreed agreed more l a r g e l y . completely completely; disagreed; more than than d i s - agreed; agreed; very many many or agreed; agreed; some few or no doubts or very strong more than doubts weak strong doubts some doubts doubts doubts 15. A f t e r the session was f i n i s h e d , to what extent did you have questions or doubts about the proposed solutions that you e i t h e r didn't t a l k about or didn't f e e l comfortable i n st r e s s i n g as much as you would have l i k e d during the session? 1 2 3 4 5 6 very many many doubts more than some doubts few doubts very few doubts or or questions some doubts or questions or questions doubts or questions not raised or ques- not rai s e d not rai s e d questions not raised or not tions not or not or not not rai s e d or not stressed raised or stressed stressed or not stressed not stressed stressed 16. Did you f e e l that the group exerted pressure upon some members when they r a i s e d opposing ideas? 1 2 3 4 5 6 very l i t t l e l i t t l e some more than much very much pressure pressure pressure some pressure pressure pressure - 141 -17. Did you f e e l that the group exerted pressure upon you when you rai s e d opposing ideas? 1 2 3 4 5 6 very much much more than some l i t t l e very l i t t l e pressure pressure some pressure pressure pressure pressure 18. To what extent do you see the solutions reached by the group as being the most e t h i c a l l y or morally correct ones possible? 1 2 3 4 5 6 to a very to a large to more to some to a small to a very large degree than some degree degree small degree degree degree 19. How o p t i m i s t i c are you about having produced the best possible solutions to the problems, within the time that you had available? 1 2 3 4 5 6 very o p t i m i s t i c somewhat somewhat p e s s i m i s t i c very o p t i m i s t i c o p t i m i s t i c p e s s i m i s t i c p e s s i m i s t i c 20. We are interested i n the r i s k associated with the proposed solutions to these problems. I f you were i n a s i t u a t i o n where you were to use these solutions as a basis for making some personal d e c i s i o n , how r i s k y do you think i t would be? 1 2 3 4 5 6 very l i t t l e a l i t t l e somewhat more than much very r i s k r i s k r i s k y somewhat r i s k r i s k y r i s k y Comments: - 142 -Notes on the Post-Experimental Questionnaire The following questions were designed to assess the groupthink preconditions: Questions 1 - 5 — assess the l e v e l of group cohesiveness 1. the degree of friendship ratings were adapted from "The d e l i n -eation and measurement of some key va r i a b l e s i n the study of frie n d s h i p " , P.H. Wright, Representative Research i n S o c i a l  Psychology, 1974, _5, 93-96. 2. p r i o r contact i n a cohesive group 3. a t t r a c t i o n to previous group 4. a t t r a c t i o n toi present* group 5. interpersonal a t t r a c t i o n among members Question 6 — assesses the degree of concurrence seeking by group members (based on the theory of Janis, 1972) Questions 7 & 8 — assess leadership s t y l e based upon the cohesive-ness l i t e r a t u r e based on the theory of Janis, 1972 7. leader stress upon own point of view 8. leader timing of proposing own solutions. (The same leadership s t y l e v a r i a b l e s are also rated by two independent raters under the section f o r the assessment of the q u a l i t y of the decision making process.) Question 9 — assesses the l e v e l of motivation f o r producing solutions to the problems (based on the theory of Jan i s , 1972) Questions 10 - 12 — assess the l e v e l of stress 10. personal importance of problem 11. stress due to time l i m i t a t i o n 12. o v e r a l l r a t i n g of personal stress based upon the work of Robinson, J . A. & Snyder, R.C. Decision making i n i n t e r -n a t i o n a l p o l i t i c s . . In - 143 -International behaviour: A  s o c i a l psychological a n a l y s i s . Kelman (Ed.) N.Y.: Holt, Rinehart & Winston, 1965. The following questions were designed to assess the groupthink symptoms: (Questions 13 - 20 are based upon the theory of Janis, 1972.) Questions 13 & 14 — assess the i l l u s i o n of unanimity 13. perception of agreement with f i n a l s o l u t i o n among other members. 14. personal l e v e l of agreement with f i n a l s o l u t i o n (ratings on the questionnaire scored i n reverse). i = l — — - | x i (#13) - X n o t i (#14)1 I l l u s i o n of unanimity = 4  4 Question 15 — assesses the degree of self-censorship Questions 16 & 17 — assess the l e v e l of conformity pressure 16. perceived pressure upon other deviant group members 17. perceived pressure upon oneself Question 18 — assesses the perception of the moral l e v e l of the decision making. Questions 19 & 20 and Quality of the Decision Outcome, Rating #6 — assess the i l l u s i o n of i n v u l n e r a b i l i t y 19. l e v e l of optimism concerning the proposed solutions 20. estimated l e v e l of r i s k concerning the proposed s o l u t i o n , as seen by i n d i v i d u a l s i n s i d e the group Quality of the Decision Outcome, Rating #6. estimated l e v e l of r i s k concerning the proposed s o l u t i o n , as seen by i n d i v i d u a l s outside the group. - 144 -Group Leader Questionnaire c Questions 1 - 10 — assess the group atmosphere. They are summed, averaged, and then used to describe leader-member r e l a t i o n s as one aspect of the " s i t u a t i o n a l favourableness" dimension of F i e d l e r ' s contingency theory. Based upon the theory of F i e d l e r , 1967; F i e d l e r and Chemers, 1974. (Median group atmosphere score = 64.9 r e a l l i f e groups; 67.0 laboratory groups; —Posthuma, 1970, c i t e d i n F i e d l e r and Chemers, 1974.) - 145 -Appendix K Debriefing Information - 146 -Debriefing Information The f i r s t thing that I'd l i k e to mention i s that i n doing t h i s study we did not attempt to deceive you. We were interested i n exactly the things that I stated we were interested i n when I f i r s t explained t h i s study to you. We have had groups of students, businessmen and c i v i l servants involved i n these sessions. Some of the groups have worked together before and these we l a b e l the cohesive groups. Other groups have not worked together before, these we l a b e l the less cohesive groups. Then, we have some of the groups operating under leaders who are a s s e r t i v e about t h e i r point of view and some operating under leaders who are d e l i b e r a t e l y not assertive about t h e i r point of view. We are interested i n comparing how these d i f f e r e n t groups operate. Do they go through the same decision making processes? If they are d i f f e r -ent processes, what are the differences? Do they produce the same q u a l i t y of group decision making outcome? I f not, how d i f f e r e n t are they? The questions we're asking stem from several books on decision making and leadership by Janis and F i e d l e r . Janis reviewed case studies of i n t e r -n a t i o n a l c r i s e s and proposed that cohesive, somewhat i s o l a t e d groups, under as s e r t i v e leaders may engage i n what he c a l l s groupthink. That i s , they become more interested i n "not rocking the boat" or maintaining group s o l i -d a r i t y and tend to become less interested i n f i n d i n g the best possible so l u t i o n . He stated that groups under these conditions w i l l put pressure on dissident members to tone down t h e i r c r i t i c i s m s , i n d i v i d u a l members may s t i f l e t h e i r own doubts about the proposed s o l u t i o n , and the groups may tend to be over o p t i m i s t i c and overly r i s k y about the decisions they make. The r e s u l t of a l l t h i s i s that the decision making process w i l l be poorer and the decisions reached w i l l also tend to be poorer. Since the basis for h i s ideas came from case studies, i t i s hoped that the present research w i l l be - 147 -able to test out h i s ideas i n r e a l l i f e and see the extent to which h i s theory i s correct. So, what you've p a r t i c i p a t e d i n today w i l l help contribute to our knowledge i n t h i s area. Thank you very much for helping. Do you have any questions? (I would l i k e the nature of the problems and the explanation of the study to remain c o n f i d e n t i a l so that we can study each group under the same conditions.) (Arrange a feedback session,, i f desired.) - 148 -Appendix L Example of the Feedback Given to a Group - 149 -Feedback B i l l R's Group This i s a copy of the information that I prepared concerning the group research that you recently p a r t i c i p a t e d i n . I hope that you f i n d i t i n t e r -e sting, i f you have any questions as we're going through the material j u s t ask me. No other copies of t h i s information have been made and as mentioned during the research session a l l information w i l l remain c o n f i d e n t i a l . F i r s t , before I continue on, I would l i k e to thank you a l l f o r your time and help i n doing t h i s research, I appreciate i t very much. Now, the report w i l l be i n three stages. F i r s t I ' l l o u t l i n e some of the c h a r a c t e r i s t i c s of the group, such as i t s l e v e l of cohesivenessv Then I ' l l mention what the average f e e l i n g s of the group members were concerning the experience, such as fe e l i n g s of conformity pressure, etc. Last, on the basis of my review of the tape of the session, I w i l l note a few character-i s t i c s of the group decision making process, such as the number of a l t e r -native solutions considered by the group. (The questionnaire that you have i s the same as the one that you com-pleted and i t w i l l help you to i d e n t i f y the average responses of the group as I mention them.) I Group C h a r a c t e r i s t i c s A. Group Cohesiveness The measures of group cohesiveness indicated that, as planned, i t was a group with a non-cohesive background. The average degree of r e l a t i o n s h i p between members was rated as being that of an acquaintance ( Q l — 5 . 1 ) , On the average, the members had not worked with each other before i n a problem - 150 -solving or decision making s i t u a t i o n . However, members were very interested i n p a r t i c i p a t i n g together i n the research (Q4—1.25) and rated the others as being moderately enjoyable and i n t e r e s t i n g to work with (Q5—4.25). This indicated that the i n i t i a l steps for cohesiveness i n a group were developing to some extent. In summary, the degree of f r i e n d s h i p , p r i o r contact, a t t r a c t i o n to the previous and present groups with these members, and the degree of i n t e r -personal a t t r a c t i o n a l l i n d i c a t e the operations of a b a s i c a l l y non-cohesive group. B. Leadership Style The group leader was perceived by the others i n the group as very strongly s t r e s s i n g h i s own point of view (Q7—1.33) and as proposing h i s ideas to a large degree before the others had much opportunity to discuss the problems (Q8—2.0). My impression was that I wouldn't have rated B i l l quite as strongly, however, as you know, t h i s a s s ertive s t y l e of leadership was the approach that I had asked B i l l to take during the session. From your ratings and the tape I would say that he took on that r o l e quite s u c c e s s f u l l y . As a second way of describing leadership s t y l e , I asked B i l l to complete a leadership s t y l e questionnaire during the time that we met before the session. This questionnaire was based on F i e d l e r ' s Contingency Theory of Leadership. He suggests that group performance i s dependent upon the i n t e r -action between leadership s t y l e and the group s i t u a t i o n . The leadership s t y l e may be e i t h e r task oriented or interpersonal r e l a t i o n s oriented. Both of these approaches are e f f e c t i v e under d i f f e r e n t conditions. The group s i t u a t i o n i s analyzed i n terms of three components: (1) leader-member r e l a t i o n s , (2) task structure and (3) leader power. In terms of F i e d l e r ' s theory, based upon B i l l ' s responses, he tends to operate i n an interpersonal r e l a t i o n s oriented fashion. D i f f e r e n t theories may describe the s t y l e that - 151 -a person t y p i c a l l y uses, i n d i f f e r e n t fashions, since the measurements and ideas proposed by any one theory have not yet gained acceptance by everyone. The group s i t u a t i o n from the questionnaire ratings was one of favourable leader-member r e l a t i o n s , i n v o l v i n g an unstructured task, and inv o l v i n g a s i t u a t i o n where the leader had l i t t l e power. According to F i e d l e r ' s theory, i n t h i s kind of group s i t u a t i o n a leader operating i n an interpersonal r e l a t i o n s oriented fashion w i l l tend to be more e f f e c t i v e than w i l l a task oriented leader. Part of the purpose of the present research i s to test t h i s aspect of F i e d l e r ' s theory to see i f i n fact t h i s i s correct. C. Other Group C h a r a c t e r i s t i c s The group was strongly motivated to produce solutions to the problems (Q9—2.0), and the attempt to reach a group consensus on the solutions was f e l t by the members to be a very important consideration i n solving the problems (Q6—5.5). The problems were thought to be very important (Q10— 5.5) and some time pressure was f e l t i n t r y i n g to respond to them (Qll-2.75). However, the o v e r a l l l e v e l of stress was not high, with the group members f e e l i n g , on the average, somewhat relaxed (Q12—3.25). The range of answers indicated some stress was f e l t by some of the members. However, i n t h i s s i t u a t i o n i t i s usual for even high l e v e l s of motivation, problem importance and time pressure not to be translated into strong f e e l i n g s of s t r e s s . Normally they would, but here the s i t u a t i o n was a r t i f i c i a l enough that these factors do not seem to operate i n the same fashion. I I . The Experiences of the Group A. Accuracy of Perception and I l l u s i o n of Agreement or Disagreement The accuracy of the group member's perceptions of the amount of agree-ment or disagreement that existed i n the group over the proposed solutions was rated on a scale ranging from 0 to 5; the normal range of scores i s from - 152 -0 to 1. This i s done by taking each member's perceptions of what the group feeli n g s were (Q13) and comparing that to what the other members a c t u a l l y were f e e l i n g (Q14). When t h i s was done, the group scored 0.42 on the f i v e point scale. This indicated that there was some but r e l a t i v e l y l i t t l e inaccuracy concerning the amount of agreement or disagreement i n the group, p a r t i c u l a r l y for a non-cohesive group. Next i t was analyzed to see what type of i l l u s i o n (or bias) there might be (even though the score was small). Here I use a scale of -5 to +5, where (-) means an i l l u s i o n of agreement and (+) means an i l l u s i o n of disagreement; the normal range of scores i s from -1 to +1. On t h i s scale the group scored -.25. This means that the group members f e l t that there was a b i t more agreement on the solutions than there a c t u a l l y was. These scores are moderate to small and o v e r a l l , the members appeared to have a r e l a t i v e l y good idea of where everyone else stood with respect to the proposed solutions to the problems. B. Degree of Self-Censorship It appears that on the average there were quite a few doubts or ques-tions about the proposed solutions which the members did not f e e l free to ra i s e or stress as much as they would have l i k e d to have done during the session (Q15—3.5). There was a wide range of responses, some f e l t t h i s more strongly than others. One could i n t e r p r e t t h i s and the previous information concerning the accuracy of the group's perceptions as i n d i c a t i n g that communications within the group were r e l a t i v e l y accurate even though they weren't completely open. C. Conformity Pressure This area also provided some i n t e r e s t i n g r e s u l t s . When the group mem-bers were asked to rate to what extent they thought that the others were f e e l i n g pressure to conform i n t h e i r ideas, they stated that they f e l t the group exerted some conformity pressure (Q16—2.75). However, when the mem-bers were asked to rate the extent to which they personally f e l t conformity pressure i t was rated as being a b i t lower, closer to f e e l i n g l i t t l e pressure (Q17—4.75). Thus i t seems as though the members were perceiving a b i t more conformity pressure e x i s t i n g i n the group than the members themselves were act u a l l y f e e l i n g . My own observation was that while t h i s may be true, there was a f a i r amount of tension i n the group not r e f l e c t e d i n these scores. I think that t h i s tension and the newness of the group was probably more the cause of the self-censorship that existed than time pressures or conformity pressures. D. Perceived Morality of the Solutions The group members perceived the solutions that were developed as being to a moderately large degree the most e t h i c a l l y or morally correct ones possible (Q18—3.0). This i s about an average degree of perceived morality and seems to be consistent with other s i m i l a r groups. E. Feelings ( I l l u s i o n ) of Security i n the Solutions The f e e l i n g s ( i l l u s i o n ) of s e c u r i t y are measured by considering the l e v e l of optimism about the solutions and the perceived r i s k associated with the solutions. The group members were somewhat p e s s i m i s t i c that they had produced the best possible solutions to the problems, within the time they had a v a i l a b l e (Q19—4.25). They f e l t that there would be much r i s k asso-ciated with using these decisions as a basis for making a personal decision i n an appropriate s i t u a t i o n (Q20—4.5). On the whole i t appears as though the group f e l t pretty p e s s i m i s t i c about the solutions that they proposed. It i s d i f f i c u l t to estimate the extent to which these f e e l i n g s may be j u s t i f i e d or not u n t i l I receive the ratings of the panel of experts con-cerning the extent to which they f e e l that there may be r i s k associated with the proposed solutions. - 154 -I I I . C h a r a c t e r i s t i c s of the Group Decision Making Process I ' l l s t a r t o ff by t a l k i n g about some of the processes basic to the decision making task. One item that I look at i s to what extent did the group consider the objectives of the session? In both problems although the group did not formally mention that they were reviewing the objectives they i n f a c t did so informally i n terms of the extensive discussion and i n t e r -p r e t a t i o n of what the nature of the problem r e a l l y was. For example, i n the c a p i t a l punishment case the group spent considerable time discussing whether the p r o - c a p i t a l punishment views of 60% of the population r e f l e c t e d a b e l i e f i n deterrence e f f e c t s or were simply a sign of the present s o c i a l conditions, i . e . , our poor economic times. I t i s not often that a group spends so much time not j u s t dealing with the problems but c l a r i f y i n g and r e i n t e r p r e t i n g the problems. This c l a r i f i c a t i o n of the goals and problems i s a valuable exercise since the members of a group may sometimes see the " r e a l " problem as being somewhat d i f f e r e n t than what i s stated as being the problem. At times i n group discussions c o n f l i c t s may a r i s e i n the decision making process simply because i t becomes incr e a s i n g l y cle a r that everyone i s not t r y i n g to reach the same objectives. In general, the basic approach taken to both problems was the same. There was an early agreement among the group members concerning the general outline of the best s o l u t i o n f o r each of the problems. However, then the group proceeded to review an extensive amount of information r e l a t e d to the problems and to consider i n some d e t a i l the opposite points of view. This was an approach which led to a very good review of the pros and cons on each of the two issues. Now, l e t me t e l l you why I said "however". What t y p i -c a l l y happens i n a group reaching a quick consensus on the general s o l u t i o n to a problem i s that they do not spend too much time considering the p o s s i -ble negative aspects of the o v e r a l l package. Instead they usually consider - 155 -the p o s i t i v e a s p e c t s of the d e c i s i o n or spend time making m o d i f i c a t i o n s t h a t a r e not major but l o o k l i k e they would be u s e f u l . I n c o n t r a s t , the most e x t e n s i v e d i s c u s s i o n s of t h e p r o s and cons a s s o c i a t e d w i t h a l t e r n a t i v e s u s u a l l y occur when t h e r e i s a s p l i t i n the group w i t h two s t r o n g l y opposing p o i n t s of view. What was f a s c i n a t i n g w i t h t h i s group was the r e a c h i n g of a q u i c k consensus, y e t the e x t e n s i v e d i s c u s s i o n and r e v i e w of s o l u t i o n s and p r o s and cons. T h i s has o n l y happened w i t h one o t h e r group and i s an approach t h a t i s not u s u a l l y found i n groups u n l e s s they have r e c e i v e d p r i o r i n s t r u c t i o n s t o b r a i n s t o r m o r group members have had s p e c i f i c t r a i n i n g i n problem s o l v i n g . F o r example, d e l i b e r a t e l y t a k i n g on a d e v i l ' s advocate r o l e i s something r a r e l y done i n a group and i t i s c e r t a i n l y a t e c h n i q u e which i s taught i n t r a i n i n g s e s s i o n s as a method o f c r e a t i n g a b e t t e r r e v i e w of the i s s u e s w i t h i n a group. I noted as I was l i s t e n i n g t o t h e tape t h a t t h e r e was a w e l l rounded p a r t i c i p a t i o n i n t h e d i s c u s s i o n s w i t h a l a r g e i n p u t by a l l of the members. A b a l a n c e d d i s c u s s i o n of t h i s n a t u r e i s g e n e r a l l y more e f f e c t i v e and l e s s f r u s t r a t i n g t o t h o s e i n v o l v e d than t h e s i t u a t i o n s where the c o n v e r s a t i o n i s dominated by one or two members. The f i n a l i t e m a n a l y z e d was the e x t e n t to which the group gave a l a s t r e v i e w t o t h e p o s s i b i l i t i e s b e f o r e making a f i n a l d e c i s i o n . I n t h e p r e s e n t group you were q u i t e e f f e c t i v e i n r e v i e w i n g the arguments and i s s u e s w h i l e making your f i n a l summation, b e f o r e moving on t o t h e next problem or the next a s p e c t o f t h e problem (as was the case w i t h t h e i m m i g r a t i o n problem where you d e a l t w i t h i t i n t h r e e p a r t s ) . T h i s p r o c e s s of making a l a s t r e v i e w of t h e a l t e r n a t i v e s i s u s u a l l y found to be a v a l u a b l e p r a c t i c e i n d e c i s i o n making s i t u a t i o n s s i n c e i t h e l p s t o a v o i d jumping to h a s t y d e c i s i o n s or s i m p l y c h o o s i n g t h e l a s t a l t e r n a t i v e c o n s i d e r e d . I t b r i n g s back the - 156 -major p o i n t s d i s c u s s e d e a r l i e r and h e l p s t o ensure t h a t e v e r y t h i n g i s g i v e n a f i n a l w e i g h t i n g b e f o r e the d e c i s i o n i s made. O v e r a l l , i t was one of t h e b e s t d i s c u s s i o n s t h a t I've s a t i n on; you f u n c t i o n e d w e l l t o g e t h e r and came a c r o s s as a s o p h i s t i c a t e d problem s o l v i n g group. I t was v e r y i n t e r e s t i n g t o watch and I a p p r e c i a t e your time because i t c e r t a i n l y h e l p s i n i n c r e a s i n g my u n d e r s t a n d i n g o f groups. - 157 -Appendix M Rating Scales f o r Assessing the Quality of the Decision Making Process - 158 -RATINGS OF TAPERECORDED GROUP SESSIONS Instructions: You are to l i s t e n to the taperecording of the group session and make the following ratings. You may reverse the tape to go back and l i s t e n to parts of the tape more than once, i f desired. The group sessions l a s t e i t h e r 30 minutes or approximately one hour. Use Section B as the work sheets to obtain the average ratings on each question; the eight questions l i s t e d under Section A, the "Average Group Rating", should be used to help you make the ratings under Section B. Section A: Average Group Ratings 1. Indicate the extent to which the objectives of the group were discussed or considered, (discussion, i n t e r p r e t a t i o n , r e i n t e r p r e t a t i o n , or c l a r i f i c a t i o n of the problems and/or goals). 1 2 3 4 5 6 7 very much much more than some a l i t t l e very l i t t l e no d i s -discussion discussion some discus^- discussion discussion cussion or consid- or consid- discussion sion or or consid- or consid- or consid-eration eration or consid- consid- eration, e r a t i o n , or eration eration eration or reads i m p l i c i t goals from mention paper 2. Indicate the number of possible solutions that were considered by the group. 1 - 2; 3 - 4; 5 - 6; 7 - 8; 9 - 10; >10. 3. Indicate the extent to which the p o s i t i v e and negative consequences of each a l t e r n a t i v e were considered. 1 2 3 4 very much much more than some consider- consider- some consid- consider at i o n ation eration a t i o n given given given given 5 6 a l i t t l e very - consid- consi eration ation given l i t t l e not der- consid-given ered - 159 -Indicate the amount of general information r e l a t e d to the problem which was discussed by the group. 1 2 3 4 5 very large large 'more than moderate less than am't of am't -.of moderate am't of moderate informa- i n f o r - am't of informa- am't of ti o n used mation informa- t i o n used informa-used t i o n used t i o n used a small am' t of informa-J a very small am't of information t i o n used used Indicate the extent to which a l l the group members provided information or p a r t i c i p a t e d i n the discussion. 1 2 very much much involve-ment by a l l members involve-ment by a l l members some involve-ment by two members, l i t t l e by the rest some involve-ment, by only one member, l i t t l e by the rest l i t t l e involve-ment by any of the members 3 4 more than some some i n - " involve-volvement ment by by most most mem-members bers ( 3 - 4 ) ( 3 - 4 ) or much involve-ment by only a few members (1-2) Indicate the extent to which the a l t e r n a t i v e s and previous information were reviewed during the discussion and at the end before the f i n a l d e c i s i o n was made. 1 2 3 4 5 6 7 very strong strong more than some re- a l i t t l e very l i t t l e no r e -review review some review, view, j u s t review review viewing during & at near or at the end of the end of the discus- the discus-sion or sion or not thorough thorough To what extent did the group leader stress h i s or her own point of view concerning the best solutions? 1 2 3 4 very strongly more than some strongly asserted some assertion asserted own view assertion of own own view of own view view 5 6 7 a l i t t l e very l i t t l e no asser-assertion assertion t i o n of of own of own own view view view - 160 -8. To what extent did the leader propose h i s or her own solutions before the other group members had much opportunity to discuss the problems? 1 2 3 4 5 6 7 to a very to a to more to some to a to a very not at a l l large large than some degree small small degree degree degree degree degree SECTION: B GROUP: RATER: Problem 1: Immigration Proposed Solutions Immigration Objectives Solutions Pros & Cons Gen Info Partici- Review Assertion Timing Ratings: pation SECTION: B G R 0 U P : RATER: Problem 2: C*P1r»l P, m«r=w, n r Proposed Solutions Capital Punishment Ratings: Objectives Solutions Pros & Cons Gen.Info. Pa r t i c i p a t i o n Review Assertion Timing iwo rroblem Average: Obj ectives Solutions Pros & Cons Gen.Info. P a r t i c i p a t i o n Review Assertion Timing - 163 -Appendix N Scales for Expert Ratings of the Quality of the Decision Outcome - 164 -Leathers P r o d u c t i v i t y Rating Instrument Group # : R a t e r : Please rate the solutions produced by each group on the following scales: c i r c l e your response. 1. Effectiveness. The degree to which the ideas, which are part of the major decision or s o l u t i o n , help the group achieve the objective of developing a r e a l i s t i c s o l u t i o n i n terms of i t ' s e f f e c t i v e n e s s . E f f e c t i v e T I n e f f e c t i v e 3 2 _1 0 I 2 3 2. F e a s i b i l i t y . The degree to which the major decision or s o l u t i o n r e f l e c t s a p i c t u r e of s o c i a l r e a l i t y which i s consistent with relevant p u b l i c a t t i t u d e s . Feasible Unfeasible 3 2 1 0 1 2 3 3. C r e a t i v i t y . The degree to which the major decision or s o l u t i o n r e f l e c t s markedly o r i g i n a l ideas not previously applied to the problem under discussion. Creative Uncreative 3 2 1 0 1 2 3 4. Si g n i f i c a n c e . The degree to which the major decision or s o l u t i o n appears to be based on relevant and s i g n i f i c a n t information as opposed to non-relevant and i n s i g n i f i c a n t information. S i g n i f i c a n t I n s i g n i f i c a n t 3 2 1 0 1 2 3 - 165 -Comprehensiveness. The degree to which the group's major decision or s o l u t i o n r e f l e c t s a response to a l l the dimensions of the problem under consideration. Comprehensive Noncomprehensive 3 2 1 0 1 2 3 Riskiness. The degree to which there would be r i s k associated with using the major decision or so l u t i o n proposed. Risky Non-Risky 3 2 1 0 1 2 3 Overall Quality. The o v e r a l l degree of q u a l i t y associated with the major decision or so l u t i o n proposed. Excellent Non-Excellent 3 2 1 0 1 2 3 - 166 -Appendix 0 Additional Analyses of the Quality of the Decision Outcome Ratings - 167 -Additional Analyses of the Quality of the Decision Outcome Ratings Within normal s t a t i s t i c a l considerations i t i s inappropriate to continue analyses once the o v e r a l l MANOVA i s found to be non- s i g n i f i c a n t . However, due to the consistency of r e s u l t s i n the ANOVA and simple main e f f e c t s analyses i t was considered worthwhile to report them i n appendix form. Immigration: MANOVA and ANOVA Analyses The MANOVA's of the ratings of the two Immigration experts did not y i e l d s i g n i f i c a n t r e s u l t s . However, of i n t e r e s t because of i t s consistency with l a t e r r e s u l t s , the ratings of the f i r s t expert demonstrated a s i g n i f i c a n t i n t e r a c t i o n e f f e c t on the ANOVA of the f e a s i b i l i t y measure (F ,„ = 4.96, p_ X , 4 / <.05). When the antecedent conditions of cohesiveness and time stress were present, the groups under a s s e r t i v e leaders produced decisions that were rated by the experts as being poorer than the decisions produced by the groups under non-assertive leaders. This s i t u a t i o n was reversed when the antecedent condi-tions were absent; the decisions made by groups under assertive leaders were rated as better than the decisions made by groups under non-assertive'leaders. The MANOVA of the second expert's ratings shows an i n t e r a c t i o n e f f e c t between antecedent conditions and leadership s t y l e that approaches s i g n i -ficance (F^ = 1-93, p_ = .09). The ANOVA's show that there i s a s i g n i -f i c a n t i n t e r a c t i o n e f f e c t on the measure of comprehensiveness (F_^ ^  = 9-88, p_ <.005). The comprehensiveness i n t e r a c t i o n e f f e c t was i d e n t i c a l to that found for the f e a s i b i l i t y measure by the f i r s t expert. That i s , when the antecedent conditions were present, decisions by groups under assertive leaders were rated poorer than decisions by groups under non-assertive leaders. In contrast, when the antecedent conditions were absent, the decisions made by:,groups under non-assertive leaders were rated poorer than - 168 -those made by groups under assertive leaders. The MANOVA of the combined data from the two immigration experts did not demonstrate s i g n i f i c a n t main or i n t e r a c t i o n e f f e c t s . However, as was found with the analyses of the data from the i n d i v i d u a l experts, the ANOVA's indicated s i g n i f i c a n t i n t e r a c t i o n e f f e c t s f o r the measure of f e a s i b i l i t y ^—1 44 = ^'33, p_ <.05) and the measure of comprehensiveness (F_^ ^  = 6.02, p_ <.05). The nature of the int e r a c t i o n s were i d e n t i c a l to those previously discussed concerning the data from the i n d i v i d u a l experts. C a p i t a l Punishment and Immigration: MANOVA and ANOVA's on the Combined Data The MANOVA r e s u l t s indicated no s i g n i f i c a n t e f f e c t s . However, the ANOVA's demonstrated s i g n i f i c a n t i n t e r a c t i o n e f f e c t s on the f e a s i b i l i t y measure (F_^ ^  = 4.40, p_ <.05) and on the comprehensiveness measure (F_^ ^ = 4.02, p_ <.05). This outcome i s somewhat remarkable given the low r e l i a b i -l i t i e s of the ratings across the two problems (see Tables 7 and 8), and supports the findings shown i n the analyses of the immigration data. Since these r e s u l t s demonstrated consistency across the immigration data and the two-problem combined data, i t was decided to perform a simple main e f f e c t s analysis on the s i g n i f i c a n t i n t e r a c t i o n terms of the f e a s i b i l i t y and com-prehensiveness measures. The information i s taken from the combined data across the immigration and c a p i t a l punishment problems. Figures 9 and 10 i l l u s t r a t e the i n t e r a c t i o n e f f e c t s on the comprehensiveness and f e a s i b i l i t y measures respectively. The simple e f f e c t s analysis on the comprehensiveness measure yielded a main e f f e c t that approached s i g n i f i c a n c e f o r the assertive versus the non-assertive groups at the point where the antecedent conditions were present (F^ ^  = 2.78, p_ - .10). S i m i l a r l y , the simple main e f f e c t s analysis on the f e a s i b i l i t y measure yielded two main e f f e c t s that approached - 169 -Figure 9 Interaction E f f e c t s : Comprehensiveness Measure Outcomes High i n Quality 3 r 2 X 1 + 0 4-Outcomes Low i n Quality 2 1 + 3 + (-.2) (-.9) + Antecedent Conditions Present Non-Assertive Leadership Style Assertive Leadership Style Antecedent Conditions Absent - 170 -Figure 10 Interaction E f f e c t s : F e a s i b i l i t y Measure Outcomes High i n Quality 3 T 2 X 1 X 1 X Outcomes Low i n Quality 2 X 4 - 1 • Antecedent Conditions Antecedent Conditions Present Absent Non-Assertive Leadership Style Assertive Leadership Style - 171 -s i g n i f i c a n c e . One simple e f f e c t was antecedent conditions present versus absent on the assertiveness v a r i a b l e (F_^ ^ = 3.89, p_ <.10). The second simple e f f e c t was the same as that found f o r the comprehensiveness measure, s p e c i f i c a l l y , an e f f e c t for the assertive versus non-assertive groups at the point where the antecedent conditions were present (F_^ ^ = 2.99, p_ <.10). Groupthink theory predicts that when the antecedent conditions of cohesive-ness and stress are present, groups led by non-assertive leaders w i l l per-form better than groups led by assertive leaders. The above simple, e f f e c t s analyses of both the f e a s i b i l i t y and comprehensiveness measures are not s t a t i s t i c a l l y s i g n i f i c a n t . However, the trends displayed are consistent and are i n conformity with the predictions of groupthink theory. In summary, the consistent but s t a t i s t i c a l l y n o n-significant pattern of i n t e r a c t i o n e f f e c t s and simple main e f f e c t s provides tentative support for the predictions of groupthink theory on two of the seven q u a l i t y of decision outcome measures. I t should be noted that these r e s u l t s are also consistent with the research of Courtright (1978) where a s i g n i f i c a n t i n t e r a c t i o n e f f e c t was found for one of the group process measures (the "statements of disagreement"). These consistencies provide an argument for considering the continued t e s t i n g of groupthink theory. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0095476/manifest

Comment

Related Items