Looking for Groupthink: Detecting Problems
in Small Group Decision Making.

by
Scott Stone

Thesis Sp«Com
c. 2
Stone, Scott.

1994 S8801

Looking for
Detecting problems m
1994

Submitted in Partial Fulfillment of the Requirements for
the Master of Arts in Communications Studies Degree

Approved by:

Dr. Kathleen Golden
Chairperson, Thesis Committee
Edinboro University of Pennsylvania

7/'r/w
Date

D/ Mary Alice Dye
Committee Member

O
Dr. Timothy Thompson
Committee Member

7f
Date

J\6C-&!28
C-"2-

Looking for groupthink:
Detecting problems in small group
decision-making.
Scott Stone
Dr. Kathleen M. Golden
April 22, 1994

ACKNOWKlanewaiNTS
Sincere appreciation is extended to the many people who
participated or assisted in the preparation and completion of
this study. Special thanks to Thomas Egloff for his time and
effort in coupleting this study.
Special appreciation is paid to Dr. Kathleen M. Golden,
thesis director, for her excellent guidance, encouragement,
continuous wise counsel, and friendship.
Gratitude for their interest and advice is expressed to the
members of the thesis committee: Dr. Mary Alice Dye and Dr.
Timothy Thompson.

This study was influenced greatly by Dr. Randy

Hirokawa, whose advice, information, and time are greatly
appreciated.
To Rose, my father, and my mother, the writer acknowledges
his indebtedness for the many sacrifices and never-ending
encouragement which brought a mutually-shared goal to
realization.

This study would not have been possible to complete

without their help and support.

ABSTRACT
Hie purpose of this investigation was to examine the reasons why
small groups may make effective or ineffective decisions. This
research combined two related studies to examine both the effects
of group cohesion on small group decision quality and the
groupthihk phenomenon effect on small groups. A qualitative
approach was used in examining the data. The attempt to create
cohesion among group members proved highly successful, and as a
result many of the problems that are often associated with
cohesive groups appeared. The results of this study lend support
to Hirokawa and Pace's theory that group decision-making quality
is dependant upon key aspects of interaction between group
members. Furthermore, the propositions that proved true continue
the research into Janis' theory of groupthink. Overall this study
offers further insight into the reasons why small groups make
effective or ineffective decisions.

TABLE OF CONTENTS
Page
I. INTRODUCTION

1

A. Research Problem

2

B. Terms

2

II. REVIEW OF THE LITERATURE .

5

A. Irving Janis .......................

5

B. Data to Theory Researchers

5

C. Theory to Data Researchers

7

D. Summary...............................

11

E. Table of Researchers . . .

13

III. METHOD

14

A. Judging

16

B. Propositions

21

C. Verification

22
22

IV. RESULTS
A. Propositions .......................................

22

B. Promotive and Disruptive Influences

23

C. Demographics

...................................

25

.......................................

26

D. Limitations
V. DISCUSSION

...........................................

VI. RECOMMENDATIONS FOR FUTURE RESEARCH

28
29

Introduction
The concept of groupthink originated from the research
of Irving Janis, who was trying to determine why groups
comprised of highly intelligent and effective decision­
making individuals often combined to formulate some of the
most ill-conceived decisions of our time.

Janis states

"Groupthink refers to a deterioration of mental efficiency,
reality testing, and moral judgment that results from in­
group pressures." (Janis, 1972, p.28) Janis has spent a
great deal of time examining some of recent history's worst
decision-making events, trying to summarize how the groups
found themselves making such terrible decisions.
Hirokawa is one of many people who has examined some of
Janis's ideas.

Hirokawa and Pace (1983) researched what

elements go into making an effective or ineffective
decision.

This team did an extensive study on small group

decision-making.

The major emphasis of this study comes

from an experiment Hirokawa and Pace conducted in which they
examined effective and ineffective groups to find out what
aspects of decision making separated them.

The current

study continues the experiments of Hirokawa and Pace with a
closer emphasis on Janis's symptoms of groupthink.

These

symptoms are described in the terms section of this paper.
By replicating the experiment Hirokawa and Pace directed,
changing only the four propositions to compare to groupthink
symptoms, experimental results lend additional support to
Janis's theory.

Looking for groupthink

2

Research Problem
By replicating Hirokawa and Pace (1983) and by changing
the propositions to include some of the symptoms of
groupthink, the hope is to add a greater degree of substance
to the premise that Janis's symptoms of groupthink do indeed
lead to low quality decisions.

The research question asks:

Do the symptoms of groupthink really exist in small group
decision-making and if so, are they contributors to
ineffective decision-making?
Some of the problems examined were:

Why do some small

groups make ineffective, low quality decisions? Since this
problem is very general, the scope of the research was
narrowed to examine these problems: Are there any key
elements that ineffective groups share that are not found
within effective groups? Are there any key elements that
effective groups share that are not found in ineffective
groups?
Terms
Below are the key terms which guide this study.
Groupthink - "Groupthink refers to a deterioration of mental
efficiency, reality testing, and moral judgment that results
from in-group pressures." (Janis, 1972, p.28)
Effective group - a group whose decision received a rating
of at least a "6," on a scale from 1 to 7, on all four
criteria by all three judges (Hirokawa and Pace 1983) .

Looking for groupthink

3

Ineffective or-mip - a group whose decision received a rating
of no higher than "2," on a scale from 1 to 7, on all four
criteria by all three judges (Hirokawa and Pace 1983).
Symptoms of groupthink 1. An illusion of invulnerability, shared by most or
all members, which creates excessive optimism and
encourages taking extreme risks;
2. Collective efforts to rationalize in order to
discount warnings which might lead members to
reconsider their assumptions before they recommit
themselves to past policy decisions;
3. An unquestioned belief in the group's inherent
morality, inclining the members to ethical and moral
consequences of the decisions;
4. Stereotyped views of enemy leaders as too evil to
warrant genuine attempts to negotiate, or as too weak
and stupid to counter whatever risky attempts are made
to defeat their purpose;
5. Direct pressure on a member who expresses strong
arguments against any of the group's stereotypes,
illusions, or commitments, making clear that this type
of dissent is contrary to what is expected of all loyal
members;
6. Self-censorship of deviations from the apparent
group consensus, reflecting each member's inclination

Looking for groupthink

4

to minimize to himself/herself the importance of
his/her doubts and counter-arguments;
7. Shared illusion of unanimity concerning judgements
conforming to the majority view (partly resulting selfcensorship of deviations, augmented by the false
assumption that silence means consent);
8. The emergence of self-appointed mindguards - members
who protect the group from adverse information that
might shatter their shared complacency about the
effectiveness and morality of their decisions. (Janis
1972, Montanari & Moorhead 1989)

Looking for groupthink

5

Review of literature
Tiie study of group decision-making has been around as
long as people have joined in an attempt to create a higher
quality decision.

Much of the literature reviewed is dated,

but necessary to understand the concepts that came about
from this early work.
Janis (1972, 1982) has been a pioneer in the study of
group decision-making.

Janis believes there are distinct

elements which hinder a group's ability to make high-quality
decisions.

In the books Groupthink and Victims of

Groupthink, Janis illustrates his theory by applying it to
some of the worst decisions ever made by high-powered
groups.

He examines the Kennedy administration's errors

that led to the Bay of Pigs fiasco and Nixon's Watergate
coverup.

In a counterpoint, he shows us how Kennedy' s

administration learned from its mistakes and avoided
groupthink during the Cuban Missile Crisis.

These books

were the catalysts from which many other studies and
theories of group decision-making arose.
Many researchers, like Janis, took the data-to-theory
approach and applied groupthink to obvious decision-making
fiascoes creating case studies.

These data-to-theory

researchers include Raven (1974), who examined the Nixon
administration and the Watergate coverup focusing on the
administration's inability to accept that they had been
caught; Huseman & Drive (1979), who applied Janis' theory to

Looking for groupthirik

6

both small and large business groups to show both were
susceptible to groupthink; and Manz & Sims (1982), who
attempted to elaborate on Janis' theory to include
autonomous work groups.

These early studies were marked by

a constant emphasis on the need for further research into
group decision-making.
Continuing to examine flawed decisions, Smith (1984)
looked at the possibility of groupthink playing a large part
in the failure to liberate hostages from Teheran, and
Hensley & Griffin (1986) studied the Kent State University
Board of Trustees' decisions during the 1977 gymnasium
controversy.

This study was continued through many articles

and has still not reached a complete resolution.

The

Challenger Shuttle disaster left N.A.S.A. under the
microscope of many researchers.

Among the first to examine

this catastrophe were Esser & Lindoerfer (1989), who
questioned the decision of N.A.S.A. to launch the shuttle,
using recent studies conducted by others in the field as
support for their research.

This study thoroughly examined

verbal interactions recorded just before the shuttle launch
and tried to tie in the symptoms of groupthink with the
interaction of the people involved.
The most recent case study reviewed was by Moorehead,
Ference, & Neck (1991), and continued the analysis of
decision fiascoes by reexamining the Space Shuttle
Challenger decision and revising the groupthink framework.

Looking for groupthink

7

Much of Moorehead, Ference, & Neck's research was based on
McCauley's (1989) research when he reviewed seven
independent cases for possible groupthink trying to
establish a clearer definition of the groupthink phenomenon.
Whyte (1989) proposed a new look at groupthink by combining
the pressures for uniformity with the aspects of framing,
risk, and group polarization.

This unique view provides a

wider, but often more confusing, way of examining fiascoes.
In 1990 Hart (1990) applied his revised model of groupthink
to Irangate and the failure to prevent the 1940 Nazi
invasion of Holland.
These resources give the reader a good basis of
knowledge into the possible reasons why these fiascoes
occur.

However, this only helps us to understand why these

poor decisions were made.

Very little information is

supplied on how to keep from falling victim to the symptoms
of groupthink.

We need to better understand the ways in

which groups go about making high-quality decisions, and
what they are doing differently from groups making
predominantly low-quality decisions.

This is the approach

taken by the theory-to-data analysts.
The theory-to-data approach was led by Flowers (1977)
who used 120 undergraduates to examine how group
cohesiveness and leadership style affect a group's decision­
making process.

He found that group cohesiveness was not a

significant factor in the quality of the decision, but that

Looking for groupthink
leadership style was.

8

The combination of leadership style

and group cohesiveness was also not found to be a
significant enough factor to influence the decision-making
process.

Since this is an aspect of particular importance

to this study, his findings were examined very closely.
Close attention was also paid to Courtright (1978), who used
96 undergraduates to study the effect of group cohesion and
leadership style on the quality of decisions reached by
groups.

He could not prove any significance between group

cohesion or leadership style when associated with the
quality of decisions but used many interesting techniques of
gathering the large amount of data.

For example, Courtright

told twelve groups that they were highly cohesive, and told
another twelve groups that they would probably not get
along.

Using these two different sets of data, he was able

to compare and contrast the interpersonal reactions of the
groups.

Since Courtright could not prove any significance

of group cohesion or leadership style on the quality of the
decisions, this experiment attempted to create cohesion
among the members of the groups.

Janis believed cohesion is

a key element in the groupthink phenomenon and therefore
this experiment will attempt to introduce that element.
Using content analysis, Tetlock (1979) researched
American foreign policy decisions looking for possible
groupthink situations.

By evaluating the complexity of the

speaker, the political group with which the speaker

Looking for groupthink

9

identifies, and the opponents of the speaker, Tetlock
observed the effects of these elements on the quality of the
decisions made.

This study has come under scrutiny because

of the limited scope of information examined by Tetlock.

In

the wake of Tetlok's experiment Fodor and Smith (1982)
reexamined Flowers study regarding group cohesion and
leadership style using 80 undergraduates.

Again group

cohesion showed little significance to the quality of the
decisions reached by the group, while leadership style
proved to be a significant factor.

In another large

experiment Callaway & Esser (1984), using 128
undergraduates, examined the problem-solving procedure
undertaken by these groups, along with the effects of group
cohesion.

The students were randomly assigned to 32 four-

person groups.

Half the groups were given a decision-making

problem concerning the amount of money a horse trader made
during a series of trades, while the other half were given a
problem regarding ranking 15 items by importance they would
want if lost at sea.

Only the lost at sea problem showed

any significance in group cohesion, while the problem
solving procedures showed no significance at all.

However,

when combined, the group cohesion and the problem solving
procedure showed significance in both the lost at sea and
horse trader problems.

This experiment proved very

informative in both its results and the way in which the
experiment was carried out.

Many of the minor changes made

Looking for groupthink 10
from Hirokawa and Pace's study can be linked to aspects of
Callaway and Esser's experiments,

Some of these changes

include the use of the personality questionnaire to create
cohesion and the post experiment questionnaire to measure
cohesiveness.
Continuing their research Callaway, Marriott, and Esser
(1985) used 112 undergraduates to examine the effects of
dominant members and procedural guidelines in decision­
making groups.

The results showed that dominant members

made higher quality decisions, exhibited less anxiety, took
more time to reach a decision, verbally agreed and disagreed
more, and reported more group influence on members.

The

decision-making procedures affected only the amount of time
it took to reach a decision.

In the same year, Leana (1985)

used 52, four-person groups comprised of undergraduates and
recreated Flowers' experiment with a few more dependant
variables.

Group cohesiveness showed significance on one-

fourth of the decision process while leadership style showed
significance on three-fourths of the decision process.

The

combination of leadership style and group cohesion again
proved insignificant but was judged to be worthy of further
study.

Moorhead and Montanari (1986) had 45 teams of three

to five college seniors, and used factor and path analysis
to study group cohesion, insulation, and leadership style.
By attempting to introduce symptoms of groupthink into the
group, they studied its effects on the quality of the

Looking for groupthirik 11
decisions produced.

The results showed that the effects of

groupthink on the quality of decisions were not as
substantial as Janis had implied.

Since it is such a

difficult task to introduce any factor consistently into a
group's decision-making, especially factors dependant on
intergroup dynamics, this experiment left considerable room
for further research into groupthink.
Hirokawa (1980, 1983, 1987, 1989) began examining
Janis's theory of groupthink and eventually branched off to
examine other possible problems with small group decision­
making.

Hirokawa has published numerous articles on the

topic of small group decision-making, looking more toward
why groups make effective or ineffective decisions.

In his

first publications, Hirokawa examined the elements used by
both effective and ineffective groups while coming to a
decision.

He found a strong correlation between group

interaction and the quality of the decision reached.

These

articles cover his reasoning, procedures for gathering data,
findings, and reactions.

Hirokawa's later work focuses more

on how effective groups go about making high-quality
decisions and what people can do to avoid the problems
associated with low quality decision groups.
A substantial amount of research still must be done
before any concrete answers will develop as to why some
groups make high-quality decisions while others cannot.
Hopefully, by examining as many aspects of group decision-

Looking for groupthink 12
making as we can, we will be able to focus in on those
characteristics that keep some groups from achieving highquality decisions.

The concepts of cohesion, groupthink,

leadership style, group interaction, and problem solving
procedures have been analyzed from many different points of
view.

This experiment adds another piece to the puzzle of

group decision-making by combining ideas to examine group
interaction from yet another angle, adding to the general
pool of information from which the next researcher will
gather his or her ideas.

The table on the following page

presents the focus and the approach of the research reports
that were studied during the planning of this experiment.

Looking for groupthink 13

Authors by
Focus and Approach

Data
Theory
to
to
Theory Data

Groupthink

Group
Cohesive
-ness

Janis (1972, 1982)
Raven (1974)
Flowers (1977)
Courtright (1978)
Huseman & Drive (1979)
Tetlock (1979)
Hirokawa (1980, 1983,
1987, 1989)
Fodor (1982)
Manz & Sims (1982)
Callaway & Esser
(1984)
Smith (1984)
Callaway, Marriott &
Esser (1985)
Leana (1985)
Hensley & Griffin ■
(1986)
Moorhead & Montanari
(1986)
Esser & Lindoerfer
(1989)

:kkk

kkk

kkk

kkk

McCualey (1989)

kkk
kkk
kkk
kkk

kkk
kkk
kkk
kkk

Whyte (1989)
Hart (1990)
Moorehead, Ference &
Neck (1991)

kkk
kirk

kkk
kkk
kkk
kkk
kkk

kkk

kkk
kkk
kkk

kkk
kkk
kkk
kkk

kkk
kkk

kkk

kkk
kkk

kkk

kkk

kkk
kkk

kkk

kkk

kkk

kkk

Looking for groupthink 14
Method
This study began by attempting to obtain samples of
"effective" and "ineffective" groups.

To do this,

undergraduate communication classes at a North-eastern
University were used.

The volunteers were told that they

would be participating in a study of group decision-making
including a 30-minute, group discussion that would be
videotaped.

A portion of the 16PF (Schuerger, 1992)

personality factor instrument was administered to three
classes of approximately 25 students.

A copy of this

personality questionnaire is included in Appendix E.

It was

explained to the students that the personality factor
questionnaire would be the basis for assigning each of them
to groups.

They were also told that by using these

questionnaires to examine each of their individual
personalities, we would be able to create groups who should
work well together and be able to reach high-quality,
effective decisions.

A written copy of what the students

were told can be found in Appendix A.
Janis believed that groupthink is a result of the
cohesiveness of the group.

To create cohesion among the

group members, a technique used by Callaway and Esser (1984)
was used.

The classes participating in the study were

administered the personality questionnaire and the subjects
were led to believe that the questionnaire was the basis for
assigning them to groups with similar personalities.

In

Looking for groupthink 15
reality, the questionnaires were no more than a tool to
allow the students the freedom to believe in their group's
ability to come up with a high-quality decision.

The only

piece of information collected from the questionnaire was
the student's name.

Each of the questionnaires collected

was numbered one through the number of people in the class.
A random number generating computer program was used to
randomly assign each of these numbers to one of four groups
for each class.

Each of these numbers was used to find the

corresponding student in each class and set up the groups.
Students who were not assigned to one of the four groups per
class were listed as alternates for absent class members.
After the groups had been formed, half of the groups
were given the task of deciding on a recommendation for
dealing with a journalism student who was caught
plagiarizing on a class assignment.

The other half of the

groups were asked to produce a recommendation for dealing
with a dormitory resident who maliciously damaged her
roommate's stereo system.

Each group was given written

background information pertaining to their assigned case,
and each had 30 minutes to reach the best decision they
could.

All the groups were videotaped while making their

decision.
A post-test was administered to all group members to
check the cohesiveness manipulation.
can be found in Appendix D.

A copy of this test

The subjects responded to a

Looking for groupthink 16
seven-point Likert-type (Likert, 1932) scale to questions
concerning different aspects of their interpersonal
communication within the group.

These data provided a

better understanding as to the subjects' confidence in the
group's decision quality, agreement with the group's
decisions, extent of attempted influence by the group, and
extent of actual influence by the group.

A small

demographic questionnaire was placed on the back of the
post-test.

Although this information was not used in this

experiment, it may be helpful for future reference and
further study.
After completing the questionnaire, the students were
debriefed as to the actual purpose of the study.
copy of the debriefing can be found in Appendix B.

A written
After

the debriefing the students were allowed to ask questions
pertaining to the study and were asked not to discuss any of
the experiment with individuals outside of their class until
the next week.
The next step was to evaluate the decisions reached by
the groups in order to identify both high and low-quality
decisions for further study.
of judges were utilized.

To do this, two separate sets

Each set of judges consisted of

three individuals who have the knowledge and expertise to
properly evaluate the decision reached by the groups.

The

judges for the plagiarism case consisted of two journalism
professors and an English professor while three senior

Looking for groupthink 17
resident assistants judged the vandalism case,

The judges

were asked to evaluate the group's decisions in four
categories: appropriateness, warrantedness, reasonableness,
and fairness.

The judges used the scaling system below.

Appropriate _1_:_2
Warranted _1_:_2'
Reasonable _1_:_2'
Fair 1 : 2

3_:_4_:_5_:_6_:_7_
'3_:_4_:_5_:_6_:_7_
3_:_4_:_5_:_6_:_7_
3 : 4 : 5 : 6 : 7

Inappropriate
Unwarranted
Unreasonable
Unfair

The next step was to identify "effective" and
"ineffective" groups.

Hirokawa and Pace believed that the

four evaluation criteria were equal in importance, and
therefore the researcher will not try to weight one category
more than another.

Using the judges' evaluations as a

basis, and by defining an "effective" group as a group whose
recoirmendation received at least a "6" on all four criteria
by all three judges, and an "ineffective" group as a group
whose recommendation received no higher than a "2" on all
four criteria by all three judges, "effective" and
"ineffective" groups were selected.

By these criteria one

effective group and one ineffective group were chosen.
These group's written decisions as well as judges I
evaluations can be found in Appendix C.
Analysis of the group discussions created a small
problem for Hirokawa and Pace.

The decision to use a

structured or unstructured approach for analyzing the data
took much consideration.

For this study, an unstructured

Looking for groupthink 18
approach was used to analyze the discussions in a more free­
form manner, without the aid of any a priori observational
categories.

The reason this approach was taken was that an

unstructured format was deemed more appropriate for this
kind of study; since the discussions were to look at the
problem as broadly and as deeply as possible, Hirokawa and
Pace did not want to restrict the judges' opinions,

In this

experiment the judges were given checklists of observational
categories to the judges for two main reasons,

First, since

the experiment was looking for specific aspects of the
decision-making process, it was deemed unnecessary for the
judge's to have to labor over each communication utterance.
Second, since we have the benefit of Hirokawa and Pace's
findings, we already have an extensive list of promotive and
disruptive influences to create checklists for use by the
judges.
Procedures for analyzing the group discussions were set
up by creating two separate research teams.

Each team

consisted of two principal observers, who were carefully
instructed as to the objective of the study and the
procedures that would be errployed for analyzing the
discussions.

One team began by examining the "effective"

groups independently of each other.

The goal of the team

was to "identify all observable aspects of the discussions
that might have accounted for the groups' high-quality
decisions"(Hirokawa & Pace, 1983).

Each judge was

Looking for groupthink 19
instructed to carefully read through the list of promotive
influences and independently view the videotape looking for
an instance of promotive influence - that is, "communication
utterances which function to help the group arrive at a
high-quality decision" (Gouran & Hirokawa, 1980) .

Once the

judges had looked at the video tape, they met as a team and
compared their respective observational lists, reexamining
any discrepancies.

By comparing these lists, the team came

up with one general list of communication characteristics
which appear to account for the decision-making success of
the "effective" groups.

To formulate this team list the

judges had to agree that the promotive influence actually
occurred during the discussion and the promotive influence
had to be common to both judges.
The judges' next job was to observe and analyze the
"ineffective" groups.

The judges watched the video tape

independently of one another and attempted to "identify all
observable aspects of the discussions that might have
accounted for the groups' low-quality decisions"(Hirokawa &
Pace, 1983).

Each judge was instructed to carefully read

through the list of disruptive influences and independently
view the videotape, looking for any instance of disruptive
influence - that is, "communication utterances which
function to prevent the group from arriving at a highquality decision" (Gouran & Hirokawa, 1980).

Once the

judges had looked at the video tape they met again as a team

Looking for groupthink 20
and compared their respective observational lists,
reexamining any discrepancies.

By comparing these lists,

the team came up with one general list of communication
characteristics which appear to account for the decision­
making failure of the "ineffective" groups.

To formulate

this team list the judges had to agree that the disruptive
influence actually occurred during the discussion and the
disruptive influence had to be common to both judges.
The next step was to repeat the judging in reverse with
the second team of judges first examining the "ineffective"
groups for disruptive influences and then studying the
"effective" groups for promotive influences.

Then the two

teams met and compared their respective team lists for the
purpose of reducing results to two final lists: a list of
disruptive influences not found in effective groups, and a
list of promotive influences not found in ineffective
groups.

Looking for groupthink 21
Propositions
Hirokawa and Pace developed four propositions that they
believed would be proven by their study,

The researcher

changed these to more closely represent symptoms of
groupthink and tested these propositions to add substance to
the idea that Janis's theory of groupthink does affect the
quality of decision-making.
Proposition 1: The quality of a group's decision may be
dependent upon the manner in which group members examine and
evaluate the risks involved with each decision (illusion of
invulnerability) .
Proposition 2: The quality of a group' s decision may be
dependent upon the manner in which group members allow each
other to present and evaluate alternatives (direct pressure
on dissenters).
Proposition 3: The quality of a group's decision may be
dependent upon the group's ability to assess the ethical and
moral consequences of their decisions (unquestioned belief
in the group's inherent morality).
Proposition 4: The quality of a group's decision may be
dependent upon the group's willingness to rationalize
decisions and discount warnings that might lead the members
to reconsider their assumptions (collective
rationalization).

Looking for groupthink 22
Verification
The research was verified by turning the four
propositions into four seven-point rating scales, and then
having other observers use those scales to rate the quality
of the group discussions.

Hirokawa and Pace reasoned that

if the observations were valid, the raters should rate the
discussions in a manner consistent with the predictions of
the four propositions.

The researcher believes this to be a

valid conclusion, so he too used this method of
verification.

The results of the verification showed that

there are large differences in "effective" and "ineffective"
groups.
Results
The goal of this experiment was to add substance to
Janis' theory that highly cohesive groups sometimes make
low-quality decisions, not because of any flaw in the
individuals themselves, but in their interaction with each
other and their method of decision-making.

The researcher

set about this by establishing four propositions, directly
related to groupthink symptoms, that he felt would separate
• effective decision-making groups from ineffective ones.

Of

these four propositions, three were found to exist in
ineffective groups but not in effective groups.

The three

propositions found to be valid were:
Proposition 1: The quality of a group's decision may be
dependent upon the manner in which group members examine and

Looking for groupthink 23
evaluate the risks involved with each decision (illusion of
invulnerability).
Proposition 3: The quality of a group's decision may be
dependent upon the group's ability to assess the ethical and
moral consequences of their decisions (unquestioned belief
in the group's inherent morality).
Proposition 4: The quality of a group's decision may be
dependent upon the group's willingness to rationalize
decisions and discount warnings that might lead the members
to reconsider their assumptions (collective
rationalization).
The second proposition was rejected because it was
found in both the effective and ineffective groups.

This

proposition stated:
Proposition 2: The quality of a group's decision may be
dependent upon the manner in which group members allow each
other to present and evaluate alternatives (direct pressure
on dissenters).
After the judging process, lists of promotive
influences common only to effective groups and disruptive
influences common only to ineffective groups were formed.
These lists are supplementary to the propositions but proved
to be very enlightening. These list are:
Pramotive influences not found in ineffective groups.
1. The group determined the difference between facts,
evidence, and speculation.

Looking for groupthink 24
2. The group had a member who controlled communication
channels and promoted evenness of participation.
3. The group asked for clarification on vague ideas
presented by other group members.
4. The group avoided rationalizing their decisions with
irrelevant information, sticking with the facts to back up
their conclusions.
5. The group was able to assess the ethical and moral
consequences of their decision.
6. The group was able to examine and evaluate the risks
involved with each decision.
Disruptive influences not found in effective groups.
1. The group had a member(s) who withdrew from group
discussion by refusing to participate and/or engaged in
irrelevant conversations.
2. The group made no attempt to determine the difference
between facts, evidence, and speculation while making their
decision.
3. The group had a member who monopolized group time with
long, drawn-out monologues.
4. The group had a member or members who went along with one
member1s ideas without examining other possibilities.
5. The group went along with one member's vague ideas
without asking for clarification.
6. The group had a member who goes along with other members'
ideas without questioning them or offering any ideas of
their own.
7. The group rationalized their decisions with information
other than facts.
8. The group failed to examine the ethical and moral
consequences of their decision.
9. The group failed to examine and evaluate the risks
involved with each group.
A key variable in this experiment was the group
members' perceived cohesiveness with one another.

The post-

Looking for groupthink 25
experiment evaluation was administered to measure how
cohesive the group members themselves felt within their
group.

With the copy
Post Experiment Evaluation IVfean Scores

of the post experiment
evaluation that can be
found in Appendix D
and the chart here, we
can see that the

is
Q1

Q2

Q3

Q4

□ Overall
■ Ineffective

□ Effective

Q5

overall cohesion level among group members was quite high.
Questions three and four are deceiving due to the nature of
the scaling, but are showing highly-cohesive groups. Both
the effective and ineffective groups show a higher than
average cohesion level in four of the five questions.

The

groups' mean cohesion levels were well above the acceptable
mean of four on all five questions.

Ages of Students

Demographic information
is presented in charts as
seen here.

The average age

of the students and the average
class rank were surprising with

□
■
□
■
□
■
□
ED

18
19
20
21
22
23
24
25+

eight non-traditional students and
more juniors than any other class rank.

The more advanced

age of the students could have contributed to the overall
high-quality of the decisions reached.

With the older

students bringing more group decision-making experience into
the resolution process, the overall conclusions reached were

Looking for groupthink 26
enhanced.

Groups containing at least one member of the over

24 age group scored an average of 1.66 points better in the
overall decisions rated by the judges.

This calculates out

to a 5.9% increase in their overall score.

Further research

into the role of age and experience in the group decision­
making process might provide more insight into another
aspect of quality decision-making.

Since the average age of

undergraduate students at large Universities is under age
21, the more mature average age of students in this study
could have produced higher-quality decisions than might
occur in similar studies done at other large Universities.
Women outnumbered men in this study as shown below
reflecting the demographics of the University where female
students outnumber male students.

Women were not the

majority in every group due to factors involving random
placement into groups.

Demographic information may prove

useful in further research on this subjectveer inSchool
Number of Females to Males

0 2 4 6 8 1012141618 20 22 24 26 28 30 32 34

□ Freshman
■ Sophmore
□ Junior
■ Senior

Limitations
With this experiment, the attempt was to test Janis's
theory of groupthink by replicating a previous study with
minor changes,

The limitations that exist are that it took

at least fifty volunteer subjects and eleven judges to
replicate Hirokawa and Pace's study.

A study using this

Looking for groupthink 27
many people takes considerable planning, resources, and time
to accomplish.

Time proved to be a considerable limitation

in the collection of data.

The experiment hinged on

misleading the students into believing the groups were
formed to allow the highest quality decisions.

In the post­

experiment debriefing the students were told about the
random group assignments and were asked not to discuss this
with others outside of class until the following week.

To

reduce the chance of the students discussing the experiment
with other individuals, the entire data collection phase of
the experiment was conpleted in just three days.

Another

limitation was that Janis's theory of groupthink is very
broad, and it was difficult to define the symptoms
precisely.

To define these symptoms as influences of

communication was also not an easy task and took much
consideration.

Looking for groupthink 28
Discussion
This experiment was originally conceived to examine the
effects of groupthink on small group decision making.

As

the research into current literature progressed, the idea of
what aspects allow some groups to create higher quality
decisions than other groups came as a natural extension of
the groupthink problem.

When it came time to begin

collecting data, the overriding goal was to try to find the
aspects of decision making used by effective groups that
were not found in ineffective groups.

The propositions

relating to groupthink came as extensions to the
superordinate goal of the decision making aspects.

The

experiment confirmed three propositions and rejected one
while the attempt to create cohesion among group members
proved effective.
The attempt to create cohesion among group members was
a key element in establishing a successful environment for
groupthink to occur.

Cohesion usually takes place over time

and thus time is often thought of as a necessary ingredient
of groupthink.

Since time was limited for this experiment

another option had to be found in the way of creating
cohesion.

After carefully examining the literature it was

found that another research group (Callaway & Esser, 1984)
had administered, with outstanding success, a bogus
personality questionnaire in an attempt to create cohesion.
This method was chosen since it allowed the study to remain

Looking for groupthink 29
in a limited time frame.

Callaway & Esser's experiment also

addressed the aspect of zero-history groups sunmarizing that
with the cohesive manipulation, the lack of prior group
interaction was insignificant.
A wide variety of decisions was formulated by the
groups and can be found in Appendix C.

Although many of the

decisions were similar in the individual ideas the group
presented, the final judgment often differed substantially.
The fact that the judges had to rate the decisions on the
overall conclusion reached, and not the elements within each
decision, frustrated some of the judges.

The expert judges

who examined the written decisions expressed a feeling that
many of the conclusions that were rated as poor or mediocre
included excellent elements and ideas, but were not
altogether fair when brought together as a package.

This

eliminated some decisions that may have been excellent
except for one or two elements.

An example of this can be

found on the second page of Appendix C.

Group 4 had many

good ideas but two of the judges felt that asking the young
woman's parents to attend counseling was beyond their right.
Unfortunately, there is no easy way to avoid this problem
since the total decision must be examined.
Recommendations for Further Research
When this study was originally conceived it was not
intended to be such a vast undertaking for the researcher.
Many unexpected and interesting problems presented

Looking for groupthink 30
themselves along the way.

This study could have been

improved by increasing the number of groups and judges used.
This increased amount would allow a more narrow scope of
research while lending more credibility to the data.
Allowing the students to meet a few times in their groups
and make decisions similar to the type used in this
experiment, prior to videotaping groups, could prove useful
in future research.

The extended time would allow the

groups to become more comfortable with one another, and
possibly could improve group cohesion.

In turn, this could

increase the interval between high-quality and low-quality
decisions by eliminating many of the interpersonal problems
associated with first-time meetings of group members.
The experimenter gained a great deal of knowledge into
both the subject matter reviewed for formulating this
experiment and the planning, effort, and time it takes to
conduct such an investigation.

The data generated by this

study should allow better insight into small group decision­
making and the problems associated with it.

The researcher

will gain much more information than he expected concerning
the effort and time constraints of such an experiment, the
generosity with time and information of many University
Professors across the country, the political process that
researchers must wade through, and the patience of friends
and family.

Bibliography
Barge, J.K. & Hirokawa, R.Y. (1989). Toward a communication
competency model of group leadership. Small Group Behavior.,
20 (2) ,167-189.
Callaway, M.R. & Esser, J.K. (1984) . Groupthink: Effects of
cohesiveness and problem-solving procedures on group
decision-making. Social Behavdor and Personality. 12 (2),
157-164.
Callaway, W.R., Marriott, R.G. & Esser, J.K. (1985). Effects of
dominance on group decision-making: Toward a stress
reduction explanation of groupthink. Journal of Personality
and Social Psychology. 49.949-952.
Cline, R.J.W. (1990).

Detecting groupthink: Methods

for observing the illusion of unanimity.

Comrnum cation

Quarterly. 28. (2),112-126.
Courtright, J.A. (1978) . A laboratory investigation of
groupthink. Communication Monographs. 45,229-246.
Esser, J.K. & Lindoerfer, J.S. (1989) . Groupthink and the space
shuttle challenger accident: Toward a quantitative case
analysis. Journal of Behavioral Decision-making, 2,167-177.
Flowers, M.L. (1977). A laboratory test of some implications of
Janis's groupthink hypothesis. Journal of Personality and
Sonia1 Psychology. 22,888-896.
Fodor, E.M. & Smith, T. (1982) . The power motive as an influence
on group decision-making. Journal of Personality and Social
Psychology. 42,178-185.

Hart, P. (1990) . Groupthink In government-: A study of small
.groups and policy failurer Published doctoral dissertation,
Rijksuniversiteit, Leiden, Netherlands.
Hensley, T.R. & Griffin, G.W. (1986) . Victims of groupthink: The
Kent State University Board of Trustees and the 1977
gymnasium controversy. Journal of Conflict Resolution. 30.
497-531.
Hirokawa, R.Y. (1987). Why informed groups make faulty decisions:
an investigation of possible interaction-based explanations.
Small Group Behavior,. 18 (l),3-29.
Hirokawa, R.Y. (1980). A comparative analysis of communication
patterns within effective and ineffective decision-making
groups. Communication Monographs. 47,314-321.
Hirokawa, R.Y. (1988). Group communication and decision-making
performance. Human Communication Research. 14 (4),487-515.
Hirokawa, R.Y. & Johnston, D. (1989). Toward a general theory of
group decision-making: Development of an integrated model.
.Small Group Behavior. 20 (4),500-523.
Hirokawa, R.Y. & Pace, R. (1983) . A descriptive investigation of
the possible communication-based reasons for effective
and ineffective group decision-making. Communication
Monographs. 50.363-379.
Huseman, R.C. & Driver, R.W. (1979). Groupthink: Implications for
small group decision-making in business. Readings in
organizational, behavior. Boston: Allyn and Bacon.

Janis, I.L. (1982) . Groupthink.

Boston: Houghton Mifflin

Company.
Janis, I.L. (1972) .

Victims of groupthink.

Boston:

Houghton Mifflin Company.
Leana, C.R. (1985) . A partial test of Janis' groupthink model:
Effects of group cohesiveness and leader behavior on
defective decision-making. Journal on Management, 11,5-17.
Likert, R. (1932) . A technique for the measurement of attitudes.
Archives of Psychology, No. 140.
Manz, C.C. & Sims, H.P. (1982) . The potential for groupthink in
autonomous work groups. Human Relations. 15.,773-784.
McCauley, C. (1989) . The nature of social influence in
groupthink: Compliance and internalization. Journal of
Personality and Social Psychology, 57 , 250-260.
Montanari, J.R. & Moorhead G. (1989) . Development of the
groupthink assessment inventory. Educational and
Psychological Measurementf 49,209-219.
Moorhead, G., Ference, R. & Neck C.P. (1991) . Group decision
fiascoes continue: Space shuttle Challenger and a revised
groupthink framework. Human Relations, M (6),539-550.
Moorhead, G. & Montanari, J.R. (1986) . An empirical investigation
of the groupthink phenomenon. Human Relations, 38 (5),
399-410.

Park, W. (1990) . A review of research on groupthink. Journal of
Behavioral Decision-makincr, a (4),229-245.

Posner-Weber, C. (1987) . Update on groupthink. Small Group
Behavior, 13. (1),H8-125.
Raven, B.H. (1974) . The Nixon group. Journal of Social Issues.
aa, 117-126.
Smith, S. (1984) . Groupthink and hostage rescue mission. British
Journal of Political Science, 15 f117-126.
Schuerger, J.M. (1992) . The Sixteen Personality Factor
Questionnaire and its junior versions. Journal of Counseling;
& Developement, 21,231-244.
Tetlock, P.E. (1979) . Identifying victims of groupthink from
public statements of decision makers. Journal of Personality
and Social Psychology, 37r1314-1324.
Whyte, G. (1989) . Groupthink reconsidered. Academy of Management!
Review. 14 (1),40-56.

Looking for groupthink

Appendix A

Appendix A
Pre-experiment:
Congratulations, you have been selected as members of the
Honors and Ethics Committee at Edinboro University. Each
committee will consist of five members and be given a different
case on which to render a decision. These groups have been
carefully selected to ensure the highest quality decisions based
on the individual personality traits of the committee members.
Please read over the case thoroughly, paying close attention to
all the relevant information. You will be given approximately 30
minutes in which to deliberate the case among yourselves. At the
end of that time you will be asked to render a written decision
regarding the action to be taken. The time remaining will be
announced at the 15 and 25 minute marks. Take a few minutes to
carefully read over the case. Only the written information in the
case is available. Any questions about other elements of the case
will not be answered. Please do not write on the cases and good
luck. (Hand out cases and start recording)

Looking for groupthink

Appendix B

Appendix B
Post-experiment:
(Collect written decisions by the group. Turn off the video
recorders. Administer the post experiment evaluation and collect
it . )

_ Now for the truth. The personality questionnaire was not the
basis _for assigning people to the groups. Though the
guestionnaire is a legitimate instrument in communications study,
its only purpose here was an attempt to create the illusion of
cohesiveness within the group. In other words to allow the group
members to believe in their committee's ability to get along and
make good decisions. The quality of the decisions reached will
determine which video tapes are to be studied further. The
highest quality and lowest quality decisions will be examined to
find out what elements of decision making were used by high
quality groups that were not used by the low quality groups. The
post experiment evaluation will be used to determine how well the
false personality questionnaire created cohesion in the group.
Any questions about any aspect of this experiment can be directed
to your professor or Scott Stone. All information gathered during
this experiment will be held in strict confidentiality. We ask
that you please do not discuss this experiment with anyone
outside of class until Friday since this experiment also involves
other classes. Thank you for your time.

Looking for groupthink

Appendix C

Appendix C
Vandal ism Case

2/1/94
Group 2:
Ms. Lewis and Ms. Belovorak should be written up for
excessive noise.
Ms. Lewis should also receive a one year probation and pay
Ms. Belovorak $170 in damages for the stereo.
- Ms. Belovorak should be moved to an available room while
Ms. Lewis remains in her present room since she was there
first.
Appropriate
Warranted
Reasonable
Fair

1..2..3..★..5..6..7
1..2..3..4.*5..6..7
1..2..3+.4..5..6..7
1..2..3..4*.5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Group 4:
(1) Ms. Lewis needs to pay for the stereo system.
(2) Ms. Lewis will be put on probation and several weeks of
counseling.
(3) The two girls will be separated. They will both be given
the choice of who will stay or leave. In the event that
neither would like to leave, Ms. Lewis will remain in
the room due to seniority. Proper placement will be
given to Ms. Belovorak.
Appropriate
Warranted
Reasonable
Fair

1..2.★3..4..5..6..7
1..2..3..4.*5..6..7
1.*2..3..4..5..6..7
1..2.*3..4..5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

2/2/94
Group 2:
- Ms. Lewis should have mandatory counseling if available.
- Ms. Lewis should give restitution for the stereo by
getting employment through the University.
- Ms. Lewis should no longer be eligible for on carrpus
housing beginning next semester.
- For the Remainder of the semester Ms. Lewis should room
alone.
Appropriate
Warranted
Reasonable
Fair

1..2..4..5..6..7
1..2. .3. .4..*..6..7
1..2..4..5..6..7
1..2..3..4.*5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Group (4: ^ ^13 wm reimburse the $170 plus tax by the end of
the semester.
_
.
_
(2) Ms. Lewis will be on probation for one year.

Looking for groupthink

Appendix C

(3) If Ms. Lewis receives one more write up she is out of
the dorms.
(4) A written apology to Ms. Belovorak.
(5) A letter to Ms. Lewis' parents.
(6) Ms. Lewis has 24 hours to find a new roommate or one
will be appointed.
Appropriate
Warranted
Reasonable
Fair

1..2..3..4.^5..6..7
1..2..3*.4..5 6..7
1..2..3..4*.5 6..7
1..2..3..4.*5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

2/3/94
Group 2:
Ms. Belovorak _ should receive a semester of probation and has
the option of staying in the room or moving to a new one. She
will get a private room.
Ms. Lewis should receive a semester of probation, 30 hours
of community service and the choice to pay for a private room or
to get another roommate.
Appropriate
Warranted
Reasonable
Fair

l..*..3..4..5..6..7
1.+2..3..4 5..6..7
l.*2..3..4 5..6..7
1.*2..3..4..5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Group 4:
Our recommendation is that Ms. Lewis have mandatory
counseling concerning her behavior. We also request that her
parents be notified as well as be present at the first counseling
session. (This may be stemming from some past family problems or
events.)
Also we suggest that she participate in community service
activities such as working with the handicapped or young children
in order to instill the value of patience.
As a final request, Ms. Lewis shall be on career probation
in order that she be reprimanded for the belligerent behavior eg.
arguing, fighting, disrupting others, etc. She shall be expelled
from school permanently.
We also suggest that she pay for the damage to the stereo,
regardless of whether Ms. Belovorak presses charges.
We also suggest that the R.A. be spoken to by her superiors
for her inability to take charge in the past situations involving
the two roommates (their fighting). Also, she should not have
immediately accused Ms. Lewis for having broken the stereo
without concrete evidence. If she already has an explosive
personality this will immediately put her on the offense.
Appropriate
Warranted
Reasonable
Fair

1..2..*..4..5..6..7
1..2..3*.4..5..6..7
1..2..*..4..5..6..7
1..2..★..4..5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Looking for groupthink

Appendix C

Plagiarism Case
2/1/94
Group 1:
_
Our conclusion is based on admitting Mr. Jordan plagiarized
his assigranent, and he is a senior in the journalism department.
We feel that these two factors play a large role in our decision.
We wish we could have viewed the students and William F.
Buckley's paper. Our recommendation is failure of this class,
although plagiarism will not appear on his final transcript.
Appropriate
Warranted
Reasonable
Fair

l..2..3..★..5..6..7
1..2..3..★..5..6..7
1..2..3..4*.5..6..7
1..2..3..4..5*.6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Group 3:
We, the members of the Ethics Committee, find that since Mr.
Jordan did have some knowledge that what he was doing was
plagiarism, we feel that he should receive a zero for this
assignment. Also, his future professors should be made aware. If
this were to occur again, further and more severe action will be
taken against Mr. Jordan.
Appropriate
Warranted
Reasonable
Fair

1..2..3..★..5..6..7
1..2..3.★4..5..6..7
1..2..3.★4..5..6..7
1..2..★..4..5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

2/2/94
Group 1:
James Jordan receives an "F" for the class and must retake
the course with a different professor. Professor Adamson should
be lectured by the department head on how to respect students and
the proper way to handle this type of situation.
Appropriate
Warranted
Reasonable
Fair

1..2..3..4..5..6*.7
1..2..3..4..5..6*.7
1..2..3..4..5..6*.7
1..2..3..4..5..6*.7

Inappropriate
Unwarranted
Unreasonable
Unfair

Looking for groupthink

Appendix C

Group 3:
We decided to give James Jordan a failing grade on the paper
with no make-up. James Jordan should give an oral apology to his
professor and classmates and also a written apology to the
University. The reason for this conclusion is because of his
academic background and he has not had any previous incidents. If
this same event occurs again he will be immediately expelled from
the University.
Appropriate
Warranted
Reasonable
Fair

1..2..3..4..5i.6..7
1..2..3..4..5i.6..7
1..2..3..4.*5..6..7
1..2..3..4.*5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

2/3/94
Group 1:
We felt that James Jordan deserves an "F" for the course. He
has the option to retake the course. If he continues to
plagiarize, he will be expelled from the University. Also, there
will be no trace of this circumstance on his record.
Appropriate
Warranted
Reasonable
Fair

1..2..3..★..5..6..7
1..2..3..★..5..6..7
1..2..3..★..5..6..7
1..2..3..*..5..6..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Group 3:
The course of action to be taken against the student will
result in a failing grade for the paper and lowering of the final
grade by one letter. Remark will be made on transcripts to read
that he was brought up on plagiarism charges.
Appropriate
Warranted
Reasonable
Fair

1..2..3. .4..5..★..7
1..2..3. . 4..5.*6..7
1..2 3..4..5..★..7
1..2 3..4..5..★..7

Inappropriate
Unwarranted
Unreasonable
Unfair

Appendix D

Looking for groupthink

Appendix D
Post Experiment Evaluation
1. How responsive were other group members to differences of
opinions and criticism from you?
Resentful |_1_|_2_|_3_|_4_|_5_|_6_|_7_| Encouraging
2. How attractive, interesting, fun to be with, and enjoyable was
the group?
Not Attractive |-1-|-2-|-3-|-4-|-5-|-6-|-7-| Very Attractive
3. How would you rate the ability and competence of the group?
Highly Competent |-1-|-2-|-3-|-4-|-5-|-6-|-7-| Incompetent
4. How successful do you feel the experimenter was in matching
individuals who worked well together?
Very Successful [-1-|-2-|-3-|-4-|-5-|-6-|-7-| Unsuccessful
5. How willing would you be to participate in future research as
a member of the same group?
Not Interested |-1-|-2-|-3-|-4-|-5-|-6-|-7-

Delighted

Background Information
Name:

__________________

Home Town or City:
Major:____________

Age:____

Sex:
State:

Year in School:

Appendix E

Looking for groupthink

Appendix E

Personality Questionnaire
Date:

Name:

Please circle the number that you feel is closest to your own
personality trait. Answer truthfully because this will be the
basis for assigning groups that work well together.
Reserved

123456789

10 Warm

Concrete Thinking

12345678910 Bright, Abstract

Emotional

123456789

10 Calm

Adaptable

123456789

10 Assertive

Serious

123456789

10 Enthusiastic

Expedient

123456789

10 Conscientious

Shy

123456789

10 Bold

Tough-Minded

12345678

9

10 Tender-Minded

Trusting

12345678

9

10 Suspicious

Practical

12345678

9

10 Imaginative

Forthright

12345678

9

10 Socially Proper

Composed

12345678

9

10 Worrisome

Conservative

12345678

9

10 Progressive

Group-Oriented

12345678

9

10 Self-Sufficient

Spontaneous

123456789

10 Self-Disciplined

Relaxed

123456789

10 Tense