~ - ~ ‘~ "‘"u‘wfl‘m‘

   

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 
     

 

 

 

     

 
       

 
 

 

 

 

 
 
       

 

 

     

     
 

   

     

 

 

 

 

 

 

   

 

 

 
 

 

 

ZQL?L203

«am sure umve names ‘

lillflllllzlll‘lllllillllll llllmum

93 00784 1277

 

LIBRARY
[Michigan State
L University

 

 

This is to certify that the

dissertation entitled

Motivation in Groups: A Control Theory Model

presented by

Anne Marie O'Leary

 

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Management

%

Major professy
Date My

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 ‘

 

PLACE IN RETURN BOX to remove thle checkout from your record.
TO AVOID FINES return on or betore due due.

DATE DUE DATE DUE DATE DUE
”6361261 3‘ [AW 22152003

 

  
 

 

 

 

 

 

 

 

 

 

—-- J—J
“—1.; JET—ii
_____ l _____ l ______ L:
____ ‘

MSU Ie An Afimdtve Action/Equal Opportunity lnetttulon
mum

 

#
f

 

 

 

 

 

 

 

m1

 

 

 

 

MOTIVATION IN GROUPS:
A CONTROL THEORY MODEL

By

Anne Marie O'Leary

A DISSERTATION

submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Management

1990

ABSTRACT
MOTIVATION IN GROUPS: A CONTROL THEORY MODEL
BY
Anne Marie O'Leary

This research developed and tested a model of the psychological
processes involved in the motivation of individuals operating within
groups. In line with control theory, the model suggested that
individual group members compare the group's referent standard to the
group's performance feedback information and a decision is made about
whether a discrepancy exists between the two. Group members were
hypothesized to undergo a social comparison process aimed at determining
the degree of their individual responsibility for the group's
performance outcome. These perceptions were expected to influence the
individual's choice of attributional source (i.e., attributions to self,
group, task difficulty, or luck). The locus of causality of the
attributional source chosen was hypothesized to influence the affective
reactions (satisfaction with self performance, satisfaction with the
group) of the group member, and the stability of the source was expected
to influence the future commitment of the individual to the group's
referent standard. Finally, the commitment of the group member was
exPected to relate to the future performance of the group member.

This model was tested in a sample of nineteen college intramural
basketball teams. One hundred and forty four individuals completed
queationnaires at each of three basketball games. Individual

performance in the games served as the performance measure.

The results provided general support for the model. The individual

team member's perception of his performance level influenced his
perception of responsibility for the group's performance outcome. In
addition, this perception of felt responsibility influenced the
attributional source chosen by the individual. The locus of causality
of the attributional source influenced the individual's satisfaction
with self performance, however there was no support for the hypothesis
that the locus of causality of attributions would predict satisfaction
with the group. There was mixed support for the prediction that the
stability of attributions would influence commitment to the group's
standard on the next performance trial. Although there was no support
for the hypothesis that commitment to the group's standard would be
positively related to individual performance, there was a positive
relationship between individual commitment to the group's standard and

group performance.

Copyright by
ANNE MARIE O'LEARY

1990

ACKNOWLEDGEMENTS

There are a few wonderful people who have greatly influenced the
person that I am, both personally and professionally, and to them I owe
a great debt. I can offer only gratitude in return, along with the
promise that I will try to make them proud of this person they have
shaped.

Most of all I thank my parents, whose opinion of me was always much
higher than my own, for seeing my options as so limitless. I also thank
the rest of my family (Margo, Joseph, Eileen, Mary, Sharon, Joanne,
Colleen, Patrick, Becky, Terry, Bob, Tim, Betsy, Jenny) for being my
most treasured friends; there is no group I would rather belong to.

I thank my teachers at MSU for helping me develop a new view of the
world. Professionally, I owe them all that I am. On a personal level,
however, they have also given to me. I thank Dan Ilgen for being a
truly genuine, gentle person; he is my standard. I thank Georgia Chao
for the role modeling; she has shown me the meaning of poise. I thank
John Hollenbeck for making this "work" so much fun; his enthusiasm is
infectious.

And, of course, to all of my friends in the program I am grateful.
Each of them has expanded my view in some unique way, and that is the
greatest gift they could give.

It's time to move on. I am ready. I am sorry.

Yet all experience is an arch wherethro'
Gleams that untravell'd world whose margin fades

For ever and for ever when I move.
-Tennyson

TABLE OF CONTENTS

List of Tables.

List of Figures .

CHAPTER 1: INTRODUCTION.
Introduction .
Research at the Group Level.
Outline.

CHAPTER 2: APPLICATION OF CONTROL THEORY .
Control Theory .
Application of Control Theory.
Control Theory and Goal Setting Theory .
Control Theory and Performance Feedback.
Control Theory and Individual Motivation .
Control Theory and Affective Responses .
Summary of Control Theory Applications .

Model.

CHAPTER 3: A CONTROL THEORY MODEL OF MOTIVATION IN GROUPS.

Overview .

Group Referent Standard.
Group Performance.

Group Performance Feedback .

The Comparator .

vi

.xiii

15

16

21

27

33

37

38

40

40

42

45

45

46

The Social Comparison Process and Felt Responsibility.

Hypothesis 1 .
Attributions .
Hypothesis 2.
Hypothesis 3 .
Hypothesis 4.
Hypothesis 5.
Self-Serving Bias.
Affective Reactions.
Attributions to External Factors.
Attributions to Self.
Attributions to the Group .
Summary and Hypothesis.
Hypothesis 6 .
Hypothesis 7 .
Commitment to the Group Goal .
Stable Attributions for Successful Performance.
Unstable Attributions for Successful Performance.

Stable Attributions for Unsuccessful Performance.

Unstable Attributions for Unsuccessful Performance.

Hypothesis 8.
Individual Performance .

Hypothesis 9.
Social Loafing and Free-Rider Effects.
Feedback Loops .

Model Assumptions.

vii

48

51

51

58

58

58

59

59

60

63

63

65

66

66

68

70

72

72

74

74

75

75

76

76

79

81

Group Goal Setting Research and the Control Theory Model . . . 82
Empirical Investigation. . . . . . . . . . . . . . . . . . . . 89
CHAPTER 4: METHOD. . . . . . . . . . . . . . . . . . . . . . . . 91
Power Analysis . . . . . . . . . . . . . . . . . . . . . . . . 91
Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Pilot Study. . . . . . . . . . . . . . . . . . . . . . . . . . 93
Measures Used in the Sample of Intramural Basketball Teams . . 94
Information Collected Prior to First Game . . . . . . . . . 94
Information Collected At Each Game. . . . . . . . . . . . . 96
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
CHAPTER 5: RESULTS . . . . . . . . . . . . . . . . . . . . . . . 102
Hypothesis 1 . . . . . . . . . . . . . . . . . . . . . . . . . 102
Hypothesis 2 . . . . . . . . . . . . . . . . . . . . . . . . . 114
Hypothesis 3 . . . . . . . . . . . . . . . . . . . . . . . . . 117
Hypothesis 4 . . . . . . . . . . . . . . . . . . . . . . . . . 119
Hypothesis 5 . . . . . . . . . . . . . . . . . . . . . . . . . 119
Attributions . . . . . . . . . . . . . . . . . . . . . . . . . 119
Hypothesis 6 . . . . . . . . . . . . . . . . . . . . . . . . . 131
Hypothesis 7 . . . . . . . . . . . . . . . . . . . . . . . . . 142
Hypothesis 8 . . . . . . . . . . . . . . . . . . . . . . . . . 148
Hypothesis 9 . . . . . . . . . . . . . . . . . . . . . . . . . 158
Other Hypothesized Relationships . . . . . . . . . . . . . . . 162
Feedback Loop. . . . . . . . . . . . . . . . . . . . . . . . . 165
CHAPTER SIX: DISCUSSION. . . . . . . . . . . . . . . . . . . . . 167

Predicting Felt Responsibility . . . . . . . . . . . . . . . . 167

viii

Attributions at the Group Level. . . . . . . . . . . . . . . . 174

Influence of Felt Responsibility Perceptions on Attributions . 178

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Influence of Attributions on Satisfaction with Self
Performance. . . . . . . . . . . . . . . . . . . . . . . . . . 183
Influence of Attributions on Satisfaction with the Group . . . 187

Influence of Attributions on Commitment to the Group's

Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Influence of Commitment to the Group's Standard on

Performance. . . . . . . . . . . . . . . . . . . . . . . . . . 194

Influence of Other Variables . . . . . . . . . . . . . . . . . 196

Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . 196

Summary of Findings Related to the Model . . . . . . . . . . . 198
APPENDIX A: Questionnaire Used in Pilot Study. . . . . . . . . . 203
APPENDIX B: Questionnaires Completed By Study Participants . . . 211
List of References. . . . . . . . . . . . . . . . . . . . . . . . 235

ix

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

10:

11:

12:

13:

14:

15:

LIST OF TABLES

Timing of Measures.

Internal Consistencies of Measures.

Means and Standard Deviations of Variables in the Model .

Intercorrelations Among Variables in the Model.
Regression Results for Felt Responsibility Perceptions.
Regression Results for Felt Responsibility Perceptions.

Correlations Between Likert Scale Measures of Attribu-
tions and Ipsative Scale Measures of Attributions
for Game #1.

Correlations Between Likert Scale Measures of Attribu-
tions and Ipsative Scale Measures of Attributions
for Game #2.

Correlations Between Felt Responsibility Perceptions
and Self Attributions (Measured on Likert Scale) .

Correlations Between Felt Responsibility Perceptions
and Self Attributions (Measured on Ipsative Scale) .

Correlations Between Felt Responsibility Perceptions
and Team.Attributions (Measured on Likert Scale) .

Correlations Between Felt Responsibility Perceptions
and Team Attributions (Measured on Ipsative Scale) .

Correlations Between Felt Responsibility Perceptions
and Task Difficulty Attributions (Measured on
Likert Scale). . . . . . . . . . . . .

Correlations Between Felt Responsibility Perceptions
and Task Difficulty Attributions (Measured on
Ipsative Scale). . . . . . . . . . . . .

Correlations Between Felt Responsibility Perceptions
and.Luck.Attributions (Measured on Likert Scale) .

95

. 103
. 104
. 105
. 106

. 110

. 112

. 113

. 115

. 116

. 118

. 120

. 121

. 121

. 122

Table
Table
Table
Table
Table
Table

Table

Table

Table
Table

Table

Table

Table

Table

16:

17:

18:

19:

20:

21:

22:

23:

24:

25:

26:

27:

28:

29:

Correlations Between Felt Responsibility Perceptions
and Luck Attributions (Measured on Ipsative Scale) .

Rotated Factor Loadings for Principal Axis Factor
Analysis of Attribution Items Related to Game #1 .

Rotated Factor Loadings for Principal Axis Factor
Analysis of Attribution Items Related to Game #2 .

Correlations Between Factor Scores for Game #1 and
Game #2.

Means of Subject Perceptions of the Locus of Causality
of Attributions (6 Attributional Categories) .

Means of Subject Perceptions of the Locus of Causality
of Attributions (3 Attributional Categories) .

T-tests of the Differences Between Mean Perceptions of
the Locus of Causality of Attributions (6 Attribu-
tional Categories) .

T-tests of the Differences Between Subject Mean Percep-
tions of the Locus of Causality of Attributions
(3 Attributional Categories) .

Means of Subject Perceptions of the Stability of
Attributions . . . . . . .

T-tests of the Differences Between Mean Perceptions of the
Stability of Attributions.

Regression of Satisfaction with Self Performance on Group
Standard Attainment, Locus of Causality of Attributions,
and Their Interaction (With Locus of Causality Measured
Using an.A Priori Framework) . . . . .

Regression of Satisfaction with Self Performance on Group
Standard.Attainment, Locus of Causality of Attributions,
and Their Interaction (With Locus of Causality Measured
Using Russell Scale)

Regression of Satisfaction with Self Performance on Six
Attributional Sources.

Regression of Satisfaction with Self Performance on Group
Standard Attainment, Locus of Causality of Attribution,
and Their Interaction (With Locus of Causality Measured
Using Least Squares Weights) .

xi

. 122

. 123

. 124

. 126

. 128

. 128

. 129

. 129

. 130

. 130

. 133

. 136

. 138

. 139

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

30:

32:

33:

34:

35:

36:

37:

38:

39:

40:

41:

Regression of Satisfaction with the Group on Group Standard
Attainment, Locus of Causality of Attributions, and Their
Interaction (With Locus of Causality Measured Using an A
Priori Framework). . . . . 143

: Regression of Satisfaction with the Group on Group Standard

Attainment, Locus of Causality of Attributions, and Their
Interaction (With Locus of Causality Measured Using

Russell Scale) . 145
Regression of Satisfaction with Group Performance on Six
Attributional Sources. . 147

Regression of Satisfaction with the Group on Group Standard
Attainment, the Locus of Causality of Attributions, and
Their Interaction (Using Least Squares Estimates to
Determine Locus of Causality). . 149

Regression of Commitment to the Group's Goal on the Next
Performance Trial on Group Standard Attainment, the
Stability of Attributions, and Their Interaction (With
Stability of Attributions Measured Using an A Priori
Framework) . . . . . . . . . . . 153

Regression of Commitment to the Group's Goal on the Next
Performance Trial on Group Standard Attainment, the
Stability of Attributions, and Their Interaction (With
Stability of Attributions Measured Using Russell Scale). 155

Regression of Commitment to the Group's Standard on the Next

Performance Trial on Five Attributional Sources. . 159
Regression of Commitment to the Group's Goal on Group

Standard Attainment, the Stability of Attributions, and

Their Interaction (Using Least Squares Estimates of

Stability of Attributions) . . 160
Correlations Between Individual Goal Commitment and

Individual Performance . . 163
Correlations Between Commitment to the Group Standard in

the Next Game and Group Standard Attainment in the

Next Game. . 163
Relationship Among Other Variables in the Model . . 164
Feedback Loop Relationships . . 166

xii

Figure
Figure
Figure
Figure
Figure
Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

10:

11:

12:

13:

14:

15:

16:

LIST OF FIGURES

Miller, Galanter, and Pribram (1960) TOTE Sequence.
Carver and Scheier (1981) Control Theory Model.
Campion and Lord (1982) Control Theory Model.
Klein (1989) Model of Individual Motivation .
Hollenbeck and Brief (1988) Control Theory Model.
Hollenbeck (1989) Model .
Overview of Control Theory Model of Motivation in
Groups . . . . . . . . . . . . . . . .

Model of the Motivation of Individuals Operating in
Groups . . . . . . . . . . . . .

Hypothesized Interaction Effect of Group Standard
Attainment and Self Performance Perceptions on Felt
Responsibility Perceptions .

Locus of Causality of Group Level Attributions.

Stability of Group Level Attributions .

Weiner's Model of the Cognition-Emotion Process .

Hypothesized Interaction Effect of Group Standard
Attainment and Locus of Causality of Attributions
on Satisfaction With Self Performance.

Hypothesized Interaction Effect of Group Standard
Attainment and Locus of Causality of Attributions
on Satisfaction With the Group .

Hypothesized Interaction Effect of Group Standard

Attainment and Stability of Attributions on Commitment
. 73

to the Group's Standard.

Regression Results for Felt Responsibility Perceptions
(Game #1). . .

xiii

. 10

. 14

. 28

. 34

. 36

. 41

. 43

. 52

. 55

. 57

. 61

. 67

. 69

.108

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

17:

18:

19:

20:

21:

22:

23:

24:

25:

26:

27:

28:

29:

Regression

Results

(Game #2).

Regression
Weights

Regression
Weights

Regression
Weights

Regression
Weights

Regression
Weights

Regression
Weights

Regression
Weights

Regression

Game #1)

Regression
in Game

Regression
in Game

Regression
in Game

Revised Model of the Motivation of Individuals Operating

Results
in Game

Results
in Game

Results
in Game

Results
in Game

Results
in Game

Results
in Game

Results
in Game

Results
Results
#1).

Results
#2).

Results
#1).

in Groups.

for Felt Responsibility Perceptions
for Self Satisfaction (A Priori

#1). . . . . . . . . . . . . .

for Self Satisfaction (A Priori

#2). . . . .

for Self Satisfaction (Least Squares
#1). . . . . . .
for Self Satisfaction (Least Squares
#2). . . . . . .
for Group Satisfaction (Russell

#2). . . . .

for Group Satisfaction (Least Squares
#1).

for Group Satisfaction (Least Squares
#2).

for Commitment (A Priori Weights in
for commitment (Russell Scale Weights
for

Commitment (Russell Scale Weights

for Commitment (Least Squares Weights

xiv

.109

.134

.135

.140

.141

.146

.150

.151

.154

.156

.157

.161

.199

CHAPTER ONE: INTRODUCTION

Landy and Becker (1987), in their recent review of research and
theory in the work motivation area, suggest that the search for one
"right theory” of work motivation is misguided. They point out that no
one theory developed to date is able to explain all motivated human
behavior, nor are they optimistic that such a theory will ever be
developed. They suggest, in line with Pinder (1984), that existing
theories of work motivation might be considered "middle range" theories
(i.e., theories that are useful in the prediction and understanding of a
subset of motivated behavior). The authors further point out that
viewing existing theories of work motivation as competitors is an
ineffective strategy for knowledge acquisition. They suggest instead
that there is value in devising frameworks that link these existing
theories, and they refer to such frameworks as "metatheories" (Landy,
1985; Landy & Becker, 1987; Landy & Becker, 1982). One of the central
points of the review was the need for a new synthesis of middle range
theories into metatheoretic frameworks.

The synthesis of existing theories of work motivation into
metatheoretic frameworks is useful for several reasons. First,
metatheories allow researchers to cumulate the findings related to
existing theories. It is possible, therefore, to obtain a clearer
picture of what is known and what is not yet known regarding a specific
area of study than would be possible from examination of the findings
related to only one theory. For example, Campion and Lord (1982)
suggested that one reason for the lack of progress in regard to

understanding the processes underlying goal setting was that researchers

2
concerned with goal setting were rarely concerned with feedback, despite
the well known finding that knowledge of results is a necessary
condition for the use of goal setting techniques. By placing goal
setting theory within a metatheoretic framework (control theory) the
authors were able to point to an area of goal setting theory which has
not been fully explained. Second, the development of a metatheory
allows researchers to discover the strengths and weaknesses of each of
the middle range theories (i.e., it may suggest boundary conditions for
the middle range theories). For example, Landy and Becker (1985)
proposed that existing theories of work motivation are differentially
suited for predicting different dependent variables. They suggested
that expectancy theory is most appropriate for predicting occupational
and/or organizational choice, while equity theory is better suited to
the prediction of affective reactions. Finally, providing a metatheo-
retic framework to existing theories serves the useful function of
pointing to areas which are promising for future research.

It will be argued here that control theory (Wiener, 1948) is a
useful metatheoretic framework for providing a synthesis of existing
theories. At the individual level of analysis, control theory has been
used to explain the relationships among goal setting, feedback, task
performance, and future goal setting (Hollenbeck, 1989). In addition,
control theory has been suggested as an integrative metatheory for the
synthesis of existing theories of work motivation at the individual
level (Klein, 1989). Control theory can cumulate previous research
findings (Klein, 1989), and is useful for pointing out areas requiring

future research (Klein, 1989; Taylor, Fisher, & Ilgen, 1984). In

3
addition control theory is a dynamic theory, an important trait lacking
in many existing theories of work motivation. Finally, control theory
is flexible enough to incorporate constructs from other areas of
psychology and to explain findings related to these constructs.
e e a e G 0 ve

Over fifteen years ago, Steiner suggested that, "the group seems to
have gone out of fashion" (1972, p. vii). He pointed out that
researchers had shown a great deal of interest in groups until the late
1950's, at which time focus was shifted to the individual level of
analysis. After that time, the group as a task-performing system was
largely ignored. Austin and Bobko (1985) similarly suggested that here
has been relatively little research conducted on group processes and
group performance given the fact that the productivity of work groups is
a critical component of the productivity of an organization.

Presently, there is a renewed interest in the study of groups and
group variables in the organizational behavior literature. This renewed
interest is likely due to the recognition that much of the work
accomplished in organizations is accomplished in groups (Austin & Bobko,
1985). Locke, Shaw, Saari, and Latham (1981), in their review of the
goal setting literature, called for more research dealing with goal
setting at the group level of analysis. Ilgen, Shapiro, Sales, and
Weiss (1987) have also discussed the importance of exploring goal
setting within the context of groups. Nadler (1979) reviewed research
related to the effects of feedback on the task behavior of groups and
called for future research in this area, especially research directed at

understanding the process of feedback use in groups. While there is not

4
yet a comprehensive body of knowledge available regarding work
motivation in groups, there is clearly a growing interest in conducting
research on this topic at the group level.

As research on work motivation in groups develops, it is necessary
that some integration of existing findings take place. For example, as
the focus on group goal setting increases, it is important that issues
of group feedback also become salient to researchers because of the
necessity of feedback in the use of goal setting techniques (Locke,
et al., 1981; Erez, 1977). Several researchers have recognized this
increased interest in group level issues, and the need for some
integration of research, and have called for the development of a
comprehensive framework to guide the cultivation of this interest.
Ilgen, et al., (1987) point to the need for an overarching model to
serve as a framework for the examination of goal setting in group
contexts. They suggest that such a comprehensive model would provide a
useful guide for pointing to future research needs. Similarly, Klein
(1989), who provided an integrated control theory model of work
motivation at the individual level, has suggested that control theory
might be adapted to explain work motivation at the group level. In
addition, Steers and Porter (1974), in their review of goal attributes
at the individual level, suggested that it is insufficient to simply
accept that goal setting programs "work"; these authors encouraged
research which explains how and why goal setting has been so successful
in attempts to influence human behavior.

The present paper suggests that control theory serves as a useful

metatheoretic framework for motivational issues at the group level. The

5

usefulness of control theory for explaining motivational phenomena at
the individual level is well documented (Campion & Lord, 1982; Carver &
Scheier, 1982; Carver & Scheier, 1981; Hollenbeck, 1989; Ilgen, et al.,
1987; Klein, 1989; Scheier & Carver, 1982; Taylor, et al., 1984). This
study will develop and test a control theory model of group motivation
and performance. It is argued that the proposed group level control
theory model is useful for: a) cumulating empirical findings of group
level research, b) pointing out the strengths and weaknesses in existing
theories of motivation applied to group settings, and c) providing a
framework for future research on group motivation and performance. The
necessity for such a model has been established by previous researchers
(Ilgen, et al., 1987; Klein, 1989) and appears especially vital given
the renewed interest in groups in both the goal setting and feedback
areas.
M11112

The following chapter will review control theory and its previous
applications to the organizational behavior area. Specifically, the
chapter discusses the use of control theory for explaining goal setting
findings at the individual level (Campion & Lord, 1982; Hollenbeck &
Williams, 1987) and for explaining the use of feedback at the individual
level (Taylor, et al., 1984). In addition, the chapter discusses the
model developed by Klein (1989) that presents an individual level
control theory model of motivation. Finally, the research by Hollenbeck
(1989) and Hollenbeck and Brief (1987), that addresses behavioral and
affective reactions within the control theory perspective, will be

discussed.

6

The third chapter will present and discuss the proposed control
theory model of motivation at the group level, as well as specific
hypotheses related to this model. Chapter Four will outline the method
used to test these hypotheses, and Chapter Five will provide the results
of this test. Finally, Chapter Six will discuss the findings in
relation to the model, and will make suggestions regarding
interpretation and areas requiring future theoretical and empirical

research attention.

CHAPTER TWO: APPLICATIONS OF CONTROL THEORY

Wm

Most theorists trace the origins of control theory to the work of
Wiener (1948). Control theory has since proved useful as a theoretical
framework in a variety of fields of study (e.g., engineering, applied
mathematics, economics, and medicine; Carver & Scheier, 1982). Powers
(1973), Carver and Scheier (1982), and Miller, Galanter, and Pribram
(1960) are usually credited with extending control theory into the realm
of psychology and individual behavior. Since these early works, control
theory has been used as a framework for explaining and predicting a
variety of human behaviors (Campion & Lord, 1982; Carver & Scheier,
1981; Carver & Scheier, 1982; Hollenbeck, 1989; Hollenbeck & Brief,
1988; Hollenbeck & Williams, 1987; Klein, 1989; Taylor, et al., 1984).

The basic unit of organization in control theory is the negative
feedback loop. Miller, et al., (1960) referred to this loop as the TOTE
(test-operate-test-exit) sequence (see Figure 1). Within this
framework, the system engages in self-monitoring, testing input
information about the current state of operation against some desired
state of operation (represented by the "Test" box in Figure 1). When an
incongruity between the current state and the desired state is detected,
the system takes some action to eliminate this incongruity (represented
by the "Operate" box in Figure 1). The system continues to test and
operate until the incongruity is eliminated.

The example typically provided in demonstrating the use of control
theory for mechanical systems is the functioning of a thermostat (i.e.,

a system which senses and monitors environmental temperature).

 

 

 

~ >'
TeSt (Cong-trig;

 

 

 

(Incongru fly)

 

 

;.

Opera te

 

 

 

 

Figure 1. Miller. Galanter, 8 Pribram (1960) TOTE sequence.

9
Thermostats are set at some desired temperature (which represents the
desired state of operation). The thermostat system has a mechanism for
sensing the existing temperature of the environment (current state of
operation), and this temperature is compared to the desired temperature
(the "test" box in Figure 1). If a discrepancy is discovered between
the desired and actual temperatures, the system is activated to reduce
this discrepancy. The system continues to monitor the environment and
take corrective action until the discrepancy reaches some acceptable
level.

Other control theorists have described a cycle similar to that
proposed by Miller, et al., (1960), although they have provided more
detail regarding process. Carver and Scheier (1982, 1981) proposed the
control theory model presented in Figure 2. Again, they contend that
the basic unit of control is the negative feedback loop. They refer to
this as a "negative" loop because its purpose is to negate, or decrease,
the discrepancies between the desired and actual states. Within their
model, the input function involves the sensing of the present condition
of the system. This condition is then compared to some predetermined
reference value via a mechanism called a comparator. If the comparator
senses some discrepancy between the present condition and the reference
value, the system undertakes some action, referred to as an output
function, in order to eliminate or decrease this discrepancy. This
action is hypothesized to impact upon the system's environment rather
than upon the discrepancy directly. Finally, the disturbance component
represented in the model indicates that forces external to the system

may impinge upon the system. These forces may be discrepancy reducing

10

 

[REFERENCE VALUE 1

I .7

 

 

 

 

 

 

 

 

 

 

 

COMPARATOR
INPUT OUTPUT
FUNCTION FUNCTION
(PERCEPTION) _ (BEHAVIOR)
IMPACT ON

ENVIRONMENT

t—Lousrunsmce]

 

 

Figure 2. Carver and Scheier (1981) Control Theory Model

11
or discrepancy creating (Carver & Scheier, 1981; Carver & Scheier,
1982).

Carver and Scheier (1981) suggested that a feedback loop regulates
behavior with respect to some comparison value. They also suggested
that multiple comparison values exist at various levels of abstraction.
These authors (Carver & Scheier, 1982, 1981; Scheier & Carver, 1982) and
others (Powers, 1973) have suggested that these multiple control systems
are connected hierarchically. They hypothesized that subordinate
standards must be attained in order for standards more superordinate in
the hierarchy to be attained. In addition, they suggest that
superordinate standards serve as guides for the setting of subordinate
standards.

Applying control theory to human systems is a complex undertaking.
In human systems, the sensor function is a perception (i.e., it involves
the sensing of existing conditions; Hollenbeck, 1989). Carver and
Scheier (1982) provided the example of an individual driving a car.

When driving a car, the driver maintains his or her course on the road
by maintaining, visually, just the right distance between the hood of
the car and the edge of the road (referent standard). If there is a
curve in the road, the driver will begin to notice that this distance is
changing and is different than the desired distance (i.e., the sensor
information and the referent standard are different). This difference
motivates the driver to reduce the discrepancy, and he or she will then
turn the wheel in order to accomplish this (effector function).

Hollenbeck (1989) suggested that, when dealing with human systems,

there are two general components to control theory. First, there is a

12

cognitive component in which the comparison process takes place.
Second, there is an affective component in which the discrepancy
produces some negative affect (dissatisfaction, frustration, anger,
etc.) which the individual is motivated to reduce. Lord and Hanges
(1987) point out that motivation has been defined as a multidimensional
construct which involves the arousal, direction, and persistence of
behavior (Locke et al., 1981; Mitchell, 1982; Naylor, Pritchard & Ilgen,
1980), and they suggest that these dimensions might be linked with
specific control systems components. Specifically, they suggest that
arousal is related primarily to the comparator component because
discrepancies discovered by the comparator create effective reactions
which motivate behavior. They further suggest that direction depends
primarily upon the decision mechanism which determines the system's
response to a discrepancy (i.e., whether the system will respond by
changing behavior or by altering the referent). In addition, the
authors suggest that persistence is related primarily to the effector
function and whether it is able to create the desired change in the
environment which will eventually lead to changes in the input function.

Carver & Scheier (1981) suggest that, when dealing with human
systems, it is important to recognize the inherent information
processing limitations of people. That is, they propose that an
individual is incapable of engaging in monitoring behavior for all
possible control loops in which he or she is engaged. In addition, they
suggest that there are individual differences in ability to attend to
this self-monitoring cycle. According to the authors, some individuals

tend to focus attention inward upon the self (high "self-focus"), while

13
others generally focus attention outward to the environment (low
"self-focus"). They propose that the negative feedback loop process
described previously operates best when individuals are highly
self-attentive (i.e., they focus their attention inward upon the self).

Campion and Lord (1982) presented a control theory model of the
goal setting process (see Figure 3), which was modeled after Powers
(1973). They suggested that the sensor accepts information from the
environment regarding the actual state of the system. The sensor then
produces a signal that is compared to the referent standard (desired
state) via the comparator mechanism. If an error is detected, the
system is motivated to eliminate this error. In order to eliminate the
error, the system must make a decision between: a) initiating some
action in order to alter the environment via the effector function, or
b) cognitively changing the referent. According to the theory, regard-
less of which decision is made, the system will attempt to align the
existing condition with the desired condition (the referent standard).
In the interest of clarity, the component names used by Campion and Lord
(i.e., referent standard, comparator, sensor, decision mechanism,
effector) will be used throughout the remainder of this paper.

These models (Campion & Lord, 1982; Carver & Scheier, 1982; Miller,
et al., 1960) provide somewhat different versions of the self-
regulation process. The differences between these models have been
discussed elsewhere (Klein, 1989) and are not the central focus of this
paper. Although these models are somewhat different in their depiction
of self-regulation procedures, the basic processes remain the same. All

of the models suggest that the environment is monitored via some sensor

14

   
   
 
 

te erent
(900')

‘F

   
 

(O nitive

rejerent ‘ cage
ugnol
error
7 decision
temporal" mechanism

   

: 9'23; behavior
change
SYSTEM
m I eflector I
A\
output
"'“m‘ ENVIRONMENT (0-0;.
when)

 

 

—- .nyfioflmgn' emu—.4

Figure 3. Campion and Lord (1982) Control Theory Model

15
function, that some comparison between the sensor signal and the
referent standard occurs, and that a discrepancy triggers some
activation on the part of the system in order to reduce this
discrepancy.

An important feature of the control theory perspective is its
dynamic nature. This perspective suggests that a self-attentive system
is continually monitoring the environment and making adjustments in
order to establish or maintain an alignment between existing and desired
conditions. It has been suggested that advances in work motivation
research will occur only if theories become more dynamic in nature
(Kanfer, 1989).

t o eo

Control theory has been applied to several specific issues in the
field of organizational behavior. Campion and Lord (1982) and
Hollenbeck and Williams (1987) have applied control theory to the goal
setting area and have empirically tested the resulting predictions.
Taylor, et al., (1984) have provided a control theory framework to
previous research on performance feedback. Hollenbeck (1989) and
Hollenbeck and Brief (1987) have used control theory to investigate
affective and behavioral reactions to work. Finally, Klein (1989) has
used a control theory framework for the development of a general
motivational model, which integrates previous theoretical work in this
area. Each of these applications of control theory will be discussed in

more detail below.

16
0 Goal Sett n eo

Campion and Lord (1982) suggested that previous research has not
answered the question of why goal setting works, and they pointed to
three primary factors limiting the ability of researchers to expand
their knowledge of the goal setting process. First, goal setting
researchers have not shown enough interest in feedback research, and
have not integrated these two areas. Second, they suggested that most
goal setting research focuses on static models of motivation. Finally,
they suggested that the goals used in goal setting research are
typically static, specific, and isolated. The authors contended that
the integration of control theory and goal setting theory provides a way
of overcoming these limitations.

The authors pointed out that the linkage between goal setting,
feedback, and control systems was first suggested by Powers (1973). As
previously discussed, control theory postulates a process in which a
referent value is compared (via a comparator) to environmental
information, and discrepancies between the two activate the system in
order to diminish the discrepancy. Powers suggested that the referent
value in this process could be thought of as a goal, and the
environmental information could be thought of as feedback information.

Campion and Lord (1982) suggested that control theory can be used
to explain many of the findings central to goal setting theory.
Researchers in the goal setting area have suggested that both goals and
performance information are essential in order for performance to be
affected positively (Erez, 1977; Locke et al., 1981), and Campion and

Lord suggested that this finding is predictable from control theory. In

17
control theory, the comparator must monitor both the referent value
(goal) and sensor information (feedback) in order to activate any self-
correcting responses. Both goals and feedback, therefore, are necessary
components of the process. Control theory also can be used to explain
the goal setting principle which suggests that specific, difficult goals
lead to higher performance. In relation to specificity, ambiguous goals
would not provide clear referent standards. If the referent standard is
unclear, the comparison process central to control theory would be
unable to consistently detect discrepancies between the referent and
sensor information and, therefore, self-correcting responses would not
be consistently activated. It is likely, therefore, that ambiguous
goals would lead to greater variance in performance because of the
increased latitude provided in the comparison process.

Control theory is also useful in explaining previous research
findings which suggest that difficult goals lead to higher performance
than do easy goals. Campion and Lord (1982) suggested that difficult
goals are more likely to trigger a discrepancy between the referent
standard and the sensor signal than are easy goals (i.e., the referent
and the sensor signal are more likely to be discrepant if the referent
standard is high than if it is comparable to average performance
levels). The goal setting literature also suggests that goals must be
accepted in order to be effective. Campion and Lord pointed out that
control theory accounts for this finding by suggesting that goal
acceptance indicates an individual's decision to use a goal as a
referent standard. If no referent standard is established, the goal

will not influence subsequent effort.

18

In their study, Campion and Lord examined students' academic
performance on tests and in the course from a control theory perspec-
tive. One hundred eighty-eight undergraduates enrolled in a psychology
course were asked to set goals at the beginning of the 10-week term and
after each of four exams during the term. Student effort toward the
goal was measured by five factors: a self-report of effort expended on
each test, a self-report of the number of hours spent studying for each
test, a record of the number of hours spent studying each day (collected
via a daily log which students were required to keep), a self-report of
the expected amount of effort exerted for the next exam, and the number
of classes missed during the two weeks prior to each exam. Performance
was measured by the exam and course grades. Past performance was
measured by previous overall grade point average and by grade point
average for the previous term. Scholastic ability was measured by the
American College Test (ACT) scores in English, Mathematics, Social
Science, Natural Science, and the composite ACT score.

The authors found that performance impacted on later goals set by
students. Specifically, initial grade goals were positively and
significantly related to both ability and past performance. In
addition, students showed a positive past performance-future grade goal
discrepancy for all time periods (except for those students who received
an A on the previous exam). That is, students consistently set goals at
a level somewhat higher than that of their past performance. The
authors predicted that small discrepancies (one letter grade or less)
between goals and performance on classroom exams would not significantly

affect future goals, while larger discrepancies would lead to lowered

19
goals. This hypothesis was not supported for exam goals, but
supplementary analyses provided some support for this hypothesis for
course grades. The results also suggested that successful students were
more likely to increase than to decrease subsequent goals, while
unsuccessful students both increased and decreased subsequent test goals
(although they were more likely to decrease subsequent course goals).

Campion and Lord (1982) also examined the impact of past
performance on subsequent effort. Specifically, the results showed that
the magnitude of failure was positively associated with subsequent
increases in effort. The frequency of failure, however, was not clearly
associated with subsequent increases in effort or subsequent lowering of
goals (i.e., only partial support was found for both of these
relationships). Finally, the study found that the proportion of
students who decreased their grade goals increased with consecutive
failures. In sum, the data from the study conducted by Campion and Lord
were supportive of the control theory model.

Hollenbeck and Williams (1987) also addressed the relationship
between goal setting theory and control theory. They focused upon areas
of control theory which were not addressed by Campion and Lord (1982).
Specifically, they examined the role of goal importance and the role of
individual self-focus in a control theory model of the goal setting
process. The authors pointed out that the self-regulation process, as
described by Powers (1973) and Carver and Scheier (1982, 1981), allows
for the existence of multiple goals organized in an hierarchical
priority system. The construct of goal importance, therefore, provides

some insight into this priority system by specifying how closely one

20
goal is being regulated compared to others in the hierarchy. Second,
the authors suggested that self-focus should play an important role in
investigations of control theory. Self-focus is construed as a
relatively stable individual difference construct indicating the extent
to which an individual tends to direct his or her attention inward
toward the self, versus outward toward the environment (Carver &
Scheier, 1981; Fenigstein, Scheier, & Buss, 1975). Previous research
(Carver, 1974, 1975; Scheier & Carver, 1980a, 1980b; Scheier,
Fenigstein, & Buss, 1974) has shown that high self-focus individuals
show stronger goal-performance relationships than do low self-focus
individuals.

In the Hollenbeck and Williams (1987) study, three hypotheses were
tested in a sample of 88 retail salespersons. First, it was
hypothesized that task performance would be a function of goal level,
self-focus, and perceived goal importance (with performance being
highest for those individuals with difficult goals, high self-focus, and
high goal importance perceptions). Second, goal level was expected to
be a function of perceptions of past performance, self-focus, and
perceived goal importance (with goal level being highest for those
individuals with high past performance, high self-focus, and high goal
importance perceptions). Finally, it was hypothesized that perceptions
of past performance would be a function of actual past performance,
self-focus, and goal importance perceptions (such that perceptions of
past performance would be more accurate for individuals with high

self-focus and high goal importance perceptions).

21

The results supported the first two hypotheses. That is, a
significant interaction between goal level, self-focus, and goal
importance was found, such that individuals with difficult goals, high
self-focus, and high perceptions of goal importance tended to be higher
performers. In addition, the results also showed a significant
interaction between past performance, self-focus, and perceived goal
importance, with individuals high in self-focus and goal importance
perceptions showing stronger past performance-goal level relationships.
Hollenbeck and Williams (1987) provided evidence of the importance of
two individual difference variables (goal importance and self-focus) in
control theory investigations.

eo and Performance Feedback

Control theory was used by Taylor, Fisher, and Ilgen (1984) as a
framework for examining individuals' responses to performance feedback
in organizational settings. This conceptual review integrated previous
findings regarding performance feedback into the control theory
framework. Taylor, et a1. focused on three processes inherent in
control theory: 1) the establishment of referent standards, 2) the
comparison of actual behavior to the standard, and 3) responses to
feedback. Each of these areas will be discussed below.

The authors suggested that behavioral standards originate from
three primary sources. First, behavioral standards may develop from the
individual's values or attitudes. For example, an individual may
ascribe to a value such as the Protestant work ethic, and this value
will have implications for the behavioral standards set by the

individual. Secondly, standards may also be obtained from others

22
through observation or via direct communication. For example, a
supervisor may establish a certain performance level which a subordinate
must meet and this level may serve as the individual subordinate's
standard. Finally, standards may be derived from higher-level
standards. As mentioned earlier, previous researchers (Carver &
Scheier, 1982, 1981; Scheier & Carver, 1982) have suggested that
standards are organized into hierarchies, with higher-level standards
having implications for lower-level standards.

The authors also discussed the comparison process in which the
actual behavior is compared to the established standard. They proposed
that two factors are important in this process. The first was the
frequency of this comparison. Based upon previous research, frequency
of comparison was hypothesized to be a function of three factors: 1)
the degree of self-focus of the individual, 2) the individual's degree
of certainty regarding ability to attain the standard, and 3) the
importance of the standard to the individual. The authors cite moderate
support for the suggestion that individuals who are high in self-focus
(i.e., individuals who attend to information originating from and
concerning the self) will make more frequent comparisons (Carver &
Scheier, 1981). A second factor that is hypothesized to affect the
frequency of testing behavior is the individual's degree of certainty
regarding his or her own ability to meet the standard. The authors cite
previous research that provides indirect evidence in support of the
hypothesis that increased uncertainty is associated with increased
testing behavior (Brief, Aldag, & VanSell, 1977; Weiss, Ilgen, &

Sharbaugh, 1982). The final factor that is hypothesized to influence

23
testing frequency is the importance of the standard to the individual,
with more frequent testing being associated with more important
standards. Although at the time of their review there were no direct
empirical tests of this hypothesis, recent research by Hollenbeck and
Williams (1987) suggests goal importance as a critical variable, both in
the establishment of goal level and in individual performance.

A second factor that is important in understanding the process by
which performance feedback is compared to standards is the conscious
versus unconscious nature of the process. Previous research suggests
that individuals, when confronted with certain situations that have
occurred frequently in the past, are able to automatically and
unconsciously process information regarding this situation (Logan, 1979;
Shiffrin & Schneider, 1977). When, however, individuals are confronted
with situations with which they have little experience (i.e., they have
not developed expectations regarding such situations), information is
more likely to be processed consciously and in a systematic, controlled
manner. Taylor, et al. suggested that these previous research findings
relate to the processing of performance feedback information.
Specifically, they hypothesize that feedback consistent with the
expectations of the recipient has a lower probability of being
consciously processed by the individual. They also suggest that
feedback that is processed unconsciously is more likely to reflect the
subjective expectations of the recipient (i.e., their assessment of the
situation based on past experience), rather than an objective
assessment. Based on previous research, they further hypothesize that

conscious processing is more likely to occur when: a) individuals are

24
unfamiliar with a situation (Bargh, 1982); b) the performance feedback
received is inconsistent with the recipient's expectancies (Weiss,
et al., 1982), and c) others cue the individual to consciously search
for feedback (Bargh, 1982). The authors also suggest that conscious
processing leads to more accurate information being obtained because:
a) feedback seeking behavior (Ashford & Cummings, 1982) increases;
b) increased feedback seeking leads to greater attention to the feedback
received; and c) individuals may weight the feedback received according
to its perceived accuracy.

The third general area discussed by Taylor, et al., (1984) relates
to the individual's response to feedback. The authors suggest three
categories of responses: a) cognitive, b) behavioral, and c) affective.

An individual's behavioral response to feedback cannot be predicted
solely on the basis of the objective content of the feedback received
(Ilgen, et al., 1979). Cognitive responses reflect the individual's
perception of the feedback information and are more directly related to
his or her behavioral response to the feedback. Cognitive responses are
influenced by various factors. The individual's perception of the
accuracy of the feedback, his or her perception of the credibility of
the source of feedback, and his or her perception of the fairness of the
feedback system all influence cognitive responses to feedback. In
addition, the individual's beliefs regarding whether the standard is
attainable has an influence on cognitive response. Carver and Scheier
(1982) have suggested that a discrepancy between the referent and the
sensor signal leads an individual to reassess the probability that he or

she will ever be able to reach the standard. These beliefs, termed

25
expectancy beliefs, are likely to be affected by both personal
characteristics (such as self esteem and locus of control) and
situational factors. Situational factors which influence expectancies
include causal attributions for past performance on similar tasks. For
example, if past success is attributed to a stable cause (such as
ability), the individual is likely to expect that he or she will succeed
again in the future. If past success is attributed to an unstable cause
(such as effort or luck), he or she is less likely to believe that
success will occur again in the future (Weiner, Nierenberg, & Goldstein,
1976). In addition to causal attributions, Taylor, et al., (1984)
suggest that two other situational determinants of expectancy beliefs
are persuasion and modeling (Bandura, 1977). They suggest that
persuasion and modeling are likely to be more important to the
development of expectancies when the individual has had no previous
experience with the task at hand.

Taylor, et al., (1984) also suggested that an individual's decision
regarding whether to change the behavioral standards established in the
past is an important determinant of cognitive response to feedback. An
individual can react to a discrepancy between performance feedback and
the established standard by Changing the level or the nature of the
standard in order to bring it more in line with the feedback information
received. Taylor, et al., (1984) suggested that standards may increase,
decrease, stay the same, or change in nature in response to either
positive or negative performance feedback. The valence of the goal to
the individual appears to be of importance here. Campion and Lord

(1982) suggest that commitment is a function of the valence of the goal.

26
An individual who highly values the established standard (goal) would
not be expected to change this goal easily (i.e., he or she is committed
to the standard). In their research, Mento, Cartledge, and Locke (1980)
found that goal valence was a determinant of initial goal acceptance.
Changing behavioral standards is one cognitive response available to
individuals, however Campion and Lord (1982) suggest that this response
is a longer term solution and will occur less readily in the short term
than will the behavioral and affective responses discussed below.

Taylor, et al., (1984) suggest that behavioral responses to
performance feedback are determined by the cognitive changes in
expectancies and behavioral standards discussed above. Specifically,
they suggest four behavioral responses: 1) changing the direction of
behavior, 2) altering effort, 3) changing the degree of persistence, and
4) responding against the feedback system. These behavioral responses
are hypothesized to be influenced by three factors: initial performance
standards, expectancies, and perceptions of reward contingencies. The
authors provide a detailed discussion of how the individual's
perceptions of performance standards, expectancies regarding future
performance, and perceptions of rewards will influence his or her
decision regarding whether to change the direction of behavior, alter
effort, persist, and/or react against the system.

Finally, Taylor, et al., (1984) suggest that individuals may
respond to discrepancies between performance feedback information and
established standards with a change in effect. Factors that impact on
the individual's effective response to feedback include the sign of

feedback (Ilgen & Hamstra, 1972; Locke, 1967; Locke, Cartledge, & Knerr,

27

1970; Morran 8 Stockton, 1980), the consistency between the feedback
information and the individual's expectations (Bernstein 8 Lecomte,
1979; Ilgen 8 Hamstra, 1972), and the amount of feedback received (Kim 8
Schuler, 1979). In addition, several researchers have suggested that
affective responses to feedback are dependent upon causal attributions
(Carver 8 Scheier, 1981; Weiner, 1982). For example, an individual who
attributes success to an internal cause (such as ability or effort) is
likely to feel pride and/or achievement, while an individual who
attributes success to an external cause (task difficulty or luck) would
be more likely to feel gratitude.

o d I d'v dua ot vation

Klein (1989) employed control theory as a framework for the
integration of existing theories of individual work motivation. He
suggested that a control theory model of motivation could incorporate a
variety of existing theories of individual motivation, including goal
setting theory, expectancy theory, attribution theory, and social
learning theory. He proposed that an integration of these theories
based on control theory was useful because of the emphasis which the
control theory framework places on the cognitive processes underlying
motivation, as well as for reasons of parsimony.

Klein's model is presented in Figure 4. The author suggested that
his model incorporates the work of other control theorists such as
Carver, Scheier, Lord, Taylor, and Hollenbeck (Campion 8 Lord, 1982;
Carver, 1979; Carver, Blaney, 8 Scheier, 1979; Carver 8 Scheier, 1981,
1982; Hollenbeck, 1989; Hollenbeck 8 Brief, 1988; Hollenbeck 8 Williams,

1987; Lord 8 Hanges, 1987; Taylor, et al., 1984). In his model, goals

28

 

l3
Behavior

1
\ Change /

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l0- l2 1‘ . 2
individual SubIective Expected Goal Choice
and 51'90150001 UUlth at Cool and Cognitive (5‘6"?) d) Behavior
Characteristics Attainment Change a or (Eliector)
- _ 7 l ,
9' ’ 8 3‘ i
. . Unconscious
Au bub t .
gem?“ RSC"P‘ed Performance
esponse
I
m l
6 ‘ ’ 5‘ ‘ _
Error? vigompom‘o, l ' I figbsgsk
NO ' I ' ’

 

 

 

7
Continuotiol
ol Previous
\ Behavior j

Figure 4. Klein (1989) Model of Individual Motivation

29
are equated with the referent component, as defined by Campion and Lord
(1982), behavior with the effector component, feedback with the sensor
component, and the comparator component remains the same as described by
Campion and Lord (1982). As in previous models, the central focus in
Klein's model is on the negative feedback loop. Performance feedback
(Box 4) is compared with the goal (Box 1) and the resulting equality or
discrepancy is expected to have implications for the cognitive and
behavioral responses of the individual. This model, however, extends
previous control theory models by focusing on the cognitive processes
occurring between the comparator and behavior (i.e., the effector).

Inspection of the model shows that it incorporates the goal setting
proposition which suggests that goals are the immediate determinants of
behavior (Locke, et al., 1981; Mento, Steel, 8 Karren, 1987).
Specifically, the goal (Box 1) is hypothesized to influence behavior
(Box 2), which influences performance (Box 3), and performance results
in performance feedback (Box 4). The model also recognizes the control
theory proposition which suggests that behavior is activated by a
perceived discrepancy between the actual and the desired state of the
system. That is, feedback information is compared to the goal in Box 5,
the comparator.

In the model, there are three possible outcomes of the comparator
mechanism (Box 6). The individual may perceive that he or she is on
target, behind schedule, or ahead of schedule. If the individual is on
target, no discrepancy exists and he or she will most likely continue
previous behavior (Box 7). If, on the other hand, the individual is

behind schedule or ahead of schedule, a discrepancy does exist. This

30
model suggests two alternative paths following the perception of a
discrepancy. The choice between these paths depends upon whether the
discrepancy is processed consciously or unconsciously, and whether the
individual's response is conscious or unconscious.

Research in the area of cognitive psychology suggests that
information processing can occur in an automatic or a controlled manner
(Shiffrin 8 Schneider, 1977). Klein's model reflects these two
alternatives. He suggests that discrepancies between the goal and
performance feedback may be construed consciously or unconsciously,
depending upon the salience of the error (Taylor 8 Fiske, 1978) and
whether or not the feedback itself was consciously processed (Taylor,
et al., 1984). Similarly, responses to discrepancy can occur in a
conscious or unconscious manner, depending upon whether the discrepancy
was consciously perceived (Taylor, et al., 1984) and upon the
individual's familiarity with the situation (Lord 8 Hanges, 1987).

Responses of which the individual is relatively unaware are
expected to be driven by cognitive scripts, which are overlearned
performance programs that drive an individual's actions in familiar
situations (Lord 8 Kernan, 1987; Schank 8 Abelson, 1977). Behavior,
therefore, may be activated automatically and unconsciously by cognitive
scripts. It has been suggested that scripts may be tied to goals
(Lichtenstein 8 Brewer, 1980; Lord 8 Kernan, 1987). One possible path
following a discrepancy between the individual's goal and performance
feedback, therefore, is an unconscious, automatic, scripted behavioral

response (Box 8).

31

The second path stemming from a perceived discrepancy (Box 9)
suggests a conscious awareness on the part of the individual. If no
behavioral script exists for a particular situation, the individual must
engage in conscious decision making (Carver 8 Scheier, 1981). The
individual must assess the probability that he or she will attain the
established goal. In order to do this, the individual is likely to
search for causal explanations for the discrepancy (i.e., engage in an
attributional search). Weiner (1985) suggested that individuals engage
in causal searches in order to better understand and manage their
environments and proposed that causal ascriptions are central to
individual motivation. Previous research suggests that individuals do
undertake attributional searches and that these searches are most likely
to occur when unexpected situations are confronted (Pyszczynski 8
Greenberg, 1981; Wong 8 Weiner, 1981).

The causal attributions made by an individual influence his or her
expectations regarding goal attainment (Carver 8 Scheier, 1981). The
dimension of central interest in the determination Of outcome
expectancies is stability (McMahan, 1973). If a perceived negative
discrepancy is attributed to a stable cause, such as lack of ability,
the expectation that the goal will be reached in the future is likely to
be very low. On the other hand, if the individual attributes the
negative discrepancy to an unstable cause, such as bad luck, his or her
expectancy regarding goal attainment is likely to remain high. Other
factors, in addition to attributions, which are likely to influence
outcome expectancies are past performance, locus of control, and social

influence (Carver 8 Scheier, 1981).

32

Klein (1989) suggested that the subjective expected utility of goal
attainment (Box 10) is a multiplicative function of the expectancy that
the goal may be attained and the attractiveness of goal attainment. The
subjective expected utility, in turn, is partially determined by
individual characteristics (such as ability and past experience) and
situational characteristics (such as social comparisons and performance
constraints), as presented in Box 11.

The individual's perception of the subjective expected utility of
goal attainment determines the individual's choice of response. If the
subjective expected utility is high (i.e., if he or she expects that the
goal is attainable), the individual will continue in his or her efforts
toward the goal. If the subjective expected utility is low, however,
the individual is expected to react to this low expectancy, either
through a behavioral response or a cognitive response.

The individual may cognitively respond (Box 12) by altering either
his or her goal commitment, the level of the goal, or the goal itself
(Campion 8 Lord, 1982; Taylor, et al., 1984). For example, an
individual who expects that goal attainment is impossible will likely
feel decreased commitment to the goal, will decrease the level of goal
he or she is striving toward, and/or will decide that the goal is not
appropriate and establish a new goal.

Alternatively, the individual may respond behaviorally. This
behavioral response may or may not be a result of some earlier cognitive
change. Several researchers (Carver 8 Scheier, 1981; Lord 8 Hanges,
1987; Taylor, et al., 1984) have suggested that there are two primary

ways in which behavior may be changed: the intensity of effort may

33
change, or the direction of behavior may change. Klein (1989) suggested
that the choice between these two is likely to be dependent upon causal
ascriptions for previous performance.

Klein's (1989) control theory model of work motivation incorporates
existing theories of motivation in order to explain previous empirical
findings in this area. The model parallels previous control theory
models, but provides an explanation of the cognitive processes which may
occur between the comparator mechanism and the response mechanism.

C t eor a d Affective es onses

Hollenbeck and Brief (1987) suggested that the negative feedback
loop described in control theory applies only to situations in which the
individual is able to exert self-corrective action when a discrepancy is
detected. However, they pointed out that there are times when the
individual may be unable to initiate these self-corrective actions, or
when the actions are ineffective in reducing the discrepancy. In such
circumstances, an adaptive individual must recognize that continued
efforts toward discrepancy reduction are not beneficial. The authors
presented an extended model of the basic negative feedback loop in order
to explain such situations. This model is presented in Figure 5. The
model is hypothesized to apply to individuals who are high in
self-focus.

The dotted line in this figure represents the situation in which
the individual is unable to reduce the discrepancy between his or her
standard and the feedback received (a situation referred to as an
”interruption”). It is hypothesized that such an interruption triggers

a cognitive process aimed at discovering the reason for the individual's

34

Standards tor
controlled quantities

 

r > Comparator

 

V
Feedback Behavior

 

 

 

Environment <

I l

 

 

 

 

 

L___.__.._.__..._._...____.._.___.__

 

 

 

 

 

 

l

' Unstable Stable

: attributions attributions

| . 7 \

| I \

, I x

I Positive outcome Negative Outcome

'— - expectations expectations - - -

(Hope) (Dissatrstaction)
i
Withdrawal

motivation and
organizational
commitment

l

t

Voluntary
turnover and
absenteeism

 

 

 

 

 

 

 

Figure 5. Hollenbeck and Brief (1988) Control Theory Model

35
inability to reduce the discrepancy (an attributional search). If the
individual attributes the failure to unstable causes which could be
expected to change in the future (for example, lack of effort or bad
luck), he or she is expected to generate positive outcome expectancies,
and to persist at discrepancy reduction efforts. The model also
predicts that the individual will feel increased arousal, which will be
experienced as hope or anticipation.

The second possibility is for the individual to attribute the
discrepancy-reduction failure to stable causes (for example, lack of
ability or task difficulty), in which case the individual is expected to
experience negative outcome expectancies. The authors suggest that one
possible outcome of the negative expectancy is behavioral and/or mental
withdrawal (e.g., decreased commitment). A second possibility is for
the individual to change the existing standard. The effective reaction
hypothesized to occur in this situation involves increased arousal,
experienced by the individual as dissatisfaction.

Hollenbeck (1989) suggested that there has been little attention
directed toward examining either affective work reactions or behavioral
work withdrawal from a control theory perspective. He presented a model
of affective and behavioral reactions, which appears in Figure 6.
Essentially, the model suggests that the relationships between perceived
discrepancies and affective reactions, and between outcome expectations
and affective reactions (job satisfaction and commitment in this study)
are stronger for high self-focus individuals. In addition, the model
proposed similar relationships for behavioral withdrawal responses

(organizational and job turnover, in this study).

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sell-feces
ltlectlve leactlen:
Percelvrd
Jet Sellstectlee
Discrepancies ornammul Commitment
lehnterel Immune!

Outcome .—. Oreeelzweeel tereever

hl'fl'm" Jel Iereever

 

 

Figure 6. Hollenbeck (1989) Model

37

The model was tested using 140 retail salespeople. The results
provided support for the part of the model dealing with outcome
expectations. That is, the model suggested that individuals who are
high in self-focus show a stronger relationship between negative
expectations and negative attitudes, and between negative expectations
and withdrawal behaviors than do individuals who are low in self-focus.
Summary of Control Theory Applications

The major applications of control theory to the organizational
behavior literature were discussed in the previous sections. This
discussion focused on control theory applications to four general areas:
goal setting theory, feedback systems, general individual motivation,
and individual affect. While these models differ somewhat in emphasis,
they share a focus on several key components of control theory. First,
all of the models described in the previous sections (Campion 8 Lord,
1982; Carver 8 Scheier, 1982; Carver 8 Scheier, 1981; Hollenbeck, 1989;
Hollenbeck 8 Brief, 1988; Hollenbeck 8 Williams, 1987; Klein, 1989;
Taylor, et al., 1984) incorporated a comparison process in which a
standard is compared to input information. Second, all models suggested
that the existence of a discrepancy leads to a self-correcting
motivation, which results in the system taking some action to eliminate
the discrepancy. Several authors (Campion 8 Lord, 1982; Hollenbeck 8
Brief, 1988; Taylor, et al., 1984) pointed out that this action may take
the form of behavior, or may involve a change in the referent standard
itself. Hollenbeck and Brief (1988) and Hollenbeck (1989) expanded the
discussion of reactions to discrepancy by focusing upon affective

reactions to discrepancies and discrepancy-reducing behaviors. Third,

38

all models incorporated an input function, although this component was
given different labels by different authors ("input function" by Carver
and Scheier, 1981, and by Hollenbeck, 1989; "sensor signal" by Campion
and Lord, 1982; "feedback" by Taylor, et al., 1984, by Klein, 1989, and
by Hollenbeck 8 Brief, 1988). Taylor, et al., (1984) placed the most
emphasis on this component in their presentation of feedback issues from
a control theory perspective. A fourth key component of control theory
is the referent standard. Again, all authors incorporated some form of
this component. Taylor, et al., (1984) provided a detailed discussion
of the origins of standards, as did Carver and Scheier (1981). Campion
and Lord (1982) equated referent standards with goals in their empirical
investigation, and Hollenbeck and Williams (1987) placed special
emphasis on the referent standard by investigating the effect of goal
importance on performance and goal level. A final central component of
control theory has received relatively less research attention. Self-
focus, an individual difference variable discussed by Carver and Scheier
(1981), has been investigated primarily by Hollenbeck (Hollenbeck, 1989;
Hollenbeck 8 Williams, 1987).
I The components of control theory reviewed above are central to the
development of a control theory model of motivation. These components
will be discussed in relation to the proposed model of motivation in
groups which will be presented in the next chapter.
Model

The present chapter has presented and discussed various individual
level control theory models and applications. The next chapter will

describe a model of group motivation, using control theory as a

39
framework. This model will incorporate existing knowledge regarding
control theory, such as that described in this chapter. The model will
also incorporate research findings in the social psychological
literature related to group functioning. As the model is described, the

empirical hypotheses resulting from the model will be presented.

CHAPTER THREE: A CONTROL THEORY MODEL OF MOTIVATION IN GROUPS
229nm

Previous discussions of control theory in the organizational
behavior area have examined its Operation at the individual level. This
chapter outlines a control theory model of group motivation, building
upon the work of previous control theorists who have focused exclusively
upon individuals (Campion 8 Lord, 1982; Carver 8 Scheier, 1982; Carver 8
Scheier, 1981; Hollenbeck, 1989; Hollenbeck 8 Brief, 1987; Hollenbeck 8
Williams, 1987; Klein, 1989; Taylor, et al., 1984). The model maintains
the basic components of control theory that were discussed previously
(i.e., the negative feedback loop, the referent standard, the sensor
information, the comparator, the effector). In addition, the model
incorporates cognitive variables and variables unique to group
situations which expand and modify the operation of the basic negative
feedback cycle. Each of the elements of the model will be described in
detail. In addition, the hypotheses resulting from the various
components of the model will be presented. First, however, a general
overview of the model is provided in Figure 7.

In the model it is suggested that group standards impact on group
performance level and that group performance is evaluated via group
performance feedback. After group performance is evaluated and reported
to the group by means of group performance feedback, a comparison
process between the group's standard and the group feedback is expected
to occur within the perceptions of each individual group member. Based
on this comparison, the individual group member decides whether there is

a discrepancy between the group standard and group performance

40

41

03.032". .Eo_>ucom

3:032. 3:022

.2596 5 5.32.05 .o
.32: >32: .2250 .o 32220 K 2:2“.

canton—Eco
ace—23.32 ill 3.00m [agave—goo

.25.. .3222: .26... 9.0.5 II.

xoeaooom
355.23..
830.0

0235.2an
aaoco

_

Eco—.3»
cacao

42

(i.e., did the group performance level approximate its established
standard?). Once this decision is made, the individual is expected to
be concerned with the question Of "why" the group was successful or
unsuccessful. A social comparison process is expected to occur, in
which the group member assesses his or her contributions in relation to
those of other group members. This assessment process is expected to
have implications for the degree of responsibility the individual feels
for the group's success or failure. The responsibility perceptions of
the individual are expected to influence the individual group member's
causal attributions regarding the group's performance, and these
attributions are expected to influence the individual's affective and
behavioral responses. Figure 7 outlines the central processes of the
model that are presented in more detail in Figure 8. Figure 8 will be
discussed in detail in the remainder of this chapter.
G e ndar

At the individual level, the basic principles of control theory
suggest that when individuals begin a behavioral cycle they hold
predetermined referent standards for behavior that are useful in
self-regulatory activities. Similarly, referent standards could be said
to be relevant at the group level. Zander (1971) has pointed to the
necessity of group goals for the effective functioning and, ultimately,
survival of work groups. Essentially, what Zander has emphasized is the
importance and existence of referent standards in groups. Referent
standards at the group level are inherently more complex than those at
the individual level. Zander (1980) has suggested that, when dealing at

the group level, there are at least four different types of goals

43

maoouo CH wcaumuwao mesvH>HvsH we coaum>fiuoz sou me Home: .w ouowam

 

«3 8 :o 3.2....
.\_ : /

02:252an

Emoc Leora—om
coco—coaxm
>::o<

 

3:222: i

a

 

300 5.0.0 3
“cos“.EEoo

030.0 USS

20w 2:3
cozouEZum

2.033232] >:__o_ocoanom
:0“.

coo—EeeEoo .200”

 

 

xounoooa
oocuseotoa
aaoco
.fl
2:25:32
Econ—3m 9.0.2
EBEeQEoo eoceeaotoa
9.95
2.55.5

 

8:20

 

44
operating. First, the group has a goal for itself. Second, each
individual group member may have a goal for the group. Third, a goal
may be set for each individual group member by the group and, finally,
each individual group member may have a goal for him- or herself. The
interplay between these different goal types and goal levels has been
discussed elsewhere (Ilgen, et al., 1987; Nadler, 1979; Zander, 1971;
Zander, 1980), and is not a primary focus of this paper. This paper
will focus upon referent standards for the group (i.e., the first type
of goal discussed above).

The fact that the model presented here addresses goals at the group
level raises an important question. If, as Zander has suggested, a
group holds a goal for itself the question arises as to where the goal
exists. It would appear that the goal can exist only in the perceptions
of individual group members. A group has no common information
processing component which can hold perceptions independently of
individual group members. It is likely, however, that the perceptions
of individual group members overlap in regard to the group goal. For
example, members of a sales team who are committed to achieving a group
sales volume goal are likely to have common perceptions of the goal of
the group. It is probable that these perceptions overlap a great deal,
assuming that the goal has been well communicated and that individual
group members perceive the goal relatively accurately. For the purposes
of this model, therefore, it is assumed that a group goal exists because
individual group members perceive the goal and share similar perceptions
of the group's goal. In other words, group goals will be assumed to

exist when a high degree of consensus exists among the group members as

45

to the level and nature of the group's goal.
W

The model presented in Figure 8 suggests that the group referent
standard influences group performance. Support for this proposition
stems from the goal setting literature. At the individual level, goal
setting proponents have shown that goals (referent standards) direct
attention, mobilize effort, and sustain the persistence of effort
(Locke, et al., 1981). It has also been suggested that goals are the
direct precursors of behavior (Locke, et al., 1981; Mento, Steel, 8
Karren, 1987). Previous research on goal setting at the group level
(Buller 8 Bell, 1986; Ivancevich, 1974; Latham 8 Locke, 1975; Lawrence 8
Smith, 1955; Pearson, 1987; Smith, Locke, 8 Barry, 1989; Weingart 8
Weldon, 1988) has shown that the existence of a difficult group goal
does result in higher levels of group performance.
Q1932 Performance Feedback

Because individuals frequently work in groups, evaluative feedback
on group performance is often provided by organizations. Austin and
Bobko (1985) suggested that most work in organizations is performed in
groups, or at least in group settings. In addition, they pointed out
that often only aggregate, group performance information is available
for use as performance feedback. Therefore, much of the feedback
information received by individuals is received at the group level
(i.e., as group performance feedback). In the model in Figure 8, group
level performance feedback is presumed to result primarily from group
level performance (although it is recognized that situational variables

may moderate the relationship between group performance and group

46
feedback).
W

At the individual level, the negative feedback lOOp outlined by
control theory involves the comparison of the referent standard and
performance feedback (sensor information). This process is represented
by the comparator component. Control theory principles, when extended
to the group level, would suggest a similar process. Specifically, this
model proposes that individuals compare the group level referent
standard to the group level performance feedback received. Zander
(1971) made a similar prediction, although he did not specifically apply
control theory. He suggested that group members use the group's level
of aspiration (i.e., their referent standard) as their criterion for
assessing the performance of the group.

It is important to note that the model outlined in Figure 8
transfers from a group level of analysis to an individual level of
analysis at the point of this comparison between the group referent
standard and the group performance feedback. This is a necessary
transition. The referent standard and sensor information of the group
necessarily exist at the group level. The comparison between the
referent standard and sensor information, however, cannot be made by the
group. This comparison is a cognitive function which occurs within the
perceptions of individual group members and drives their behavioral and
affective reactions. It might be argued that the comparison process
could occur at the group level via some group discussion (e.g., group
members meet and, as a group, compare the group referent standard to the

group performance information). The result of such a discussion would

47
be a group consensus that reflects the comparison process. The group
consensus, however, would not necessarily have direct implications for
the behavioral and affective reactions of the individual group members.
For example, any individual group member might disagree with the group
consensus regarding the comparison and, therefore, his or her behavior
may not be predictable from the group consensus. It is suggested here
that it is only through examination of the comparison processes of
individuals that understanding and prediction of group member behavior
and affect can occur. This model, therefore, is concerned with group
level phenomena (group referent standards, group performance feedback)
but operates primarily at the individual level. In addition, it is
presumed that the transition from a group level of analysis to an
individual level of analysis occurs at the point of comparison between
the predetermined referent standard of the group and the group
performance feedback which has been received. This comparison process
is comparable to the process which occurs within the comparator compo-
nent described in control theory models at the individual level.

The outcome of the comparison process between the group referent
standard and group sensor information suggested by this model is a
decision (made by each individual group member) regarding the degree of
discrepancy which exists between the two. In other words, the
individual group member assesses whether the group has succeeded or
failed in its attempts to attain the group's referent standard. The
individual group member who decides that there is no discrepancy between
the group referent standard and the actual performance of the group is

likely to perceive that the group has been successful. In other words,

48
the referent standard represents the performance level toward which the
group is striving, and attainment of that performance level (i.e., group
performance at a level comparable to the desired level) is perceived as
a group success. 0n the other hand, the individual group member who
decides that there is a discrepancy between the group's referent
standard and the actual performance of the group is likely to perceive
that the group has been unsuccessful.
o o d s s t

Nadler (1979) suggested that the feedback information received by
groups is different than that received by individuals for two primary
reasons. First, individual group members are constrained in their
ability to act on the group feedback information received (Steiner,
1972). Second, any individual group member will have difficulty
determining the extent to which the feedback information reflects his or
her own performance, because the feedback information received by the
group is a function of the actions of other group members. In addition,
previous research that examines group motivational effects, such as
social loafing and free-riding (Albanese 8 VanFleet, 1985; Harkins 8
Petty, 1982; Kerr, 1983; Kerr 8 Bruun, 1983; Kerr 8 MacCoun, 1985;
Latane, Williams, 8 Harkins, 1979; Price, 1987; Williams, Harkins, 8
Latane, 1981), implies that individual group members compare their
efforts and performance to that of other group members in order to
determine their degree of responsibility for the successful or failed
outcome. Underlying all of this research is the presumption that
individuals operating within groups engage in a social comparison

process when assessing performance and interpreting group performance

49
feedback.

The model of group motivation presented here suggests that, after
determining whether a discrepancy between the referent and the sensor
information exists, the individual group member engages in a social
comparison process. This process establishes, within the individual's
own perceptions, the manner in which responsibility for the outcome is
distributed among group members. Nadler (1979) suggested that group
feedback has motivational effects that influence the effort expended by
individual group members. It is assumed here that before individual
group members decide whether their effort should increase or decrease,
they must assess their past effort and performance in relation to that
of other group members (who were also partially responsible for the
group performance feedback received). Equity theory (Adams, 1965)
suggests that individuals engage in social comparisons between their own
and a referent other's inputs (e.g., effort) and outcomes (e.g., perfor-
mance feedback), and that the result of this comparison determines the
individual's motivation. In the group, it appears likely that
individual group members will engage in social comparison processes
because of the importance of knowledge of others' performance to the
evaluation of their own performance. In individual level control theory
models of motivation there is no equivalent to the social comparison
process that takes place in groups. As Nadler (1979) pointed out,
individuals working alone are not presented with the dilemma which is
faced by individuals working within groups; that is, performance
feedback received is not confounded by the performance of other

individuals and it is more clear to the individual that the performance

50
information received is relevant to his or her own performance.

The outcome of the social comparison process described above, in
conjunction with group standard attainment, will determine the
individual's perception of the source of responsibility for the group
performance feedback. For example, the individual who is a member of a
successful group (i.e., high group standard attainment) and who
perceives that his or her individual performance level was high is
likely to feel more responsible for the group's success than is the
individual whose performance level was low. On the other hand, the
individual who is a member of an unsuccessful group (i.e., lbw group
standard attainment) and who perceives that his or her individual
performance level was low is likely to feel more responsible for the
group's failure than is the individual whose performance level was high.

The assignment of responsibility is also likely to be influenced by
the type of task in which the group is engaged. For example, if the
task is such that all group members are expected to contribute equally
(additive task; Steiner, 1972), the individual's assignment of
responsibility for success will be different than if the task requires
one individual to contribute more than others (conjunctive or
disjunctive tasks; Steiner, 1972). The model described in Figure 8,
however, is applicable to all types of tasks. Because the test of this
model described in the next chapter involves a task wherein all members
of the group have some responsibility for group performance, the
individual's perception of felt responsibility for group performance is

expected to be related to his or her previous individual performance.

51

Based upon the above discussion, the following hypothesis is
presented. The proposed interaction is shown in Figure 9.

HYPOTHESIS 1: Self performance perceptions and group standard

attainment will interact to affect felt responsibility perceptions

such that the relationship between self performance perceptions and

felt responsibility will be positive for high group standard

attainment individuals and negative for low group standard

attainment individuals.
Attribgtions

Attribution theory, which is rooted in the work of Heider (1958),
suggests that individuals act as "naive psychologists" who are motivated
to determine the causes of specific events. This motivation is based on
a need to develop a cognitive understanding of the environment.
Attribution theory suggests that individuals develop and maintain sets
of beliefs (schemes) regarding how causes and effects are related
(Kelley 8 Micela, 1980). Kelley (1967) suggested that people use three
types of information in their causal attributions: consistency,
consensus, and distinctiveness. Consistency information relates to the
stability of performance over time, while consensus information involves
comparing the performance of several individuals, and distinctiveness
information represents the degree to which present performance is
different from performance on other tasks.

Weiner (1972) initially suggested that individuals in achievement
situations assign causality primarily to four sources (effort, ability,
task difficulty, and luck). In addition, he suggested that these

sources vary along two primary dimensions: internal versus external,

52

High ngh GSA

Felt Respon-
sibility
Perceptions

LOW Low GSA

Low High
Self Performance Perceptions

Figure 9. Hypothesized interaction effect of
group standard attainment and
self performance perceptions on
felt responsibility perceptions.

53
and stable versus unstable. Attributions to effort and ability are
presumed to be internal (i.e., the causes of behavior are within the
person), while attributions to task difficulty and luck are external
(i.e., the causes of behavior are centered in factors in the environ-
ment). Attributions to ability and task difficulty are presumed to be
stable (i.e., relatively unchanging over time), while attributions to
effort and luck are expected to be unstable (i.e., relatively variable
over time). Although additional dimensions have been suggested since
Weiner's 1972 article (Weiner, 1979; Weiner, 1985), the two dimensions
described here (internal/external, stable/unstable) have received the
most research attention and are the most relevant to the current discus-
sion.

At the group level, causal attributions are more complex than at
the individual level. As discussed above, attributions can be made to
four sources at the individual level: ability, effort, task difficulty,
luck. At the group level, however, additional sources are applicable.
It seems reasonable that an individual who is attempting to discern the
cause of the group's performance might attribute performance to his or
her own ability or effort, but it is also possible that he or she will
attribute performance to the overall ability or effort of the group as a
whole. For example, an individual who is a member of a successful group
may attribute the success primarily to the overall ability of the other
group members, rather than to his or her own ability. As at the
individual level, task difficulty and luck are viable attributional
options for individuals searching for the cause of group performance.

At the group level, therefore, this model suggests that there are six

54
important attributional sources: self-ability, self-effort, group-
ability, group-effort, task difficulty, and luck.

As discussed above, at the individual level a group member may
attribute success to internal factors (effort, ability) or to external
factors (task difficulty, luck; Weiner, 1972). However, at the group
level the internal/external distinction is less clear. An individual
who attributes success to his or her group (i.e., to all group members)
is not ascribing causality to factors internal to him- or herself. On
the other hand, this individual is also not attributing causality to a
purely external cause, because the individual values the group's goal
and may be able to exert some control over other group members regarding
the attainment of this goal. Although previous research has not
addressed this issue, at the group level it seems reasonable to view
internal and external attributions as two poles on a continuum. Within
this framework, attributions to self (effort, ability) might be placed
at the extreme ”internal" pole, attributions to factors outside the self
and group (task difficulty, luck) might be placed at the extreme ”exter-
nal" pole, and attributions to the set of individuals who make up the
group might be expected to fall between the extremes. It is suggested
here that the internality continuum may be represented as depicted in
Figure 10. This distinction is in accord with the discussion by
Schlenker and Miller (1977) who suggested that individuals, when making
attributions for group performance, may view other group members as
either environmental variables, or as extensions of the self.

It is suggested here that the stability dimension could also be

represented as a continuum which encompasses the six group level

55

.wcozzntzm
_o>m_ 965 .6 3:338 Lo £60.. .9 059".

 

 

>:=omr:om 3:398:95 v.21.
tote-zow tote-anew 3.22:5 gem...
wZO_.__.Dm_m.—:_.< wZO_._.Dm_m—._L.<

._<zmm._.z_ I_<Zm_m._.xm

56

attributions. Weiner (1972) suggested that attributions to ability and
task difficulty are relatively stable, while attributions to effort and
luck are relatively unstable. Both group-ability and self-ability
sources, therefore, are likely to represent stable attributional
sources. It could be argued, however, that the group-ability source is
more stable than the self-ability source because it represents multiple
stable factors, rather than a single stable factor (i.e., the self-
ability source). The most stable end of the continuum is likely,
therefore, to be anchored with group-ability attributions, followed by
self-ability attributions. The next most stable attributional source
would appear to be self-effort. While both effort sources (self-effort
and group-effort) are considered less stable than ability sources, self-
effort is less likely to change in the future than group-effort because
group-effort will change if any one of a number of individuals shows a
change in effort level, while self-effort will change only if the focal
individual shows a change in effort level. In agreement with Weiner
(1972), luck represents a highly unstable attributional source and
therefore is assigned to the unstable end of the continuum. It is
expected here that the stability of task difficulty attributions depends
upon the type of task faced. In some cases, the task could be expected
to remain relatively stable, while in others the task is highly change-
able. For this reason, task difficulty attributions are not included in
the stability continuum outlined in Figure 11.

In summary, the model presented here suggests that individuals use
performance information to make causal attributions to six sources:

self-ability, self-effort, group-ability, group-effort, task difficulty,

57

.wcogottm .05. 8:05 “—0 3:5me .3 08:9“.

3:32:55 3:33.?” :23le 3:08:20 v.2...

wZO_._.Dm_m._:_.< w20_._.Dm_m._:_.<
m._m<._..w w..m<._..wZD

58

and luck. The presumption that performance information influences
causal attributions is supported by several researchers. Nadler (1979)
pointed to the importance of feedback for helping group members develop
attributions regarding the relative contributions of individual group
members to group performance. In addition, previous empirical research
has shown that individuals use performance outcome information as a
primary cue for making causal attributions (Downey, Chacko, 8 McElroy,
1979; Staw, 1975). Further research (Checks 8 McElroy, 1983), supports
the relationship between performance information and causal attributions
by showing that knowledge of results has a stronger impact on
attributions than on goal aspirations (i.e., the individual's report of
the effort he or she will exert in order to reach the goal). In
addition, the control theory model of individual motivation presented by
Klein (1989) included an attributional search process.

Based on the above discussion, the following predictions are
presented:

HYPOTHESIS 2: An individual's perception of his or her

responsibility for the group's performance will be signifi-

cantly positively correlated with self attributions.

HYPOTHESIS 3: An individual's perception of his or her

responsibility for the group's performance will be signifi-

cantly negatively correlated with group attributions.

HYPOTHESIS 4: An individual's perception of his or her

responsibility for the group's performance will be signifi-

cantly negatively correlated with task difficulty attributions.

59

HYPOTHESIS 5: An individual's perception of his or her

responsibility for the group's performance will be signifi-

cantly negatively correlated with luck attributions.
Se - as

One attributional option is for the individual to ascribe the cause
of the group's successful performance to him or herself (i.e., either to
self-ability or self-effort). Previous research (Forsyth 8 Schlenker,
1977; Johnston, 1967; Schlenker, 1975; Wolosin, Sherman, 8 Till, 1973)
suggests the existence of a self-serving bias (i.e., individuals
assigning more responsibility to self for a group success than for a
group failure). Other research places boundaries on the existence of
the self-serving bias. For example, Schlenker and Miller (1977)
suggested that group cohesiveness might moderate the extent to which
group success was attributed to self. Other moderating variables
(individual need for achievement and self-esteem) have been suggested as
well (Adler, 1980; Fitch, 1970; Kukla, 1972). Miller and Ross (1975)
reviewed the relevant literature and suggested that there is some
evidence that individuals engage in self-enhancing attributions under
successful conditions, but that individuals do not engage in self-
protective attributions under conditions of failure. The question of
the existence of a self-serving bias in attributions for success is far
from resolved. However, it is important to recognize that the existence
of a self-serving bias could influence the model by making it more
likely that individual group members would make self attributions for
successful group performance and external attributions for failed group

performance. Support for a self-serving bias would be evidenced by a

60
significantly positive correlation between group standard attainment and
self attributions.

t v ea 0 s

The model suggests that individuals' causal attributions for the
group's performance have implications for their affective reactions.
Specifically, it is predicted below that group members' satisfaction
with their own performance and satisfaction with the group will be
influenced by the causal attributions they have made. Previous authors
(Carver 8 Scheier, 1981; Hollenbeck 8 Brief, 1987; Taylor, et al., 1984;
Weiner, 1982, 1985) have suggested that effective responses to
performance feedback are dependent upon causal attributions. The
direction of this affective reaction (i.e., whether it is positive or
negative), however, is expected to depend upon the degree of the group's
success in attaining its referent standard as well as upon the
attributions made by the individual.

Weiner (1985) suggested that causal attributions play a central
role in the emotion process (Figure 12 outlines his cognition-emotion
model of motivation). Specifically, he suggested that outcome
evaluation influences causal attributions and that the dimensions
underlying these attributions (locus of causality, stability,
controllability) determine which emotional response results. In other
words, each attributional dimension is expected to be associated with a
unique set of emotions. Weiner also proposed that the success of the
outcome is important in the determination of emotional reactions. In
his model, different emotional reactions are associated with success and

failure.

61

.8895 coroEmrcoonoo of Lo .258 c.8525 .NF 2:9".

mcozoEm A mcoficoéo use
8555 8.35:2 .830

/

cozeagw eEooSO aloEooSO

\

mcozoEm o>3mmez
to o>_:mon_ .9256

62

When dealing with the attributions of individuals in an achievement
situation, Weiner suggested that locus of causality determines the
impact of these ascriptions on self-related emotions. Specifically, he
proposed that the attribution of success or failure to internal causes
(ability, effort) is associated with changes in the individual's feeling
of self-esteem. Following the same reasoning, Weiner suggested that
attributions for success or failure to causes outside the self (i.e.,
"external” causes such as task difficulty and luck) would not influence
self-related emotions. These predictions are based upon the logic that
attitudes toward the source of the outcome are influenced by the
direction (i.e., success/failure) of the outcome. Weiner provided
research support for his propositions (Stipek, 1983; Weiner, Russell, &
Lerman, 1978, 1979).

As mentioned earlier, there has been little previous discussion of
the causal attributions which result from group level outcomes. It
appears, however, that Weiner's line of reasoning regarding the impact
of locus of causality on emotional reactions could be extended to the
group level. If the attribution of causality to internal sources leads
to changes in self-related emotions, then it seems likely that group-
related emotions would be influenced by attributions to causes
associated with the group. While not addressed by Weiner, this
prediction is based upon the logic he presented in relation to
attributions for individual performance.

Figure 10 presented a continuum which is expected to represent the
locus of causality dimension when individuals make attributions

regarding group outcomes. This continuum is useful at this point for

63
predicting affective reactions to causal attributions.

Attributions to External Eactors. As suggested by Weiner (1985),
attributions to external factors (task difficulty, luck) are not
expected to influence self-related emotions because the locus of
causality is perceived to be external to the focal individual.
Following the same logic, attributions to external factors are not
expected to influence group-related emotions. In other words,
individuals who make attributions to either task difficulty or luck are
not expected to show changes in either satisfaction with self
performance or satisfaction with the group.

Attributions to Self. Weiner (1985) cited previous research which
supports the proposition that individuals who attribute success or
failure to internal sources (self-ability, self-effort) will show
changes in self-related emotions. In other words, the individual who
attributes a success to him- or herself is likely to show an increase in
self-esteem, while the individual who attributes failure to him- or
herself is likely to show a decrease in self-esteem. In the study to be
described in the next chapter, satisfaction with self-performance
rather than general self-esteem is measured in order to more directly
measure affect related to the focal task.

Weiner's (1985) predictions regarding the impact of causal
properties on emotional responses were based upon individuals making
attributions for indigiggal performance. Given that the present model
addresses the group as well as the individual level of analysis,
Weiner's line of reasoning must be extended. In other words, it is

important to ask what impact attributions to internal factors

64
(self-ability, self-effort) have on group-related emotions.

Previous research in the social psychology literature is helpful in
making predictions regarding the influence of internal attributions on
group-related emotions. Kerr (1983) examined the situation in which one
individual group member feels primarily responsible for successful group
performance and termed it the "sucker" effect. Kerr found empirical
evidence of a sucker effect such that individuals who feel that other
capable group members are "free-riding" on their efforts showed
decreased effort and task motivation. The author attributed the sucker
effect to the fact that people find it aversive to play this role and,
therefore, subsequently reduce their efforts. Although research on the
sucker effect has not specifically addressed affective reactions, the
presumption that it is due to aversive feelings on the part of the
individual implies that a group-related affective reaction has occurred.
It seems likely, therefore, that individuals who attribute group success
primarily to their own efforts or ability are likely to experience a
decrease in positive feelings toward the group.

Predictions regarding the group-related affective response of the
individual group member who attributes poor group performance (i.e.,
group performance that does not meet the referent standard) primarily to
him- or herself are uncertain. Following the logic of Weiner's (1985)
reasoning, an attribution to an internal source (i.e., the self) should
not influence emotional responses to more external sources (i.e.,
satisfaction with the group). On the other hand, previous research
(Kerr, 1983) suggests that attributions to internal sources can

influence group-related emotional reactions. Although there is clearly

65
uncertainty associated with this issue, it is predicted (in line with
Weiner, 1985) that attributions for poor group performance to the self
are not expected to influence group-related affect.

Attributions to the Group. Another option is for the individual to
attribute the group’s performance to the other members of the group.
Following the logic of Weiner's (1985) cognition-emotion model, it is
hypothesized here that attributions to the group are expected to
influence group-related emotions.

Previous research supports this prediction. Staw (1975) suggested
that members of high performing groups were likely to attribute positive
characteristics to the group. He found that individual members of high
performing groups rated their group as more cohesive, more open to
change, more motivated, having better communication, having higher
ability, having higher satisfaction, and having higher role clarity.
Similarly, Downey, et al., (1979) found that group performance
influenced individual group members' perceptions of group
characteristics. Specifically, members of high performing groups rated
their group as: more cohesive, having higher quality communication;
having greater openness; having group members with higher ability,
motivation, satisfaction, and role clarity; having greater enjoyment in
working together; having a higher degree of confrontation. Still other
research suggests that members of groups that receive positive
performance feedback will feel more satisfied with their membership in
the group than will individuals who are not members of successful groups
(Schlenker, Soraci, & McCarthy, 1976). Additional research suggests

that group performance feedback, via its effect on self-evaluation,

66
plays an important role in the development of interpersonal attitudes
(Lott & Lott, 1966, 1960) and group morale (Berkowitz & Levy, 1956).
This research suggests that group success is associated with increased
positive affect toward the group on the part of group members.

0n the other hand, if the individual perceives that other group
members are the cause of the group's failure to reach the referent
standard, it is expected that he or she will experience decreased
positive affect toward the group. This, again, represents an extension
of the reasoning presented by Weiner (1985) which suggested that if
causal ascriptions are made to a particular source, emotional responses
to this source are influenced.

Again following Weiner's (1985) logic, attributions to the group
(group-ability, group-effort) are not expected to influence self-related
emotions. In other words, the individual who believes that others in
the group are responsible for the success or failure of the group is not
expected to show a change in satisfaction with self performance.

Summary and flypotheses. The discussion presented in the previous
sections leads to the following predictions.

HYPOTHESIS 6: The locus of causality of attributions and group

standard attainment will interact to affect satisfaction with self

performance such that the relationship between locus and satisfac-
tion with self performance will be positive for high group standard
attainment individuals and negative for low group standard attain-

ment individuals. (See Figure 13.)

The interaction outlined in Figure 13 predicts that individuals who

perceive the cause of group success (i.e., high group standard

67

High

High GSA
Satisfaction
with Self-
Performance
LOW LOW GSA
External Internal

Locus of Causality

Figure 13. Hypothesized interaction effect of
group standard attainment and

locus of causality of attributions
on satisfaction with self
performance.

68
attainment) as primarily internal will show greater satisfaction with
their own performance than will individuals who perceive the cause of
group success as more external. In addition, the interaction predicts
that individuals who perceive the cause of group failure (i.e., low
group standard attainment) as primarily internal will show less
satisfaction with their own performance than will individuals who
attribute group failure to more external factors.

It is further expected that the locus of causality of attributions
will interact with group standard attainment to influence satisfaction
with the group as follows:

HYPOTHESIS 7: The locus of causality of attributions and group

standard attainment will interact to affect satisfaction with the

group such that the relationship between locus of causality and
satisfaction with the group will be negative for high group
standard attainment individuals and positive for low group standard
attainment individuals.

Figure 14 presents the proposed interaction. The interaction
predicts that individuals who make highly internal attributions (to
self-ability or self-effort) for group success (i.e., high group
standard attainment) are expected to show less satisfaction with the
group (as predicted by research dealing with the "sucker" effect) than
are individuals who attribute group success to more external sources.
0n the other hand, individuals who make more external attributions
(i.e., attributions to the group) for group failure (i.e., low group
standard attainment) are expected to show less satisfaction with the

group than are individuals who attribute group failure to internal

69

High

Satisfaction

with Group High GSA
Low GSA
Low
Group Self
Attribution Attribution

Locus of Causality

Figure 14. Hypothesized interaction effect of
group standard attainment and
locus of causality of attributions
on satisfaction with the group.

70
sources.
C e t e u Goa

Carver and Scheier (1981) suggested that individuals who are
uncertain about whether they will be able to reach their referent
standard will engage in an assessment process, the result of which is an
outcome expectancy. In addition, Klein (1989) used a subjective
expected utilityconcept (Edwards, 1961) in his control theory model of
individual work motivation. Subjective expected utility was defined as
the product of the individual's expectancy regarding the goal and the
valence of this goal. Hollenbeck and Klein (1987) integrated goal
setting theory (Locke, et al., 1981) and expectancy theory concepts
(Vroom, 1964) through their conceptualization of goal commitment. These
authors presented a model of goal commitment that viewed this construct
as a factor composed of expectancy and valence judgments. The model
presented here suggests that the individual's commitment to the group
goal is influenced by the individual's causal attributions. This
assumption is consistent with the propositions of Carver and Scheier
(1981), as these authors proposed that attributions are one of several
factors influencing outcome expectancies, and goal commitment beliefs
entail expectancy judgments.

It is important to note that in this model goal commitment is a
belief which exists at the individual level (i.e., within the
perceptions of the individual group member). It would be impossible for
a group to develop a collective goal commitment belief for the same
reason that the comparison process between the group standard and group

performance feedback (discussed earlier) could not be carried out at the

71
group level. Groups do not have perceptual capabilities that exist
outside of the perceptual abilities of individual group members. A goal
commitment belief, therefore, represents an individual group member's
judgment regarding the probability that the group standard can be
reached in the future and his or her perception of the attractiveness of
this standard.

The group level of analysis complicates the assessment of the
expectancy that the standard may be achieved in the future. At the
individual level, any individual may hold expectations regarding the
probability of reaching a standard that depend upon his or her causal
attributions for past performance. For example, the individual who
attributes past individual performance success to effort must assess the
likelihood that he or she will exert this level of effort in the future.
However at the group level the effort component is multidimensional.

The individual must base expectancy judgments about the likelihood of
goal attainment on his or her own efforts as well as the efforts of
other group members. The same logic holds true for ability
attributions. While these judgments are more complex than at the
individual level, the model assumes that the individual group member is
capable of developing such judgments and cumulating them into an overall
expectancy perception.

The valence component of the goal commitment variable reflects the
individual group member's desire to reach the group's standard. This
component must also exist at the individual, rather than the group,
level. Note that the standard may be attractive to the group as a whole

and that this may drive the individual group member to view the standard

72
as attractive. In this case, the opinion of the group is a possible
reason for the individual's perception that the standard is attractive,
however the perception still occurs at the individual level.

McMahan (1973) suggested stability as the dimension of central
importance in the determination of outcome expectancies and, as
mentioned earlier, stability (i.e., consistency) of performance is
believed to be one of the primary sources of performance attribution
information (Heider, 1958). In order to predict an individual's
perception of the expectancy that the group can reach the goal in the
future (one component of the goal commitment construct), it is also
important to know the level of the group's previous performance (i.e.,
did the group attain it's referent standard?). Assuming a constant
valence for the goal, goal commitment is expected, therefore, to be a
function of the interaction of previous group standard attainment and
the stability of attributions for previous group performance. Possible
combinations of these variables and their effect on an individual's
commitment to the group referent standard will be discussed below. The
proposed interaction is presented in Figure 15.

o ucces u er a c . An individual group
member who attributes the cause of the group's previous successful
performance to stable factors should hold a high expectancy that the
group can be successful in the future (assuming that task difficulty
does not change). Such individuals are expected, therefore, to hold
high levels of goal commitment (given a stable valence).

Wesse- An individual

group member who attributes the cause of the group's previous successful

73

High

l-llgh GSA
Commitment
to Group
Goal
Low Low GSA

Low High
Stability of Attributions

Figure 15. Hypothesized interaction effect of
group standard attainment and

stability of attributions on

commitment to the group’s
standard.

74

performance to unstable factors should hold a low expectancy that the
group can achieve its standard in the future (given that task difficulty
remains stable). If the expectancy judgment is low, the individual's
commitment to the group's goal should also be low given a stable
valence.

§t§ble Atgributions for Unsuccessful Performance. An individual
who is a member of a group that has failed to attain its standard is
expected to hold a low expectancy that the group can attain this
standard in the future if the individual attributes the failure to a
stable cause (given a stable level of task difficulty). If the
expectancy belief of the individual is low, it is expected that the
individual's commitment to the group's goal will also be low given a
stable level of valence for the goal.

t e t ibu 'o s fo Unsuccessfu Pe anc . If an
individual group member attributes previous group failure to attain its
goal to an unstable cause, the individual is expected to hold a high
expectancy that the goal can be reached in the future. If the
individual's expectancy judgment is high, commitment to the goal is
expected to be high (assuming a stable level of valence for the goal).

As mentioned in the above discussion, commitment to the group's
goal is also likely to be influenced by the difficulty of the task
presently faced by the group. Because task difficulty is not always
stable, it is important to recognize the potential impact of task diffi-
culty level on the individual's expectancy judgment and, therefore, upon
his or her commitment to the group's goal. Holding task difficulty

constant, the following hypothesis is generated from Figure 15:

75

HYPOTHESIS 8: Group standard attainment and attribution

stability will interact to affect goal commitment such that

there will be a positive relationship between attribution

stability and goal commitment for high group standard attainment

individuals and a negative relationship between attribution
stability and goal commitment for low group standard attainment
individuals.

The model also posits a relationship between an individual group
member’s perception of the difficulty of the group's task on the next
performance trial and commitment to the group's goal on the next
performance trial. Task difficulty is expected to influence commitment
via its effect on expectancy judgments. For example, if the group
member perceives that the task will be very difficult to perform in the
future, his or her expectancy that the group can reach its goal
(standard) will be low, leading to lower commitment.
W

The individual's goal commitment beliefs are expected to have
implications for his or her subsequent behavior. Carver and Scheier
(1981) suggested that favorable expectancies lead to a continuation of
previous behavior and that unfavorable expectancies result in an impetus
to withdraw from further efforts toward the referent standard (unless
withdrawal itself is associated with aversive consequences). Previous
research on goal commitment at the individual level (Hollenbeck & Klein,
1987; Hollenbeck, Williams, & Klein, 1987; Locke, et al., 1981) points
to the importance of the goal commitment construct for predicting

performance in a goal setting context. This research suggests that

76
goals do not have a positive influence on individual performance unless
the individual is committed to the goal. This model, therefore,
predicts that the level of the group standard and commitment to the
standard will interact to influence individual performance. In the test
of the model described in the next chapter the level of the group
standard remains constant and, therefore, the following hypothesis is
presented:

HYPOTHESIS 9: There will be a significant positive relationship

between commitment to the group standard and individual performance

level.

While the model posits that commitment to the group goal influences
individual performance, it also recognizes that commitment is not the
only determinant of individual performance. The difficulty of the task
faced by the group, as well as the experience and ability of group
members on this task is expected to be related to individual
performance. In addition, individual group members' self-set goals
regarding the task at hand are also expected to influence individual
performance.

afi d ee-Rider f ec s

It is important to address social loafing and free-rider effects,
group-level phenomena which have been investigated in the social
psychology literature (Albanese & VanFleet, 1985; Harkins 6 Petty, 1982;
Kerr & Bruun, 1983; Latane, Williams, & Harkins, 1979; Price, 1987;
Williams, Harkins, & Latane, 1981). Social loafing is a phrase that
refers to the possibility that individuals exert less individual effort

when working in groups than when working alone, while free-riding refers

77
to the possibility that an individual group member who perceives that
the group will succeed without his or her efforts will decrease indivi-
dual effort toward the group goal.

Social loafing and free-riding provide further explanation of the
effects hypothesized in an earlier section. The previous discussion
suggested that individuals who attribute previous successful group
performance to stable factors will hold positive expectancies that the
group's goal may be attained in the future by maintaining present levels
of performance. Note that this present level of effort may, for any one
individual, represent either high or low performance. An individual may
perceive that he or she performed poorly individually on the task but
also perceive that the group was successful in reaching its goal despite
this poor individual performance. In this case, the social loafing
research suggests that the individual has little incentive to increase
his or her individual performance level, because the valued goal has
been attained without increased performance. Latane, Williams, and
Harkins (1979) found empirical evidence that individuals who performed
physical tasks in groups exerted less individual effort than did
individuals performing alone. Williams, Harkins, and Latane (1981)
suggested identifiability of individual performance as a boundary
condition for the social loafing phenomenon, however, by showing that
the effect occurred only if an individual's outputs were unidentifiable.
Harkins and Petty (1982) provided empirical evidence to suggest that
identifiability was not the sole boundary variable. They found that
subjects who felt that they were making a unique contribution to the

group endeavor were less likely to engage in social loafing than

78
individuals who felt that their efforts were redundant. In addition,
groups members operating within a group which faced a difficult task
were less likely to engage in social loafing than members of groups with
easier tasks. Kerr and Bruun (1983) and Albanese and VanFleet (1985)
have suggested that as group size increases, free-riding behavior
increases. Price (1987) found that type of task moderated the
relationship between identifiability and social loafing. Specifically,
identifiability did decrease social loafing behavior when individuals
were asked to express an opinion, but did not impact social loafing
behavior when individuals were asked to make a decision. Most research,
therefore, assumes the existence of social loafing and free-rider
effects in some circumstances.

Research on social loafing and free-rider effects in groups points
to the importance of the expectancy judgment (part of the goal
commitment variable) for understanding individual behavior. An
individual who perceives that the group has been successful in the past
despite his or her low level of effort may not feel that his or her
individual contribution is required in order for the group to succeed in
the future. Such an individual appears to value attainment of the
group's referent standard, but to hold a high expectancy that the group
will be able to reach the referent standard without his or her efforts.
This individual, therefore, might choose to engage in social loafing and
free-ride on the efforts of other group members. 0n the other hand, an
individual who values the goal, but does not expect that the group will
be able to reach the goal without his or her efforts would be unlikely

to free-ride. This is evidenced by research which found that

79

individuals who perceived that the task was difficult or that their
individual contributions were required to reach the group's goal were
unlikely to engage in free-riding behavior (Harkins & Petty, 1982).
Egedbgck Lgops

The model presented in Figure 8 suggests two feedback loops. The
first feedback loop is represented by the relationship between
individual performance and felt responsibility. As suggested in
Hypothesis 1, individual performance is expected to be positively
correlated with felt responsibility perceptions. Second, individual
performance is expected to influence subsequent group performance,
because the performance level of the group depends upon the performance
level of individual group members. It is important to note that this
second feedback 100p involves a transition from the individual level of
analysis to the group level of analysis. As discussed at the beginning
of this chapter, the referent standard, group performance, and feedback
which drive this model exist at the group level, while the cognitive
processes described occur at the individual level. Once these cognitive
processes have occurred, however, the behavioral reaction which results
is expected to influence the group's subsequent performance (via the
performance of individual group members). This feedback loop,
therefore, involves a transition from the individual level to the group
level.

The model also suggests that group performance may influence
subsequent group standards. This relationship is supported by previous
research dealing with level of aspiration (Forward & Zander, 1971;

Zander, Forward, & Albert, 1969; Zander & Medow, 1963; Zander & Newcomb,

80
1967). Although an examination of the processes through which group
standards are established is beyond the focus of this paper, the
influence of such processes is recognized in the model (by the multi-
directional arrow shown between group performance and group standard).

An additional feedback loop should be mentioned at this point,
although it is not included in the model. Previous control theorists
(Campion & Lord, 1982; Hollenbeck & Brief, 1987; Klein, 1989; Taylor,
et al., 1984) have suggested that individuals may respond to a
discrepancy between group performance and the group referent standard
via a behavioral change (which has already been discussed) or by means
of a cognitive change in the referent standard itself. It is important
to note that this second option is not directly available to individuals
operating within group settings. In most cases any one group member
does not have the power to change the group's referent standard. It is
possible that an individual group member may succeed in changing the
group standard if he or she is able to convince other group members to
agree to the change. This change, however,'is expected to occur over a
much longer time period than is the change in the referent standard
which occurs at the individual level.

It is likely, however, that the individual group member's feelings
toward the group may influence his or her feelings about the group's
goal. If the individual is very dissatisfied with his or her group, it
is less likely that he or she will greatly value the group's referent
standard. From the perspective of Hollenbeck and Williams (1987), the
individual may experience a change in goal importance. This negative

affect toward the group's referent standard may, over time, influence

81
the level or nature of the referent standard. If enough group members
decrease their commitment and efforts toward the group referent
standard, the standard will have to be changed if the group ever hopes
to attain it. Dissatisfaction with the group may, therefore, eventually
lead to actual changes in the nature or level of the goal. However,
because this relationship is expected to be neither direct nor
short-term, it is not depicted in Figure 8.
e sum 0 s

It is important to note that the control theory model presented in
the previous sections departs from traditional control theory
specifications. Previous presentations of control theory have suggested
that when there is no discrepancy between the standard and the sensor
information, the system continues its previous actions without
interruption. However, the control theory model presented here suggests
that even when the standard has been attained, the individual will
engage in a self-regulatory assessment process which has implications
for affect and behavior. The reason that the present model differs from
previous models involves the fact that the model presented here focuses
on group standards and feedback information. As Nadler (1979)
suggested, feedback information received by groups is different than
that received by individuals in that the group members may have
difficulty in determining the extent to which the feedback information
reflects their individual performance. Therefore, in order to determine
the relevance of group feedback to themselves, individual group members
must engage in the social comparison and attribution processes described

in this model even when there is no discrepancy between the referent

82
standard and performance feedback. In addition, as discussed
previously, individual group members are constrained in their ability to
directly change group referent standards. For these reasons, it is
expected that the processes which occur at the group level are
inherently more complex and involve more cognitive processing than those
occurring at the individual level. Consistent with this expectation,
Taylor, et a1. (1984) suggested that individuals engage in more frequent
monitoring behavior when they are uncertain about their ability to meet
the desired standard. Individuals who are operating in groups, compared
to individuals functioning alone, are more likely to be uncertain about
their ability to meet the standard (because of the uncertainty regarding
the performance of other group members). This suggests that individuals
working in groups are more likely to engage in conscious monitoring than
are individuals who are working alone.

The model described above makes several assumptions. First, the
model assumes that individuals initially value the group referent
standard. Second, the model is descriptive of only situations involving
evaluative feedback information. Finally, following the work of
Hollenbeck (Hollenbeck, 1989; Hollenbeck & Brief, 1987; Hollenbeck &
Williams, 1987), the processes described here are assumed to be more
descriptive of individuals high in self-focus than of individuals low in
self-focus.

G Go et 1 Re e c d the Contro eo odel

An explanation of the proposed model of motivation in groups (shown

in Figure 8) has been provided in this chapter. In addition to an

explanation of the model, however, it is also important to establish the

83
relationship between the proposed model and research findings dealing
with group motivation. Previous research on group goal setting will now
be presented and related to the proposed model.

Empirical research on goal setting in groups can be divided into
four general types: 1) research which examines the group goal
difficulty-group performance relationship, 2) research which examines
the group goal specificity-group performance relationship, 3) research
dealing with how groups establish goals, and 4) research focusing upon
the effects of group attributes on the goal setting process. Each of
these areas will be discussed and related to the model presented in
Figure 8.

Several research studies have examined the relationship between
group goal difficulty and group performance. Becker (1978) compared the
energy conservation behavior of families in the following experimental
treatments: 1) those assigned difficult conservation goals, 2) those
provided with feedback on their conservation performance, 3) those
assigned difficult conservation goals and provided with performance
feedback, and 4) those who were neither assigned conservation goals nor
provided with performance feedback. The results of this study indicated
that the families that were most successful in conserving energy were
those that were assigned difficult goals and provided with performance
feedback. Research conducted by Klein and Mulvey (1988) on the group
performance of undergraduate students working on a class project found
that goal level was positively related to group performance. Recent
research by Smith, Locke, and Barry (1990) examined the performance of

two types of groups working on a longitudinal business simulation:

84

1) eight groups assigned specific and moderately difficult goals, and
2) eight groups not assigned such goals. Their results indicated that
the existence of group goals was positively associated with performance
and the quality of work plans that the groups developed. Research by
Latham and Locke (1975) found that as time restrictions were placed on
the activities of logging crews (a restriction that might be equated
with increasing the difficulty of the goals of these crews), rate of
output increased. Further research, which examined the performance of
garment workers, indicated that groups that set production goals (and
were encouraged to set these goals at levels higher than their present
production levels) showed significantly higher subsequent performance
than did groups which did not set production goals (Lawrence & Smith,
1955). In addition, a recent investigation by Pritchard, Jones, Roth,
Stuebing, and Ekeberg (1988) involved the setting of group goals in five
organizational units at a U.S. Air Force base. Their results indicated
that the setting of group goals (along with the provision of performance
feedback) was associated with an increase in group performance of
approximately 75% over baseline. The authors encouraged the
organizational units to set difficult goals by making the goals
nonreportable to higher management. Finally, research by Ivancevich
(1974) which examined the impact of a Management by Objectives program
in a manufacturing corporation suggested that the institution of a
formal goal setting program was associated with higher subsequent
departmental performance.

The above research provides support for the proposition that group

goal difficulty is positively associated with group performance. The

85
model presented in Figure 8 is able to account for this relationship via
its suggestion that individual performance is related to group
performance directed toward group goal attainment. The model suggests
that individual performance will be directed toward group goal
attainment (via group performance) as long as the individual members of
the group are committed to the group goal. If the group's goal is set
at a higher (more difficult) level and the individual group members are
committed to attaining the group goal, group performance directed toward
achieving this goal should be higher than if the group's goal is set at
a lower (easier) level. Evidence in support of this position is
provided by Matsui, Kakuyama, and Onglatco (1987). These authors found
that subjects working in two-person groups showed the most performance
increase when they received feedback suggesting that their performance
was below both the group goal level and their individual goal level
(compared with individuals receiving feedback indicating that their
performance was: above standard on individual goals but below standard
on group goals, below standard on individual goals but above standard on
group goals, and above standard on both individual and group goals). In
addition, high ability subjects who had already achieved their
individual goal continued to exert effort aimed at achievement of the
group goal.

One previous study addressed the relationship between group goal
specificity and group performance. In an investigation of groups of
sewing machine operators, Koch (1979) found that the assignment of
specific goals and performance feedback was associated with higher group

performance. It should be noted, however, that these assigned goals

86
appeared to be both specific and difficult for the majority of work
groups. The goals were assigned by management at a level equal to that
of the most productive groups, suggesting that this goal was moderately
difficult for those who were members of average work groups. The
finding that goal specificity, when combined with goal difficulty, might
be associated with higher group performance is also predictable given
the model outlined in Figure 8. More specific goals allow individual
group members to make more accurate and consistent comparisons between
the group's goal and group performance and, therefore, to target their
individual performance toward group goal attainment in a more efficient
manner.

Research by Raven and Rietsema (1957) suggested that greater goal
clarity was associated with higher perceptions of group involvement and
higher attraction to the group's task. While attraction to the group
task is not a variable addressed in the model, the finding regarding
feelings toward the group is explainable from the model in Figure 8.
The model suggests that the comparator component is related to the
individual's affective responses to the group. If the goal is unclear,
individuals will be unable to make consistent comparisons between the
group standard and group performance, and affective reactions toward the
group would not be predictable.

The third major area of group goal setting research involves the
determination of how group goals are established. Zander and his
colleagues have provided the bulk of the research in this area. Zander
and Medow (1963) examined the goal setting behavior of high school boys

working on a motor skills task in order to determine whether previous

87
findings regarding level of aspiration applied to groups as well as
individuals. They found that both groups and solo individuals raised
their aspirations following a performance success, but that groups more
often lowered their aspirations following a performance failure. In
addition, they found that individual group members tended to judge their
own performance less harshly in the group failure than in the individual
failure condition. Zander and Newcomb (1967) examined the levels of
aspiration of United Fund campaign teams. Their findings suggested that
teams that failed to reach their goals tended to set future goals at a
higher level (compared to previous goals) than did teams that succeeded
in reaching their campaign goals. A follow-up study by Zander, Forward,
and Albert (1969) examined the reasons for this finding. The authors
concluded that failed groups tended to show a greater dislike for
lowering the campaign goal, tended to be more conscious of the needs of
the community which they represented, and tended to place pressure upon
themselves to meet these needs. A third study by Forward and Zander
(1971) focused upon further explaining this phenomenon by conducting a
laboratory study involving high school boys which explored three
possible reasons for the previous findings: 1) group failure arouses
strong dispositions among members to avoid the negative consequences of
failure and failed groups therefore make the goal irrelevant as a
criterion for group achievement, 2) the goal failure highlights unmet
needs, so members are hesitant to set low goals because they indicate a
lesser need for community welfare, and 3) external pressures force the
setting of a higher goal despite failure. The authors found that all

three alternatives had additive parts in explaining the phenomenon. A

88
final study in this area was conducted by Pearson (1987) who compared
the performance of groups involved in participative goal setting with a
control group. He found that participation in goal setting was
associated with higher group performance and with greater job
satisfaction.

The above research addresses the factors which contribute to the
establishment of group goal levels. While the model in Figure 8 does
recognize that previous group performance is one variable involved in
the setting of future goals (represented by the multidirectional arrow
between group performance and group standard), the model is not
concerned specifically with where the group referent standard comes
from. In other words, the model is assumed to operate given an
established group standard (group goal); determination of the reasons
why this standard was chosen are beyond the focus of this paper.

A final area which has received attention in the group goal setting
research is the influence of various group attributes on the goal
setting process. For example, in their investigation of groups of
undergraduates working on a class project, Klein and Mulvey (1988) found
that group cohesion was positively associated with both goal level
chosen and goal commitment. In addition, the authors found that group
performance norms were positively associated with goal level chosen and
goal commitment. A study by Medow and Zander (1965) which examined the
perceptions of high school boys working on a simple group task found
that individuals who perceived themselves to be more central to the
group felt more responsible for group performance than did peripheral

members when the group failed but not when the group succeeded in

89

reaching its goal. In addition, central members rated the group's
performance more favorably regardless of whether the team failed or
succeeded.

These findings related to the influence of various group attributes
on the goal setting process are not directly addressed by the model.
For example, the findings of Klein and Mulvey (1988) regarding the
relationships between group cohesion, group norms, and group goal level
are relevant to the establishment of group goal standards, however the
model presented in Figure 8 is not intended to address fig! group goals
are established. Klein and Mulvey's (1988) findings which suggest that
group cohesion and group norms are positively related to group goal
commitment are not inconsistent with the model presented in Figure 8.
The two independent variables involved in the Klein and Mulvey study
(group cohesion, group norms) are not, however, variables of central
interest in an initial test of this model. The findings of Medow and
Zander (1965) which suggest that central and peripheral group members
hold different perceptions of responsibility for group performance is
easily incorporated by the model in that central and peripheral members
could be assumed to perceive differing levels of felt responsibility.
W

The propositions presented in this chapter will be investigated in
an empirical study described in the following chapter. The research is
longitudinal in nature (i.e., there are multiple performance trials) and
deals with evaluative feedback information only (i.e., descriptive
feedback is not considered). The investigation utilizes sports teams in

order to examine motivation in groups. The suggestion that goal setting

90
programs may operate in sports teams and work groups in similar ways was
made by Locke and Latham (1985), and previous research on goal setting
at the individual level has examined motivation in sports teams
(Anderson, Crowell, Doman, & Howard, 1988). The procedures and measures
involved in this investigation will be described in the following

chapter.

CHAPTER FOUR: METHOD

The groups used to test the model presented in the previous chapter
were college men's intramural basketball teams. Sports teams were
chosen because they represented a group of individuals working together
toward a common group goal (i.e., winning the game), and because the
task faced by these teams involved interdependence of the efforts of
individual team members.

P w s

As suggested by Cohen (1977), a power analysis was conducted to
determine the sample size necessary to acquire a desired level of power
for the statistics used to test the hypotheses presented in the previous
chapter. The analysis requiring the greatest statistical power is that
which tests the hypotheses suggesting moderated regression.

Assuming that the standardized regression coefficient for the
moderator explains at least 5% of the variance in a complete regression
equation that explains a small amount of variance (R? - .10; Cohen,
1977) 114 subjects would provide a power of .80 at the .10 level.
Assuming that seven players per sports team would be involved in this
research, a sample of 114 subjects would require the participation of
sixteen teams. The participation of twenty teams was solicited to
insure against loss of power due to attrition.

529.1212

The final sample consisted of 19 college men's intramural basketball
teams at Michigan State University (one team decided at the last moment
not to participate). There were two intramural leagues: a competitive

league which played to a final league playoff tournament, and a

91

92
non—competitive league which did not include a playoff. Only teams from
the competitive intramural league were included in the study in order to
ensure that subjects were committed to winning the team's games. A
total of 153 subjects were involved in the research. The subjects were
all male and ranged in age from 18 to 36 years. Almost all (98%) of the
subjects had previous experience playing on basketball teams.
Progeduzg

The team managers were asked during an organizational meeting to
volunteer the team to participate in the study. Over ninety team
managers expressed interest in the study and twenty were chosen with
preference given to those teams that had played together in the past (on
average, teams had been together for almost two years), and those teams
that had between six and ten players (in order to ensure some player
consistency across games). In exchange for their participation, each
team was paid $75, and provided with game statistics (information on
points, rebounds, and assists) following each of the games in which data
were collected. For their part, team members were asked to take time
prior to the first game, and after each of their first three games to
complete a series of short (15-20 minute) questionnaires. Player
responses were seen by no one other than the experimenter and her
assistant.

The group goal was to win each of the games played. Game
statistics were collected by a trained rater during each game, and these
statistics served as a measure of individual performance. In order to
check the reliability of these statistics, a second trained rater was

sent to a subsample of the games and asked to collect statistics

93

independently. Raters were recruited from an undergraduate management
class, and ten individuals were hired. All raters had previous
experience either as basketball coaches, basketball statisticians, or
basketball referees. A local high school basketball coach was hired to
provide a one hour training session to these raters, in which the
scoring of statistics was explained and videotapes were viewed and
scored. The coach who trained the raters was paid $50, and raters were
paid $15 per game for their services. One rater failed to attend the
training session and was therefore not further involved in the research.

Prior to the first game, players completed a questionnaire
collecting information on demographics and the Russell Causal Dimension
Scale (1982), which will be described below. Immediately following each
of their first three games, players also completed a second
questionnaire that focused on their perceptions of events in the game
they had just played. This second questionnaire was collected by the
trained raters. This procedure was repeated for each of the three
games. Therefore, three types of information were collected: 1) game
statistics, 2) questionnaire data from each team member prior to the
first game, and 3) questionnaire data from each team member following
each of the first three games.
W

A pilot study was conducted with 139 college juniors and seniors
who had been working in groups as part of a management class. The
questionnaire used in the pilot study is provided in Appendix A. The
pilot study was conducted to check the reliabilities of the scales to be

used in the test of the model described above. The results of this

94
study indicated satisfactory alphas for all scales, with the lowest
equal to .71. The scales were refined based upon these results, and
certain items dropped if their deletion did not result in significant
decreases in the reliability of the scale. The resulting scales, which
were used in the actual test of the model, are presented in Appendix B.
e 5 Used n the Sam 1e of tramural Baske ba 1 eams

Table 1 provides a summary of the measures used in this study, as
well as the timing of completion of the measures. This section provides
specific information on these scales.

ormat o 01 e ted r o to 5 Game. Players were asked to
provide the following information on a one-time-only basis. Team
members were asked to arrive one half hour early to their first game in
order to provide the following information:

Demographic Information. Subjects were asked to provide general
demographic information (age, student number, team name, and year in
college).

Self-focus. The Fenigstein, Scheier, and Buss (1975) scale was
used to assess the self-focus of individual team members. The scale
contains items such as "I reflect about myself a lot," and ”I'm
constantly examining my own motives," and is rated on a 5 point scale
ranging from "Strongly Disagree" to "Strongly Agree.”

Experience. Each player was asked to indicate the number of
previous seasons he had played on a basketball team (pre-high school,
high school, intramural, college, or other).

Ability. Each player was also asked to provide a rating, on a 5-

point scale ranging from "Very Poor" to "Very Good," of each other

95

Table 1: Timing of Measures

 

Prior to After After After
Eirst Game Game #1 Game #2 Game #3

Demographic Information X

Self-focus X

Experience X

Ability X X
Russell Causal Dimension Scale X

Self Performance Ratings X X

Felt Responsibility Ratings X X
Attributions (Likert & Ipsative) X X

Task Difficultyl X X X
Task Difficultyz X X

Self-set Goals X X X
Satisfaction with Self Performance X X
Satisfaction with Group X X
Commitment to Group Standard X X

Group Standard Attainment X X X
Individual Performance X X X

 

96
player on the team. These ratings were then averaged for each player to
arrive at an estimate of the ability of each player. Because some
players had difficulty rating other members of the team at the beginning
of the season, players were asked to provide a second rating of ability
after their third game.

Russell Causal Dimension Scale (1982). Players were asked to
provide ratings of the locus of causality and stability of six
attributional sources (self ability, self effort, team ability, team
effort, task difficulty, and luck). The locus of causality of each
source was measured with three items, each involving a 9-point rating
scale. Similarly, subject perceptions of the stability of each
attributional source was determined via three items, each involving a 9-
point rating scale.

Information Collected At Each Game. Team members were asked to
complete a questionnaire immediately following each game. Each team
member was asked to provide ratings on the following dimensions:

Self Performance Ratings. After each game, 3 items assessed each
team member's perceptions of his own performance compared to that of
others on the team. All questions were answered on a 5-point Likert
scale, with anchors ranging from "Strongly Disagree" to "Strongly
Agree."

Felt Responsibility Ratings. Each player completed 5 items
designed to measure perceptions of his own responsibility for the team's
performance. All questions were answered on a 5-point Likert scale with

anchors ranging from "Strongly Disagree" to "Strongly Agree.”

97

Attributions - Likert Scale. For each goal, team members completed
items related to six categories of attributions: self ability
(4 items), self effort (5 items), team ability (5 items), team effort
(5 items), task difficulty (3 items), and luck (3 items). All items
were rated on a 5-point "Strongly Disagree" to "Strongly Agree" scale.

Attributions - Ipsative Scale. Team members were asked to allocate
100 percentage points across the six attributional sources, indicating
the importance of each to the team's performance outcome in the game
that was just played.

Task Difficulty1. 3 items assessed team member perceptions of the
ability of the team just played, in order to provide a measure of the
difficulty of the task just completed. All items were rated on a 5-
point scale ranging from "Strongly Disagree" to "Strongly Agree."

Task Difficultyz. Team members responded to 2 items designed to
measure their perceptions of the ability of the team which would be
played in the next game, in order to provide a measure of the difficulty
of the team's next task. Again, items were rated on a 5-point scale
ranging from "Strongly Disagree" to "Strongly Agree."

Self-Set Goals. Each player was asked to indicate whether he had
set a goal related to each of the three performance dimensions (points,
rebounds, assists) prior to the game just played. If a goal had been
set, the player was asked to indicate the level of that goal.

Satisfaction with Self Performance. Each player answered 4
questions related to his satisfaction with his own performance in the
game just played. All questions were in the form of a 5-point Likert

scale, with anchors ranging from "Strongly Disagree” to "Strongly

98
Agree.”

Satisfaction with Group. 6 items assessed the team member's
satisfaction with his team. A five point Likert scale, with anchors
ranging from ”Strongly Disagree" to "Strongly Agree" was used. Items
related to three dimensions of satisfaction with the team:

1) satisfaction with the skill level of the team, 2) satisfaction with
the effort level of the team, and 3) general satisfaction associated
with being a member of the team.

Commitment to Group Standard. Commitment to the group's standard
(i.e., commitment to winning the next game) was measured with the seven
item version of the Hollenbeck, Klein, O'Leary, and Wright (1989) goal
commitment scale, adapted to the group level.

Group Standard Attainment. Group standard attainment was
represented by the outcome of the game (i.e., win or loss). A team win
was scored as ”1" while a team loss was scored as "2.”

Individual Performance. The performance of individual team members
was measured by raters trained to collect performance statistics during
the games. Information on three performance dimensions was collected
during each game: 1) number of points scored, 2) number of rebounds,
and 3) number of assists.

a s 8

Two types of analyses were used to test the nine hypotheses
presented in Chapter Three. Hypotheses #2, #3, #4, #5, and #9 presented
similar predictions. All hypothesized correlational associations
between two variables. The statistic used to test these hypotheses was

a Pearson Product Moment correlation, calculated for each hypothesis for

99
both Games #1 and #2. Support for the hypotheses would be represented
by a significant correlation in the predicted direction.

For Hypotheses #2 through #5, two correlations were examined. The
first related to the relationship between felt responsibility and
attributions measured using a Likert scale measurement method. The
second indicated the relationship between felt responsibility and
attributions measured using an ipsative method of attribution
measurement.

The remaining hypotheses (#l, #6, #7, and #8) tested the
interactive effect of two variables on a dependent variable. The
analysis used to test these predictions was moderated regression.

For Hypothesis #1, a moderated regression analysis was conducted
for each game. In this analysis, felt responsibility perceptions were
regressed on group standard attainment (i.e., team win or loss), self
performance perceptions and their interaction. Support for this
hypothesis would be indicated by a statistically significant beta
coefficient corresponding to the interaction variable.

Hypotheses #6 and #7 both involved three moderated regression
analyses per game. Because the locus of causality variable (an
independent variable in both Hypothesis #6 and #7) is a linear composite
of the locus of causality of the individual attributional sources (self
ability, self effort, team ability, team effort, task difficulty, luck),
and because no previous research has addressed the manner in which these
sources should be weighted in order to arrive at a locus of causality
value, three different weighting schemes were used. The first method

weighted the individual attributional sources according to the

100
theoretical a priori framework predicted in Figure 10. The second
method used the subject's responses on the Russell Causal Dimension
Scale (i.e., their perception of the internality/externality of each
source) as weights for each attributional source. The final method
involved the empirical derivation of the best (i.e., least squares)
weights corresponding to each attributional source (determined by
regressing the relevant dependent variable on the attributions to each
of the six attributional sources, and using the resulting beta
coefficients as weights).

Hypothesis #6 was tested by regressing satisfaction with self
performance on group standard attainment, the locus of causality of
attributions (measured via the three different methods discussed above),
and their interaction. Support for this hypothesis would be represented
by a significant beta coefficient for the interaction variable.

Hypothesis #7 was tested by regressing satisfaction with group
performance on group standard attainment, the locus of causality of
attributions (again measured via the three different methods discussed
above), and their interaction. The locus of causality of attributions
involved only the self ability, self effort, team ability, and team
effort attributional sources. Support for this hypothesis would also be
represented by a significant beta coefficient for the interaction
variable.

The testing of Hypothesis #8 involved the use of "stability of
attributions" as an independent variable. The stability of attributions
variable is a linear composite of the stability of individual

attributional sources and the manner in which stability of attributions

101
should be measured is not well established. Therefore, stability of
attributions was measured in three ways: 1) using the a priori
theoretical framework depicted in Figure 11, 2) using responses on the
Russell Causal Dimension Scale to weight the individual attributional
sources, and 3) using empirically derived (least squares) weights
(determined by regressing commitment to the group's goal on the six
individual attributional sources). Hypothesis #8 was tested using three
moderated regression analyses per game, each regressing commitment to
the group's standard on group standard attainment, the locus of
causality of attributions (measured in one of the three ways), and their
interaction. Support for Hypothesis #8 would be provided by a

significant beta coefficient for the interaction variable.

CHAPTER FIVE: RESULTS

The internal consistencies of all measures were determined by
calculating the coefficient alphas corresponding to each scale. These
values are provided in Table 2. All alphas were acceptable (all greater
than .70; Nunnally, 1978), with the lowest value equal to .75. The
means and standard deviations of these variables are presented in Table
3 (sample sizes vary across scales because there were varying numbers of
team members attending each game), and intercorrelations among the key
variables in the model are presented in Table 4.

In 10 of the 57 games in which data were collected, two trained
raters provided ratings of individual performance. The interrater
reliability across these two raters was equal to .99 for number of
points, .98 for number of rebounds, and .59 for number of assists,
leading to an average interrater reliability across performance
dimensions equal to .86.
fiypgrhgsis #1

Hypothesis 1 predicted a significant interaction between group
standard attainment and self performance perceptions in the prediction
of felt responsibility perceptions. Table 5 contains the results of two
moderated regression analyses used to test this hypothesis. The first
regression analysis related to Game #1 and regressed felt responsibility
perceptions on group standard attainment (i.e., team win or loss), self
performance perceptions, and their interaction. The second analysis
related to Game #2 and regressed felt responsibility perceptions on
group standard attainment, self performance perceptions, and their

interaction. For both Game #1 and Game #2, the interaction of group

102

103

Table 2: Internal Consistencies of Measures

 

GAME #1

Eumber

figale Alpha of Items
Self Performance Perceptions .90 3
Felt Responsibility .75 5
Self Ability Attributions .89 4
Self Effort Attributions .85 5
Team Ability Attributions .90 5
Team Effort Attributions .91 5
Task Difficulty Attributions 82 3
Luck Attributions .84 3
Satisfaction with Self Performance .94 4
Satisfaction with Team Performance .87 6
Perceptions of Task Difficultyl .85 3
Perceptions of Task Difficultyz .97 2
Goal Commitmentz .83 7

GAME #2

Number

Scale Alpha pf Items
Self Performance Perceptions .91 3
Felt Responsibility .82 5
Self Ability Attributions .90 4
Self Effort Attributions .91 5
Team Ability Attributions .87 5
Team Effort Attributions .91 5
Task Difficulty Attributions .85 3
Luck Attributions .93 3
Satisfaction with Self Performance .97 4
Satisfaction with Team Performance .89 6
Perceptions of Task Difficultyl .88 3
Perceptions of Task Difficulty2 .94 2
Goal Commitmentz .84 7

 

104

Table 3: Means and Standard Deviations of Variables in the Model

 

GAME #1
Enrinhlg Mean Standarn Deviation fl
Group Standard Attainment 1.31 .46 148
Self Performance Perceptions 2.65 .82 138
Felt Responsibility 3.25 .70 137
Self Ability Attributions 2.72 .75 139
Self Effort Attributions 2.82 .67 139
Team Ability Attributions 3.80 .68 139
Team Effort Attributions 3.76 .72 139
Task Difficulty Attributions 2.89 .85 139
Luck Attributions 2.30 .93 139
Self Satisfaction 3.06 .99 137
Group Satisfaction 3.76 .76 137
Commitment to Group Standard 4.23 .61 137
Individual Performance 11.58 9.08 139

GAME #2
V ab e Mean Standard Deviarion fl

 

.51 .50 155
.55 .86 135
.17 .77 135
.76 .80 134
.94 .76 134
.60 .63 134

Group Standard Attainment
Self Performance Perceptions
Felt Responsibility

Self Ability Attributions
Self Effort Attributions
Team Ability Attributions

ObWWNUWWNNUNH
O‘
on

Team Effort Attributions .60 134
Task Difficulty Attributions .06 .81 134
Luck Attributions .16 1.05 134
Self Satisfaction .04 1.03 135
Group Satisfaction .56 .77 135
Commitment to Group Standard .08 .69 135
Individual Performance 1 .96 8.63 135

 

Means and Standard Deviations are divided by the number of items in the
scale.

105

we menu 0» vcoqmouuou Hmcowmuv on»

scams occuuoaouuoo one L‘

omoq

Each I N van .:«3 Each I _ umxu zoom vououu nu <mo

a...
m.

mo.
cu.

case on vcoemouuoo Accomoav ozu o>ono uncauuaeuuou

 

o.~ La. 00. canN. co. co.l
oo. o.~ «ecu. so. «co—.1 «0.1
«new. «emu. o.~ seen. 0~.I no.
«com. no. «can. a.“ can—.1 no.
No.1 ca¢~.l aann.f cawn.l o.~ cc.
No. no. ecu. «sun. ¢¢o~.1 c.~
«eon. mo. can“. «can. c«o~.1 no.
«o. co. «soc. «use. «enn.l «can.
eaod. co. «IMN. «ewe. no.1 o_.
eemd. we. no. «cow. co. no.1
eedu. no. no. No. and. «cm—.1
eewq. «0. sea”. echo. 0—.1 «NA.
eew~.l caw~.l «can.l ecnn.l «Im~. ¢¢o~.l
hump chm OH umuh<m hmHH<m x04 ah
Hxhuxtou msomu hamm

no.

«nu.

«Ice.

ecew.
ee-.l
ecu.
o.~
aces.
«and.
«o.
no.1
no.
ca~n.t

uh

ac.

eco_.

ease.

ecdw.
eenN.I
«cam.
sens.
o.—
«Na.
00.
een~.t
oo.
aecn.1

<9

ecwe.

co.

«chm.

«can.
no.1

see—.1

«cs—.

ecn~.

«ecu.

econ.

«use.
e¢~.l

um

«awn.

uo.

sewn.

«ewe.
No.1
o—.l

seed.

e¢~—.
demo.
o.—

«ace.

«awe.
oo.

<m

sewn.
o~.
«e—w.

«saw.
no.

ea-.l
no.
No.1

sews.

«can.

«can.
ecu.

«m

«can.

00.

e¢~.|

«coo.
co.
lea~.
no.
mo.
sewn.
«can.
came.
0.—
cea~.1

an

ace—.1 hum;
Ohm 0k
cc¢~.l athmxxou
hm~h<m
eehc.l anamo
hm~h<m
ec~n.t mama
came. :04
no. ah
ceen.l wk
ecmn.I <9
e¢o~.l mm
can~.1 <m
no.1 as
an~.l mm
o._ <mc
(mu

 

cocoEuOHuom Hmsvu>wvcH I mama
uumvcmum n:55 ob ucwsuaeaoo I new op ezyazxoo

oucqatoutaa agate can: coauoucmauam Ismue<m cacao
oocasuocuoa «sum gun: coauuocaauam I cmsa<m anew
. acoasancuuu< seas I gas
Icouusnatuu< sussuauuao smug I as

Humor 0:» :« uoaaeque> mcoa< acouuaaouuouuuucu

ncouusbuuuu< shaman aqua
ecoauanauuu< scaaup<.
uccwusnuuuu< uuouum
acouusnauuu< huuaun<
ccowucouscm auuaabaoconoox
ncoqueouuom oucqauouuom
uccesuouu< unavcwum noose

Boos
unom
«Hum
udoh
«How

a canes

106

Table 5: Regression Results for Felt Responsibility Perceptions

 

 

GAME #1
Step Variable Beta R Adj R2 R2 Change
1. Group Standard Attainment1 .850 .140 .012 .012
2. Self Performance Perceptions .970 .381 .132 .120**
3. GSA X Self Perf. Perceptions -.824 .436 .171 .039**
GAME #2
Step Variable Beta R Adj R2 R2 Change
1. Group Standard Attainment .504 .075 .002 .002
2. Self Performance Perceptions .995 .431 .174 .172**
3. GSA x Self Perf. Perceptions -.741 .463 .196 .022**
* p < .10
** p < .05

1Group Standard Attainment is scored such that 1 - Team Win and 2 - Team
Loss

107

standard attainment and self performance perceptions explained a
significant amount of additional variance in felt responsibility
perceptions. These effects are presented in Figures 16 and 17, and show
that the interactions are different than predicted. Contrary to the
interaction hypothesized, for members of both successful and
unsuccessful teams there was a positive relationship between self
performance perceptions and felt responsibility perceptions, however
this relationship was stronger for high group standard attainment
individuals than for low group standard attainment individuals.

Hypothesis #1 was also tested by examining the interactive effect
of group standard attainment and actual individual performance level on
felt responsibility perceptions. Individual performance levels for
number of points, number of rebounds, and number of assists were
converted to z-scores and z-scores were summed across these three
performance dimensions to yield one standardiZed performance score per
individual per game. Felt responsibility perceptions were then
regressed on group standard attainment, the standardized performance
score, and the interaction of group standard attainment and the
standardized performance score for both Game #1 and Game #2. As shown
in Table 6, the group standard attainment by standardized performance
score interaction did not explain a significant amount of additional
variance in either Game #1 or Game #2.

To summarize, a significant interaction was found when felt
responsibility perceptions were predicted via group standard attainment
and self performance perceptions, although the form of this interaction

was different than predicted. A significant interaction was not found,

108

Felt Responsibility ’

 

25
254 _. ....... _ ......................................... . ............. I .......................... . .................. . ........................................... . ....... . ...................................
2K3 _mme-mwmmwwmmmmwmeM”Inn--m.--.w-_n--uu.www-.flw .wwfl-.-flmmmmmmw
252 _ ............... . ............. . ...................................................... . ............. . ................. , ................. .....................................
221 _. .......................................... _ ..... _ ......................... -H--_ ................... I ........... , ....................................... m. ...............................
ch)._ ................................................................ _ ................. , ................................ , ............ ,. ............... U ....................... . ....................
1s; _- ........................................................................ , ...................................................................... , ............ . ..................
1‘3 _ ......................... _ ......................................................................................................... .m .................................................................
1:7 _ ......................... . ................................................ m. .................................................................................... . ...............
16-” .................................................................................................................................................. . .............. , ..................................
1:; p .......................................................................................................................... _ ................................................. . .........................
1“ _. ............ I .............................................................................................. . .........................................................................................
1:3 _. ............................... . ......................................... .m. ........ .mwm ....... V. ........... , ................................................................ .m.
1:! _- .................... , ...... _ ................................................................................................................................ -m. ....................... .

1 1 _ ................................................................................ . ........................... . . ..........................................................................................

1o 1 l l l l l l l l l l l 1

2345678910111213141516

 

 

 

Self Performance Perceptions

—“" High GSA —*— Low GSA

Figure 16. Regression Results for Felt Responsibility Perceptions
(Game #1)

25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10

Felt Responsibility

109

 

l J

 

l_ J l l

l

J

L

l

l l

 

 

2345678910111213

14 15 16

Self Performance Perceptions

Figure 17.

—‘— High GSA

—1— Low GSA

Regression Results for Felt Responsibility Perceptions

(Game #2)

110

 

Table 6: Regression Results for Felt Responsibility Perceptions
GAME #1
Step Variable Beta R Adj R2 R? Change
1. Group Standard Attainmentl .179 .140 .012 .012
2. Standardized Performance .303 .279 .063 .051**
3. GSA x Standardized Performance -.062 .279 .056 -.007
GAME #2
Step Variable Beta R R2 R? Change
1. Group Standard Attainment -.023 .075 .002 .002
2. Standardized Performance .259 .342 .104 .102**
3. GSA x Standardized Performance .083 .343 .097 -.007

 

* p < .10
** p < .05

1 Group Standard Attainment is scored such that 1 - Team Win and 2 -
Team Loss

111
however, when felt responsibility perceptions were predicted via group
standard attainment and actual individual performance, suggesting that
felt responsibility is more predictive from self performance perceptions
than from actual performance levels.

Hypothesis 2 through Hypothesis 5 predicted correlational
associations between felt responsibility perceptions and various
attributional sources. Attributional sources were measured in two
different ways in each of these hypotheses. First, six different Likert
scales were used to assess the extent to which individuals attributed
group outcomes to each of the six attributional sources (i.e., self
ability, self effort, team ability, team effort, task difficulty, luck).
Second, subjects were asked to indicate the importance of each
attributional source to the group's outcome by allocating 100 percentage
points across the six attributional sources (i.e., an ipsative scale).

Tables 7 and 8 present the intercorrelations among attributional
sources measured via these two different methods in each of the games.
There was a significant correlation between the two measures for all but
the team effort attributions in Game #1, and all but the team ability
attributions in Game #2. Despite the fact that most correlations were
significant, the values of these correlations were relatively low given
that the two scales were measuring the same construct (the highest
correlation was equal to .481). The correlations were especially low
for team attributions (team ability and team effort), suggesting that
the method used to measure attributions has an influence on the

attributions made.

112

 

 

Table 7: Correlations Between Likert Scale Measures of Attributions and
Ipsative Scale measures of Attributions for Game #1
LIKERT SCALE
Self Self Team Team Task Luck
Ability Effort Ability Effort Difficulty

Self .395** .372** -.086 -.018 .006 .067
I Ability
P
S Self .260** .276** -.O41 -.034 .013 .044
A Effort
T
I Team -.210** .204** .186** .148** -.016 .091
V Ability
E

Team -.046 .042 .058 .089 -.019 .113*
S Effort
C
A Task -.211** .174** .046 -.175** .291** .078
L Difficulty
E

Luck .018 .075 -.342** -.183** -.160** .481**

* p < .10

** p < .05

113

 

 

Table 8: Correlations Between Likert Scale Measures of Attributions and
Ipsative Scale Measures of Attributions for Game #2
LIKERT SCALE
Self Self Team Team Task Luck
Ability Effort Ability Effort Difficulty

Self .395** .377** .122* -.083 .101 .030
I Ability
P
S Self .293** .388** .011 -.122* .196** .051
A Effort
T
I Team -.214** .224** .108 .098 .053 -.201**
V Ability
E

Team -.189** .116* .270** .291** .073 -.014
S Effort
C
A Task -.208** .258** .133* -.130* .275** -.014
L Difficulty
E

Luck .067 .020 .295** -.222** .101 .430**
* p < .10

** p < .05

114
ct s #

Hypothesis 2 predicted a significant positive relationship between
felt responsibility perceptions and attributions to the self. Table 9
presents the findings related to this hypothesis when attributions were
measured using the Likert scales. The Pearson Product Moment
correlations between felt responsibility perceptions and attributions to
self ability were significant for both games (r - .635 and r - .699, p
<.001 for Games #1 and #2 respectively). The correlations between felt
responsibility perceptions and self effort attributions were also
significant (r - .501 and r - .715, p < .001 for Games #1 and #2
respectively). Felt responsibility perceptions were therefore
significantly positively related (r - .602, p <.001 for Game #1;

r - .735, p < .001 for Game #2) to the total attributions to the self
(created by adding self ability and self effort attributions).

The correlations between felt responsibility perceptions and self
ability and self effort attributions as measured using an ipsative scale
also revealed significant positive correlations (see Table 10).
Specifically, felt responsibility perceptions were significantly
correlated with attributions to self ability (r - .346, p < .001 for
Game #1; r - .288, p < .001 for Game #2), attributions to self effort
(r - .215 p < .01 for Game #1; r - .176, p < .05 for Game #2), and total
attributions to the self (r - .337, p < .001 for Game #1; r - .266,

p < .001 for Game #2). The data provide strong support for Hypothesis
#2, regardless of the method used to measure attributions (Likert scale

versus ipsative scale).

115

Table 9: Correlations Between Felt Responsibility Perceptions and Self
Attributions (Measured on Likert Scale)

 

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Self Ability Attributions r - .635 (p < .001)
Self Effort Attributions r - .501 (p < .001)
Sum of Self Ability and Self Effort
Attributions r - .602 (p < .001)
GAME #2
Variable Correlation With
Felt Responsibility Perceptions2
Self Ability Attributions r - .699 (p < .001)
Self Effort Attributions r - .715 (p < .001)
Sum of Self Ability and Self Effort
Attributions r - .735 (p < .001)
1 n - 137

2 n - 134

116

Table 10: Correlations Between Felt Responsibility Perceptions and Self
Attributions (Measured on Ipsative Scale)

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Self Ability Attributions r - .346 (p < .001)
Self Effort Attributions r - .215 (p < .01)
Sum of Self Ability and Self Effort
Attributions r - .337 (p < .001)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Self Ability Attributions r - .288 (p < .001)
Self Effort Attributions r - .176 (p < .05)
Sum of Self Ability and Self Effort
Attributions r - .266 (p < .001)

 

1 n - 134
211-135

117
W

Hypothesis 3 proposed a significant negative correlation between
felt responsibility perceptions and attributions to the group. Table 11
presents the findings related to this hypothesis. When attributions
were measured with Likert scales, there was little support for this
hypothesis. Specifically, the correlation between felt responsibility
perceptions and group ability attributions was significant for Game #1
(r - -.l47, p < .05), but not for Game #2 (r - -.018, p > .10). The
relationship between felt responsibility perceptions and group effort
attributions was nonsignificant for both games (r - -.025, p > .10 for
Game #1; r - .083, p > .10 for Game #2). The correlation between felt
responsibility perceptions and total attributions to the group (created
by adding attributions to group ability and attributions to group
effort) was not significant for either Game #1 (r - -.092, p >.10) or
Game #2 (r - .034, p > .10).

When group attributions were measured using an ipsative scale, a
significant negative correlation between felt responsibility perceptions
and group ability attributions was found for Game #1 (r - -.185, p <
.05) and a marginally significant correlation was found for Game #2
(r - -.141, p < .10). However, felt responsibility perceptions were not
significantly correlated with attributions to group effort for either
game (r - -.081, p > .10 for Game #1; r - -.110, p > .10 for Game #2).
When felt responsibility perceptions were correlated with total
attributions to the group (created by adding group ability and group
effort attributions), a significant negative correlation was found in

both games (r - -.205, p (.01 for Game #1; r - -.175, p < .05 for Game

118

Table 11: Correlations Between Felt Responsibility Perceptions and

Team Attributions (Measured on Likert Scale)

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Group Ability Attributions r - -.147 (p < .05)
Group Effort Attributions r - -.O25 (p > .10)
Sum of Group Ability and Group Effort
Attributions r - -.092 (p > .10)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Group Ability Attributions r - -.018 (p > .10)
Group Effort Attributions r - .083 (p > .10)
Sum of Group Ability and Group Effort
Attributions r - .034 (p > .10)

1n-137
2 n - 134

119
#2). Table 12 presents the results related to this hypothesis.
W
Hypothesis 4 proposed a significantly negative relationship between
felt responsibility perceptions and attributions to task difficulty.
This hypothesis was strongly supported (see Table 13). The correlations
between felt responsibility perceptions and task difficulty attributions
(measured with Likert scales) were significant for both Game #1
(r - -.l89, p < .05) and Game #2 (r - -.225, p < .01). As shown in
Table 14, these correlations were also significant when measured using
ipsative scales (r - -.176, p < .05 for Game #1; r - -.186, p < .05 for
Game #2).
o s s #5
Hypothesis #5 predicted a significantly negative relationship
between felt responsibility perceptions and attributions to luck. This
hypothesis was not supported. As shown in Table 15, the correlation
between felt responsibility perceptions and luck attributions measured
using Likert scales was significant and positive for Game #1 (r - .119,
p (.10), and nonsignificant for Game #2 (r - .076, p > .10). As
presented in Table 16, the correlation between felt responsibility
perceptions and luck attributions measured using ipsative scales was
nonsignificant for both Game #1 (r - .061, p > .10) and Game #2
(r - .087, p > .10).
mm
A principle axes factor analysis of the 25 Likert scale items
measuring attributions produced the rotated factor structure presented

in Table 17 for Game #1 and Table 18 for Game #2. As shown in Table 17,

120

Table 12: Correlations Between Felt Responsibility Perceptions and

Team Attributions (Measured on Ipsative Scale)

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Group Ability Attributions r - -.185 (p < .05)
Group Effort Attributions r - -.081 (p > .10)
Sum of Group Ability and Group Effort
Attributions r - -.205 (p < .01)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Group Ability Attributions r - -.l4l (p < .10)
Group Effort Attributions r - -.110 (p > .10)
Sum of Group Ability and Group Effort
Attributions r - -.175 (p < .05)

 

1n-l34
2 n - 135

121

Table 13: Correlations Between Felt Responsibility Perceptions and Task
Difficulty Attributions (Measured on Likert Scale)

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Task Difficulty Attributions r - -.189 (p < .05)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Task Difficulty Attributions r - -.225 (p < .01)

 

1 n - 137
2n-l34

Table 14. Correlations Between Felt Responsibility Perceptions and Task
Difficulty Attributions (Measured on Ipsative Scale)

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Task Difficulty Attributions r - -.l76 (p < .05)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Task Difficulty Attributions r - -.l86 (p < .05)

 

1n-l34
2 n - 135

122

Table 15: Correlations Between Felt Responsibility Perceptions and Luck
Attributions (Measured on Likert Scale)

 

GAME #1
Correlation With
Variable Felt Responsibility Perceptions1
Luck Attributions r - .119 (p < .10)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Luck Attributions r - .076 (p > .10)

 

111-137
2 n - 134

Table 16: Correlations Between Felt Responsibility Perceptions and Luck
Attributions (Measured on Ipsative Scale)

 

GAME'#1
Correlation With
Variable Felt Responsibility Perceptions1
Luck Attributions r - .061 (p > .10)
GAME #2
Correlation With
Variable Felt Responsibility Perceptions2
Luck Attributions r - .087 (p > .10)

 

1 n - 134
2m-135

123

Table 17: Rotated Factor Loadings for Principal Axis Factor Analysis of
Attribution Items Related to Game #1

 

lrpm Eagrpr 1 Eactor 2 Factor 3 Eacrpr 4 Engror 5
25 .875 .078 .008 .025 -.124
19 .854 .046 -.112 -.004 .001
17 .801 .138 -.177 .058 .098
23 .781 -.075 -.l74 .110 .060

4 .741 -.089 -.013 -.104 -.043
10 .718 .034 -.087 .106 .368
12 .701 .040 -.164 .104 .439
22 .674 .016 .012 .062 -.319
15 .593 .046 -.121 .052 .397

6 .577 .072 -.l64 .094 .382
18 -.054 .846 .004 -.055 -.032
11 .044 .819 -.014 .048 -.131
14 -.010 .816 .022 -.022 -.112
16 -.O4O .801 .097 -.O66 -.007
20 .002 .771 -.O99 .105 .057

l -.000 .725 .026 -.057 -.O37

8 .156 .688 -.032 .007 .204

2 -.027 .664 .102 .059 .054

S .254 .585 -.249 .079 .102
24 -.118 .027 .923 -.006 -.142
21 -.131 -.007 .793 -.135 -.018
13 -.l94 -.001 .747 -.152 -.063

3 .115 -.032 -.128 .832 -.009

9 .032 .090 -.129 .790 .274

7 .057 -.033 -.102 .264 .405

 

Self Ability Items - (1, ll, 16, 18)
Self Effort Items - (2, 5, 8, 14, 20)
Team Ability Items - (6, 10, 12, 15, 23)
Team Effort Items - (4, 17, 19, 22, 25)
Task Difficulty Items - (3, 7. 9)

Luck Items - (13, 21, 24)

124

Table 18: Rotated Factor Loadings for Principal Axis Factor Analysis of
Attribution Items Related to Game #2

 

tem Factor 1 Factor 2 Factor 3 Eactor 4
14 .848 .069 -.005 .026
2 .840 .029 -.028 -.069
18 .839 .100 .054 .039
11 .833 .120 .001 -.038
16 .831 .031 .036 -.078
8 .816 .036 -.077 -.057
20 .806 .149 -.106 -.104
1 .806 .067 .000 .015
5 .732 .119 .011 -.178
22 .084 .777 -.O93 .078
10 .114 .757 -.O82 .112
12 .175 .735 -.126 .017
19 .120 .727 -.012 .113
17 .095 .715 -.146 .085
23 -.025 .695 -.105 .146
25 .102 .692 -.072 .067
15 .060 .656 -.O71 .021
6 .012 .634 .013 .003
4 -.009 .578 -.108 -.026
21 -.004 -.235 .927 -.004
24 .020 -.153 .872 .007
13 -.062 -.161 .800 .029
9 -.033 .109 -.010 .944
7 -.O65 .151 .019 .631
3 -.133 .077 .015 .609

 

Self Ability Items - (1, ll, 16, 18)
Self Effort Items - (2, 5, 8, 14, 20)
Team Ability Items - (6, 10, 12, 15, 23)
Team Effort Items - (4, 17, 19, 22, 25)
Task Difficulty Items - (3, 7, 9)

Luck Items - (13, 21, 24)

125
the attributional items were represented by five factors for Game #1.
The first factor represented team attributions (including both team
ability and team effort attributions), the second factor represented
self attributions (including both self ability and self effort
attributions), the third factor corresponded to luck attributions, and
the fourth and fifth represented task difficulty attributions. These
five factors accounted for 62.9% of the variance.

For Game #2, a similar factor structure was found. As shown in
Table 18, the first factor represented attributions to the self
(including both self ability and self effort attributions), the second
factor corresponded to team attributions (including both team ability
and team effort attributions), the third factor represented attributions
to luck, and the fourth corresponded to task difficulty attributions.
These four factors accounted for 61.5% of the variance.

In order to determine the degree of similarity among these two
factor structures, the correlations between factor scores for Game #1
and Game #2 were determined (Nunnally, 1978). These results are
presented in Table 19. As shown in this table, the factors representing
team attributions, self attributions, and luck attributions were
significantly correlated, indicating high similarity between the factors
in Game #1 and Game #2. The correlation between factor scores for task
difficulty attributions across the two games, however, was not
significant. This was likely due to the fact that task difficulty was
represented by two factors in Game #1, but by only one factor in Game

#2.

126

Table 19: Correlations Between Factor Scores for Game #1 and Game #2

 

 

Team Self Luck Task Difficulty
Factor Factor Factor Factor
Team
Factor .19** -.14* -.06 -.08
Self
Factor .02 .31** -.04 .01
Luck
Factor -.12 .13 .40** .20**
Task Diffi-
culty Factor .06 -.20** -.18** .09
(2 items)
Task Diffi-
culty Factor .07 -.18** .01 .02
(1 item)
* p < 10

127

Figure 10 presented a theoretical a priori view of the way in which
the six attributional sources might be aligned along the locus of
causality continuum. In this research, the Russell Causal Dimension
Scale (1982) was used to measure each subject's perceptions of the way
in which these attributional sources are aligned along the locus of
causality continuum (the scale asks subjects to indicate their
perceptions of the internality/externality of each attributional
source). Table 20 presents the mean ratings across subjects, and the
resulting alignment of each of the six attributional sources along the
locus of causality continuum. When these six attributional sources are
combined into the three theoretical categories discussed earlier (self
attributions, team attributions, external attributions), they fall along
the locus of causality continuum as shown in Table 21. Tables 22 and 23
present t-tests of the differences among the locus of causality means of
the attributional sources. There were significant differences between
all means. The data, therefore, support the a priori framework for the
locus of causality of attributions.

Figure 11 presents the theoretical a priori view of the alignment
of the five relevant attributional sources along the stability
continuum. The Russell Causal Dimension Scale (1982) was also used to
measure subject perceptions of the stability of the attributional
sources. The mean ratings across subjects on this scale are provided in
Table 24, and the results of t-tests of these means are provided in
Table 25. In line with the a priori framework, subjects perceived both
self ability and team ability to be stable sources, while they perceived

self effort and team effort to be unstable sources. Attributions to

128

Table 20: Means of Subject Perceptions of the Locus of Causality of
Attributions (6 Attributional Categories)

 

LOCUS 0F CAUSALITY

 

External Internal
Task Team Team Self Self
Difficulty Luck Ability Effort Ability Effort

8.47 11.20 12.99 13.50 20.20 21.93
n - 144

Table 21: Means of Subject Perceptions of the Locus of Causality of
Attributions (3 Attributional Categories)

 

LOCUS 0F CAUSALITY

 

External Internal
External Team Self
Attributions1 Attributionsz Attributions3
19.63 26.44 41.99

n - 144

1External Attributions - Task Difficulty Attributions + Luck
Attributions

2Team Attributions - Team Ability + Team Effort Attributions
3Self Attributions - Self Ability + Self Effort Attributions

129

Table 22: T-tests of the Differences Between Mean Perceptions of the
Locus of Causality of Attributions (6 Attributional
Categories)

 

 

Standard E 2-ta1l
Var1ab1e n Mean Deviation Va1ue Prpbabiliry
Self Effort 21.93 4.19
Self Ability 143 20.17 3.77 -4.53 .000
Self Ability 20.17 3.76
Team Effort 142 13.50 5.49 12.67 .000
Team Effort , 13.57 5.54
Team Ability 143 12.95 4.59 -1.92 .057
Team Ability 12.96 4.57
Luck 144 11.17 5.30 3.72 .000
Luck 11.17 5.30
Task Difficulty 144 8.46 5.45 -5.35 .000

 

Table 23: T-tests of the Differences Between subject Mean Perceptions
of the Locus of Causality of Attributions (3 Attributional

 

Categories)
Standarg p 2 rail
Enrinplg n Mean Dev1ation V§1ne oba t
Self1 41.99 6 72
Team2 144 26 44 9 43 17 so .000
Team 26.44 9.43
External3 144 19.63 8.86 9.35 .000

 

1Self-Self Ability Attributions + Self Effort Attributions
zTeam-Team Ability Attributions + Team Effort Attributions
3External-Task Difficulty Attributions + Luck Attributions

130

Table 24: Means of Subject Perceptions of the Stability of Attributions

 

STABILITY
Unstable Stable
Team Self Team Self
Effort Effort Luck Ability Ability
10.00 10.36 10.50 11.40 12.44

 

n - 144

Table 25: T-tests of the Differences Between Mean Perceptions of the
Stability of Attributions

 

 

Standard r 2-ra11
Var1ab1e n Mean Deviation Va1ue £robab111ty
Self Ability 12.44 4.67
Team Ability 143 11.38 5.05 2.72 .007
Team Ability 11.37 5.04
Luck 144 10.50 6.04 1.60 .113
Luck 10.36 5.73
Self Effort 144 10.50 6.04 -.23 .816
Self Effort 10.40 5.74

Team Effort ' 143 10.00 5.15 1.21 .227

 

131
luck, however, were viewed as neither highly stable nor highly unstable,
a finding that conflicts with the a priori framework which suggests that
luck attributions are highly unstable.
W5

Hypothesis 6 predicted a significant interaction between group
standard attainment and the locus of causality of attributions in their
effect on satisfaction with self performance. The locus of causality of
attributions was measured in three ways, both theoretically and
empirically, and the results related to each method of measurement will
be discussed below.

One method for determining the locus of causality of attributions
was to weight the six attributional sources (self ability, self effort,
team ability, team effort, task difficulty, luck) according to the
theoretical a priori framework presented in Figure 10. In order to
simulate this framework, attributions to the self (self ability, self
effort) were assigned a weight of 3.0 to indicate high internality.
Attributions to the team (team ability, team effort) were assigned a
weight of 2.0 to indicate a moderate level of internality, and external
attributions (task difficulty, luck) were assigned a weight of 1.0 to
indicate a low level of internality. These three scores (one for self
attributions, one for team attributions, and one for external
attributions) were then summed for each individual and this sum was used
as the subject's locus of causality score. A moderated regression
analysis was then conducted, regressing satisfaction with self
performance on group standard attainment, the locus of causality of

attributions, and their interaction. The findings related to this

132
analysis are presented in Table 26. The results show that group
standard attainment and the locus of causality of attributions did
interact to influence satisfaction with self performance, however in a
manner different than predicted. The interactions are depicted in
Figures 18 and 19. In Game #1 and Game #2, the data suggested a
positive relationship between the locus of causality of attributions and
satisfaction with self performance for members of both successful and
unsuccessful teams, but this relationship was stronger for the former
group. In Game #2, this interaction was crossed such that highly
external attributions were associated with lower self satisfaction for
high group standard attainment individuals than for low group standard
attainment individuals.

A second way to determine the locus of causality of attributions
involved using subjects' responses to the Russell Causal Dimension Scale
(1982) as weights for each attributional source. In Russell's scale,
subjects were asked to provide ratings (using 3 items) of the extent to
which each of the six attributional sources was considered to be
external or internal to the self. The mean (across the 3 items) of the
subject's ratings for each attributional source was used as a weight for
the source, and the six scores (corresponding to the six attributional
sources) were summed to arrive at one value indicating the subject's
perception of the locus of causality of attributions. A moderated
regression equation then regressed satisfaction with self performance on
group standard attainment, the locus of causality of attributions, and
their interaction. The results of this regression analysis are provided

in Table 27, and provide no support for Hypothesis #6. Main effects

133

 

 

Table 26: Regression of Satisfaction with Self Performance on Group
Standard Attainment, Locus of Causality of Attributions, and
Their Interaction (With Locus of Causality Measured Using an
A Priori Framework)
GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 .521 .533 .279 .279**
2 Locus of Causality .777 .642 .403 .124**
3 GSA X Locus of Causality -.853 .654 .415 .012*
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment .832 .321 .096 .096**
2 Locus of Causality 1.016 .616 .371 .275**
3 GSA x Locus of Causality -.931 .630 .383 .012*
*p < .10
**p < .05

1 Group Standard Attainment is scored such that 1 - Team Win and 2 -
Team Loss

134

Satisi with Sell Peri

 

 

 

 

 

20

15 -

10 r

5 1 l l l l l J l L 1 l

100 110 120 130 140 150 160 170 180 190 200 210
External ' Internal

Locus of Causality

““ High GSA “‘1‘“ Low GSA

Figure 18. Regression Results for Self Satisfaction (A Priori
Weights in Game #1)

25

20

15

10

135

Satisi with Self Peri

 

 

l l l L l J l 4

 

 

80 100 120 140 160 180 200 220 240

External . Internal
Locus of Causality

*“ High GSA —1— Low GSA

Figure 19. Regression Results for Self Satisfaction (A Priori

Weights in Game #2)

136

Table 27: Regression of Satisfaction with Self Performance on Group
Standard Attainment, Locus of Causality of Attributions, and
Their Interaction (With Locus of Causality Measured Using

Russell Scale)

 

 

GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 -.345 .533 .279 .279**
2 Locus of Causality .296 .568 .312 .033**
3 GSA X Locus of Causality -.l62 .579 .307 -.005
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment -.122 .321 .096 .096**
2 Locus of Causality .300 .367 .120 .024**
3 GSA x Locus of Causality -.173 .369 .114 -.006
*p < .10
**p < .05

1 Group Standard Attainment is scored such that l - Team Win and 2 -

Team Loss

137
were found such that high group standard attainment (i.e., team win) was
associated with greater satisfaction with self performance than was low
group standard attainment (i.e., team loss), and internal attributions
for the group's performance outcome were associated with greater
satisfaction with self performance than were external attributions.

The third way in which the locus of causality of attributions was
calculated involved the empirical determination of the most advantageous
weighting scheme, using least squares estimation. In order to do this,
satisfaction with self perceptions was regressed on self ability
attributions, self effort attributions, team ability attributions, team
effort attributions, task difficulty attributions, and luck
attributions. The results of this regression analysis are presented in
Table 28. The beta coefficients resulting from this analysis were then
used as weights for each of the six attributional sources in a moderated
regression analysis, in which satisfaction with self performance was
regressed on group standard attainment, the locus of causality of
attributions (determined using the weights developed in the first
regression equation), and their interaction. The results of this
analysis are presented in Table 29. These findings suggest support for
Hypothesis #6 in both Game #1 and Game #2. The form of these
interactions is depicted in Figures 20 and 21. The interactions suggest
a positive relationship between the locus of causality of attributions
and satisfaction with self performance for both high and low group
standard attainment individuals, but a stronger relationship for the

former group.

138

Table 28: Regression of Satisfaction with Self Performance on Six
Attributional Sources

 

GAME #1
Step Variable Beta
1 Self Ability Attributions .0127
2 Self Effort Attributions .3311
3 Team Ability Attributions .2502
4 Team Effort Attributions .1147
5 Task Difficulty Attributions .1634
6 Luck Attributions -.1393

GAME #2
Step Variable ,Beta
1 Self Ability Attributions .1684
2 Self Effort Attributions .1420
3 Team Ability Attributions .1303
4 Team Effort Attributions .1369
5 Task Difficulty Attributions .1194
6 Luck Attributions .0926

 

139

Table 29: Regression of Satisfaction with Self Performance on Group
Standard Attainment, Locus of Causality of Attributions, and
Their Interaction (With Locus of Causality Measured Using
Least Squares Weights)

 

 

GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 .466 .533 .279 .279**
2 Locus of Causality .885 .672 .444 .165**
3 GSA X Locus of Causality -.656 .683 .454 .010*
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment .933 .321 .096 .096**
2 Locus of Causality .996 .589 .336 .240**
3 GSA x Locus of Causality -1.061 .602 .348 .012*
*p < .10
**p < .05

1 Group Standard Attainment is scored such that 1 - Team Win and 2 -
Team Loss

20
19
18
17
16
15
14
13
12

140

Satisi with Self Peri

 

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

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

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

CD1 ‘4 00 «D

 

 

 

6 7 8 91011121314151617

External Locus of Causality Internal

““ High GSA —*_ Low GSA

Figure 20. Regression Results for Self Satisfaction (Least Squares
Weights in Game #1)

141

Satisf with Self Peri

 

40

35

30

25

 

20

15

10

 

 

 

 

5 6 7 8 9 10 11 12 13 14 15
External . Internal
Locus of Causality

—“— High GSA —1— Low GSA

Figure 21. Regression Results for Self Satisfaction (Least Squares
Weights in Game #2)

142
mm

Hypothesis 7 predicted that group standard attainment and the locus
of causality of attributions (measured using self and team attributions
only) would interact to influence satisfaction with the group. This
hypothesis was tested using three separate methods for calculating the
locus of causality of attributions, as described in Hypothesis #6 above.
Each of these analyses will be discussed below.

The first method for calculating the locus of causality of
attributions weighted subject ratings of each of the four relevant
attributional sources (self ability, self effort, team ability, team
effort) based upon an a priori theoretical framework (see Figure 10).
Attributional sources which had been hypothesized to be highly internal
(self ability, self effort) were assigned a weight of 3.0 and sources
which had been hypothesized to be moderately internal (team ability,
team effort) were assigned a weight of 2.0. These four values were then
summed to arrive at one value indicating the subject's perception of the
locus of causality of attributions. A moderated regression analysis was
then conducted, regressing satisfaction with the group on group standard
attainment, the locus of causality of attributions, and their
interaction. The results of this regression analysis are presented in
Table 30. No significant interactions were found. Significant main
effects were found across both games for both group standard attainment
(with higher group standard attainment associated with higher group
satisfaction) and the locus of causality of attributions (with more

internal attributions associated with higher satisfaction with the

group).

143

 

 

Table 30: Regression of Satisfaction with the Group on Group Standard
Attainment, Locus of Causality of Attributions, and Their
Interaction (With Locus of Causality Measured Using an A
Priori Framework)

GAME #1

Step Variable Beta R Adj R2 R2 Change

1 Group Standard Attainment1 .155 .568 .317 .317**

2 Locus of Causality .552 .610 .362 .045**

3 GSA X Locus of Causality -.628 .619 .368 .006

GAME #2

Step Variable Beta R Adj R2 R2 Change

1 Group Standard Attainment -.460 .465 .210 .210**

2 Locus of Causality .225 .532 .272 .062**

3 GSA x Locus of Causality .093 .532 .267 -.005

*p < .10
**p < .05
1 Group Standard Attainment is scored such that 1 - Team Win and 2 -

Team Loss

144

A second method used to calculate the locus of causality of
attributions, involved the Russell Causal Dimension Scale (1982). The
subject's rating of the locus of causality of each of the four
attributional sources was used as a weight for the source. The
resulting values were summed to arrive at a total locus of causality
value. A moderated regression analysis was then conducted, regressing
satisfaction with the group on group standard attainment, the locus of
causality of attributions, and their interaction. The results of this
analysis are presented in Table 31. Evidence for a significant group
standard attainment by locus interaction was found only for Game #2, and
this interaction is depicted in Figure 22. This interaction suggests
that for members of both successful and unsuccessful groups, more
internal attributions are associated with higher group satisfaction than
are external attributions, but that this association is stronger for
members of unsuccessful groups.

The final method used to compute the locus of causality of
attributions involved an empirical determination of the ”best" (i.e.,
least squares) attributional weights. An hierarchical regression
analysis was conducted, regressing satisfaction with the group on
attributions to self ability, self effort, team ability, and team
effort. The resulting beta coefficients (shown in Table 32) were used
to weight each of the four attributional sources, which were summed to
arrive at a total locus of causality value. A moderated regression
analysis was then conducted, regressing satisfaction with the group on
group standard attainment, the locus of causality of attributions, and

their interaction. Significant interaction effects were found for both

145

 

 

Table 31: Regression of Satisfaction with the Group on Group Standard
Attainment, Locus of Causality of Attributions, and Their
Interaction (With Locus of Causality Measured Using Russell
Scale)

GAME #1

Step Variable Beta R Adj R2 R2 Change

1 Group Standard Attainment1 -.414 .568 .317 .317**

2 Locus of Causality .267 .596 .345 .028**

3 GSA X Locus of Causality -.122 .597 .341 -.004

GAME #2

Step Variable Beta R Adj R2 R2 Change

1 Group Standard Attainment -l.008 .465 .210 .210**

2 Locus of Causality -.338 .474 .212 .002

3 GSA x Locus of Causality .616 .495 .226 .014*

*p < .10
**p < .05
1 Group Standard Attainment is scored such that l - Team Win and 2 -

Team Loss

30
29
28
27
26
25

23
22
21
20
19

18 °
17_ --

16
15

146

Satisf with Group

 

 

 

 

l L l 1 1 1 l l

 

400 600 800 1000 1200 1400 1600 1800 2000 2200

Group Locus of Causality Self

“‘— High GSA —1— Low GSA

Figure 22. Regression Results for Group Satisfaction (Russell
Weights in Game #2)

147

Table 32: Regression of Satisfaction with Group Performance on Six
Attributional Sources

 

GAME #1
Step Variable Beta
1 Self Ability Attributions -.1088
2 Self Effort Attributions .2298
3 Team Ability Attributions .2213
4 Team Effort Attributions .3448
GAME #2
Step Variable Beta
1 Self Ability Attributions .0106
2 Self Effort Attributions .1790
3 Team Ability Attributions .3269
4 Team Effort Attributions .1693

 

148
Game #1 and Game #2. The results of these analyses are presented in
Table 33 and Figures 23 and 24. In both games a crossed interaction was
found such that there was a positive locus-satisfaction with group
relationship for members of successful and unsuccessful teams, but a
stronger relationship for those in the former group.

In general, there is little consistent support for Hypothesis #7.
Significant interaction effects were found in the data, however these
effects were not consistent across locus of causality measurement
methods (i.e., a priori versus Russell scale versus least squares
weighting).

Hypothes1s #8

Hypothesis 8 predicted a significant interaction between group
standard attainment and the stability of attributions in their effect on
commitment to the group's referent standard on the next performance
trial. Stability of attributions was calculated using three different
methods. Each method and the results associated with it will be
discussed below.

The stability of attributions was first calculated using the a
priori theoretical framework shown in Figure 11. A weight of 2.0 was
assigned to attributional sources that were hypothesized to be highly
stable (self ability, team ability), and a weight of 1.0 was assigned to
attributional sources that were hypothesized to be highly unstable (self
effort, team effort, luck). These weighted attributional scores were
then added to arrive at one stability of attributions value. A
moderated regression analysis was conducted, regressing commitment to

the team's standard on the next performance trial on group standard

149

Table 33: Regression of Satisfaction with the Group on Group Standard
Attainment, the Locus of Causality of Attributions, and Their

Interaction (Using Least Squares Estimates to Determine Locus
of Causality)

 

 

GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 1.254 .568 .317 .317**
2 Locus of Causality 1.228 .647 .409 .092**
3 GSA X Locus of Causality -l.344 .690 .463 .054**
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment 1.156 .465 .210 .210**
2 Locus of Causality 1.089 .587 .335 .125**
3 GSA x Locus of Causality -l.304 .611 .359 .024**
*p < .10
**p < .05

1 Group Standard Attainment is scored such that 1 - Team Win and 2 -
Team Loss

150

Satisf with Group

 

 

 

 

 

30
1255 L , .HMHH
20 r:
15 r
10 1 l L l 1

0 5 10 15 20 25

Group . Self
Locus of Causality
—‘— High GSA "i“ Low GSA
Figure 23. Regression Results for Group Satisfaction (Least Squares

Weights in Game #1)

30

151

Satisf with Group

 

 

 

 

 

so
20 -
15 _ _.
10 ' L 1 L 1 ‘
o -5 10 15 20 25 so 35

Gm" Locus of Causality 39”

—‘— High GSA "1— Low GSA

Figure 24. Regression Results for Group Satisfaction (Least Squares
Weights in Game #2)

152
attainment, the stability of attributions, and their interaction. The
results of this analysis are shown in Table 34. There was partial
support for Hypothesis 8, with a significant interaction found in Game
#1 but no evidence of an interaction effect in Game #2. The interaction
for Game #1 is presented in Figure 25. This crossed interaction
suggests a positive relationship between stability of attributions and
commitment to the group's standard for members of successful teams and a
negative relationship for members of unsuccessful teams.

A second method for determining the stability of attributions was
also employed. Each subject's ratings on the Russell Causal Dimension
Scale (1982) regarding the stability of the five relevant attributional
sources were used as weights applied to the actual attributions made by
subjects, and these values were summed to arrive at an overall stability
of attributions score. A moderated regression analysis was then
conducted, regressing commitment to the group's standard on the next
performance trial on group standard attainment, the stability of
attributions, and their interaction. The results of this analysis are
presented in Table 35. The data provided evidence of interaction
effects for both Games #1 and #2 (see Figures 26 and 27). The
interaction found in Game #1 is similar to that found in Figure 25
(i.e., interaction effect when the stability of attributions was
measured using the a priori framework). For Game #2 the form of the
interaction was reversed; that is, there was a positive relationship
between the stability of attributions and commitment to the group's
standard for members of unsuccessful teams and a negative relationship

for members of successful teams.

153

Table 34: Regression of Commitment to the Group's Goal on the Next
Performance Trial on Group Standard Attainment, the Stability
of Attributions, and Their Interaction (With Stability of
Attributions Measured Using an A Priori Framework)

 

 

GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 1.701 .151 .015 .015*
2 Stability of Attributions .785 .152 .008 -.007
3 GSA x Stability -l.725 .293 .064 .056**
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment -.674 .243 .052 .052**
2 Stability of Attributions -.198 .243 .045 -.007
3 GSA x Stability .416 .249 .041 -.004
*p < .10
**p < .05

1 Group Standard Attainment is scored such that l - Team Win and 2 -
Team Loss

154

Commitment to Standard

 

35

34 b » . .............. . ............................... . .................................... .... ......................................................................... .

3 1 h— ........................................................................................................ . ....... . ............. , ................. . ............................................ .

29._ .......... , .............................. , ........ . ................................................................................. . ....................... . ........... .

 

 

 

25 l l l l l l l
50 60 7O 80 90 100 110 120 130 140

L H' 11
W Stability of Attributions '9

'—“‘ High GSA —+— Low GSA

Figure 25. Regression Results for Commitment (A Priori Weights in Game #1)

155

Table 35: Regression of Commitment to the Group's Goal on the Next
Performance Trial on Group Standard Attainment, the Stability
of Attributions, and Their Interaction (With Stability of
Attributions Measured Using Russell Scale)

 

 

GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 .422 .152 .015 .015*
2 Stability of Attributions .517 .152 .008 -.007
3 GSA X Stability -.729 .236 .033 .025**
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment -.660 .243 .051 .051**
2 Stability of Attributions -.515 .253 .048 -.003
3 GSA x Stability .584 .297 .065 .017*
*p < .10
**'p < .05

1 Group Standard Attainment is scored such that l - Team Win and 2 -
Team Loss

156

Commitment to Standard

 

35

 

 

 

 

25 l l l l L L
0 250 500 750 1000 1250 1500 1750

. ' 11
1°“ Stability of Attributions “'9

—““ High GSA + Low GSA

Figure 26. Regression Results for Commitment (Russell Scale Weights in
Game #1)

157

Commitment to Standard

 

35
34 __. ......................................................... ...................... . .............................. _. ............................................................................ ..
32 1— .......................................................................................... . ..................... V ............................................................
30 __ ................ , ................................................................................................................................................................................
29 ._ .................................................. , ................................... . ..............................................................................................................
28 _ .................... . .............................. , ........................................................................................... . ...............................................
27 .. ............................................... . .................. ............................. . ........... . .................................... .

2K5 _ ................................................... , ....................... --- ................. . ....... - ......................... , ....................................... _ ......

 

 

 

25 l l l 1 l J l l
0 250 500 750 1000 1250 1500 1750 2000 2250

Low . _ . . High
Stability of Attributions

"“ High GSA “—1— Low GSA

Figure 27. Regression Results for Commitment (Russell Scale Weights in
Game #2)

158

A final method for determining the stability of attributions
involved the deve10pment of least squares estimates to be used as
weights for the attributional sources. In order to do this, each
subject's rating of commitment to the group's standard was regressed on
self ability attributions, self effort attributions, team ability
attributions, team effort attributions, and luck attributions. The
resulting beta coefficients (see Table 36) were used as weights applied
to the corresponding attributional sources. A total stability score was
then computed by summing these five weighted values. A moderated
regression analysis was conducted, regressing commitment to the group's
standard on the next performance trial on group standard attainment, the
stability of attributions, and their interaction. The results of this
analysis are provided in Table 37, and show a significant interaction
for Game #1 only. This interaction is presented in Figure 28 and
suggests a positive relationship between the stability of attributions
and commitment to the group's standard for members of successful teams
and a negative relationship for members of unsuccessful teams.

In general, there was evidence of a pattern in the interactions
across the measurement methods for Game #1, but little evidence of a
consistent interaction pattern for Game #2.

te #

Hypothesis 9 predicted a significant positive relationship between
individual commitment to the group's goal and individual performance.
This hypothesis was not supported for either game. That is, no
significant relationship was found between individual commitment and

individual performance for either Game #1 or Game #2. There was,

159

Table 36: Regression of Commitment to the Group's Standard on the Next
Performance Trial on Five Attributional Sources

 

GAME #1
Step Variable Beta
1 Self Ability Attributions .0206
2 Self Effort Attributions .0286
3 Team Ability Attributions -.0230
4 Team Effort Attributions .0502
5 Luck Attributions -.1826

GAME #2
Step Variable Beta
1 Self Ability Attributions -.1337
2 Self Effort Attributions .1411
3 Team Ability Attributions .1393
4 Team Effort Attributions -.0098
5 Luck Attributions -.1199

 

160

 

 

Table 37: Regression of Commitment to the Group's Goal on Group
Standard Attainment, the Stability of Attributions, and Their
Interaction (Using Least Squares Estimates to Stability of
Attributions)

GAME #1
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment1 -.116 .152 .015 .015*
2 Stability of Attributions .743 .213 .030 .015*
3 GSA X Stability -.620 .283 .058 .028**
GAME #2
Step Variable Beta R Adj R2 R2 Change
1 Group Standard Attainment .335 .243 .052 .052**
2 Stability of Attributions .657 .264 .056 .004
3 GSA x Stability -.533 .291 .063 .007

*p < .10

**p < .05

1 Group Standard Attainment is scored such that l - Team Win and 2 -

Team Loss

35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20

161

Commitment to Standard

 

 

 

 

 

Stability of Attributions

—“" High GSA "1— Low GSA

Figure 28. Regression Results for Commitment (Least Squares Weights
in Game #1)

162
however, a significant positive relationship between individual
commitment to the group's goal and group standard attainment (which is a
measure of group performance). Tables 38 and 39 present the results
related to this hypothesis.
er t e ed elationshi s

The model presented in Figure 8 also hypothesized that several
other variables would be related to individual performance level. The
difficulty of the task was expected to be related to individual
performance, however a significant correlation between these variables
was not found for either Game #1 or Game #2 (see Table 40).

Individual performance was also expected to be significantly
related to previous basketball experience and to ability to play
basketball (as rated by other players on the team). Experience was not
significantly related to performance in either Game #1 or Game #2,
however ability was significantly related to individual performance in
both games (see Table 40).

The model also predicted that self-set goals would be correlated
with individual performance. As shown in Table 40, in Game #1 the
correlation between the self-set goal for number of points and
individual performance related to number of points scored was
significant and equal to .296 (p < .05). For Game #2, the correlation
between the self-set goal for number of points and individual
performance related to number of points scored was also significant
(p < .05) and equal to .232, as was the correlation related to Game #3
(r-.241; p < .05). For goals related to number of rebounds, there was a

significant correlation between the goal and the performance related to

163

Table 38: Correlations Between Individual Goal Commitment and
Individual Performance

 

GAME #2
Variable Correlation n
Number of Points .024 117
Number of Rebounds .070 117
Number of Assists .089 117
GAME #3
Variable Correlation n
Number of Points .052 114
Number of Rebounds -.004 114
Number of Assists -.004 114

 

All r's nonsignificant; p > .10

Table 39: Correlations Between Commitment to the Group Standard in the
Next Game and Group Standard Attainment in the Next Game

 

 

Game #21 Game #32
Correlation Between Individual Commitment
and Group Standard Attainment3 r - -.l62 r - -.243
(p - .029) (p - .002)
1 n - 137
2 n - 135

3 Group Standard Attainment is scored such that 1 - Team Win and 2 -
Team Loss

164

Table 40: Relationships Among Other Variables in the Model

 

GAME #1
Variables Correlation
Task Difficulty: and Individual Performance r --.107
Task Difficultyz and Individual Performance r - .071
Task Difficulty; and Commitment to Group's Standard r - .058
Experience and Individual Performance r --.108
Ability and Individual Performance r - .555**
Self-set Goal for Points and Point Performance r - .296**
Self-set Goal for Rebounds and Rebound Performance r - .222**
Self-set Goal for Assists and Assist Performance r --.008
GAME #2
Variable Qorre1ation
Task Difficultyl and Individual Performance _ r --.048
Task Difficulty; and Individual Performance r --.063
Task Difficultyz and Commitment to Group's Standard r - .082
Experience and Individual Performance r --.046
Ability and Individual Performance r - .519**
Self-set Goal for Points and Point Performance r - .232**
Self-set Goal for Rebounds and Rebound Performance r - .313**
Self-set Goal for Assists and Assist Performance r - .260**
GAME #3
Variable ' Correlation
Task Difficultyi and Individual Performance r --.027
Self-set Goal for Points and Point Performance r - 241**
Self-set Goal for Rebounds and Rebound Performance r - .061
Self-set Goal for Assists and Assist Performance r --.026

139
120
137
139
130
137
135
134

1:1

135
114
135
120
120
131
130

.132

ID

139
123
122
121

 

165

rebounds for both Game #1 (r - .222, p < .05) and Game #2 (r - .313,
p < .05), but not for Game #3 (r-.06l; p > .10). Self-set goals related
to number of assists were significantly correlated with performance in
relation to assists for Game #2 (r - .260, p < .05), but not for Game #1
(r - -.008, p > .10) or Game #3 (r--.026; p > .10).
Feedbagk Loop

The model presented in Figure 8 also hypothesized a feedback loop,
with individual performance on each performance trial feeding back to
influence perceptions of felt responsibility. In relation to Game #1, a
significant correlation was found between individual performance level
(summed across performance on points, rebounds, and assists) and felt
responsibility perceptions (r - .212, p < .05). A similar correlation
was found for the relationship between individual performance level and
felt responsibility perceptions following Game #2 (r - .341, p < .05).

These results are presented in Table 41.

166

Table 41: Feedback Loop Relationships

 

GAME #1

Va a s Cprrelntions

Felt Responsibility Perceptions & Individual Performance r - .212**

GAME #2

Var1ab1es Cprrglntions

Felt Responsibility Perceptions & Individual Performance r - .34l**

 

CHAPTER SIX: DISCUSSION

The model describing the motivation of individuals operating in
groups which was presented in Figure 8 received general support in the
test of the data described in the previous chapters. It appears that
individuals who are functioning in groups do undertake a social
comparison process aimed at determining their individual responsibility
for the group's performance outcome. In addition, the data suggest that
these felt responsibility perceptions are related to the attributions
which are made, and that attributions do have implications for certain
of the individual's affective responses. There was partial support for
the prediction that attributions, in combination with group standard
attainment (i.e., team win or loss), would influence commitment to the
group standard on the next performance trial. Although there was no
support for the prediction that commitment to the group's standard would
influence performance at the individual level, there was evidence that
commitment was associated with group performance. Each of these
relationships will be discussed in more detail below.

d c e Res 5

The model predicted that self performance perceptions and group
standard attainment would interact to influence felt responsibility
perceptions such that individuals who were members of winning teams and
who perceived that their performance level was high compared to that of
other team members would feel more responsible for group success than
would individuals who perceived that their performance level was
relatively low. For individuals who were members of losing teams, low

self performance perceptions were expected to be more highly related to

167

168
felt responsibility perceptions than were high self performance
perceptions (see Figure 9). This hypothesis was supported in that a
significant interaction effect was found; this interaction, however, was
of a different form than predicted. The data showed a positive
relationship between self performance perceptions and felt
responsibility perceptions for members of both successful and
unsuccessful teams, although this relationship was stronger in the
former group. This interaction suggests that individuals whose groups
are successful but whose own individual performance levels are low are
able to recognize their own minimal contributions to the group's
success. Individual members of unsuccessful groups, however, do not
appear to claim higher responsibility for the group's failure when their
performance levels are low rather than high.

The interaction effect found in these data has interesting
implications. As expected, individuals whose perceived performance
levels were high and who were members of successful teams felt highly
responsible for the group's success. In addition, as expected,
individuals whose perceived performance levels were low and who were
members of successful teams felt very little responsibility for the
group's success. Such individuals are free-riding on the efforts of
other group members; they receive the benefits of belonging to a
successful team, without bearing the responsibility of creating that
success. The findings of this research make several contributions to
previous research on social loafing and free riding.

First, the fact that individuals who are free-riding on the efforts

of other group members recognize their own lack of contributions to the

169
team's success is noteworthy. Previous research on free riding and
social loafing has not addressed the perceptions of the free-riding
individual (i.e., it has not examined the extent to which the individual
is aware that he or she is free-riding). It is interesting that such
individuals do not delude themselves into believing that they have made
a significant contribution and therefore are responsible for the group's
success; rather, they are cognizant of the fact that their contributions
were low and that they do not share responsibility for the group's
positive outcome. Future research in this area should now address the
extent to which such individuals consciously phppng to free-ride versus
simply recognizing that they have been free-riding on the efforts of
others.

This research also adds to the free-riding/social loafing
literature because of the type of task used. In previous research in
this area, the task faced by the groups has typically been either a
physical exertion task in which individuals work alone (blowing air,
clapping hands, turning a knob; Kerr, 1983; Kerr & Bruun, 1983; Latane,
Williams, & Harkins, 1979; Williams, Harkins, & Latane, 1981), or a task
which did not involve group interaction (individuals working alone to
generate uses for an object, evaluation of the solution to a business
problem; Harkins & Petty, 1982; Price, 1987). The findings reported
here show evidence of social loafing in a sample of individuals working
on a task in which there is a great deal of group interaction and
interdependence.

The fact that individuals who were members of unsuccessful teams

felt highly responsible for the group's failure even when they also

170
perceived their individual performance level to have been high points
out the importance of group standards to individuals. Despite the fact
that the individual team member had performed well, he continued to feel
responsible for group failure. This finding is in line with the
research by Matsui, Kakuyama, and Onglatco (1987) which found that
individual group members continued to exert effort toward the group's
goal even after their own individual goal had been reached.

Perhaps the group standard had such a significant impact on group
members' perceptions of felt responsibility in this sample because the
group standard was highly volitional and public. Individuals chose of
their own volition to belong to the basketball team. In addition, team
membership was a highly public event as many team members wore shirts or
other clothing that clearly labelled them as team members. Combining
this with the fact that the team standard (goal) was also clear and
public (all teams wanted to win their games), previous research
(Salancik, 1977) would suggest that individual commitment to the group's
standard would be very high. It would be interesting to see whether the
results obtained here generalize to the situation in which team
membership is less volitional and public, or in which the team goal is
less public.

1 The findings of this study are also interesting in regard to the
priorities individuals set regarding individual versus group goals.
Recent research is beginning to address the interaction of group and
individual goals (Matsui, Kakuyama, & Onglatco, 1987; Mitchell & Silver,
1990). It may be that the priorities established between these two

types of goals (i.e., the respective levels of commitment) are partly a

171
function of the publicness and volition of group membership and of group
goals compared to that of individual goals.

This finding also has implications for previous research on the
self-serving bias. The interaction found here suggests that individual
group members are willing to accept responsibility for group failure,
while the self-serving bias suggests that individuals are more likely to
accept responsibility for group success than for group failure. A
direct test of the existence of a self-serving bias, a t-test between
the mean self attribution ratings for individuals who are members of
winning versus losing teams, shows a significant (p < .05) difference
between means for Game #2, but not for Game #1 (p > .10). The data
provide mixed evidence regarding the existence of a self-serving bias.

It is important to recognize that there may be an alternative
explanation for the findings in relation to Hypothesis #1 which
necessitates caution when interpreting these results. The model
presented in Figure 8 suggested that the individual team member's self
performance perceptions would interact with group standard attainment
(i.e., the team's performance outcome) to influence his perceptions of
his individual responsibility for the team's win or loss. The items
which assessed felt responsibility perceptions, however, generally were
phrased such that they assessed the individual's feeling of
responsibility for the team's pgrfprmpngg (rather than for the team's
win or loss). Four of the five items used to assess felt responsibility
perceptions asked about responsibility related to team performance ("I
feel pretty responsible for the team's performance in this game;" ”I

personally had very little to do with the team's performance in this

172
game;" "I personally had a lot to do with the team's performance in this
game;" "I don't think that I was at all responsible for the team's
performance in this game'), while only one item asked about
responsibility related to the team's outcome ("If I hadn't performed the
way I did, this game would have turned out very differently”). It is
impossible to discern whether subjects interpreted these questions in
the manner intended (i.e., as expressing perceptions regarding
responsibility for the team's win or loss) or whether they interpreted
them in an alternative manner (i.e., as expressing perceptions regarding
responsibility for the team's performance).

This alternative interpretation of the felt responsibility items
may have influenced the results obtained here. For example, a subject
who scored 35 of the team's 50 points is likely to feel very responsible
for the team's pergormance (i.e., the 50 points), but may not feel
responsible if the team loses the game (i.e., he does not feel
responsible for the team's outcome). If subjects were rating their
perceptions of their responsibility for the team's performance, then the
conclusions drawn in relation to Hypothesis #1 may be incorrect. As
mentioned above, a positive relationship was found between self
performance perceptions and felt responsibility perceptions for high
group standard individuals as predicted. Contrary to predictions,
however, a positive relationship was also found for members of low group
standard attainment teams. The alternative interpretation of the felt
responsibility items discussed above may explain this finding in that
individuals who are members of losing teams who feel that their own

performance level was high may feel responsible for the team's

173
pprfiprmnnnp (for example, the number of points the team scored), but not
for the team's outcome (i.e., the loss of the game). The extent to
which this alternative interpretation influences the results and
conclusions drawn here is not known, however this alternative should be
recognized and the findings interpreted with care.

Hypothesis #1 was also tested by using actual performance levels as
a substitute for subjects' perceptions of their own performance, and no
interaction effect was found between performance and group standard
attainment on felt responsibility perceptions. There are several
possible explanations for the fact that a significant interaction was
found when performance was measured via self performance perceptions,
but that none was found when actual performance was used.

First, it may be that a significant effect was found when
performance was measured with self performance perceptions solely
because of response-response bias. This explanation is unlikely,
however, given the interactive effect of group standard attainment and
self performance perceptions on felt responsibility perceptions. In
other words, a response-response bias would be expected to inflate the
correlation between self performance perceptions and felt
responsibility, but would not be expected to create a spurious
interactive effect.

A second explanation suggests that individual group members'
feelings regarding their own performance levels may influence felt
responsibility perceptions, but that their self perceptions of
performance may not correspond directly to actual performance levels.

In other words, individuals may not hold realistic perceptions of their

174
own performance levels. An examination of mean self performance
perceptions and individual ability ratings (as rated by other players on
the team) supports this explanation; individuals tended to underrate
their performance levels in games (t--l4.73, p < .000 for Game #1;
t— -l4.46, p < .000 for Game 2). It should be recognized, however, that
self performance perceptions related to specific games, while ability
ratings related to ability in general.

It is also important to note that the model depicted in Figure 8
predicts an interactive effect of self erce tions of e ance and
group standard attainment on felt responsibility perceptions. The model
is concerned with the cognitive processes occurring within the
perceptions of individual group members. Therefore, while it is
interesting that self perceptions of performance and actual performance
levels are not highly related, the lack of a strong relationship between
these two variables does not argue against the model. The results
related to Hypothesis #1 support the model in suggesting that
individuals do hold perceptions regarding their efforts versus those of
other group members (social comparison process) and that these
perceptions influence felt responsibility perceptions.

t e ve

The predictions made regarding the model depicted in Figure 8 are
based upon the premise that, when dealing at the group level, there are
at least six important attributional sources: self ability, self
effort, group ability, group effort, task difficulty, and luck. A
factor analysis of the attributional items used in this research

indicates support for at least four attributional sources when

175
individuals are functioning in group contexts: l) attributions to self
(including both self ability and self effort), 2) attributions to the
group (including both group ability and group effort), 3) attributions
to task difficulty, and 4) attributions to luck. The results of this
factor analysis provide support for the general attributional framework
suggested in previous chapters. Although subjects did not make the fine
distinctions between self ability and self effort attributions, or
between group ability and group effort attributions, they did view
attributional sources as being divided along the lines of self, group,
task difficulty, and luck.

These results are interesting in that very little research has
addressed the types of attributional sources which operate when
individuals are functioning within a group context. Based on the
results presented here, it appears that individuals do make attributions
to sources beyond the traditional four discussed in previous research
(i.e., ability, effort, task difficulty, and luck). This supports the
suggestions of Weiner (1982) that researchers move beyond the original
four sources and consider additional attributional sources which may be
specific to the context of the situation. In the case of an individual
operating within a group setting, it appears that individuals will make
attributions to the group as well as to the other four attributional
sources.

The intercorrelations among the six attributional sources were also
instructive. Not surprisingly, there was a high correlation between
attributions to self ability and self effort, and between attributions

to team ability and team effort. Attributions to luck were generally

176
found to be negatively related to attributions other than self (i.e.,
luck attributions were significantly negatively related to team ability,
team effort, and task difficulty attributions). In other words,
individuals who attributed group performance outcomes to sources other
than the self tended to make fewer attributions to luck, while there was
no significant relationship between luck attributions and attributions
to the self. It is not clear why luck attributions were not
significantly related to self attributions but were consistently related
to other attributional sources. Perhaps luck becomes a more central
attributional source for individuals working within groups because they
feel a lack of control over group outcomes. Individuals who are working
alone are likely to feel stronger control over their individual
outcomes, whereas the addition of other group members adds a great deal
of uncertainty to the attributional process for group outcomes.

In this research, subjects did not distinguish between either self
ability and self effort or group ability and group effort attributions.
Individuals were able only to make distinctions between general self and
general group attributional categories. There are various explanations
for this phenomenon. Effort/ability distinctions may have been more
difficult to make in this sample. In sports teams, effort and ability
are perhaps more closely entwined than in work settings. Another more
interesting explanation might be that when subjects make broad
self/group distinctions, they are less likely to make the more specific
ability/effort distinctions. In other words, there may be a ”level of
focus" factor which influences the attributions made by individuals.

This possibility would be interesting to explore in a study in which

177
some subjects were working alone while others were operating in groups,
but all were asked to provide attributions in a free-response format.
Along the same lines, perhaps the introduction of a relevant other
(i.e., the group) as a possible attributional source allows the
individual a new, broader scapegoat. In other words, individuals who
previously might have made attributions for personal failure to low self
effort in order to protect their self esteem (e.g., "My ability level is
high enough that I could have performed well if I had wished to, but I
didn't want to exert the required effort") now have the option of
attributing failure to the relevant other (i.e., the group) in order to
protect their self esteem.

Attributions were measured using two different measurement
techniques. First, separate Likert scales addressed subject perceptions
of each of the six attributional sources. Second, subjects were asked
to allocate 100 percentage points across the six attributional sources
(i.e., an ipsative scale). The correlations between attributions made
using these two different methods were generally significant (only team
effort attributions for Game #1 and team ability attributions for Game
#2 failed to reach significance). Despite the fact that the
correlations tended to reach Statistical significance, however, the
values of these correlations were low. Given that these two methods
were supposedly measuring the same phenomena, the correlations which
were found (ranging from .089 to .430) were not impressive. The fact
that these two measurement methods are not highly related suggests that
care should be taken in the interpretation of results using different

measurement methods. Because of the difficulties associated with the

178
use of ipsative scales (Kerlinger, 1964) and because the ipsative method
involved one-item measures, the Likert scale measures of attributional
sources were used to test most of the hypotheses in this research.

The relationship between Likert scales and ipsative scales in the
measurement of attributions points to an issue which is not resolved in
the literature on attributional processes. The Likert scale measurement
method assumes that the attributional sources are independent of each
other; that is, attributions to one source have no direct implications
for attributions to another source. The ipsative measurement method
assumes that attributions are dependent; as attributions to one source
increase, attributions to another source must decrease. The extent to
which attributions should be viewed as independent versus dependent is
not resolved in previous literature. It is important to recognize,

however, that the ipsative measure forces attributional sources to be

 

dependent while the Likert scale measurement allows for dependence pr
independence.
f e e ons b t ce 0 s n u

The model presented in Figure 8 suggested that felt responsibility
perceptions would be related to the attributions made by individual
group members. In these data felt responsibility perceptions were
associated with some, however not all, of the attributions made by
subjects.

The data clearly supported the proposition presented in Hypothesis
#2, that individual group members who feel highly responsible for the
group's performance outcome are likely to attribute this outcome to

themselves (either to their own ability or their own effort). This

179
finding was supported regardless of the measurement method used for
determining attributions (i.e., Likert scales or ipsative scales). This
result is not surprising in that individuals who feel a responsibility
for an outcome are likely to perceive themselves to be the source of
that outcome.

Hypothesis #3 suggested that individuals who feel responsible for
the group's performance outcome are less likely to make attributions to
the group. This hypothesis received no support when group attributions
were measured with the Likert scales, but did receive support when group
attributions were measured with the ipsative scales. It should be noted
that, in the case of an ipsative scale, a negative relationship between
felt responsibility perceptions and group attributions is quite likely
(although not assured) because of the positive relationship between felt
responsibility and self attributions and the ipsative nature of the
scale (i.e., as attributions to self increase it is likely, although not
certain, that attributions to the group will decrease). In other words,
the fact that there was a significant negative relationship between felt
responsibility perceptions and group attributions measured via the
ipsative scale is not taken as strong evidence that such a negative
relationship does in fact exist.

The results of the test of Hypothesis #3 provide interesting
questions for future research. The fact that individuals who feel
highly responsible for the group's outcome still make attributions to
the group suggests that the distinction between self and group
attributions is not clearcut (in the same way that responsibility for

group outcomes was not clearly a self/group issue in the results

180
associated with Hypothesis #1). Schlenker and Miller (1977) suggested
that some individuals perceive the group to be a direct extension of
themselves, while other individuals consider the group to be an
environmental variable (i.e., external to the self). Perhaps the
individual group member's perceptions regarding the group influences the
attributions which are made.

Some evidence for this suggestion is found when the sample is
broken down (using the Russell Causal Dimension scale) into: 1) those
individuals who reported team attributions (the sum of team ability and
team effort attributions) as highly internal to self, and 2) those
individuals who perceived team attributions as highly gnrprnnl to self.
A comparison of the felt responsibility-team attribution correlations
for these subsamples shows the correlation for "internals" (i.e., #1
above) to be nonsignificant for Game #1 (r - -.l3, p > .10), and
significantly positive for Game #2 (r - .31, p < .05). For "externals"
(i.e., #2 above), the correlation for Game #1 was marginally significant
and negative (r - -.18, p -.104), while the correlation for Game #2 was
significantly negative (r - -.27, p < .10). These data, therefore,
provide some evidence that individuals who perceive attributions to the
team as external attributions show a more negative relationship between
felt responsibility and team attributions than do individuals who
perceive team attributions as internal attributions. Future research
might address the influence of additional factors (such as individual
perceptions of group cohesiveness and the centrality of individuals in
the group) on individuals' perceptions of the locus of causality of

group attributions and on the felt responsibility-team attribution

181
relationship.

Hypothesis #4 suggested a negative relationship between felt
responsibility perceptions and attributions to task difficulty. This
hypothesis was strongly supported for both the Likert and ipsative
scales. It appears, therefore, that as an individual group member's
feelings of responsibility for the group's performance outcome
increases, his or her likelihood of attributing the group outcome to
task difficulty decreases.

Hypothesis #5 suggested that attributions to luck would be
negatively related to felt responsibility perceptions. This hypothesis
was not supported by the data. As felt responsibility perceptions
increased, attributions to luck did not decrease significantly. In this
sample, subjects tended to view luck as an attributional source which
was closer to team attributions than to highly external attributions on
the locus of causality continuum (see Table 20). Attributions to luck
may not, therefore, be perceived as highly external attributions as
suggested by previous research (Weiner, 1972). Luck attributions appear
to be more complicated than previously discussed.

It is interesting to examine these results in association with the
locus of causality continuum outlined in Table 20. The data showed that
felt responsibility was positively related to the highly internal
attributions (self attributions) and negatively related to the highly
external attributions (task difficulty attributions). For attributions
falling in the middle of the locus of causality continuum (team
attributions, luck attributions), no significant relationship with felt

responsibility was found. This pattern suggests that more internal

182

attributions are likely to be positively related to felt responsibility
perceptions, while more external attributions are likely to be
negatively related to felt responsibility perceptions. Perhaps the
extent to which attributions falling in the moderate range along the
locus of causality continuum are related to felt responsibility is
dependent upon additional variables (e.g., group cohesiveness,
centrality of individual in group).

finmmnry. Hypotheses #2 through #5 addressed the relationship
between felt responsibility perceptions and various attributions. There
was strong support for the proposition that felt responsibility
perceptions are positively associated with self attributions and
negatively associated with task difficulty attributions. There was no
strong evidence to indicate a negative relationship between felt
responsibility perceptions and team attributions or luck attributions.
In general, these findings lead to two general conclusions. First, they
provide support in a group context for previous research which suggests
that performance information influences causal attributions (Chacko &
McElroy, 1983; Downey, Chacko, & McElroy, 1979; Nadler, 1979; Staw,
1975). Second, they suggest that the relationship between felt
responsibility perceptions and attributions to the group may be
influenced by additional group-related variables which were not
addressed in this research. Specifically, the extent to which felt
responsibility is associated with group attributions may be a function
of variables such as group cohesiveness and an individual's centrality

in the group.

183

It should be noted that the tests of the relationship between felt
responsibility perceptions and attributions were all correlational. It
is not possible, therefore, to determine the causality of these
variables. In other words, it is possible that individuals first
undertake attributional searches and then make determinations regarding
their own degree of responsibility for the group's performance outcome.
The order of causality postured here (i.e., that felt responsibility
perceptions influence attributional searches and not vice versa) is
logic based; in other words, it seems reasonable that individuals would
be unable to undergo attributional searches when their own degree of
responsibility for the outcome has not yet been determined. The
individual who has not yet decided his or her degree of responsibility
does not have the necessary tools to make a self attribution. It is
possible that attributions might be based on situational cues that have
little relationship to actual felt responsibility (for example, a
supervisor may instruct an individual to assign causality to a
particular attributional source which does not reflect the true cause of
the outcome). In this case, however, the model has moved out of the
realm of a self-regulating system and additional variables (for example,
supervisor perceptions) would need to be incorporated. Given a self-
regulating system where individuals are free to undergo their own
attributional search, the model suggests the order of causality depicted
in Figure 8.
.i- ‘! - . I t . t 0!‘ c; Sat - 2 0I w -_ '- equ:I e

Hypothesis #6 addressed the interactive influence of group standard

1 attainment and the locus of causality of attributions on satisfaction

184
with self performance. Because the locus of causality of attributions
variable is a linear composite of individual attributional sources and
there is not one established manner in which to weight these individual
sources, locus of causality was measured using three weighting methods:
1) the a priori framework depicted in Figure 10, 2) the Russell Causal
Dimension Scale (1982), and 3) empirically determined (least squares)
weights.

A significant interaction between group standard attainment and the
locus of causality of attributions was found in both games using the a
priori framework and the empirically derived weights. In all cases,
there was a positive locus-self satisfaction relationship for members of
both successful and unsuccessful teams, although this relationship was
stronger for those on the successful teams. This interaction,
therefore, is different than predicted for low group standard attainment
individuals. For these subjects, satisfaction with self performance was
high when attributions were made to more internal, versus external,
sources.

The fact that individuals who made internal attributions for group
failure were more satisfied with self performance than were individuals
who made more external attributions for group failure is interesting.

It suggests that individual group members can be satisfied with their
own performance despite the fact that they, as individuals, feel
primarily responsible for the group's failure. Perhaps there is
something inherently positive involved in the process of making internal
attributions. That is, individuals recognize their contribution to the

group's failed outcome, but feel good about themselves because they were

185

involved in the process. For example, in the context of the basketball
teams used in this research, this phenomenon might take the form of an
individual player who recognizes that his skills are lower than those of
other players, but who feels good about his performance because he
exerted effort in the game and did not give up. In other words, there
may be some positive affect associated with being involved in the group
which offsets the negative affect associated with a performance failure
(a situation which does not exist at the individual level).

Examination of this finding in conjunction with other results of
this research are also enlightening. Previous findings (Hypothesis #1)
suggested that individual group members who felt that their performance
levels were high were still likely to feel responsibility for a group
failure. The findings related to Hypothesis #6 suggest that individuals
are able to attribute blame for group failure to themselves, but to
still feel satisfied with their own performance levels. These two
findings together suggest that individuals are able to separate their
role as "individual" from their role as "group member." That is, a grpnp
mpmpgr may feel responsible for the group failure even when his or her
individual performance level is high, and an 1ng1v1dua1 may be satisfied
with his or her own performance level, even when the group has failed.

In one case (a priori weights used in Game #2), a crossed
interaction effect was found. The direction of this interaction
suggests that individuals who attributed group failure to external
factors were more satisfied with their own performance than were
individuals who attributed group success to external factors. This

finding is interesting in that it suggests that external attributions

186
may provide some protection for self perceptions. That is, if a team
has lost a game and the individual can attribute blame to external
factors, satisfaction with self performance does not have to suffer. On
the other hand, if external attributions are made for a team win, the
individual's satisfaction with self performance may be low because he or
she does not feel that his or her performance contributed to the team's
win.

As mentioned above, no significant interactive effect of group
standard attainment and locus of causality on satisfaction with self
performance was found when locus of causality was measured using the
Russell Causal Dimension Scale (1982). The fact that a significant
interaction was found using other methods of measurement for locus of
causality of attributions is not conclusive evidence of a problem with
the Russell scale, but does suggest that the scale requires further
investigation as a measure of the locus of causality of attributions.

In summary, the results for Hypothesis #6 suggest several
conclusions. First, the Russell Causal Dimension Scale (1982)
measurement of the locus of causality of attributions was not predictive
of satisfaction with self performance, while the two other measurement
methods were. This finding suggests the possibility that the Russell
scale may be problematic in the measurement of locus of causality.
Second, the data show a similar pattern across games using both the a
priori measurement method and the empirically determined (least squares)
measurement format. For all individuals, satisfaction with self
performance is higher when internal (versus external) attributions are

made for the group's outcome, however the locus-satisfaction

187
relationship is stronger for members of successful groups than for
members of unsuccessful groups.

e c u o W o

Hypothesis #7 predicted a significant interaction between group
standard attainment and the locus of causality of attributions on
satisfaction with the group. Again, the locus of causality of
attributions was measured using three different weighting methods (a
priori, Russell Causal Dimension Scale, and empirically derived). The
results associated with Hypothesis #7 were inconclusive. Significant
interaction effects were found in only 3 of the 6 analyses, and the
patterns of these interactions depended upon the weighting method used.
The individual interaction effects found will be interpreted below.

Significant interactions were evident for both games when the locus
of causality of attributions was measured using the empirically derived
(least squares) weighting method. These interactions showed a stronger
positive relationship between locus of causality of attributions and
satisfaction with the group for members of successful versus
unsuccessful teams.

The interactions found using the empirically derived weights have
several implications. First, they suggest that an individual is more
satisfied with the team if he or she attributes the team's success to
the self, a finding which is consistent with the results of Hypothesis
#6 (where attributions to the self were associated with higher self
satisfaction). In other words, there is positive affect associated with
making a contribution to the group's success. It is interesting that

satisfaction with the group is higher for individuals who attribute the

188
group's success to the self than for those who attribute the success to
the group as a whole. It was predicted here that individuals who
perceived the group's success to be a function of the group as a whole
would be more satisfied with the group because they perceived that all
members of the group were responsible for the group's ability to attain
the referent standard. These data, however, suggest that this is not
the case. Individuals are more satisfied with their groups if they feel
that they are the ”stars" of the group (i.e., that they are primarily
responsible for the group's success). This points to the importance to
the group member of a perception that he or she is a contributing,
involved member of the group.

It is interesting to note that the positive affect associated with
attributions to the self may extend to attributions made for group
failure also. Individuals who attributed the cause of group failure to
themselves (versus to the group as a whole) showed higher satisfaction
with the group. Again, this is consistent with the findings of
Hypothesis #6 (which suggest that attributions to the self for a group
failure are associated with higher satisfaction with the self), and
suggests that individuals are more satisfied when they perceive that
they have contributed to the group's outcome (even when this outcome is
negative). It is also possible that individuals are more satisfied with
the group when they attribute group failure to themselves because they
feel badly about their contribution to group failure. That is, such
individuals may recognize that they are among the poorest performers in
the group and therefore perceive that other members of the group must be

relatively good performers, leading to a higher satisfaction with the

189
group rating. This suggests that the individual's own performance level
may be used as an anchor around which satisfaction with other group
members and the group as a whole may be determined.

A significant group standard attainment by locus of causality
interaction was found for Game #2 when locus of causality was determined
using the a Russell Causal Dimension Scale (1982). This interaction
indicated a stronger locus-group satisfaction relationship for members
of unsuccessful teams than for members of successful teams. This
interaction suggests that if the team is successful, the locus of
causality of attributions is less critical to individuals' satisfaction
with the group than if the team is unsuccessful. This finding is in
line with previous research suggesting that group success is associated
with more positive perceptions of the group (Berkowitz & Levy, 1956;
Downey, et al., 1979; Lott & Lott, 1966, 1960; Schlenker, Soraci, &
McCarthy, 1976; Staw, 1975). The interaction also indicates, consistent
with the findings discussed above, that individuals who are members of
unsuccessful groups and who attribute this team failure to themselves
are more satisfied with the group than are those individuals who
attribute the team failure to the group as a whole.

It is important to note that the test of Hypothesis #7 provides no
evidence of a ”sucker effect.” As described in previous chapters, the
sucker effect suggests that individual group members who perceive that
they are primarily responsible for the group's success tend to show
negative reactions to the group because they feel forced to carry
responsibility for other group members. These data suggest that

individuals who attribute group success primarily to themselves

190
experience higher group satisfaction, arguing against the "sucker
effect." In fact, these data imply that individuals welcome the chance
to be central contributors to the group's success, suggesting evidence
for a "star effect" which appears to be the opposite of the "sucker
effect."

A possible explanation for these conflicting results is found in
the type of task used in each of the studies. Earlier evidence of a
sucker effect (Kerr, 1983) was based upon a task in which subjects
worked independently. In the present study, however, group members
worked together on the task. Perhaps this difference in interdependence
accounts for the conflicting findings. It may be, therefore, that the
existence of a sucker effect is dependent upon the type of task which is
faced by the group. If the task encourages a focus upon individual
contributions and standards, there may be evidence of a sucker effect;
if the task focuses upon group contributions and group standards,
however, a sucker effect may be less likely to occur.

There was strong support for the existence of both a group standard
attainment main effect and a locus of causality of attributions main
effect in the prediction of satisfaction with the group. Across all
games and measurement methods (i.e., a priori, Russell scale, empirical)
a significant main effect for group standard attainment was found, such
that individuals who were members of successful teams also showed higher
satisfaction with their teams. This finding is in line with previous
research that suggests successful group performance is associated with
more positive perceptions of the group (Berkowitz & Levy, 1956; Downey,

et al., 1979; Lott & Lott, 1966, 1960; Schlenker, et al., 1976; Staw,

191
1975).

There was substantial evidence of a significant main effect for the
locus of causality of attributions in the prediction of satisfaction
with the group. The data suggested that the more internal the
attributions made by individual group members, the higher the
satisfaction with the group. This effect was found across all games and
measurement methods, with one exception (Game #2 when locus of causality
of attributions was measured using the Russell scale). This finding
suggests that the individual's perception of his or her ability to
influence group outcomes has an important impact on satisfaction with
the group. The individual team member who perceived that he was able to
influence the group's performance outcome felt more positively toward
the group. This suggests that group members highly value the ability to
influence group outcomes, a finding which is further supported by the
interactions discussed above. Future research on attributions at the
group level might address this phenomenon more directly.

In general, the relationship between attributions and satisfaction
with the group requires further examination. As discussed above, there
was strong evidence of main effects for both group standard attainment
and the locus of causality of attributions on satisfaction with the
group, however the interactive effect of group standard attainment and
the locus of causality of attributions was not clear.

.I -‘! : . i -- '1‘ o . - me . t - ou-'; {1'3 .

Hypothesis #8 predicted a significant interactive effect of group
standard attainment and the stability of attributions on commitment to

the group's standard on the next performance trial. The stability of

192
attributions was measured using three different weighting methods:
1) a priori framework, 2) Russell Causal Dimension Scale (1982), and
3) empirically derived weights. In general, there was support for this
hypothesis in Game #1, but not in Game #2.

In Game #1, a significant interaction was found when stability of
attributions was measured using the a priori method, the Russell scale
measure, and the empirically derived (least squares) method. These
interactions suggested that for members of successful teams the more
stable the attribution the higher the individual's commitment to the
group's standard, while for members of unsuccessful teams the more
stable the attribution the lower the individual's commitment to the
group's standard.

In Game #2, the predicted interaction between group standard
attainment and the stability of attributions on commitment to the
group's standard was not supported. For both the a priori and the
empirical measurement methods, no significant interaction effect was
found. Using the Russell Scale measurement method, a significant
interaction was found for Game #2, however the form of this interaction
was the opposite of that found in Game #1. This interaction suggested
that for members of successful teams, more stable attributions were
associated with low commitment to the group's standard, while for
members of unsuccessful teams, more stable attributions were associated
with high commitment to the group's standard.

It is not clear why there was evidence of an interaction between
group standard attainment and the stability of attributions in Game #1

but not in Game #2. If a significant interaction were found for Game #2

193

but not Game #1, a possible explanation might be that players' knowledge
of teammates' abilities and efforts had increased over time and
influenced perceptions regarding attributional stability and/or
commitment to the standard. This explanation, however, works counter to
the results actually found. It is interesting to note that group
standard attainment was differentially associated with commitment to the
group's standard depending upon the game and the measurement method
used. In most cases, a team min was associated with higher commitment
to winning the next game in Game #1, while a team lpgn was associated
with higher commitment to winning the next game in Game #2. Perhaps
players felt highly motivated by a win in their first game, while in
later games (i.e., Game #2) the importance of winning games so as to
reach the playoff tournament became more salient and a loss was
associated with higher commitment to winning the next game. It may be
that this differing view of group standard attainment and commitment
depending upon time (i.e., game number) is partly responsible for the
differential results related to Hypothesis #8 in Game #1 versus Game #2.

It is also interesting to note that the significant interaction
effects found here were crossed interactions. It was hypothesized that
when group members made unstable attributions for group success or
failure, they would be uncertain regarding whether the group would be
successful in the future. The interaction effects found here, however,
suggest that when group members make unstable attributions for their
team's success or failure they expect the opposite outcome to occur in
the future. That is, if the player made an unstable attribution for the

team's success in one game, his commitment to the group standard in the

194
next game was likely to be low (presumably due to a low expectancy of
winning the next game). On the other hand, if the player's team lost
the last game and he made unstable attributions for this loss, he had
higher commitment to the group standard in the next game (presumably due
to a high expectancy of winning the next game).
I u nce of Co it e t to t e G ou Standard on e fo a ce

Hypothesis #9 predicted a significant correlation between
commitment to the group's standard in the next game and individual
performance in the next game. The data did not support this prediction
for either Game #1 or Game #2. A significant correlation was found,
however, between commitment to the group's standard in the next game and
the team's performance (i.e., win or loss) in the next game.

The lack of relationship between commitment and individual
performance may be due to several confounding factors. Perhaps effort
and performance were not highly correlated in this empirical test. In
other words, an individual who is highly committed to the group's
standard would be expected to exert a great deal of effort in order to
attain that standard; it may be, however, that the degree of effort
exerted by individuals was not captured in this study. It is possible
that the individual performance measure used here did not completely
measure effort exerted toward the goal. Individual performance level
was determined by summing performance on three dimensions: points,
rebounds, and assists. Perhaps the inclusion of additional performance
dimensions which were more related to effort level but which were not
measured in this research (number of saves, number of steals, number of

shots taken, etc.) may have given a more complete measure. It is also

195
possible that an individual player who exerted a great deal of effort
toward reaching the group's goal was unable to translate this commitment
into higher individual performance during the basketball game; a player
may be highly committed, but not have many opportunities to make points,
rebounds, or assists during any one game. In addition, an individual
may exhibit citizenship behaviors which contribute to the team's
performance, but which do not translate into higher individual
performance during the game itself (e.g., the individual who motivates
other players, acts as cheerleader on the sidelines, etc.).

As mentioned above, there was a significant relationship between
commitment to the group's standard in the next game and group standard
attainment in the next game, which suggests that high commitment is
related to high group performance. Because higher group performance is
possible only through higher individual performance, this finding
provides indirect evidence that high commitment is associated with high
individual performance.

The model depicted in Figure 8 also suggested that the anticipated
difficulty of the winning the next game (task difficultyz) would
influence players' commitment to winning the next game (commitment to
the group standard on the next performance trial). There was no
evidence of such a relationship. A possible explanation for these null
findings relates to the fact that many players were unfamiliar with the
skill level of the team they would play next. For this reason, there

was a substantial amount of missing data on this variable.

196

W

The model presented in Figure 8 suggested relationships between
several additional variables and individual performance. Specifically,
task difficulty, experience, and ability were all expected to be
positively related to individual performance. The data provided
evidence of a significant relationship between only ability and
individual performance. The model also predicted that self-set goals
would be positively related to individual performance and this
prediction was generally supported (there was a significant relationship
for goals related to points across all three games, a significant
relationship for goals related to rebounds in two games, and a
significant relationship for goals related to assists in one game).
There was also support for the existence of a feedback loop between
individual performance and individual perceptions of felt responsibility
for the outcome of the game.
Limitations

There are several limitations to the present research which should
be recognized when drawing conclusions about the results obtained here.
First, the sample consisted of sports teams, and the extent to which the
results of this research can be generalized to other achievement
situations is uncertain. Previous research on motivation at the
individual level has utilized sports teams and concluded that the
processes which occur in this context are very similar to those
occurring in work groups (Anderson, Crowell, Doman, & Howard, 1988;
Locke & Latham, 1985). Future research should examine the model

presented in Figure 8 in work settings, however, before firm conclusions

197
may be drawn about its applicability to these contexts.

A second possible limitation centers on the type of group standard
involved in this research. In this sample, the group's standard (goal)
was to win the basketball game. This standard is somewhat unique for
two reasons. First, the standard was constant for all groups; that is,
all teams were working toward the goal of winning the game. Second,
this standard was an "all or nothing” phenomenon. In other words, teams
either met the standard (won the game) or did not (lost the game); there
was no reward associated with almost reaching the standard, or with
surpassing the standard. The extent to which the results associated
with this type of goal are generalizable to other types of goals is
uncertain and should be addressed in future research.

An additional limitation involves the possibility of response-
response bias. This bias is of concern primarily for Hypotheses #2
through #5, which are tested using bivariate correlations. While the
possibility of such a bias should be acknowledged, it should also be
recognized that this bias can not be eliminated. The model is concerned
with the co tive recesses of individual subjects; objective measures
of the variables involved are therefore not relevant to the predictions
of the model. The existence of response-response bias as an alternative
explanation for the findings related to Hypotheses #1, #6, #7, and #9 is
less troublesome. The tests of these hypotheses indicated significant
interaction effects, for which response-response bias is unlikely to be
an alternative explanation.

Finally, it would be instructive to examine these processes over a

longer period of time. This research measured individual perceptions

198
and reactions over three time periods, however the results of some
hypotheses (particularly Hypothesis #8) were conflicting depending upon
the time of measurement, and a longer term investigation might shed
additional light on the "true" nature of the relationships investigated.
d n e ate t he M de

Previous research in the group goal setting area has supported the
idea that group goals (standards) are associated with higher group
performance. The manner in which group standards influence group
performance via individual performance was explored in this research.
While the results provided general support, there were several points in
the model which require future research. A revised model, based on the
findings of this research, is presented in Figure 29. In this model,
the bold lines indicate relationships which were supported in the data
collected here, while the narrow lines indicate relationships that
require additional research before firm conclusions may be drawn.

Based on previous research (Becker, 1978; Klein & Mulvey, 1988;
Latham & Locke, 1975; Lawrence & Smith, 1955; Pritchard, Jones, Roth,
Stuebing, & Ekeberg, 1988; Smith, Locke, & Barry, 1990), group referent
standards (goals) are expected to influence group performance, and group
performance level is expected to be associated with the performance
feedback information received by the group. As in the control theory
framework, individuals are expected to compare the group's referent
standard with the group's performance feedback information and to make a
determination regarding whether or not the group has attained its
referent standard. In order for this determination to influence later

affect and performance, it was expected that cognitive processing aimed

199

fi

 

mesouo afi mafiumuoao wamswa>awcH mo coaum>aucz onu m0 auto: pomfi>ox .mN unawfim

 

COG-EuOuLOL

3:322: 0’ «COESEEOO

0300 «cone—om
>::o<

0032.;

:00 930.0 3 toad—Ominzoho

a

0:023:22 55223.52.
4 :0".

20m 5:3
coaozgawm

cont-nEoo 3.00m

 

:03 use“.
005anan
e320
.!
2:25:33
vaouceum 95::
3:59:00 eon-520.com
cacao
uawucuam

 

 

cacao

 

200
at understanding previous group outcomes would be undertaken by
individual group members. This research provides some preliminary
evidence regarding the nature of these cognitive processes.

The findings from the research conducted here suggest that
individuals undergo a social comparison process wherein they make
determinations about their own performance levels as compared to that of
other group members. The data also suggest that these performance
determinations, combined with group standard attainment influence the
individual group member's felt responsibility perceptions.

Felt responsibility perceptions, in turn, were found to be
associated with the type of attributions made by the group member.
Specifically, individuals who felt highly responsible for the group's
success or failure were more likely to make internal (self)
attributions, and less likely to make attributions to task difficulty.
The relationship between felt responsibility perceptions and
attributions to the team appeared to be dependent upon the individual's
perceptions regarding the extent to which the team was regarded as an
extension of him- or herself versus being regarded as an external
factor. These perceptions are likely to depend upon group-related
variables such as the group's cohesiveness and the individual's role in
the group (i.e., whether he or she is a central figure in the group),
and future research should address the influence of these variables.
Luck attributions were not related to felt responsibility perceptions in
this sample.

There was support for the proposition that locus of causality of

attributions and group standard attainment interacted to influence

201
satisfaction with self performance. There was not, however, conclusive
evidence regarding the prediction of satisfaction with the group, and
for that reason this variable was not included in the revised model
(which is intended to present the findings related to this study).
Future research, however, should focus on examination of the variables
involved in the prediction of satisfaction with the group.

There was some evidence in these data that the stability of
attributions and group standard attainment interacted to determine
commitment to the group's referent standard on the next performance
trial. The existence of this interaction effect, however, is not firmly
established because evidence for it was found in Game #1, but not in
Game #2.

Individual commitment to the group's standard was expected to
influence individual performance levels. Although there was no
significant relationship between individual commitment and individual
performance, there was a significant association between individual
commitment and group performance on the next performance trial. Because
group performance must be a function of individual performance, this
provides some indirect evidence that commitment was associated with
individual performance. Task Difficulty} was not included in the
revised model because there was no evidence that the difficulty of the
task on the next performance trial was related to the individual team
member's commitment to the group standard in that trial.

Self-set (i.e., individual) goals and ability were generally
associated with individual performance in this sample. In addition,

individual performance level was related to later felt responsibility

202
perceptions, as predicted. There were, however, no associations between
task difficulty and individual performance or between experience and
individual performance, and therefore these variables were not included
in the revised model.

The purpose of the present research was twofold: 1) to develop a
control theory framework which might guide future research on the
cognitive processes involved in the motivation of individuals operating
in groups, and 2) to provide a first empirical test of this framework.
The control theory framework developed here involved the incorporation
of existing knowledge in the areas of individual motivation and social
psychology. The model of group member motivation received general
support in this first test of the model and serves as a useful starting

point for future research in this area.

APPENDICES

APPENDIX A

Thank you for agreeing to complete this questionnaire. The purpose of
this questionnaire is to gather your feelings regarding your experiences as a
member of a work group. As you answer the questions on the following pages,
please consider your experiences as a member of a Management 302 group working
on the case presentation which was presented orally in class. Please be honest
in your responses. NO ONE OTHER THAN THE EXPERIMENTER (ANNE M. O'LEARY) WILL
SEE YOUR RESPONSES TO THESE QUESTIONS. Neither your TA nor other members of
your group will have access to your responses.

This questionnaire is being prepared for use in a later project, and
different types of questions are being tested here. Because of this, many of
the questions on the following pages may sound similar. It is important.
however, that you carefully read and answer each question. If you don't
understand any instructions or individual items, please ask your TA. If you
have any comments regarding the format of this questionnaire. please write your

comments in the margins.

Please sign your name below. indicating your consent to participate in
this project. DO NOT PUT YOUR NAME ANYWHERE ELSE ON THIS QUESTIONNAIRE. Also
indicate your Hanagement 302 TA's name and your section number. so that you may
be given credit for this project.

THANK YOU FOR YOUR HELP.

 

Name

 

TA's Name

 

Mgt. 302 Section #

203

204

Please read each of the following questions carefully and circle the answer
which is most indicative of your feelings, using the scale below. MANY OF THE
QUESTIONS WILL SOUND SIMILAR, BUT PLEASE CAREFULLY CONSIDER AND ANSWER EACH
QUESTION.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- S
l. I feel pretty responsible for the group's performance

on the case presentation. . . . . . . . . . . . . . . l 2 3 4 5

2. I'm very happy with the way that most of the
other group members performed on the case
presentation. . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

3. If I hadn't performed the way I did, the case presen-
tation would have turned out very differently. . . . 1 2 3 4 S

4. Some of the members of this group were lazy. . . . . l 2 3 4 5

5. I would say that the skill level of this group
was way above average. . . . . . . . . . . . l 2 3 4 S

6. I personally had a lot to do with the group's
performance on the case presentation. . . . . . . . . 1 2 3 4 S

7. It makes me mad that many members of this group
didn‘t put forth 100% effort on the case
presentation. . . . . . . . . . . . . . . . . . . . . 1 2 3 4 S

8. I personally had very little to do with the group's
performance on the case presentation. . . . . . . . . 1 2 3 _4 S

9. Most of the members of this group put forth lOOt
effort on the case presentation. . . . . . . . . . . 1 2 3 4 5

10. I think that I performed very well in regard to
the case presentation. . . . . . . . . . . . . . . . 1 2 3 4 5

11. Being a member of this group was very important
to me. . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

12. I think that I did a good job on the case
presentation. . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

13. I can't imagine a group that worked harder on a
case presentation than our group did. . . . . . . . . 1 2 3 4 S

14. I think that my performance on the case presen-
tation was pretty good. . . . . . . . . . . . . . . . 1 2 3 4 5

205

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

15. At times, I seriously considered quitting this

group. . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4
16. I'm sort of disappointed in the skill levels of

most of the members of this group. . . . . . . . . . 1 2 3 4 5
17. I'm very happy with the effort levels of other

members of the group on the case presentation. . . . l 2 3 4 S
18. I'm proud of how I performed in the case presen-

tation. . . . . . . . . . . . . . . . . . . . . . . . 1 2 3 4 5
19. I'm very happy with the skill levels of most of

the other members of this group. . . . . . . . . . . l 2 3 4 5
20. I'm not at all proud of how I performed in

the case presentation. . . . . . . . . . . . . . . . l 2 3 4 5
21. I feel that being a member of this group

was a real honor. . . . . . . . . . . . . . . . . . . 1 2 3 4 S
22. I couldn't be happier with the performance of

the other members of this group on the case

presentation. . . . . . . . . . . . . . . . . . . . . l 2 3 4 5
23. I don't think that I was at all responsible for the

group's performance on the case presentation. . . . . 1 2 3 4 5

24. I felt very good about being a member of this
group. . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

206

Think about your group's case presentation. and the grade that the group
received. Did your group receive the grade you personally wanted to receive on
this assignment? (Please check one.)

Yes No Don't know what grade
we received.

Your group's performance on the case presentation is probably due to many

factors. However, please think about the most importanr reasons why your group
received or did not receive the grade you wanted. Write five of these reasons

below, IN ORDER OF IMPORTANCE.

Most important reason: 1)

 

 

2)

 

 

3)

 

 

4)

 

 

5)

 

 

207

Think about your group's performance on the case presentation. There are many
reasons why your group performed the way it did on this presentation. but we
will focus here on six possible reasons. Please think about each of these
reasons and show how important you think that each reason was to the group's
performance on the case presentation by writing a percentage in the blank.
THESE PERCENTAGES MUST ADD TO 1008.

THE GROUP'S PERFORMANCE ON THE CASE PRESENTATION WAS BECAUSE OF:

% My personal ABILITY (i.e., my intelligence and skill level)
% My personal EFFORT (i.e., how hard I tried)

% The ABILITY of other members of the group (i.e., the
intelligence and skill level of other members of the group)

% The EFFORT of other members of the group (i.e., how hard
other group members tried)

3 The DIFFICULTY of the assignment (i.e., how hard the case
presentation assignment was)

% LUCK (i.e., factors over which the group had no control)

 

TOTAL - 100 S

208

The following questions ask you to think again about the reasons why your group
performed the way it did on the case presentation. Specifically, these
questions ask you to think about the ABILITY (i.e., intelligence, skill level)
and EFFORT (i.e., energy put forth) of yourself and other group members, as
well as the DIFFICULTY of the assignment and the influence of LUCK and other
chance factors. Please read and answer each question carefully, using the

following scale.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree.
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

H

I think that our group performed the way we did on
the case presentation largely because of MY
ABILITY as a student. . . . . . . . . . . . . . . . . l 2 3 4 5

2. I don't think that LUCK had much to do with our
team's performance in regard to the case presen-
tation. . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

3. The group's performance on the case presentation
was largely due to MY PERSONAL EFFORT. . . . . . . . 1 2 3 4 5

4. It was pretty likely that our team would get the
grade it did on the case presentation because
of the TYPE OF ASSIGNMENT that we had. . . . . . . . 1 2 3 4 S

5. I feel that our group's performance on the case
presentation was greatly influenced by the EFFORTS
of OTHER MEMBERS OF THE GROUP. . . . . . . . . . . . l 2 3 4 S

6. I feel that our group's performance on the case
presentation was mostly influenced by HOW HARD
I TRIED. . . . . . . . . . . . . . ... . . . . . . . 1 2 3 4 S

7. The ABILITY level of OTHER MEMBERS OF THE GROUP was
a major factor in the group's performance on the
case presentation. . . . . . . . . . . . . . . . . . l 2 3 4 5

8. MY EFFORT LEVEL had a big impact on the group's
performance on the case presentation. . . . . . . . l 2 3 4 5

9. Our group got the grade it did on the case
presentation because of the TYPE OF
ASSIGNMENT we had. . . . . . . . . . . . . . . . . . l 2 3 4 5

10. I think that the outcome of the case presentation
had a lot to do with the SKILL LEVEL of OTHER
MEMBERS OF THE GROUP. . . . . . . . . . . . . . . . . l 2 3 4 5

209

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

11. I think that the outcome of the case presentation
had a lot to do with MY SKILL LEVEL. . . . . . . . . l 2 3 4

12. The ABILITY of OTHER MEMBERS OF THE GROUP had
a big influence on the group's performance on
the case presentation. . . . . . . . . . . . . . . . 1 2 3 4 5

13. The group's performance on the case presentation
was mostly due to the EFFORTS of OTHER MEMBERS OF
THE GROUP. . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

14. The group performed the way it did on the case
presentation largely because of MY EFFORT LEVEL. . . 1 2 3 4 S

15. The group performed the way it did on the case
presentation mostly because of the SKILL LEVELS
of OTHER MEMBERS OF THE GROUP. . . . . . . . . . . . l 2 3 4 5

16. MY ABILITY had a big influence on the group's
performance on the case presentation. . . . . . . . l 2 3 4 5

17. The group performed the way it did on the case
presentation mostly because of the EFFORT levels of
OTHER MEMBERS OF THE GROUP. . . . . . . . . . . . . . 1 2 3 4 5

18. In regard to the case presentation, our group got
the grade it did because of the DIFFICULTY level
of the assignment. . . . . . . . . . . . . . . . . . 1 2 3 4 S

19. MY ABILITY was a major factor in the group's
performance on the case presentation. . . . . . . . . 1 2 3 4 5

20. I doubt that our group could have done any better
than we did on the case presentation because of
the DIFFICULTY level of the case presentation
assignment. . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

21. The EFFORT level of OTHER MEMBERS OF THE GROUP had a
big impact on the group's performance on the
case presentation. . . . . . . . . . . . . . . . . . l 2 3 4 5

22. The outcome of the case presentation had a lot to
do with MY EFFORT LEVEL. . . . . . . . . . . . . . . 1 2 3 4 S

23. Our group performed the way it did on the case
presentation mostly because of LUCK. . . . . . . . . l 2 3 4 S

210

Strongly Neither Agree
Disagree Disagree Nor Disagree Agree
1 --------------- 2 --------------- 3 --------------- 4 ------

24. The group performed the way it did on the case
presentation largely because of MY SKILL LEVEL. . . . 1

25. I think that our group's performance on the case
presentation was largely due to the SKILL LEVEL
of OTHER MEMBERS OF THE GROUP. . . . . . . . . . . . 1

26. CHANCE had a lot to do with our group's performance
on the case presentation. . . . . . . . . . . . . . . 1

27. I would say that it was mostly FACTORS BEYOND OUR
CONTROL that caused our group to perform the way
that we did on the case presentation. . . . . . . . . 1

28. The outcome of the case presentation had a lot to do
with the EFFORT level of OTHER MEMBERS OF THE

GROUP. . . . . . . . . . . . . . . . . . . . . . . . l

Strongly
Agree

 

Page

Self Focus Scale:

214

APPENDIX B

W1

#1 - #17

Pages 215 - 220
Russell Causal Dimension Scale (1982)

Pages 221-222

Felt

Responsibility Scale:

Satisfaction with Group Scale:
Satisfaction wit Team Scale:
Social Comparison Scale:
Ipsative Scale of Locus

of Causality:

Pages 223-224

Page

Page

Self
Self
Team
Team
Task
Luck

225
Task
Task

226
Self
Goal

Ability Attributions Scale:
Effort Attributions Scale:
Ability Attributions Scale:
Effort Attributions Scale:
Difficulty Attributions Scale:
Attributions Scale:

Difficulty at Time 1
Difficulty at Time 2

#1, 3, 4, 6, 14

#2, 5, 8, 10, 12, 15

#7, 9, 11, 16

#13, 17, last item on page 222

Page 222

#1, 11, 16, 18

#2, 5, 8, 14, 20
#6, 10, 12, 15, 23
#4, 17, 19, 22, 25

#3, 7, 9
#13, 21, 24
#l - #3
#1 - #2

Set Goals for Points, Rebounds, Assists

Commitment Scale:

211

#l - #7

THANK YOU FOR ACREEING TO PARTICIPATE IN THIS RESEARCH.
LONCEST OF THE FOUR QUESTIONNAIRES YOU WILL BE ASKED TO COMPLETE OVER THE NEXT

212

FEW WEEKS. PLEASE TAKE THE TIME TO CONSIDER EACH QUESTION CAREFULLY. IF YOU

DO NOT UNDERSTAND ANY QUESTIONS. PLEASE ASK E03 AN EXPLANATION.

ALL OF YOUR ANSWERS ARE COMPLETELY CONFIDENTIAL.

Please answer the following general questions about yourself.

1.

What is your student number? (If you do not know your student number,

please provide your initials.)

 

What is the name of your team?

 

How old are you? years

What is your year in college?
___Freshman ___Sophomore ___Junior ___Senior ___Graduate Student
Before joining this team, had you ever played on a basketball team before?

(Please check one.)
Yes No

 

 

If so, please indicate here what types of teams you have played on in the
past by checking the appropriate categories. Also, for each team that you
have played on, please indicate how long you were a member of the team:

FOR HOW LONG?

Pro-high school basketball team _____ seasons
_____ High school basketball team _____ seasons
_____ Intramural basketball team _____ seasons
_____ College basketball team seasons

Other(please explain: ) seasons

THIS IS THE FIRST AND

REMEMBER THAT

213

We are interested in estimating the ability level of each member of your team.
Please fill in the names of your team members and provide a rating of this
individual's ability to play basketball. Please provide your ratings using the
scale provided below. THESE RATINGS ARE FORrRESEARCH PURPOSES ONLY AND WILL
NOT BE SEEN BY ANY OF YOUR TEAM MEMBERS. Please be honest in your responses.

Very Poor Poor Average Good Very Good
Player Player Player Player Player
1 ------------- 2 --------------- 3 ------------- 4 --------------- S

PLAYER NAME:

 

 

 

 

 

 

 

 

 

 

214

Please rate how strongly you agree or disagree with the following statements,
with respect to your life in general. Circle the appropriate number after each

item based on the scale below.

Strongly Strongly
Disagree Disagree Uncertain Agree Agree
1 --------------- 2 -------------- 3 ------------- 4 --------------- 5
1. I'm always trying to figure myself out . . . . . . . 1 2 3 4 S
2. I'm concerned about my style of doing things . . . . 1 2 3 4 5
3. Generally, I'm not very aware of myself . . . . . . l 2 3 4 5
4. I reflect about myself a lot . . . . . . . . . . . . 1 2 3 4 5
5. I'm concerned about the way I present myself . . . . 1 2 3 4 S
6. I'm often the subject of my own fantasies . . . . . l 2 3 4 S
7. I never scrutinize myself . . . . . . . . . . . . . l 2 3 4 S
8. I'm self-conscious about the way I look . . . . . . 1 2 3 4 5
9. I'm generally attentive to my inner feelings . . . . l 2 3 4 S
10. I usually worry about making a good impression . . . 1 2. 3 4 5
11. I'm constantly examining my own motives . . . . . . 1 2 3 4 S

12. One of the last things that I do before I leave

the house is look in the mirror . . . . . . . . . . 1 2 3 4 S
13. I sometimes have the feeling that I'm off somewhere

watching myself . . . . . . . . . . . . . . . . . . 1 2 3 4 S
14. I'm concerned about what other people think of me. . l 2 3 4 5
15. I'm alert to changes in my mood . . . . . . . . . . I 2 3 4 S
16. I'm usually aware of my appearance . . . . . . . . . l 2 3 4 S

17. I'm aware of the way my mind works when I work
through a problem . . . . . . . . . . . . . . . . . l 2 3 4 5

215

There are many reasons why a team might win or lose a basketball game. This
study is concerned with six causes. You are asked on the remaining pages to
answer several questions related to each of these six causes. Please carefully
read the definition of the cause which is provided and answer the questions by
circling the number which you feel is most appropriate.

gauge a] - SELF ABILITY: Definition: Your own personal ability as a
basketball player may be a primary cause of the team's win or loss.

1. Is SELF ABILITY something that is:

Changeable Unchanging

2. Is SELF ABILITY something that is:

Variable Stable
Over Time Over Time
1 ----- 2 ----- 3 ----- 4 ----- S ----- 6 ----- 7 ----- 8 ----- 9

3. Is SELF ABILITY something that is:

Temporary Permanent

1 ----- 2 ----- 3 ----- 4 ----- S ..... 6 ----- 7 ----- 8 ----- 9
4. Is SILT ABILITY something that is:

Outside Inside
of You of You

5. Is SELF ABILITY something that:

Reflects en Reflects an
aspect of the aspect of
situation yourself

I ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- 8 ----- 9

6. Is 85L! ABILITY:

Something Something
about others ' about you

1 ----- 2 ----- 3 ----- 4 ----- s ----- 6 ----- 7 ----- Va ----- 9

216

CAUSE #2 - SELF EFFORT

Qéuse e2 - SELF EFFORT: Definition: The amount of effort that you personally
exert in a game may be a primary cause of the team's win or loss.

1. Is SELF EFFORT something that is:

Changeable Unchanging

2. Is SELF EFFORT something that is:

Variable Stable
Over Time Over Time
1 ----- 2 ----- 3 ----- 4 ----- S .-... 6 ----- 7 ----- 8 ----- 9

3. Is SELF EFFORT something that is:

Temporary Permanent

4. Is SELF EFFORT something that is:

Outside Inside
of You of You

5. Is SELF EFFORT something that:

Reflects an Reflects an
aspect of the aspect of
situation yourself

I ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- 8 ----- 9

6. Is SELF EFFORT:

Something Something
about others about you

1 ----- 2 ..... 3 ----- 4 ----- s ----- 6 ----- 7 ..... a ----- 9

217

CAUSE #3 - TEAM ABILITY
Cause a} - TEAM ABILITY: Defiinitign: The ability of other members of the

team to play basketball may be a primary cause of the team's win or
loss.

I. Is TEAH ABILITY something that is:

Changeable Unchanging

2. Is TEAH ABILITY something that is:

Variable Stable
Over Time Over Time
1 ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- 8 ----- 9

3. Is TEAH ABILITY something that is:

Temporary Permanent

'4. Is TEAM ABILITY something that is:

Outside Inside
of You of You

5. Is TEAH ABILITY something that:

Reflects an Reflects an
aspect of the aspect of
situation yourself

I ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- 8 ----- 9

6. Is TEAM ABILITY:

Something Something
about others about you

1 ----- 2 ----- 3 ----- 4 ----- s ----- 6 ----- 7 ----- a ----- 9

218

CAUSE #4 -

Cause #4 - TEAH EFFORT: The amount of
team may be a primary cause

I. Is TEAH EFFORT something that is:

Changeable

2. Is TEAM EFFORT something that is:

Variable
Over Time
1 ----- 2 ----- 3 ----- 4 -----
3. Is TEAM EFFORT something that is:
Temporary
1 ----- 2 ----- 3 ----- 4 -----
4. Is TEAK EFFORT something that is:
Outside
of You
1 ----- 2 ----- 3 ooooo 4 -----

5. Is TEAH EFFORT something that:
Reflects an

aspect of the
situation

18 TEAM EFFORT:

Something
about others

TEAH EFFORT

effort exerted by other members of the
of the team's win or loss.

Unchanging

Stable
Over Time

of You

Reflects an
aspect of
yourself

Something
about you

1 ----- 2 ----- 3 ----- 4 ..... s ----- 6 ----- 7 ----- a ----- 9

219

CAUSE #5 - ABILITY OF OTHER TEAM

C u s - ABILITY OF OTHER TEAM: The ability level of the opposing team may
be a primary cause of the team's win or loss.
1. Is the ABILITY OF THE OTHER TEAM something that is:
Changeable Unchanging
l ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- 8 ----- 9
2. Is the ABILITY OF THE OTHER TEAM something that is:
Variable Stable
Over Time Over Time
1 ----- 2 ----- 3 ----- 4 ----- S ----- 6 ----- 7 ----- 8 ----- 9
3. Is the ABILITY OF THE OTHER TEAM something that is:
Temporary Permanent
1 ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- 8 ----- 9
4. Is the ABILITY OF TEE OTHER TEAM something that is:
Outside Inside
of You of You
1 ----- 2 ----- 3 ----- 4 ----- S ----- 6 ----- 7 ----- 8 ----- 9
5. Is the ABILITY OF THE OTHER TEAM something that:
Reflects an Reflects an
aspect of the aspect of
situation yourself
I ----- 2 ----- 3 ----- 4 ----- 5 ----- 6 ----- 7 ----- B ----- 9
6. Is the ABILITY OF THE OTHER TEAM:
Something Something
about others about you
1 ----- 2 ----- 3 ----- 4 ----- S ----- 6 ----- 7 ----- 8 ----- 9

220

CAUSE #6 - LUCK
a - LUCK: Luck and other factors beyond the control of the team may be

a primary cause of the team's win or loss (for example. things like
the referees. the team's schedule, conditions in the gym. etc.).

Is LUCK something that is:

Changeable Unchanging

Is LUCK something that is:

Variable Stable
Over Time Over Time
1 ----- 2 ----- 3 ----- 4 ----- S ----- 6 ----- 7 ----- 8 ----- 9

Is LUCK something that is:

Temporary Permanent

Is LUCK something that is:

Outside Inside
of You of You

Is LUCK something that:

Reflects an Reflects an
aspect of the aspect of
situation yourself

1 ----- 2 ----- 3 ----- 4 ----- S ----- 6 ----- 7 ----- 8 ----- 9
Is LUCK:
Something Something
about others about you

1 ----- 2 ----- 3 ----- 4 ..... 5 ..... 6 ----- 7 ----- a ----- 9

221
GAME '1

STUDENT NUMBER (If you don't know. fill in your initials)

TEAM NAME

 

Please read each of the following questions carefully and circle the answer

which best expresses your feelings b u u v us e , using
the scale below. MANY OF TME QUESTIONS UILI. SOUND SIMILAR. BUT PLEASE

CAREFULLY CONSIDER AND answer. my QUESTION.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5
1. I feel pretty responsible for the team's performance

in this game . . . ..... . . . . . . . . . . . . l 2 3 4 S

2. I'm very happy with the way that most of the
other team members performed in this game . . . . . . l 2 3 4 S

3. I personally had very little to do with the team's
performance in this game . . . . . . . . . . . . . . l 2 3 4 S

4. If I hadn't performed the way I did. this game would
have turned out very differently . . . . . . . . . . l 2 3 4 5

S. In the game we just played. some of the members of
this team were lazy ..... . .......... . l 2 3 4 5

6. I personally had a lot to do with the team's
performance in this game . . . ..... . . . . . . I 2 3 4 S

7. I think that I performed very well in regard to
the game we just played . . . . . ..... . . . . . l 2 3 4 S

8. I'm very happy with the effort levels of other
members of the team in this game . . . . . . . . . . l 2 3 4 S

9. I think that I did a good job in this game . . . . . l 2 3 4 5

10. After this game. I feel that being a member of this
team is a real honor . . . . . . . . . . . . . . . . l 2 3 4 5

ll. I'm proud of how I performed in this game . . . . . .»l 2 3 4 5

12. I was very happy with the skill levels of most of
the other members of the team in this game . . . . . l 2 3 4 5

13. I played better than most of the other members of my
team in the game we just played . . . . . . . . . . . l 2 3 4 . 5

222

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

I4. I don't think that I was at all responsible for the
team's performance in this game . . . . . . . . . . . l 2 3 4 S

15. After this game. I feel very good about being a
member of this team . . . . . . . . . . . . . . . . . l 2 3 4 5

I6. I think that my performance in this game was pretty
good . . . . . . . . . . . . . . : . . . . . . . . . l 2 3 4 5

17. In the game we just played. I was one of the better
players on our team . . . . . . . . . . . . . . . . . I 2 3 4 5

Think about your team's performance and the outcome of the game you just
played. There are many reasons why your team performed the way it did in this
game, but we will focus here on six possible reasons. Please think about each
of these reasons and show how important you think that each reason was to the
outcome of the game by writing a percentage in the blank. THESE PERCENTAGES
MUST ADD TO 100‘.

THE TEAM'S PERFORMANCE IN THE GAME UE JUST PLAYED HAS BECAUSE OF:
4 My personal ABILITY (i.e., my skill level)
4 My personal EFFORT (i.e., how hard I tried)

4 The ABILITY of other members of the team (i.e., the
skill level of other members of the team)

6 The EFFORT of other members of the team (i.e., how hard
other team members tried)

4 The DIFFICULTY of the game (i.e., how good the other team
was)

t LUCK (i.e., factors over which the team had no control such
as the referees, the team's game schedule. conditions in
the gym. etc.)

'ro'm.’- 100 2

How would you compare your performance in the game you just played to that of
most of the other players on your team? (Please circle below.)

Much Worse Horse The Same Better Much Better
Than Others Than Others As Others Than Others Than Others

1 ................. 2 ._ ------------ 3 ------------ 4 --------------- s

223

The following questions ask you to think again about the reasons why your team
performed the way it did in the game you just played. Specifically, these
questions ask you to think about the ABILITY (i.e., skill level) and EFFORT
(i.e., energy put forth) of yourself and other team members, as well as the
DIFFICULTY of the game and the influence of LUCK and other chance factors.
Please read and answer each question carefully. using the following scale.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

l. I think that our team performed the way we did in
this game largely because of MY ABILITY as a
basketball player . . . . . . . . . . . . . . . . . . l 2 3 4 S

2. The team's performance in this game was largely due
to MY PERSONAL EFFORT . . . . . . . . . . . . . . . . 1 2 3 4 S

3. It was pretty likely that this game would turn out
the way it did because of the SKILL LEVEL OF THE
OTHER TEAM . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

4. I feel that my team's performance in this game
was greatly influenced by the EFFORTS OF OTHER
MEMBERS OF MY TEAM . . . . . . . . . . . . . . . . . l 2 3 4 S

S. I feel that our team's performance in this game
was greatly influenced by HOH HARD I TRIED . . . . . l 2 3 4 S

6. The ABILITY level of OTHER MEMBERS OF MY TEAM was
a major factor in my team's performance in this
game . . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

7. In regard to this game. the game turned out the way
it did because of the DIFFICULTY level of the game . I 2 3 4 S

8. MY EFFORT LEVEL had a big impact on my team's
performance in this game . . . . . . . . . . . . 1 2 3 4 S

9. This game turned out the way it did mostly because of
the SKILL LEVEL OF THE OTHER TEAM . . . . . . . . . . l 2 3 4 5

10. I think that the outcome of this game had a lot to do
with the SKILL LEVEL of OTHER MEMBERS OF MY TEAM . . l 2 3 4 5

11. I think that the outcome of this game had a lot to
do with MY SKILL LEVEL . . . . . . . . . . . . . . . l 2 3 4 S

12. The ABILITY of OTHER MEMBERS OF MY TEAM had
a big influence on my team's performance in
this game . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

224

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

13. I would say that it was mostly FACTORS BEYOND OUR
CONTROL (such as the referees. the team's game
schedule. the conditions of the gym. etc.) that
caused our team to perform the way that we did in
this game . . . . . . . . . . . . . . . . . . . . . . l 2 3 4

14. The team performed the way it did in this game
largely because of MY EFFORT LEVEL . . . . . . . . . l 2 3 4

15. My team performed the way it did in this game
mostly because of the SKILL LEVELS of OTHER MEMBERS
OF MY TEAM . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

16. MY ABILITY had a big influence on the team's .
performance in this game . . . . . . . . . . . . . . l 2 3 4 S

17. My team performed the way it did in this game
mostly because of the EFFORT levels of OTHER MEMBERS
OF MY TEAM . . . . . . . . . . . . . . . . . . . . . 1 2 3 4 S

18. MY ABILITY was a major factor in the team's
performance in this game . . . . . . . . . . . . . . l 2 3 4 5

19. The EFFORT level of OTHER MEMBERS OF MY TEAM had a
big impact on my team's performance in this game . . l 2 3 4 S

20. The outcome of this game had a lot to do with
MY EFFORT LEVEL . . . . . . . . . . . . . . . . . . . l 2 3 4 5

21. Our team performed the way it did in this game
mostly because of LUCK (for example. because of things
like the referees. the team's game schedule. the
conditions in the gym. etc.). . . . . . . . . . . . . l 2 3 4 S

22. My team's performance in this game was mostly
due to the EFFORTS of OTHER MEMBERS OF MY TEAM . . . l 2 3 4 5

23. I think that my team's performance in this game
was largely due to the SKILL LEVEL of OTHER MEMBERS
OF MY TEAM . . . . . . . . . . . . . . . . . . . . . 1 2 3 4 S

24. CHANCE (things like the referees. the team's game
schedule. the conditions of the gym. etc.) had a lot
to do with our team's performance in this game . . . 1 2 3 4 5

25. The outcome of this game had a lot to do with the
EFFORT level of OTHER MEMBERS OF MY TEAM . . . . . . l . 2 3 4 S

225

Please answer each of the following questions in relation to the team which you
have just finished playing.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- S
l. The team we just played was very good . . . . . . . . l 2 3 4 S
2. The team we just played did not present us with much
of a challenge . . . . . . . . . . . . . . . . . . . l 2 3 4 S
3. I would say that the skill level of the team we just
played is very high . . . . . . . . . . . . . . . . l 2 3 4 5
Do you know which team you will play next? Yes i No

 

If you in know which team you will play next. please answer the following
two questions by circling the appropriate number.

I. How difficult do you believe that it will be to beat this team?

Very Somewhat Very
Difficult Difficult Difficult Easy Easy
1 ------------ 2 --------------- 3 ------------ 4 ------------ S
2. What are your team's chances of ginning against this team?
Very Poor Poor SOt-SOB Good Very Good
Chance Chance Chance Chance Chance

226

SELF-SET GOALS
Some people set very specific personal goals for themselves before they perform
a task. Please answer the following questions related to your personal goals.

BEFORE THE GAME THAT YOU JUST PLAYED. HAD YOU SET A GOAL FOR YOURSELF FOR:

 

 

 

 

If Yes. whg;
In his W
The number of points you wanted to score? points
The number of rebounds you wanted to make? rebounds
The number of assists you wanted to make? assists

 

 

In sports. teams have the goal of winning each game that they play. Assuming
that your team's goal is to win the next game that you play. please indicate
your feelings below by circling the most appropriate response.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5

IN REGARD TO OUR TEAM'S COAL T0 UIN THE NEXT GAME THAT HE PLAY:
1. It's hard to take this goal seriously . . . . . . . . l 2 3 4 5

2. It's unrealistic for us to expect to reach this
goal . . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

3. It is quite likely that this goal may need to be
revised. depending on how things go . . . . . . . . . l 2 3 4 S

4. Quite frankly. I don't care if we achieve this
goal or not . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

S. I am strongly committed to pursuing this goal . . . . l 2 3 4 S

6. It wouldn't take much to make me abandon this
goal . . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

7. I think that this goal is a good goal to shoot
for . . . . . . . . . . . . . . . . . . . . .

 

227

CAME l2

STUDENT NUMBER (If you don't know. fill in your initials)

TEAM NAME

 

Please read each of the following questions carefully and circle the answer
which best expresses your feelings ghggg_Shg_3gmg_xgg_h§!2_lggg_nlgx£g' using
the scale below. MANY OF THE QUESTIONS HILL SOUND SIMILAR. BUT PLEASE
CAREFULLY CONSIDER AND ANSWER ngznx QUESTION.

Strongly . Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5
l. I feel pretty responsible for the team's performance

in this game ............. . . . . . . . l 2 3 4 5

2. I'm very happy with the way that most of the
other team members performed in this game . . . . . . l 2 3 4 S

3. I personally had very little to do with the team's
performance in this game . . . . . . . . . .'. . . . l 2 3 4 S

4. If I hadn't performed the way I did. this game would
have turned out very differently . . . . . . . . . . l 2 3 4 S

S. In the game we just played. some of the members of
this team were lazy . . . . . . . . . . . . . . . . . l 2 3 4 5

6. I personally had a lot to do with the team's
performance in this game . . . . . . . . . . . . . . l 2 3 4 5

7. I think that I performed very well in regard to
the game we just played . . ...... . . . . . . . l 2 3 4 5

8. I'm very happy with the effort levels of other
members of the team in this game . . . . . . . . . . l 2 3 4 5

9. I think that I did a good job in this game . . . . . l 2 3 4 S

10. After this game. I feel that being a member of this
ten 1. ‘ t..1 honor O O O 0 O O O O O O I O O O O O 1 2 3 ‘ s

11. I'm proud of how I performed in this game . . . . . . l 2 3 4 5

I2. I was very happy with the skill levels of most of
the other members of the team in this game . . . . . l 2 3 4 5

I3. I played better than most of the other members of my
team in the game we just played . . . . . . . . . . . l 2 3 4 5

 

228

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 ............... 5

14. I don't think that I was at all responsible for the
team's performance in this game . . . . . . . . . . . l 2 3 4 S

15. After this game. I feel very good about being a
member of this team . . . . . . . . . . . . . . l 2 3 4 S

16. I think that my performance in this game was pretty
good . . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

17. In the game we just played. I was one of the better
players on our team . . . . . . . . . . . . . . . . . l 2 3 4 5

Think about your team's performance and the outcome of the game you just
played. There are many reasons why your team performed the way it did in this
game. but we will focus here on six possible reasons. Please think about each
of these reasons and show how important you think that each reason was to the
outcome of the game by writing a percentage in the blank. THESE PERCENTAGES
MUST ADD TO 1002.

THE TEAM'S PERFORMANCE IN THE GAME HE JUST PLAYED HAS BECAUSE OF:

6 My personal ABILITY (i.e.. my skill level)

 

Q My personal EFFORT (i.e.. how hard I tried)

4 The ABILITY of other members of the team (i.e.. the
skill level of other members of the team)

6 The EFFORT of other members of the team (i.e.. how hard
other team members tried)

6 The DIFFICULTY of the game (i.e.. how good the other team
was)

t LUCK (i.e.. factors over which the team had no control such
as the referees. the team's game schedule. conditions in
the gym. etc.)

TOTAL - 100 B

How would you compare your performance in the game you just played to that of
most of the other players on your team? (Please circle below.)

Much Horse Horse The Same Better Much Better
Than Others Than Others As Others Than Others Than Others

1 ................. 2 ............. 3 ...... a. ............... s

229

The following questions ask you to think again about the reasons why your team
performed the way it did in the game you just played. Specifically. these
questions ask you to think about the ABILITY (i.e.. skill level) and EFFORT
(i.e.. energy put forth) of yourself and other team members. as well as the
DIFFICULTY of the game and the influence of LUCK and other chance factors.
Please read and answer each question carefully. using the following scale.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- S

l. I think that our team performed the way we did in
this game largely because of MY ABILITY as a
basketball player . . . . . . . . . . . . . . . . . . l 2 3 4 S

2. The team's performance in this game was largely due
to MY PERSONAL EFFORT . . . . . . . . . . . . . . . . I 2 3 4 5

3. It was pretty likely that this game would turn out
the way it did because of the SKILL LEVEL OF THE
OTHER TEAM . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

4. I feel that my team's performance in this game
was greatly influenced by the EFFORTS OF OTHER
MEMBERS OF MY TEAM . . . . . . . . . . . . . . . . . l 2 3 4 S

S. I feel that our team‘s performance in this game
was greatly influenced by HG" HARD I TRIED . . . . . I 2 3 4 S

6. The ABILITY level of OTHER MEMBERS OF MY TEAM was
a major factor in my team's performance in this
game . . . . . . . . . . . . . . . . . . . . . . . . I 2 3 4 S

7. In regard to this game. the game turned out the way
it did because of the DIFFICULTY level of the game . l 2 3 4 5

8. MY EFFORT LEVEL had a big impact on my team' s
performance in this game . . . . . . . . . . . . I 2 3 4 S

9. This game turned out the way it did mostly because of
the SKILL LEVEL OF THE OTHER TEAM. . . . . . . . . . l 2 3 4 5

10. I think that the outcome of this game had a lot to do
with the SKILL LEVEL of OTHER MEMBERS OF MY TEAM . ..l 2 3 4 5

II. I think that the outcome of this game had a lot to
do with MY SKILL LEVEL . . . . . . . . . . . . . . . I 2 3 4 S

12. The ABILITY of OTHER MEMBERS OF MY TEAM had
a big influence on my team's performance in
tht. 8a.. a e e e e e e e e e e e e eeeee e e e 1 2 3 a 5

230

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 .......: ------- 5

I3. I would say that it was mostly FACTORS BEYOND OUR
CONTROL (such as the referees. the team's game
schedule. the conditions of the gym. etc.) that
caused our team to perform the way that we did in
this game . . . . . . . . . . . . . . . . . . . .'. . l 2 3 4

l4. The team performed the way it did in this game
largely because of MY EFFORT LEVEL . . . . . . . . . I 2 3 4

IS. My team performed the way it did in this game
mostly because of the SKILL LEVELS of OTHER MEMBERS

OF MY TEAM . . . . . . . . . . . . . . . . . . . . . l 2 3 4 5

16. MY ABILITY had a big influence on the team':
performance in this game . . . . . . . . . . . . . . I 2 3 4 5

17. My team performed the way it did in this game
mostly because of the EFFORT levels of OTHER MEMBERS
OF MY TEAM . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

18. MY ABILITY was a major factor in the team's
performance in this game . . . . . . . . . . . . . . I 2 3 4 5

19. The EFFORT level of OTHER MEMBERS OF MY TEAM had a
'big impact on my team's performance in this game . . I 2 3 4 S

20. The outcome of this game had a lot to do with
MY EFFORT LEVEL . . . . . . . . . . . . . . . . . . . I 2 3 4 5

21. Our team performed the way it did in this game
mostly because of LUCK (for example. because of things
like the referees. the team's game schedule. the
conditions in the gym. etc.). . . . . . . . . . . . . I 2 3 4 S

22. My team's performance in this game was mostly
due to the EFFORTS of OTHER MEMBERS OF MY TEAM . . . l 2 3 4 S

23. I think that my team's performance in this game
was largely due to the SKILL LEVEL of OTHER MEMBERS
or “Y rm 0 O O O O O O O O O O I I O O O O O O O I 1 2 3 a S

24. CHANCE (things like the referees. the team's game
schedule. the conditions of the gym. etc.) had a lot
to do with our team's performance in this game . . . I 2 3 4 S

25. The outcome of this game had a lot to do with the
EFFORT level of OTHER MEMBERS OF MY TEAM . . . . . . I 2 3 4 5

231

Please answer each of the following questions in relation to the team which you
have just finished playing.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- 5
l. The team we just played was very good . . . . . . . . I 2 3 4 S
2. The team we just played did not present us with much
of a challenge . . . . . . . . . . . . . . . . . . . I 2 3 4 5
3. I would say that the skill level of the team we just
played is very high . . . . . . . . . . . . . . . . I 2 3 4 5
Do you know which team you will play next? Yes No

 

 

If you fig know which team you will play next. please answer the following
two questions by circling the appropriate number.

I. How difficult do you believe that it will be to beat this team?

very Somewhat Very
Difficult Difficult Difficult Easy Easy
I ------------ 2 --------------- 3 ------------ 4 ------------ S

2. What are your team's chances of ginning against this team?

Very Poor Poor 504-504 Good Very Good
Chance Chance Chance Chance Chance
1 ------------ 2 ----------------- 3 ------------ 4 ------------ 5

 

232

SELF-SET GOALS
Some people set very specific personal goals for themselves before they perform
a task. Please answer the following questions related to your personal goals.

BEFORE THE CAME THAT YOU JUST PLAYED. HAD YOU SET A GOAL FOR YOURSELF FOR:

 

 

 

 

We;
mile W1
The number of points you wanted to score? points
The number of rebounds you wanted to make? rebounds
The number of assists you wanted to make? assists

 

In sports. teams have the goal of winning each game that they play. Assuming
that your team's goal is to win the next game that you play. please indicate
your feelings below by circling the most appropriate response.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 ------- . -------- 5

IN REGARD TO OUR TEAM'S GOAL TO WIN THE NEXT GAME THAT HE PLAY:
I. It's hard to take this goal seriously . . . . . . . . l 2 3 4 S

2. It's unrealistic for us to expect to reach this
goal . . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

3. It is quite likely that this goal may need to be
revised. depending on how things go . . . . . . . . . l 2 3 4 S

4. Quite frankly. I don't care if we achieve this
goal or not . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

5. I am strongly committed to pursuing this goal . . . . l 2 3 4 S

6. It wouldn't take much to make me abandon this
goal . . . . . . . . . . . . . . . . . . . . . . . . l 2 3 4 S

7. I think that this goal is a good goal to shoot
for . . . . . . . . . . . . . . . . . . . . . .

233

GAME #3

STUDENT NUMBER (If you don't know. fill in your initials)

TEAM NAME

 

Please answer each of the following questions in relation to the team which you
have just finished playing.

Strongly Neither Agree Strongly
Disagree Disagree Nor Disagree Agree Agree
1 --------------- 2 --------------- 3 --------------- 4 --------------- S
l. The team we just played was very good . . . . . . . . I 2 3 4 S

2. The team we just played did not present us with much
of a challenge . . . . . . . . . . . . . . . . . . . l 2 3 4 S

3. I would say that the skill level of the team we just
played is very high . . . . . . . . . . . . . . . . l 2 3 4 S

SELF-SET GOALS
Some people set very specific personal goals for themselves before they perform
a task. Please answer the following questions related to your personal goals.

BEFORE THE GAME THAT YOU JUST PLAYED. HAD YOU SET A GOAL FOR YOURSELF FOR:

 

 

1M
In Us W
The number of points you wanted to score? points
The number of rebounds you wanted to make? rebounds

 

 

The number of assists you wanted to make? assists

234

We are interested in estimating the ability level of each member of your team.
Please fill in the names of your team members and provide a rating of this
individual's ability to play basketball. Please provide your ratings using the

 

“819 Proflded below W
W. Please be honest in your responses.
Very Poor Poor Average Good Very Good
Player Player Player Player Player
I ---------------- 2 -------------- 3 -------------- 4 ---------------- S
PLAYER NAME:
I 2 3 4 5
l 2 3 4 5

 

 

 

 

 

 

 

 

 

LIST OF REFERENCES

LIST OF REFERENCES

Adams. J.S. (1965). Inequity in social exchange. In

L. Berkowitz (Ed.). Adyenses_in_sansrimsnta1_§osial_nsxshr
giggy. (Vol. 2, pp. 267-299.). New Yorszcademic Press.

Adler. S. (1980). Self-esteem and causal attributions for job

satisfaction and dissatisfaction. lgn:nni_2fi_npniigg
Earsholsgx. 62. 327-332.

Albanese. R.. & Van Fleet. D.D. (1985). Rational behavior in

groups: The free-riding tendency. Agnflgmy_g£iflnnnggmgn§
Review. 1Q(2). 244-255. .

Anderson. D.C.. Crowell, C.R.. Doman, M.. 6 Howard. 6.8. (1988).
Performance posting. goal setting. and activity-contingent praise

as applied to a university hockey team. lgn;nnl_nf_finniigg
Psychology. 13(1). 87-95.

Ashford, S.J.. & Cummings. L.L. (1982). Feedback as an individual
resource: Personal strategies of creating information. Unpublished
manuscript. Northwestern University.

Austin. J.J.. & Bobko. P. (1985). Goal-setting theory: unex-

plored areas and future research needs. Jgn;nnl_nfi_nnniigd
1mm. 1%. 289-308.

Bandura, A. (1977). Self-efficacy: Toward a unifying theory
of behavioral change. Egyghgingigni_figxigx. 54. 91-125.

Bargh, J.A. (1982). Attention and automaticity in the processing

of self-relevant information. I2srnaI_2f.£sra2nali£x_and_§221al
Earsholssx. 43. 425-436.

Becker, L.J. (1978). Joint effects of feedback and goal setting
on performance: A field study of residential energy conserva-

tion. I2sraal_of_622112d_£sxsholosx. 63. 428-433.

Berkowitz. L., & Levy, B.I. (1956). Pride in group performance
and group task motivation.

Earsholosx..23. 300-306.

235

236

Bernstein. B.L.. 6 Lecomte, C. (1979). Supervisory-type feedback
effects: Feedback discrepancy level. trainee psychological dif-
ferentiation. and immediate responses.

2912211212231. 2.6. 295-303.

Brief. A.P.. Aldag. R.J., 6 Van Sell. M. (1977). Moderators
of the relationship between self and supervisor evaluations

of Job performance. WW.
1Q, 124-134. .

Buller. P.F.. 6 Bell. C.H. (1986). Effects of team building and
goal setting on productivity: A field experiment. Agnggmx_2f

W21. 22(2). 305-328.

Campion. M.A., 6 Lord. R.G. (1982). A control systems concep-
tualization of the goal setting and changing process.
- €t,‘!5_ '6' : e ' . ‘ v, E, 265'287.

      

   

   

1! z! l--!15.! ’ 9!!!-_!

Carver. C.S. (1974). Facilitation of physical aggression through
obj active self-awareness. We].

231751121221. 10. 365-370.

Carver. C.S. (1975). Physical aggression as a function of objec—
tive self-awareness and attitudes toward punishment. lnnxnni

W21. 11. 510-519.

Carver. C.S. (1979). A cybernetic model of self-attention. innrnnl
WM. 3.7.. 1251-1281.

Carver. C.S., Blaney, P.H.. 6 Scheier. M.F. (1979). Reassertion and
giving up: The interactive role of self-directed attention and

outcome eXPectancy. MW.
31, 1859-1870.

Carver. C.S.. 5: Scheier. M.F. (1981). W-

n' t o v . New York:
Springer-Verlag.

Carver. C.S.. 6 Scheier. M.F. (1982). Control theory: A useful
conceptual framework for personality - social. clinical, and

health psychology. W. 22. 111-135.

Chacko, T.I.. 6 McElroy, J.C. (1983). The cognitive component in
Locke's theory of goal setting: Suggestive evidence for a

causal attribution interpretation. Agnggmy_gf_nnnnggmgn§
192ml. 26(1). 104-118.

Cohen. J. (1977). MW
figignngg. New York: Academic Press.

237

Downey, H.K., Chacko, T.I.. 6 McElroy, J.C. (1979). Attribution
of the "causes" of performance: A constructive. quasi-longi-
tudinal replication of the Staw (1975) study. Organizatignnl

W. 24. 287-299.

Edwards. W. (1961). Behavioral decision theory. Annnni_3gxigg
W. 12. 473-493.

Erez, M. (1977). Feedback: A necessary condition for the goal

setting-performance relationship. lgn:nni_n£_nnnligd
Emholm. 9.2. 624-627.

Fenigstein, A. Scheier. M.E.. 6 Buss. A.H. (1975). Public and
private self-consciousness: Assessment and theory. lgnxnnl

”MW. 43. 522-527.

Fitch, G. (1970). Effects of self-esteem. perceived performance.

and choice on causal attributions. lgn:nni_n£_£gxfinnnii§y_nng
W. 16. 311-315-

Forsyth. D.R., 6 Schlenker, B.R. (1977). Attributing the causes
of group performance: Effects of performance. quality. task

importance. and future testing. innxn§1_gfi_£g;§gngligy, 4;,
220-236.

Forward. J., 6 Zander. A. (1971). Choice of unattainable group

goals and effects on performance. Qzaanizational_flshaxior_and
MW. 6. 184-199.

Harkins, 8.6., 6 Petty. R.E. (1982). Effects of task difficulty

and task uniqueness on social Ioafing. J23;nni_nf_£g1§2nnii§y
MW. 43. 1214-1229.

Heider. F. (1958). W.
New York:Wiley.

Hollenbeck. J.R. (1989). Control theory and the perception of
work environments: The effects of focus of attention on
affective and behavioral reactions to work. ngnnizngignni

Was. (in prose).

Hollenbeck. J.R.. 6 Brief. A.P. (1987). Self-regulation in the
workplace: Towards a unified approach to understanding worker
attitudes and behavior. In R.S. Schuler (Ed.). Bending§_in

W. St Paul. MNzweat
Publishing.

Hollenbeck. J.R.. Klein. H.J.. O'Leary. A.M.. 6 Wright. P.M.
(1989). Investigation of the construct validity of a

self-report measure of goal commitment. lgn:nni_nf_finnligg
231911212311. 111(6). 951-956.

238

Hollenbeck. J.R.. 6 Williams. C.R. (1987). Coal importance, self-

focus, and the goal-setting process. J93:nni_n£_nnnligg
Earsholosx. 22(2). 204-211.

Ilgen. D.R., 6 Hamstra, B.W. (1972). Performance satisfaction as a
function of the difference between expected and reported performance

at five levels of reported performance. Qzgnnizngignn1_nennxinz_nng
Human_£erf2rman2s. Z. 359-370.

Ilgen. D.R., Shapiro. J., Salas. E.. 6 Weiss. H. (1987). Func-
tions of group goals: Possible generalizations from indivi-
duals and inferences for training. Unpublished manuscript.
Michigan State University.

Ivancevich. J.M. (1974). Changes in performance in a management by

objectives Program. Asministratixe_§sienses_92arserlx. 12. 563-574.

Johnston, W.A. (1967). Individual performance and self evalua-

tion in a simulated team. QIaanizational_Bshaxisr_and_flsman
Performanss. 2. 309-328.

Kanfer. R.C. (1989). Motivation theory and industrial/organi-
zational psychology. In M.D. Dunette (Ed.). handhnnk_nf

QIsanizstional_£§12holssx (Vol. 1). Palo Alto. CA:Conau1r
ing Psychologists Press (in press).

Kelley. H.H. (1967). Attribution theory in social psychology.
Nehraska_fizmnosism_oa_uosixation. 12. 192-238.

Kelley. H.H.. 6 Micela, J. (1980). Attribution theory and
research. In M.R. Rozenweig 6 L.W. Porter (Eds.). Annnni

Iggigg_gf_n§yghnlggy (Vol. 31. pp. 457-501). Palo Alto.CA:

Annual Reviews.

Kerlinger. F.N. (1964). E2sadations_o£_Bshaxioral_Besearsh.
Holt, Rinehart and Winston, Inc.:New York.

Kerr. N.L. (1983). Motivation losses in small groups: A social

dilemma analysis. I2srna1_2f_2eraona1i£x_ans_fiosial_£§xsholosx.
42, 819-828.

Kerr. N.L.. 6 Bruun, S.E. (1983). Dispensability of member effort
and group motivation losses: Free-rider effects. lgn:nni_nf

2ars2naIi£x_and_fiosial_2§xsholozx. 44. 78-94.

Kerr, N.L.. 6 MacCoun, R.J. (1985). Role expectations in social
dilemmas: Sex roles and task motivation in groups. inn;nni_g£

Personali£1_aad_§22ia1_£§xsbolosx. &2(6). 1547-1556.

Kim. J.S.. 6 Schuler. R.S. (I979). The nature of the task as a
moderator of the relationship between extrinsic feedback and employee

reaponaea. Assessx_2f_nanassment_lssrnal. 22. 157-162.

239

Klein. H.J. (1989). An integrated control theory model of work
motivation. Asaden1_2£_Manasemant_Bexiss. 14(2). 150-177.

Klein. H.J.. 6 Mulvey, P.W. (1988). The setting of goals in group
settings: An investigation of group and goal processes. Unpub-
lished manuscript. Ohio State University.

Koch. J.L. (1979). Effects of goal specificity and performance feed-
back to work groups on peer leadership. performance. and attitudes.

Emman_Belatigns. 32(10), 819-840.

Kukla, A. (1972). Attribution determinants of achievement-
related behavior.

Earsholosx. 21. 166-174.
Landy. F.J. (1985). Ihe_ns1sholosx_2£_xork_hshaxior (3rd ed ).

Homewood. IL:Dorsey Press.

Landy, F.J.. 6 Becker. W. (1982). Adaptive motivation theory.

In Aanaal_Bs22rt_t2_Qffise.0f.flaxal_fiesearsh. University Park.
PA:Department of Psychology.

Landy. F.J. 6 Becker. L.J. (1987). Motivation theory recon-
sidered. Beaearsh_ia_Qrsaniaatioaal_nshszior. 2. 1-38.

Latane. B., Williams. K.. 6 Harkins, S.G. (1979). Many hands make
light the work: The causes and consequences of social Ioafing.

MW. 37.. 822-832.

Latham. G.P.. 6 Locke. E.A. (1975). Increasing productivity with
decreasing time limits: A field replication of Parkinson's

law- I22Inal_2£_Annlisfl_£§1£h212£¥. £9. 524-525.

Lawrence. L.C.. 6 Smith. P.C. (1955). Group decision and

employee participation. l2srnal_2£_8221129_2s12h21221. 32.
334-337.

Lichtenstein, E.H., 6 Brewer. W.F. (1980). Memory for goal-
directed events. Qggnigixg_£nyghningy, 12. 412-445.

Locke. E.A. (1967). Relationship of success and expectation to

affect on goal-seeking tasks. l2srnal_2£_£ersonali£x_and
529151.23xsh21281. 1. 125-134.

Locke. E. A. ,Cartledge. N. D. 6 Knerr, C. S. (1970). Studies of
the relationship between satisfaction. goal setting. and per-
formance. . : . - : _ , a
135-158.

 

Locke. E.A.. Shaw, K.N.. Saari, L.M.. 6 Latham. G.P. (1981).
Coal setting and task performance: 1969-1980. Rnyghniggignl
m: E: 125-152-

240

Logan. 6.0. (1979). On the use of concurrent memory load to measure
attention and automaticity. e '

Hsaan_2erssntion.and_£srf2rmanss. S. 189-207.

Lord, R.G.. 6 Hanges. P.J. (1987). A control systems model of
organizational motivation: Theoretical development and

applied implications. fighnxigzni_§gigngg. 22, 161-178.

Lord. R.G.. 6 Kernan, M.C. (1987). Scripts as determinants of

purposeful behavior in organizations. Agnggny_gfiflnnnggn§nn
Rules. .12. 265-277.

Lott. B.E.. 6 Lott. A.J: (1960). The formation of positive
attitudes toward group members. Jnn;nni_nf_nhnnznni_§nd
Sssial_E§1£hslszx. 61. 297-300.

Lott. B.E.. 6 Lott. A.J. (1966). Group cohesiveness and indivi-
dual training. MW. 51. 61-73.

Matsui, T., Kakuyama, T., 6 Onglatco. M.L.U. (1987). Effects of
goals and feedback on performance in groups. lnnxnni_nf_Anniigd
2mm. 12(3). l107-415.

McMahan. I.D. (1973). Relationships between causal attributions

and exoectancy of success. We].
2.319.119.1232. 2.6. 108-114.

Medow, H.. 6 Zander. A. (1965). Aspirations for group chosen by

central and peripheral members. 1nn:nni_gf_£gznnnnii§y_nng
W. .1. 224-228.

Mento, A.J.. Cartledge, N.D., 6 Locke. E.A. (1980). Maryland vs.
Michigan vs. Minnesota: Another look at the relationship of
expectancy and goal difficulty to task performance. Qggnnizn-

W. 2.5.. 419-440.

Mento, A.J.. Steel, R.P.. 6 Karren, R.J. (1987). A meta-analytic
study of the effects of goal setting on task performance: 1966-1984.
'-1 ; 01-: 3‘1:V° :10 1-31:4 " ‘ 01 '9 --se. 12. 52-83.

   

Miller. G.A.. Galanter, E.. 6 Pribram. K.H. (1960). Einn§_nn§
ghg_§;;ng;n;g_nf_hghnxing. New York:Holt. Rinehart. 6 Winston.

Miller. D.T., 6 Ross. M. (1975). Self-serving biases in the
attribution of causality: Fact or fiction? Efiyghningignl

331123.13. 62(2). 213-225-

Mitchell. T.R. (1982). Expectancy-value models in organizational

psycholosy- In N.T. Feather (Ed). W
c -v e s . Hillsdale. NJzLawrence

Erlbaum Associates.

241

Mitchell. T.R.. 6 Silver. U.S. (1990). Individual and group goals
when workers are interdependent: Effects on task strategies and

performance. MIDW- 1§(2). 135-193-

Morran. D.R., 6 Stockton. R.A. (1980). Effect of self-concept on
group member reception of positive and negative feedback. lgnxnnl

Wm. 21. 260-267.

Nadler. D.A. (I979). The effects of feedback on task group
behavior: A review of the experimental research. ngnnizn-

12W. 23.. 309-338.

Naylor, J.D.. Pritchard. R.D., 6 Ilgen. D.R. (1980). A_;hgn:1
Wagons. New York=Academic Press-

Nunnally. J.C. (1978). £5yghnmg§1ig_1hggzx. McGraw-Hill:New York.

Pearson. G.A.L. (1987). Participative goal setting as a strategy
for improving performance and job satisfaction: A longitudinal
evaluation with railway truck maintenance gangs. flnmnn_figin§ign§,
4Q. 473-488.

Pinder. C. (1984). Hg;k_nn§ixn§inn. Glenview. IL:Scott.
Foresman.

Powers. WT. (1973). WWW.
Chicago:Aldine.

Price. K.H. (1987). Decision responsibility. task responsibility.

identifiability. and social Ioafing. Q1gnnizngignni_fighnginn_nng
BMW. 4.9. 330-345.

Pritchard, R.D., Jones. S.D.. Roth, P.L., Stuebing, K.K.. 6 Ekeberg, S.E.
(1988). Effects of group feedback. goal setting, and incentives on

organizational Productiflt)’. W. 13(2).
337-358.

Pyszczynski. T.A.. 6 Greenberg, J. (1981). Role of disconfirmed
expectancies in the instigation of attributional processing. Jouznal

Wm. 49(1). 31-38.

Raven. B., 6 Rietsma. J. (1957). The effects of varied clarity
of group goals and group path upon the individual and his

relation to his group. flnmnn_3gin;innn, 1Q. 29-44.

Russell. D. (1982). The causal dimension scale: A measure of how

individuals perceive causes. WWW
26121121221. 42(6), 1137-1145.

242

Salancik. C.R. (1977). Commitment and the control of organizational
behavior and belief. In B.M. Staw 6 C.R. Salancik (Eds.). flex
Diresti2na_in_Qrsanizational_nehazior (pp. 1-53). Chicago=

St. Clair Press.

Schank. R.. & Abelson. R. (1977). Sstints._nlansl_ssalsa_and
nnggzggnnging. New York:Halsted.

Scheier. M.F., 6 Carver, C.S. (1980a). Self-directed attention
and the comparison of self with standards. Unpublished manu-
script.

Scheier. M.F., 6 Carver, C.S. (1980b). Effects of dispositional
self-consciousness and prior outcomes on persistence at a
related task. Unpublished manuscript.

Scheier. M.F., 6 Carver. C.S. (1982). Cognition. affect. and
self-regulation. In M.S. Clark and S.T. Fiske (Eds.).
Affgg§_§nd_§ggnigign Hillsdale. NJ:Lawrence Erlbaum
Associates Publishers.

Scheier. C.S.. Fenigstein, A.. 6 Buss. A.H. (1974). Self-awareness

and physical aggression. I2srnal_2f_Eaparimental_fiosia1_£sysh2logy.
19, 264-273.

Schlenker. B.R. (1975). Group members' attributions of responsi-

bility for prior group performance. 3gp;gggn§n§igg_fig§g§;gh
in_§osial_2§1£hslsgx. 6. 96-108.

Schlenker. B.R., 6 Miller. R.S. (1977). Group cohesiveness as a
determinant of egocentric perceptions in cooperative groups.

Human_Belations. 32. 1039-1055.

Schlenker. B.R.. Soraci, S.. 6 McCarthy. B. (1976). Self-esteem
and group performance as determinants of egocentric percep-

tions in coopertive groups. flnmnn_3gin§ign§. 22. 1163-1176.

Shiffrin, R.M.. 6 Schneider. W. (1977). Controlled and automatic
human information processing:II. Perceptual learning. auto-

matic attending, and a general theory. Egyghgiggigni_flggigg.
84, 127-190.

Smith. K. G. .Locke, E. A. 6 Barry. D. (1990). Goal setting. plan-
ning. and organizational performance: An experimental simulation.
. : . : . ; - -- .36. 118-134.

 

Staw. B.M. (1975). Attribution of the "causes" of performance:
A general alternative interpretation of cross-sectional

research on organizations. Qraanizational_fiehazior_and_fluman
Performance. 13. 414-432.

243

Steers. R.M.. 6 Porter. L.W. (1974). The role of task-goal attributes
in employee performance. WW. 21(7). 434-452.

Steiner. I.D. (1972). W- New

Yorszcademic Press.

Stipek. D.J. (1983). A developmental analysis of pride and shame.
. 26. 42-54.

Taylor, M.S.. Fisher. C.D., 6 Ilgen. D.R. (1984). Individuals'
reactions to performance feedback in organizations: A control

theory perspective. W
W. 2. 81-124.

Taylor. S.E.. 6 Fiske, S.T. (1978). Salience. attention. and attri-
bution: Top of the head phenomena. In L Berkowitz (Ed.). Agzgnggg

WWW (Vol. 11. pp. 249-288). New York:

Academic Press.
Vroom. V.H. (1964). Wien- New YorktliileY-

Weiner. B. (1972). WWW
gngnigign. Chicagonarkham.

Weiner, B. (1979). A theory of motivation for some classroom

caperiences. W. 11. 3-25.

Weiner, B. (1982). The emotional consequences of causal ascrip-
tions. In M.S. Clark 6 S.T. Fiske (Eds.).

Wales...
Wings (pp-185-200).
Hillsdale. NJzErlbaum.

Weiner. B. (1985). An attributional theory of achievement moti-

vation and emotion. Efiyghniggigni_figxigg. 22(4). 548-573.

Weiner, B. (1985). "Spontaneous" causal thinking. Efiyghgingigni
Bulletin. 21(1). 74-84.

Weiner. B., Heckhausen, H.. Meyer. N.D., 6 Cook. R.E. (1972).
Causal ascriptions and achievement behavior: A conceptual
analysis of effort and reanalysis of locus of control.

MW. 21. 239-248.

Weiner, B., Nierenberg, R.. 6 Goldstein, M. (1976). Social learning
(locus of control) versus attributional (causal stability) inter-

pretations of expectancy of success. lgn;nni_g£_£ezagnn1i§x, 44,
52-68.

Weiner. B., Russell. D.. 6 Lerman. D. (1978). Affective consequences
of causal ascriptions. In J.M. Harvey. W.J. Ickes. and R.F. Kidd

(Eds.). WW (Vol. 2. pp. 59-88):
Hillsdale. NJzErlbaum.

244

weiner. B., Russell. D.. 6 Lerman, D. (1979). The cognition-emotion

process in achievement-related contexts. Jgnznni gf Eersgnnlitx nng
§2£1£1_£§1£h21281. 11(7). 1211-1220.

Weingart. L.R.. 6 Weldon. E. (1988). The impact of an assigned
group goal on the performance of individual group members.
Unpublished manuscript. Northwestern University.

Weiss. R.M.. Ilgen. D.R.. 6 Sharbaugh, M.E. (1982). Effects of
job stress on information search behaviors of organizational

members. MW. .61. 60-66.

Wiener. N. (1948). °
Wine. Nev YorktJohn wiley 5- Sona-

Williams, E.. Harkins. 8.. 6 Latane, B. (1981). Identifiability as
a deterrent to social loafing: Two cheering experiments. lgnznni

2f_EsIa2nal1£1_ené_§2£1sl_£§1£h212£!. 29(2). 303-311.

Wolosin, R.J.. Sherman, S.J.. 6 Till, A. (1973). Effects of
cooperation and competition on responsibility attribution

after success and failure. We].
Emhology. 2. 220-235.

Wong. P.T.P.. 6 Weiner. B. (1981). When people ask "why" questions

and the heuristic of attributional search. £23;nni_gf_£g;§gnniigy
and_§221sl_2§1£h212£1. 22. 650-663.

Zander. A. (1971). WWW. New York:

Academic Press.

Zander. A. (1980). The origins and consequences of group goals.

In L- Featinger (Ed.). WWW.
Oxford University Press.

Zander, A.. Forward. J., 6 Albert. R. (1969). Adaptation of board
members to repeated failure or success by their organization.
. . -- . - -- . 4. 56-76.

. ,

   

es. . g. e :ee 1 11.; eqn:e ;

Zander. A.. 6 Medow, H. (1963). Individual and group levels of aspira-
tion. W. 1.6. 89-105.

Zander. A.. 6 Newcomb. T. (1967). Group levels of aspiration in United

Fund campaigns. MW. 6.
157-162.