3.2.

 

 

 

 

 

‘5'? , . , :5}. I. . 3.11: .‘l‘ >31»!!! :1:
A. , , , . , , .

F...

THE 3 l8

lullllllllllllillllull

This is to certify that the
dissertation entitled
The 314wa at proczdnfl $560; in inns; An
“(Ma's «l: “Wilda“ tut, ml NM mid-«hrs

presented by

fiasco A . Comm

has been accepted towards fulfillment
of the requirements for

“M D . degree in $05M: Mn‘misx MM

 

 

Z/ZZQflL—V

Major professor

Date (0' \Squ

MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

JU_1 0 2005
321301 07an
“4th

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/90 mumps-p14

 

 

THE IMPACT OF PROCEDURAL JUSTICE IN TEAMS:
AN ANALYSIS OF INDIVIDUAL, TASK, AND TEAM MODERATORS
By

Jason A. Colquitt

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Management

1999

ABSTRACT
THE IMPACT OF PROCEDURAL JUSTICE IN TEAMS:
AN ANALYSIS OF INDIVIDUAL, TASK, AND TEAM MODERATORS
By
Jason A. Colquitt

As the use of team-based structure increases, the literature on team effectiveness
has neglected the role that managers can play in fostering team performance. One
variable which could provide managers with a means of improving team effectiveness is
procedural justice, defined as the fairness of decision making procedures (Thibaut &
Walker, 1975). Procedural justice has been shown to have many beneficial effects in
individual contexts (Greenberg, 1990). However, such findings cannot be readily
generalized to team settings because such contexts bring with them complexities that
have not been addressed in the procedural justice literature (James, 1993).

To attempt to capture the full complexity of procedural justice in team settings,
the effects of five possible procedural justice configurations were modeled using process
control as the operationalization of justice (Thibaut & Walker, 1975). Differences among
the five configurations can be expressed in the form of four orthogonal contrasts: own
process control high vs. low, teammates’ process control high vs. low, individuals’
process control consistent vs. inconsistent with teammates’ levels, and relative process
control unclear vs. clear. The effects of these process control configurations was
examined using a theoretically grounded set of outcomes based on the exit-voice-loyalty-
neglect framework for responses to dissatisfaction (Hirschman, 1970; Rusbult et al.,

1988)

eq

IE

The results indicated that individual team members were affected by the process
control that others experienced -- they did not only attend to their own treatment levels.
There were a total of 22 significant effects for the process control contrasts. Of those 22,
five were for the own process control contrast; 17 were for contrasts requiring an
individual member to consider the justice levels of others. Such effects have not been
consistently demonstrated in previous research on procedural justice.

The results also indicated that the effects of the process control contrasts on
fairness perceptions and exit-voice-loyalty-neglect variables were moderated by facets of
the task, the team, and the team member. Specifically, process control effects were
neutralized in conditions of low task interdependence, high Benevolence in terms of
equity sensitivity, and (to a lesser extent) parallel rather than non-parallel teams. These
results are discussed in terms of their relevance to team effectiveness and organizational

justice.

lon
Uni
don
teat
lfik
all“;
just

nn'e

LcPi
mien
good

mh ,
his .1

ACKNOWLEDGMENTS

This dissertation represents the conclusion of a five-year odyssey of late nights,
long exams, intense learning, tons of writing, and an astonishing number of really good
times. Any acknowledgment of the key players in my grad school life will fall far short of
doing any of them justice. Nevertheless, I thank my chair, John Hollenbeck, not just for
teaching me and motivating me, but for making me realize (from week 1) how cool of a
life this really is. I thank Ray Noe for all the opportunities he threw my way, and for
always asking me whether I was getting enough sleep. I thank Don Conlon for arriving
just in time to give me someone to talk justice with. And I thank Dan Ilgen for turning
my attention to the forest when I was stuck wondering about the trees.

I also thank my fellow doctoral students -- especially Marcia Simmering, Jeff
LePine, Mike Wesson, and Becky Luce -- for their camaraderie, support, perspective,
tolerance, and ability to make any day at the office something to look forward to. All are
good colleagues and great friends. Thanks also to the people who happily put up with the
high-maintenance data collection effort contained herein: Aleks Ellis, Brad West,
Christine Jackson, Henry Moon, and Lori Sheppard.

Lastly, thanks to Catherine. Love is not a big enough word.

iv

TABLE OF CONTENTS

List of Tables ................................................................................. x

List of Figures ................................................................................. xi

INTRODUCTION ............................................................................ 1

CHAPTER 1

THE IMPACT OF PROCEDURAL JUSTICE IN TEAMS ............................ 12
What does it mean to be procedurally just? ...................................... 12
Why is procedural justice important? ............................................. 16
Applying procedural justice in team settings ..................................... 19
Procedural justice as a relative phenomenon ..................................... 24
Summary and hypotheses ........................................................... 36

CHAPTER 2

MODERATORS OF THE IMPACT OF PROCEDURAL JUSTICE IN TEAMS. . .38

Task interdependence: A task-based moderator ................................. 39

Parallel vs. non-parallel team: A team-based moderator ........................ 42

Equity sensitivity: A member-based moderator ................................. 46
CHAPTER 3

METHOD ...................................................................................... 52

Choice of setting ..................................................................... 52

Participants ............................................................................ 53

Task .................................................................................... 53

Manipulations ........................................................................ 55

Process control .............................................................. 56

Task interdependence ...................................................... 59

Parallel vs. non-parallel ................................................... 59
Procedure .............................................................................. 60
Measures .............................................................................. 62

Equity sensitivity ............................................................ 62

Fairness perceptions ........................................................ 62

Exit ............................................................................ 63

Conflict ....................................................................... 63

Leader evaluation ........................................................... 63

Mistakes during role performance ........................................ 63

Manipulation checks ........................................................ 64
Data analysis strategy ............................................................... 65

CHAPTER 4

RESULTS ..................................................................................... 66
Manipulation checks ................................................................ 66
Descriptive statistics ................................................................ 67
Tests of process control direct effects ............................................. 67

Fairness perceptions ........................................................ 73

Exit ............................................................................ 73

Conflict ....................................................................... 73

Leader evaluation ........................................................... 74

Mistakes during role performance ........................................ 74
Tests of process control x task interdependence interaction effects ........... 74

vi

Fairness perceptions ........................................................ 76

Exit ............................................................................ 76
Conflict ....................................................................... 76
Leader evaluation ........................................................... 76
Mistakes during role performance ........................................ 76

Tests of process control x parallel vs. non-parallel team interaction effects..77

Fairness perceptions ........................................................ 79

Exit ............................................................................ 80

Conflict ....................................................................... 80

Leader evaluation ........................................................... 80

Mistakes during role performance ........................................ 80

Tests of process control x equity sensitivity interaction effects ............... 80

Fairness perceptions ........................................................ 82

Exit ............................................................................ 83

Conflict ....................................................................... 83

Leader evaluation ........................................................... 83

Mistakes during role performance ........................................ 83

Supplementary analyses ............................................................. 83
CHAPTER 5

DISCUSSION ................................................................................. 87

Discussion of direct effects of process control configurations ................. 88

Discussion of interaction effects with the process control configurations... 94

Suggestions for future research .................................................... 98

vii

Limitations ............................................................................ 101

APPENDIX A: Consent Form .............................................................. 104
APPENDIX B: Training Materials ......................................................... 106
APPENDIX C: Individual Instruction Sheets ............................................. 113
APPENDIX D: Instructions for Individual Tasks ........................................ 117

viii

LIST OF TABLES

Table l - Descriptive Statistics for Dissertation Variables .............................. 71
Table 2 - Beta-weights and Regression Results for Hypothesis 1 ..................... 73
Table 3 - Beta-weights and Regression Results for Hypothesis 2 ...................... 77
Table 4 - Beta-weights and Regression Results for Hypothesis 3 ...................... 80
Table 5 - Beta-weights and Regression Results for Hypothesis 4 ...................... 83

ix

LIST OF FIGURES

Figure 1 - Conceptual Model ............................................................... 14
Figure 2 - High vs. Low Individual Procedural Justice Contrast ....................... 27
Figure 3 - High vs. Low Teammates’ Procedural Justice Contrast .................... 28
Figure 4 - Consistency of Procedural Justice Contrast ................................... 30
Figure 5 - Clarity of Relative Procedural Justice Contrast .............................. 31
Figure 6 - Summary of Process Control Manipulations ................................. 62
Figure 7 - Screen with Individual Task .................................................... 64
Figure 8 - Cell Means for Experimental Conditions ...................................... 75
Figure 9 - Plot of Task Interdependence Moderation Effect ............................. 79
Figure 10 - Plot of Parallel vs. Non-parallel Moderation Effect ........................ 82
Figure 11 - Plot of Equity Sensitivity Moderation Effect ............................... 85

b}
(51

me

autc
tear
fore

mt

them

INTRODUCTION

The increased use of work teams in organizations has been documented in both
the academic and popular press (e.g., Cohen & Bailey, 1997; Dumaine, 1994). Much of
this attention has focused on work teams as a means of reorganizing tasks to keep pace
with foreign competition (e.g., Shaiken, Lopez, & Mankita, 1997). However, even apart
from such reorganization efforts, teams are a pervasive phenomena in everyday life
(Goodman, Ravlin, & Schminke, 1987). Some tasks are performed better by teams than
by individuals, and other tasks can never be completed by one individual working alone
(Steiner, 1972). From this perspective, team-based cooperation “justifies itself, then, as a
means of overcoming the limitations restricting what individuals can do” (Barnard, 193 8,
p. 23).

Teams in contemporary organizations are called by a variety of names, including
autonomous work groups, self-managing teams, project teams, semi-autonomous work
teams, self-directed work teams, along with more traditional designations such as task
forces or committees (Guzzo & Dickson, 1996). While some researchers have proposed
taxonomies which contrast different team manifestations (e.g., Cannon-Bowers, Oser, &
Flanagan, 1992; Cohen & Bailey, 1997), a more succinct approach is to emphasize what
all teams have in common. This approach has created some convergence in terms of what
defines a team. For example, in the most recent review of the work teams literature,
Cohen and Bailey (1997) define a team as a collection of individuals who are
interdependent in their tasks, who see themselves (and are seen by others) as a social
entity, who share responsibility for the outcome of their work, and who manage

themselves and their relations with others outside the team.

 

 

 

SUI
tea
me.
(C 0

decr

impn
team,
was:
interdc‘
(fig. 1‘
Self~ma

IICPlne

Can play

prescripn‘
let-eragjné.
and Cantu
Offer mane

fEICIOIs,

one
film

Unfortunately, the increased use of work teams has not been met with wide spread
success (e.g., Dumaine, 1994; Dunphy & Bryant, 1996). An increased reliance on work
teams has sometimes been associated with problematic outcomes on the part of team
members. Examples of such outcomes have included withdrawal from one’s work role
(Cohen & Bailey, 1997), increase in conflict among members (Jehn, 1995), and a
decreased motivation to perform one’s tasks (Sheppard, 1993).

Partly because of such team failures, a large literature has developed around
improving team effectiveness (Ilgen, 1999). This literature has tended to explore task,
team, and member characteristics as the independent variables used to predict team
success (Cohen & Bailey, 1997; Guzzo & Shea, 1992). For example, the amount of
interdependence required by a team’s tasks has been a common focus of past research
(e.g., Wageman, 1995), as has the type of team (e.g., semi-autonomous, cross-fimctional,
self-managing). Others have focused on the personalities of individual team members
(LePine, Hollenbeck, Ilgen, & Hedlund, 1995).

Unfortunately, little attention has been paid to the role that managers and leaders
can play in improving team effectiveness. This makes it difficult to give managers
prescriptions for improving their teams, particularly where they are incapable of
leveraging task, team, or member variables. Indeed, Kozlowski, Gully, McHugh, Salas,
and Cannon-Bowers (1996) note that academic scholars have very few prescriptions to
offer managers of teams, because few studies of work teams have focused on leadership
factors.

One managerial variable which could prove useful in improving team

effectiveness is procedural justice, defined as the fairness of the procedures used to make

 

tea
dit
me
mt‘t

team

JUSIII
sped}-
bfsed

l‘Hll’Sc‘

decisions (Leventhal, 1980; Lind & Tyler, 1988; Thibaut & Walker, 1975). Managers
foster procedural justice perceptions when they provide employees voice and influence,
explain decisions adequately, treat employees with dignity and respect, and employ
accurate, unbiased information in a consistent manner. Past research has linked
procedural justice to employee motivation, extra-role behavior, and commitment to
collectives (Greenberg, 1990). While such outcomes would certainly seem valuable in
team settings, all of this research has occurred in individual work contexts.

Extending procedural justice to team contexts requires the examination of factors
that do not exist in individual contexts. Specifically, the collective nature of team settings

makes procedural justice a relative phenomena, not an absolute phenomenon. Unlike

 

distributive justice (the fairness of decision outcomes), procedural justice has rarely been
examined in this manner (Cropanzano & Ambrose, 1996). The potential relative nature of
team procedural justice is critical because factors such as leader-member exchange
differences, team cross-functionality, and member diversity make it unlikely that
members’ treatment will be consistent. Indeed, in a team setting it is quite likely that one
member’s procedural justice may be higher, lower, or equal to that of his or her
teammates.

This dissertation sought to examine the impact of these different procedural
justice configurations on team member attitudes and behaviors, as shown in Figure l. The
specific attitudes and behaviors were comprised of fairness attitudes, as well as those
based on the exit-voice-loyalty-neglect typology of responses to dissatisfaction

(Hirschman, 1970; Rusbult, Farrell, Rogers, & Mainous, 1988). This typology provides a

 

 

 

1 98
Unit
Pm't
C0115
conf
Supp
Here

{It

 

Moderators
- Task: Interdependence
- Team: Parallel vs. Non-parallel
- Member: Equity Sensitivity

 

 

 

 

 

Team-Member Responses
- Fairness perceptions

”:33:ng —L_> - Exit; Withdrawal from team task
Configurations - Vorce: Conflict Within the team ‘
- Loyalty: Favorable leader evaluations
- Neglect: Poor role performance

 

 

 

 

 

 

 

Figure 1 - Conceptual Model

broad range of responses to dissatisfactory events -- responses which range in their
destructiveness and the extent to which they are active or passive.

Exit refers to leaving an organization by quitting or transferring (Rusbult et al.,
1988). In the context of this study it was operationalized as exit away from the team’s
task and toward an independent individual task. Voice refers to actively discussing
problems with a supervisor or coworker (Rusbult et al., 1988). While voice is usually
constructive, it was operationalized here in terms of conflict, given the key role of
conflict in many team failures (Jehn, 1995). Loyalty refers to giving public and private
support to the organization while waiting for conditions to improve (Rusbult et al., 1988).
Here loyalty was operationalized as favorable evaluations of the team leader. Neglect

refers to passively allowing conditions to deteriorate through reduced interest or effort

 

mC

efft

(Rusbult et al., 1988). In this study, neglect was operationalized as mistakes made during
the performance of an individual team member’s role.

This dissertation also explored how the effects of justice configurations on the
fairness and exit-voice-loyalty-neglect variables differed across levels of task, team, and
member moderators (see Figure 1). The task-based moderator was task interdependence,
perhaps the most commonly examined variable in teams research (Kiggundu, 1981;
Wageman, 1995). The team-based variable was whether the team was parallel (i.e.,
comprised of members whose work requirements were only partially fulfilled within their
team memberships) or non-parallel (i.e., comprised of members who fiinctioned solely
within the team) (Cohen & Bailey, 1997). The vast majority of teams in organizations are
parallel teams, but they have been virtually ignored in teams research. The member
moderator was equity sensitivity, a personal characteristic which could prove to be
important in staffing teams (Huseman, Hatfield, & Miles, 1987). Examining these
moderators showed when the justice conditions could exert especially strong or weak
effects as a function of the team’s task, type, and member characteristics.

Chapter 1 of this dissertation reviews past research on procedural justice, and
discusses how different procedural justice configurations can influence team member
outcomes in team settings. Hypotheses regarding main effects of different procedural
justice configurations are then advanced.

Chapter 2 of this dissertation discusses the task, team, and member moderators of
the aforementioned procedural justice effects. Hypotheses regarding interactions between
procedural justice and task interdependence, parallel vs. non-parallel team, and equity

sensitivity are then advanced.

 

I16:
ma

indi

diss:
0f III:

and \i

Chapter 3 describes a laboratory study used to test the dissertation’s predictions.
Participants took part in a team decision making task in groups of four people, one of
whom was the team leader. Procedural justice was manipulated by making that leader a
confederate and scripting his or her interactions with the team. Leaders afforded team
members varying levels of input and voice in the decision making process, consistent
with Thibaut and Walker’s (1975) original operationalization of procedural justice. Level
of task interdependence was manipulated by varying the degree to which team members
needed information from one another to complete their work. The type of team was
manipulated by giving some teams the opportunity to simultaneously engage in an
individual decision making task.

Chapter 4 describes the study results and discusses the extent to which the
dissertation’s hypotheses are supported. Chapter 5 then provides an in-depth discussion
of the findings, what they mean to the larger literature on procedural justice and teams,

and what they suggest in terms of avenues for future research.

 

sci
“0L
\ia
jusn
basi:
PIOCI
past:

1’91ch

Ofdeci
Such”)

“Var II. ,

Chapter 1
THE IMPACT OF PROCEDURAL JUSTICE IN TEAMS

As noted at the outset, the literature on team effectiveness has neglected the role
that the manager can play in improving team performance. Meanwhile, the literature on
procedural justice has shown that one of the most important things a manager can do in
individual work contexts is treat others in a procedurally just manner (Greenberg, 1990).
Understanding exactly what this means necessitates a brief review of the origin and
development of the procedural justice construct.

Woes it Mean to be Procedurallv Just?

The notion of m has become an increasingly visible construct in the social
sciences over the last three decades. Colloquially, the term justice is used to connote
“oughtness” or “righteousness.” Under the purview of ethics, an act can be defined as just
via comparison to a prevailing philosophical system. By contrast, in the social sciences
justice is socially constructed; an act is just if most individuals perceive it to be so on the
basis of empirical research (Cropanzano & Greenberg, 1997). This dissertation examines
procedural justice from a social scientific perspective. Thus “what is fair” is derived from
past research linking objective facets of decision making procedures to subjective
perceptions of fairness.

The first streams of literature in the justice domain asked “what is fair” in terms
of decision outcomes. As reviewed by Cropanzano and Randall (1993), Stouffer,
Suchman, DeVinney, Star, and Williams (1949) compared the promotion satisfaction of
officers in the army air corps with Similar officers in the military police during World

War 11. Despite the fact that army air corps officers were promoted quickly and military

poll

Sto

did

qU

police were promoted infrequently, the latter group had more promotion satisfaction.
Stouffer et a1. reasoned that officers in the air corps compared themselves with their
peers, many of whom had been promoted more quickly. Officers in the military police
did likewise, except that they were less likely to have peers who were promoted more
quickly. Out of this came the term relative deprivation, referring to the fact that
individuals compare their deprivation levels to a reference point, not to an absolute or
objective standard.

The concept of relative deprivation was extended by Homans (1961), who
contended that perceptions of distributive iu_stice were based on an implicit input vs.
output proportion. Rather than simply comparing the outcomes they receive with referent
others, individuals compare the degree to which their profits from an exchange
relationship match their investments. Homans’ (1961) theory also more fully predicted
reactions to perceived injustice. Specifically, he predicted that inequalities in input vs.
output proportions would result in anger on the part of the individual engaging in the
comparison.

Building on Homans’ (1961) ideas, Adams (1965) more fully specified reactions
to injustice in his formal version of equityI theory. Adams’ theory can be captured with
four specific propositions (Huseman, Hatfield, & Miles, 1987). (1) Individuals evaluate
their relationships by comparing their ratio of inputs to outcomes to the ratio of some
comparison standard. (2) If those input to outcome ratios are perceived to be unequal then
inequity is present. (3) Inequity results in “equity distress,” regardless of whether the
individual is advantaged or disadvantaged. (4) The greater that distress, the more the

individual attempts to restore equity or withdraw from the situation.

 

aCIS
lake
Equi
IUSIIC
Ifrills“:

Well}:

Adams (1965) predicted that overpaid individuals would experience guilt or
tension and attempt to raise their input level (usually by increasing the quality of their
work). In contrast, underpaid individuals could either lower inputs, find some ways to
increase outcomes (e. g., theft), or cognitively distort the existing input-outcome ratio.
Though much of the early equity theory research was plagued by manipulations with
questionable construct validity, more recent research has been generally supportive of the
theory (Cropanzano & Randall, 1993; Greenberg, 1990).

AS a result of the visibility of equity theory, to many researchers the term
“distributive justice” became synonymous with “equity.” However, research by
Leventhal began to dispel that perception (Leventhal, 1976; Leventhal, 1980). Leventhal
and colleagues presented research participants with various vignettes, asking them to
allocate outcomes in a way that produced distributive justice. While many situations did
prompt an allocation in accordance with equity, other situations prompted participants to
allocate rewards equally or according to need. Similar ideas have been advanced by
Deutsch and Lerner (e.g., Deutsch, 1975, 1985; Lerner, 1977).

A unifying characteristic of these approaches is that justice is a mechanism for
achieving multiple goals. Leventhal (1976) noted, “Allocation decisions are instrumental
acts through which the allocator tries to achieve various goals. Accordingly, our analysis
takes account of a wide range of goals and motives which affect his decision.” (p. 94).
Equity theory was geared primarily to achieving individual productivity via distributive
justice. However, Leventhal, Deutsch, and Lerner reasoned that other goals were
important as well, such as the fostering of cohesive social relations or protecting the

welfare of individuals in jeopardy. The former goal typically results in an equality

 

cc
to
the
to 1

f0

and We
fair (In

Show: t.

allocation, the latter results in a need strategy. While others have discussed additional
allocation strategies (e.g., Reis, 1984), most research in this domain continues to employ
the equity vs. equality vs. need distinction.

As Leventhal and colleagues were formulating their research on distributive
justice, Thibaut and Walker were beginning to examine the reactions of litigants during
courtroom proceedings (Thibaut & Walker, 1975). Courtroom proceedings inherently
consist of two phases. First, the process of presenting evidence and arguments is
conducted under the authority of the judge. Second, the outcome of the trial is decided by
the jury. Thibaut and Walker reasoned that it was entirely possible for one of these phases
to be considered fair while the other was considered unfair. That is, it is possible to have
procedural justice but not distributive justice, or vice-versa.

Thibaut and Walker’s (1975) work portrayed litigants as fundamentally concerned
with their own personal outcomes. In order to protect those outcomes litigants desire
process control and decision control. Process control refers to an opportunity to express
oneself in a decision making process, while decision control refers to actual control over
the decision outcome. Of course, in most cases actual decision control is impossible -- the
final decision rests in the hands of the judge. In such cases, the litigant seeks to attain
indirect control over the outcome by leveraging process control. Thus, process control is
valuable to individuals as a means for achieving indirect decision control.

Research has consistently demonstrated that procedures which promote Thibaut
and Walker’s (1975) notion of process control and decision control are perceived to be
fair (Lind & Tyler, 1988). Moreover, the presence of process control has even been

shown to enhance reactions to unfavorable verdicts. These findings have also been

10

 

W

moSl c.

generalized to organizational settings (Folger & Greenberg, 1985; Greenberg & F olger,
1983). For example, the ability to present arguments on one’s behalf has enhanced
reactions to performance evaluations and conflict resolution procedures (Greenberg,
1990)

Thibaut and Walker’s (1975) research on reactions to courtroom procedures was
complemented by concurrent work by Leventhal (Leventhal, 1980; Leventhal, Karuza, &
Fry, 1980). Just as he had done with distributive justice, Leventhal conducted research
which asked participants to structure procedures in a way which promoted procedural
justice. Out of this work came six criteria for fair procedures. Specifically, procedures
were judged to be fair to the extent that they satisfied six criteria:

(1) Consistengy: procedures are applied consistently across time and persons.

(2) Bias Suppression: the decision maker has no vested interest in the procedure,

or prior beliefs on the part of the decision maker do not prevent an adequate

consideration of all points of view.

(3) Accpracv of information: the procedure makes use of accurate information in
arriving at a decision.

(4) Corregt_ability: the procedure includes a mechanism for appealing bad
decisions.

(5) Representativeness: the procedure reflects the concerns of all subgroups in the
population of individuals affected by the decision, and allows those subgroups a
voice in the procedures. This latter facet of this component is similar to Thibaut
and Walker’s (1975) process control notion.

(6) Ethicalig: the procedure is consistent with prevailing ethical standards,
meaning that bribery, deception, invasion of privacy, etc. are not employed.

While Thibaut and Walker’s (1975) process control notion remains by far the

most common operationalization of procedural justice, use of Leventhal’s criteria has

11

consistently been shown to improve perceptions of procedural justice (Greenberg, 1990;
Lind & Tyler, 1988).
WhLis Procedural Justice Importa_r_l__t?

Procedural justice, whether a function of process control or decision control,
consistency, bias suppression, etc., has been shown to predict a variety of important
outcomes, from job satisfaction to organizational commitment to job performance
(Cropanzano & Greenberg, 1997; Greenberg, 1990). Moreover, high levels of procedural
justice have consistently been shown to neutralize the negative effects of distributive
injustice, leading some to remark “it’s now what you do, but how you do it” (Brockner &
Weisenfeld, 1997).

Two theoretical models have been used to explain why procedural justice is so
important to individuals in organizations. The first is the instrumental model, originally
termed the self-interest model (Lind & Tyler, 1988). This model builds on many of the
ideas advanced by Thibaut and Walker’s (1975) control theory of justice. Specifically,
the instrumental model posits that individuals are motivated to maximize their long-term
economic gains. They may even be willing to withstand short-term losses if that will help
them attain valued outcomes in the long term. Conditions are deemed fair if they help in
the achievement of such long-term goals. An examination of the justice antecedents
reviewed above illustrates the intuitive appeal of the instrumental model. It certainly
seems that one’s long-term economic interests will be protected to the degree that process
control is present, procedures are consistent over time, and accuracy and bias suppression

are evident.

l2

Issues of instrumentality and self-interest were naturally applied as theoretical
reasons for procedural justice effects, because much of the research in the organizational
justice literature has occurred in dispute resolution contexts (e.g., Thibaut & Walker,
1975). However, much of the empirical research used to test the instrumental model has
shown that procedural justice operationalizations (e. g., process control, bias suppression)
were capable of explaining variance in fairness perceptions over and above the
favorability of the actual decision outcomes. For example, Tyler (1989) examined
perceptions of procedural fairness on the part of citizens who had experiences with the
police or court system. They were surveyed regarding the favorability of the outcome
they experienced (e.g., the police solved their problem; their court case was won or lost),
along with perceptions of process control, decision control, bias suppression, and
ethicality. The results indicated that the justice operationalizations explained significant
incremental variance beyond outcome favorability using fairness perceptions and
evaluations of authorities as the dependent measure.

Tyler and colleagues (Tyler, 1989; Tyler & Lind, 1992) argued that justice is not
just important for instrumental reasons. They proposed an alternative model of justice,
termed the relational model. It posits that individuals value membership in organizations
to the extent that they receive economic rewards, but also to the extent that they achieve
high social status and reaffirm their self-esteem. These latter rewards, which are of a
socioemotional nature, are every bit as important as the economic rewards (Cropanzano
& Greenberg, 1996). The relational model contends that people derive much of their own

self-worth from the way they are treated in organizations.

13

 

H]:

CO;

lea:

3150 I

The relational model acts as a complement to the instrumental model in
explaining justice effects. The process control justice operationalization is ofien used as a
means of illustrating this point. A significant relationship between process control and the
evaluation of a decision making system or authority can be explained using either model.
Controlling for outcome favorability will reduce the size of this relationship to the extent
that it is due to instrumental reasons. The remaining variance explained is due to the
value of being able to express oneself, irrespective of outcome, because that ability acts
as a signal indicating high status in the organization (Tyler & Lind, 1992).

A l in Procedural Justice in Team Settings

While procedural justice has been linked to a variety of important organizational
outcomes, it should be noted that almost all of this work has occurred in individual
contexts. One of the only studies to examine procedural justice effects in a team setting
was Korsgaard, Schweiger, and Sapienza (1995). They examined the role of procedural
justice in fostering cooperative relations in strategic decision making teams. They found
that the degree to which leaders considered the views of the members of the team was
significantly related to member procedural fairness perceptions. This is consistent with
Thibaut and Walker’s (1975) process control notion. These effects were stronger the
more evident it became that members’ views actually influenced the final decision,
consistent with the decision control idea.

Korsgaard et al. (1995) also showed that process control and decision control
improved attachment to the team on the part of members, along with commitment to the
team’s decision and trust in the leader. Process control affected these outcomes over and

above decision control, suggesting that it had both instrumental and relational

14

implications. Moreover, these effects were mediated by perceptions of procedural
fairness. Korsgaard et a1. (1995) concluded that the fairness of the processes whereby
decisions are made do indeed impact team members.

Korsgaard et al.’s (1995) work suggests that, while procedural justice has rarely
been applied to work teams, it is capable of improving important team outcomes.
However, their study failed to capture the true collective nature of the team setting. Apart
from measuring team-relevant outcomes such as attachment to the team, there was no
aspect of the study that could not have taken place in an individual context. For example,
no attention was paid to whether one team member took another member’s level of
process control into account when forming perceptions of trust and commitment. Nor did
the authors examine what would occur if levels of process control varied within the team.

James (1993) has noted that, “Although individualistic approaches to justice have
proven valuable, a more even balance of levels of analysis in work on organizational
justice seems necessary to capture the complexity of justice processes in organizations.”
That complexity refers in part to the role that social comparisons play in forming justice
judgments. As noted at the outset, in collectives such as teams or organizations,
procedural justice is likely to be a ;e_lati_va phenomenon, not an absolute one. That is, it
may not be judged by an individual according to only personal levels of process control,
bias suppression, etc., but also treatment by others to which the individual is connected.
Interestingly, procedural justice has almost never been examined in this manner

(Cropanzano & Greenberg, 1996). This despite the fact that the very domain of

organizational justice began with the concept of relative deprivation.

15

 

ju

III
\th

itis

Pl’OC
Icanj
Ibur(
iIISl c
0110\i
PTOCed
15V€loj

pIOCedU

To “capture the complexity” of how procedural justice perceptions might be
formed in a collective, as James (1993) suggests should be done, it is instructive to
explore the simple example of a three-person team of the type explored in Korsgaard’s
(1995) study. It is possible that the team’s leader will provide each member an identical
level of process control, as occurred in their study. It is also possible that the team’s
leader will give some team members more control than others.

Indeed, from the perspective of one member within that team, five procedural
justice configurations are possible: (1) a uniformly high case, where both an individual
and his or her teammates are treated fairly; (2) a unifonnly low case; (3) a differing case
where the individual is treated more fairly than his or her teammates; (4) a differing case
where the individual is treated less fairly than his or her teammates; and (5) a case where
it is unclear whether the individual is treated more or less fairly.

These five possibilities are depicted in Figure 2. Circles represent the boundaries
of a team, the person in the left part of the circle represents the individual making the
procedural justice judgment, and the two people inside the box represent the individual’s
teammates. The true complexity of these possibilities is indicated by the fact that it takes
four orthogonal contrasts to explain how the configurations differ from one another. The
first contrast, shown in Figure 2, is whether the individual’s own level of justice is high
or low. This contrast represents the level of complexity utilized in the vast majority of
procedural justice research (i.e., focusing on one individual’s level of treatment only).

The second contrast, shown in Figure 3, is whether an individual’s teammates’
level of justice is high or low. This contrast, and the ones to follow, necessitate viewing

procedural justice as a relative phenomenon that is affected by what happens to others

16

 

 

CONTRAST ONE:
Whether one's own
justice is high. or one's
own justice is low.

 

 

High

Low

   

evel of Procedural Justice:

illlli

High High j-Iigh
UNIFORMLY HIGH

   

 

 

 

 

      
   
     
 

evel of Procedural Justice:

liliill

Low Low Low
UNIFORMLY LOW

 

 

 

 

       

            
   

evel of Procedural Justice:

illi‘ilii

High Medium Low
DIFFERING

   

 

 

       
   
     
 

evel of Procedural Jusflce:

lililil

Low N ed"ium High
DIFFERING

 

 

 

   

 

      
   

evel of Procedural Justice:

lil’ilil

Medium High Low
DI F FERING

   

 

 

 

   

 

        
 

Figure 2 - High vs. Low Individual Procedural Justice Contrast

17

evel of Procedural Justice:

it it

th Figh th
UNIFORMLY HIGH CONTRAST TWO:
Whether one's teammates'
justice is high, or one's
teammates justice is low.

 

 

 

 

 

evel of Procedural Justice:

iii. titty

UNIFORMLY LOW High High edium i-i

 

 

 

 

 

 

 

 

 

 

 

evel of Procedural Justice:

iii ttimt

High edium Low
DIFFERING

 

 

 

 

 

 

 

 

 

 

evel of Procedural Justice:

Willi

 

 

 

 

Low Medium l-igh
DIFFERING

   

eyel of Procedural Justice:

liliilii

Medium I-igh Low
DIFFERING

   

 

 

 

   

 

           
 

Figure 3 - High vs. Low Teammates’ Procedural Justice Contrast

l8

 

liq-1

within a collective. For the sake of parsimony, this contrast does not distinguish between
a case where (1) both teammates’ justice is low vs. one low and one medium; or (2) both
teammates’ justice is high vs. one high and one medium.

The third contrast, shown in Figure 4, is whether the individual member is treated
in a manner consistent with his or her teammates, or inconsistent with his or her
teammates. Consistency across persons was one of Leventhal’s original justice criteria,
and would seem to necessitate examining referents to assess their level of treatment. In
practice, however, consistency has normally been assessed only in combination with
Leventhal’s other five criteria, and has often been operationalized in terms of consistency
across time, not persons (Lind & Tyler, 1988).

The fourth contrast, shown in Figure 5, is whether it is clear what one’s level of
treatment is relative to one’s teammates, or whether it is unclear. In the first four teams in
Figure 5 it is clear that the individual’s treatment is equal to, greater than, or less than that
of his or her teammates. In the bottom team it is unclear. While the individual
experiences more justice than one teammate, he or she experiences less justice than
another. This situation is perhaps the most common one in larger teams.

Procedural Justice M Relative Phenomenon

The importance of the first contrast has been demonstrated repeatedly in the
justice literature. That is, one individual’s own procedural justice, whether a function of
high rather than low process control, lack of bias toward oneself, etc., Should be
associated with more favorable fairness perceptions, less exit, less conflict, more
favorable leader evaluations, and fewer mistakes during role performance (i.e., significant

effects for the contrast shown in Figure 2). Taking the next step by examining procedural

l9

 

 

CONTRAST THREE:

Whether justice is
consistent within the
team, or inconsistent

within the team.

 

 

        
      
    

evel of Procedural Justice:
High tgh I-figh
UNIFORMLY HIGH

 

 

 

 

  

evel of Procedural Justice:

liltilil

Low ow Low
UNIFORMLY LON

 

 

 

 

       
   

 
    
   
     
       
   
     
       
   

evel of Procedural Justice:

II i. ll
High ium Low
DIFFERING

 

 

 

 

 
 

evel of Procedural Justice:

ilillil

Low ium l-figh
DIFFERING

 

 

 

 

evel of Procedural Justice:

liliiil

Medium I-figh Low
DIFFERING

 

 

 

 

Figure 4 - Consistency of Procedural Justice Contrast

20

 

Iii

 

 

 

 

 

 

 

 

 

 

VS.
7 ii Iii
CONTRAST FOUR:

Whether it is clear how
one's own justice corrpares
to that of one's tearrmates,

or whether it is unclear.

 

 

Figure 5 - Clarity of Relative

21

    

   

evel of Procedural Justice:
High High High
UNIFORMLY HIGH

   

 

 

 

   

 

    
     
 

evel of Procedural Justice:

illilii

Low Low Low
UNIFORMLY LOW

 

 

 

 

    
 

     

  
   

evel of Procedural Justice:

High 7 q? q?

Medium Low
DI FFERING

 
   

 

  

 

    
     
  

evel of Procedural Justice:

Iiiiiii

Low edium High
DIFFERING

 

 

 

 

    
 
 
   

  

evel of Procedural Justice:

iii

Medium

 

 

 

 

High Low

DIFFERING

    

  

 
 

Procedural Justice Contrast

 

ju:

jus

reh
liu.
labi
Pan
the:
lndi

fine

one}
andi
0fth.

phen

justice as a relative phenomena can be accomplished by testing whether the second, third,
and fourth contrasts (i.e., teammates’ justice, consistency of justice, and clarity of relative
justice) also have significant effects on outcomes.

The question of whether procedural justice is best examined as an absolute or
relative phenomenon was first tested by Ambrose, Kulik, and colleagues (Ambrose &
Kulik, 1989, Ambrose, Harland, & Kulik, 1991). Ambrose and Kulik (1989) conducted a
laboratory study which used a computer simulation to manipulate process control.
Participants assumed the role of an advertising account manager, and during the course of
the simulation they were told that their expenses would be the subject of an audit.
Individuals’ own process control was manipulated by allowing or denying them input
into the audit process. They were also told in a memo whether a “coworker” had been
allowed or denied input, comprising a manipulation of others’ process control. While
one’s own process control had significant effects on fairness perceptions, satisfaction,
and intent to quit the activity, other’s process control had no such effects. Thus the results
of that study provided little support for the notion that procedural justice was a relative
phenomenon.

Ambrose, Harland, and Kulik (1991) conducted another laboratory study to
investigate the potential relative nature of procedural justice. Participants were told that
they either could choose the task they wanted to complete (high process control) or would
be assigned which task they would complete (low process control). Participants were also
told whether another participant was able to choose which task to complete, creating a
manipulation of others’ process control. As in Ambrose and Kulik (1989), no main

effects were found for others’ process control, nor were any interactions evident with the

22

variable. Again, little support was shown for the idea that participants were affected by
others’ level of procedural justice.

A third study was conducted by Grienberger, Rutte, & van Knippenberg (1997),
who did actually find some support for relative procedural justice. They followed the
paradigm of Ambrose et al. (1991) in which two participants (an individual and a referent
other) were given or denied the ability to choose which task to complete. However,
whereas Ambrose et al. (1991) attributed a no-process control situation to the luck of the
draw, Grienberger et al. (1997) attributed the situation to a decision made by the
experimenter. Their results still failed to show any main effect for others’ level of process
control, but did demonstrate an interaction between own and other control. Specifically,
the lowest level of perceived fairness was evident for participants who were denied
process control when the referent was granted control.

Taken together, the results of these studies provide little support for the notion
that the contrast shown in Figure 2 -- others’ level of procedural justice -- will predict
team member outcomes. Nonetheless, two theories in the justice domain do offer support
for examining procedural justice in a relative sense. The first is referent cognitions theog
(F olger, 1986, 1987). The primary tenet of this theory is that individuals react to events
by creating mental representations of “what might have been” (i.e., referent cognitions).
These mental simulations may be built from several ingredients, including simple
speculation, past experiences, social comparisons, and norms and rules.

While referent cognitions theory has failed to generate the research attention that
Adams’s (1965), Leventhal’s (1980), and Thibaut and Walker’s (1975) work has, it is

capable of integrating research on distributive and procedural justice (though it has most

23

 

gran
role

com]
cogn

norm

andzt

Unhné

exam]

often been applied to the former). For example, equity theory studies are an example of
creating referent cognitions using social comparisons. Studies of Thibaut and Walker’s
(1975) process/decision control criteria or Leventhal’s (1980) justice criteria are
examples of creating referent cognitions using norms or rules such as “procedures should
afford process control” or “procedures must suppress biases.” Indeed, one could argue
that these criteria have become “institutionalized” to the extent that they are taken for
granted (Scott, 1995). This might help explain why so little attention has been paid to the
role of social comparisons with procedural justice, despite the fact that: (1) such
comparisons are explicit ingredients of many distributive justice theories; and (2) referent
cognitions theory does not explicitly suggest that ingredients like social comparisons or
norms/rules are more appropriate for one type of justice than another.

Building in part on ideas embedded within referent cognitions theory, Cropanzano
and Ambrose (1996) argue for a monistic theory for procedural and distributive justice.
They suggest that the “means vs. end” distinction between the two constructs has
unintentionally resulted in artificial differences in how the two constructs are treated. For
example, they note:

“In practice, this suggests that process fairness is inferred relative to a theoretical

standard, whereas distributive justice is inferred relative to a referent standard.

Thus, the fairness of processes is evaluated against a different type of standard

than is the fairness of outcomes. Note, of course, that this differences is implicit in

the way the variables have been operationalized. It is not explicit in theories of

justice.” (p. 31).

The authors go on to argue that, in many contexts, procedural justice may “act
like” distributive justice. Specifically, what is appropriate procedurally may follow

Leventhal’s (1976) three distributive rules of equity, equality, or need. In some cases

process or decision control may be doled out in accordance to individuals’ abilities or

24

 

dc’

SO

IEI'

001
C61

sut

par
me:

We;

contributions, even if consistency across persons is violated. In other cases process or
decision control might best approximate an equality rule, especially where cohesiveness
and harmony are important goals (Cropanzano & Ambrose, 1996). These ideas have not
yet been the subject of empirical research.

Cropanzano and Ambrose’s monistic theory also suggests that the distributive
phenomenon of fratemal deprivation (J arnes, 1993) may be relevant for procedural
justice. Fraternal deprivation occurs when an individual receives fair outcomes but is
somehow connected to a group which, on the whole, does not. Fraternal deprivation with
respect to procedural justice had never been examined, until recently. Lind, Kray, and
Thompson (1999) conducted a laboratory study in which three participants took part in a
computer simulation comprised of scheduling, budgeting, and memo-routing tasks. At
certain intervals participants would encounter problems in their tasks and need to make
suggestions to the “supervisor,” who was in fact a confederate. These suggestions would
either be met by affirmative or negative text messages sent from the confederate, and
participants could monitor each other’s messages. In some conditions negative text
messages were distributed evenly among team members, while in others the injustices
were concentrated on one individual. The results showed that participants did take into
account the experiences of others when making procedural justice judgments. However,
it took over three instances of injustice toward others to affect fairness perceptions as
much as one instance toward oneself. The authors concluded that “people do consider and
respond to the injustices of others, but that response pales relative to reactions to their

own personal injustices.”

25

In summary, while research by Ambrose, Kulik, Grienberger, and colleagues has
been very equivocal, referent cognitions theory and the monistic theory of justice suggest
that the procedural justice experienced by others within a collective may be capable of
affecting the responses of an individual members. The concept of fraternal deprivation,
supported to a small degree by Lind et al. (1999), suggests a similar notion. Thus, in a
team context, it seems likely that high levels of teammates’ procedural justice will be
associated with more favorable fairness perceptions, less exit, less conflict, more
favorable leader evaluations, and fewer mistakes during role performance (i.e., significant
effects for the contrast in Figure 3).

While the first two contrasts (Figures 2 and 3) capture raw levels of procedural
justice on the part of an individual member and his or her teammates, the third contrast
captures differences between the member and the teammates. The contrast in Figure 4
compares cases where an individual’s justice is consistent with that of his or her
teammates with cases where justice is inconsistent. As mentioned above, the concept of
consistency across persons was an aspect of Leventhal’s original procedural justice
criteria. Consistency across persons would seem to necessitate a relative perspective.
Unfortunately, consistency has usually been assessed only in combination with
Leventhal’s other rules, and only in an individual setting (Lind & Tyler, 1988).

However, one unpublished study reviewed by Lind and Tyler (1988) compared
the relative importance of Leventhal’s six criteria across different contexts. Specifically,
Fry and Leventhal (1979) examined situational characteristics which made specific
justice criteria more or less important to justice judgments. Consistency across persons

was judged to be most important in task vs. socially focused interaction, formal vs.

26

 

CC

pa

informal interaction, and cooperative vs. competitive interaction. This suggests that
consistency should be especially critical in work teams, which are characterized by task-
focused, formal, cooperative interaction.

The results of this unpublished study are congruent with the monistic theory of
justice, which argues in part that procedural justice may “act like” distributive justice in
some contexts. Specifically, Cropanzano and Ambrose (1986) argued that procedural
justice may need to be created in equal levels in those contexts where Leventhal (1976)
argued for equal rather than equitable outcome allocations. Leventhal (1976) had argued
for equal allocations in situations where cohesiveness and harmony were important goals.
Those goals do seem appropriate in task-focused, formal, cooperative interaction.

However, the results of Lind et al.’s (1999) study do not support such a
contention. Recall that the computer simulation in their study included text messages sent
from a confederate which either granted or denied participant suggestions regarding
scheduling, budgeting, and memo-routing tasks. In some conditions the denial messages
were concentrated on one individual while in others they were spread around
consistently. While the latter would seem to be the most fair situation, based on the
consistency criterion, the former was associated with greater fairness perceptions on the
part of individual participants. The authors attributed this to the meager effect that others’
injustices had on individual participant’s perceptions.

Grienberger et al.’S (1997) results do offer some support for the increased
importance of consistency in team settings. Recall that their results showed an interaction
effect of own and others’ process control. The lowest level of perceived fairness was

reported by participants who were denied process control when the other participant was

27

granted control. In other words, that situation -- which was an inconsistent one -- was
deemed less fair than a consistent (albeit uniformly low) situation where everyone was
denied process control.

As with the second contrast, empirical tests of the consistency contrast in Figure 4
are equivocal. Nonetheless, Leventhal’s (1980) original conceptualization of procedural
justice, together with Fry and Leventhal’s (1979) study and the monistic theory of justice
suggest that consistency within the team may have positive effects. It therefore seems
likely that consistency in procedural justice will be associated with more favorable
fairness perceptions, less exit, less conflict, more favorable leader evaluations, and fewer
mistakes during role performance (i.e., significant effects for the contrast in Figure 4).

The final contrast, shown in Figure 5, captures whether it is clear what one’s level
of treatment is relative to one’s teammates, or whether it is unclear. In clear conditions a
team member can easily see that his or her treatment is equal to, worse than, or better
than those of his or her teammates. However, in unclear cases the individual falls
somewhere in between his or her teammates. Should the member feel good about this
situation, or bad about this situation? The answer is unclear. This situation is especially
likely in large teams, where most members are likely to fall somewhere in the middle of
the range of treatment.

To my knowledge such a contrast has never been examined in the justice
literature, so any predictions regarding it are tenable. Nonetheless, in some ways the
unclear case is the “best of both worlds.” That is, an individual member is receiving
better treatment than someone else, which has positive implications for that member’s

self-esteem and sense of self-worth. At the same time, the member avoids being in the

28

uncomfortable position of being the best treated person, a position likely to result in a
large degree of guilt and discomfort.

Indeed, if one examines the five possible justice configurations in Figure 2, the
potential benefits of the unclear condition become more clear. The only alternative which
would definitely be preferred over the unclear case is the uniformly high case. Of the
others, the uniformly low case’s only strength is that treatment is consistent. The
differing case in which the individual is treated better than his or her teammates should
create inequity distress (Adams, 1965). Adams noted the potential for such effects in the
context of over-reward cases, but Cropanzano and Ambrose’s (1996) monistic theory
would suggest that the phenomenon could generalize to procedural treatment as well.
Similarly, the differing case in which the individual is treated worse should also create
inequity distress.

Of course, the relative advantage of the unclear condition is built on the
assumption that the in-between member will focus to some extent on the member who is
treated worse than him or her. Past research in social psychology does in fact bear this
out. Most individuals instinctively view their environments in ways which protect their
self-image (Zuckerman, 1979). In the context of equity theory, comparisons which create
inequity distress often result in the changing of the comparison other or standard (Folger
& Cropanzano, 1998). In unclear conditions this should result in a tendency to compare
treatment levels with the low person, rather than the high person.

Other research suggests that the very fact that members in the unclear condition
find themselves in-between their teammates’ levels will prevent them from reacting

negatively to their treatment. Research on causal attributions explores when and how

29

individuals seek to understand the causes behind certain outcomes and behaviors (e. g.,
Kelley & Michela, 1980; Pyszcznski & Greenberg, 1981; Wong & Weiner, 1981). This
research suggests that individuals begin the attribution seeking process in the wake of
obviously negative or unexpected events. Attribution seeking is less likely to begin when
the context is ambiguous or not obviously negative. Brockner and Weisenfeld (1996)
suggest that justice perceptions are formed and acted upon during that attribution process.
If the sense-making never occurs, then neither should the reactions to injustice. Rather
individuals will react based on their default assumption, which for most individuals is
that circumstances are fair (Brockner & Weisenfeld, 1996). This suggests that members
in unclear justice configurations will hold favorable fairness perceptions, and will be less
likely to begin any sense-making process which could disconfirrn those perceptions.

Even if attribution seeking was to occur in the unclear case, the nature of the
configuration makes it unlikely that the member would attribute treatment differences to
low procedural justice on the part of the leader. Kelley’s covariation principle of
attribution suggests that individuals’ attributions are guided by the extent to which two
events (e.g., a member making a recommendation to the leader and the leader rejecting
that recommendation) covary (Kelley & Michela, 1980). Covariation is judged along
three dimensions: distinctiveness (e. g., do people besides the leader reject the member’s
recommendations?), consistency (e. g., does the leader always reject the member’s
recommendations?), and consensus (e. g., does the leader reject other members’
recommendations?)

Kelley’s theory suggests that team members will be most likely to attribute the

rejection of recommendations to the leader when high distinctiveness, high consistency,

30

and high consensus are all present. However, the unclear case precludes consensus to a
greater degree than any other justice configuration (i.e., the leader rejects one member’s
recommendations even more frequently, but usually accepts the other member’s).
Attributing the rejection outcomes to the leader is a necessary condition for forming and
acting on justice perceptions (Brockner & Weisenfeld, 1996).

Taken together, the self-serving cognitive strategies of individuals, along with the
processes governing attribution behaviors, suggest that the unclear condition may
actually result in more beneficial outcomes, at least versus the unifome low and
differing justice configurations. I should reiterate, however, that the fact that such a
contrast has never been examined in past research makes any predictions exploratory.
Summm and Hypotheses

Examining the utility of procedural justice in team settings necessitates capturing
the full complexity of justice in a collective setting. The four contrasts shown in Figures
2-5 are capable of capturing the possible justice configurations that may occur in teams.
While the first contrast has been the subject ofjustice research for two decades, the other
three represent effects that are only beginning to be considered.

This dissertation manipulated treatment within a decision making team to assume
the five configurations shown in each of the figures. Given that varying process control
has been the method of operationalizing procedural justice in past research on relative
procedural justice (Ambrose & Kulik, 1989; Ambrose et al., 1991; Grienberger et al.,
1997; Lind et al., 1999), different levels of process control was chosen as the means of
creating the justice configurations in the present study. This allows for a more direct

comparison with recent research on this topic. Moreover, the fact that process control is

31

five times more common than any other method of manipulating procedural justice
(Colquitt, Conlon, Wesson, Ng, & Porter, 1999) means that using such an
operationalization provides a bridge between this study and past research in individual
contexts. The following explicit hypotheses will be tested:
Hymthesis l: The five process control configurations will explain significant
variance in procedural fairness perceptions, exit from the team’s task, conflict, leader
evaluation, and mistakes during role performance.
Hypothesis 1a: High vs. low individual process control will lead to higher
procedural fairness perceptions, less exit from the team’s task, less conflict,
higher leader evaluations, and fewer mistakes during role performance for an
individual team member.
Hypothesis 1b: High vs. low tearnmates’ process control will lead to higher
procedural fairness perceptions, less exit from the team’s task, less conflict,
higher leader evaluations, and fewer mistakes during role performance for an
individual team member.
Hypothesis 1c: Consistent vs. inconsistent process control will lead to higher
procedural fairness perceptions, less exit from the team’s task, less conflict,
higher leader evaluations, and fewer mistakes during role performance for an
individual team member.
Hymthesis 1d: Being in an unclear vs. clear process control configuration will
lead to higher procedural fairness perceptions, less exit from the team’s task, less
conflict, higher leader evaluations, and fewer mistakes during role performance

for an individual team member.

32

Chapter 2
MODERATORS OF THE IMPACT OF PROCEDURAL JUSTICE IN TEAMS

The previous chapter discussed predictions regarding main effects for different
process control configurations on fairness perceptions, exit from the team’s task, conflict,
leader evaluation, and mistakes during role performance. This chapter examines
moderators of those effects -- moderators based on facets of the task, the team, and the
individual team member. Identifying such moderators is important scientifically,
theoretically, and practically.

Scientifically, organizational behavior has been shown to be a function not just of
individual and situational factors, but also of the interaction of the two (Roberts, Hulin, &
Rousseau, 1978). If one goal of science is to explain variance in a phenomenon, more
variance will be explained to the degree that such interactions are specified in our
predictive models. Theoretically, one criterion for good theory is Specifying when effects
should be strongest or weakest (Whetten, 1989). That is, a theory should specify the
boundary conditions for the predicted effects.

Practically, few empirical results will generalize to all organizations, and some
organizations will be incapable of altering the independent variables under study.
Applying empirical findings in such organizations often requires building contingency
frameworks. In organizations which cannot, for example, change process control levels,
such frameworks can illustrate alternative courses of action. For example, moderators
which neutralize the negative effects of certain process control configurations could be
increased. This may be especially critical in the current context, as factors such as leader-

member exchange differences, team cross-functionality, and member diversity would

33

seem to make it highly unlikely that process control could ever be consistent within a
team.
IaS_k Interdependence: A Task-Based Moderator

Task interdependence is defined by Wageman (1995) as the degree to which a
task requires collective action in order for it to be completed, and is considered by
many to be a defining characteristic of teams (Sundstrom, DeMeuse, & F utrell, 1990).
Task interdependence is an integral part of many models of team effectiveness, and
has been positively associated with team coordination and performance (e.g.,
Campion, Medsker, & Higgs; Janz, Colquitt, & Noe, 1995).

One benefit of task interdependence is that it gives individual team members a
greater sense of responsibility regarding team performance (Kiggundu, 1981; Pearce
& Gregersen, 1991). Because each member is so closely linked to other members, one
single person can have a significant impact on team-level performance. Indeed, in
such cases a team may only be as successful as its weakest member (LePine et al.,
1995; Steiner, 1972). That increased feeling of responsibility should make it more
critical to individual members that they be treated fairly, because injustices leveled at
even one member could spill over to the entire team.

A second benefit of task interdependence is that it increases members’
psychological identification with the team (Colquitt, Noe, & Janz, 1999). A variety of
research streams in social psychology indicate that individuals come to identify with
collectives and devote increased attention to the needs of their teammates (Paulus,
1980; Wegner, 1982). Moreover, teams serve as “lenses” through which the external

environment is perceived (Hackman, 1992; Salancik & Pfeffer, 1978).

34

The influence of the team lens should be greater when task interdependence is
high, because of the increased communication that occurs in those contexts. The net
effect of this phenomenon will be an increased importance accorded to other team
member’s justice in high interdependence contexts. This idea received indirect support
in Lind et al.’s (1999) study where procedural justice was manipulated via text
messages monitored by team members. Lind et al. (1999) assessed the impact of
personal (i.e., directly experienced) and others’ (i.e., observed) instances of injustice
on outcomes before and after a period of group discussion. They found that others’
injustices had much more powerful effects after group discussion, presumably after
identification with the other members had increased.

It should be noted that, while task interdependence has never been examined in
conjunction with procedural justice, it has been discussed in the literature on
distributive justice. Specifically, the construct has been discussed in terms of the
appropriateness of equitable vs. equal outcome allocations. Kabanoff (1991) proposed
that the choice between equity and equality depends in large part on the distribution of
power within a collective. He found that, where power differentials were low (as in
interdependent contexts), any injustice that was experienced as a result of equitable as
opposed to equal allocation procedures would be especially intense (as compared to
high power differential cases). Kabanoff went so far as to list task interdependence as
one factor which should prompt managers to employ an equality rather than equity
allocation strategy.

Cropanzano and Ambrose’s (1996) monistic theory would argue that

procedural justice could be subject to the same conditions of appropriateness. If

35

procedural justice “acts like” distributive justice in some contexts, then the
consistency procedural criterion should apply in the same places that the equality
distributive criterion does. This suggests that the relationship between consistency of
process control, clarity of relative process control, and member outcomes should be
stronger where interdependence is high.
Hypothesis 2: The relationship between the five process control configurations and
procedural fairness perceptions, exit from the team’s task, conflict, leader evaluation,
and mistakes during role performance will be moderated by task interdependence,
such that the process control configurations will explain less variance in the outcomes
when task interdependence is low than when it is high.
Hypothesis 2a: The relationship between high vs. low individual process control
and procedural fairness perceptions, exit from the team’s task, conflict, leader
evaluations, and mistakes during role performance will be moderated by task
interdependence, such that individual process control will be less beneficial when
task interdependence is low than when it is high.
Hymthesis 2b: The relationship between high vs. low teammates’ process control
and procedural fairness perceptions, exit from the team’s task, conflict, leader
evaluations, and mistakes during role performance will be moderated by task
interdependence, such that teammates’ process control will be less beneficial
when task interdependence is low than when it is high.
Hypothesis 20: The relationship between consistent vs. inconsistent process
control and procedural fairness perceptions, exit from the team’s task, conflict,

leader evaluations, and mistakes during role performance will be moderated by

36

task interdependence, such that the consistency of process control will be less

beneficial when task interdependence is low than when it is high.

Hypothesis 2d: The relationship between unclear vs. clear relative process

control and procedural fairness perceptions, exit from the team’s task, conflict,

leader evaluations, and mistakes during role performance will be moderated by
task interdependence, such that the lack of clarity of process control will be

less beneficial when task interdependence is low than when it is high.

Parallel vs. Non-parallel Team: A Team-Based Moderato_r

Many team-based initiatives place individuals into teams on a permanent basis,
with such units enjoying stable, long-term memberships. However, it is much more
common for a collection of individuals to be placed in a team (or teams) in parallel to
their own everyday tasks. Cohen and Bailey (1997) term these parallel teams, noting that
quality circles, task forces, and employee involvement groups are examples of this
variety. Parallel teams are the most common type of team, particularly in the exempt
segment of the workforce (Lawler & Cohen, 1992). Lawler and Cohen (1992) suggest
that parallel teams are used in 85% of Fortune 1000 companies. Indeed, it is not unusual
for a single person to be members of multiple teams, each of which is supplemental to
individual projects they must complete (Campion, Pappar, & Medsker, 1996).

Such teams create a classic dilemma for their members. At any given moment,
individuals must decide whether to devote their time and effort to their own personal in-
basket of activities, or to the deadlines and responsibilities of the team to which they
belong. Lawler and Cohen (1992) argue that, because parallel teams must compete for the

time and resources of their members they are difficult to sustain, and may have costs

37

which outweigh their benefits. Their problems include a lack of organizational legitimacy
and an inability to compete for time, money, information, and resources (Lawler &
Cohen, 1992). Parallel teams are also marked by high levels of conflict, as team members
often find themselves faced with competing objectives and serious time concerns.

Despite their widespread use and frequent problems, parallel teams have been
surprisingly ignored in the work teams literature. In a review of work teams research
between 1990 and 1996, Cohen and Bailey discuss 54 studies which have examined
teams. Only 4 of the 54 studies focused on parallel teams (Adam, 1991; Magjuka &
Baldwin, 1991; May & Schwoerer, 1994; Steel, Jennings, & Lindsey, 1990). Far more
attention has been paid to other types of work teams, such as self-managing work teams
or project teams.

In many respects, individual members of parallel teams face a classic example of
role conflict. Role conflict occurs when the demands of successfully fulfilling one role
interfere with one’s ability to successful fulfill another role (for a review see Kahn,
1981). Members of parallel teams have a set number of hours they can work on a day-to-
day basis. They must decide how to allocate those hours between their individual and
team tasks. From this perspective, time allocation is a zero-sum game. There are only so
many hours in a day, and increasing the total time worked is often impossible,
particularly in the wake of reorganization and downsizing efforts.

Research in the role conflict domain has shown that such conflict can cause
negative affect, tension, and even physical health symptoms (Kahn, 1981). Kahn and
Byosiere (1992) note that positions involving boundary spanning are particularly

susceptible to role conflict. In a parallel team, every member is a boundary spanner,

38

simultaneously dealing with the demands of both the team and individual role sets. It
therefore seems clear that one of the most critical decisions a member of a parallel team
must make is how much time to invest in the team.

That decision regarding how much time to invest should influence how strongly
individual members react to different process control configurations. The less the time
invested in the team, the less the member should care about receiving low levels of
process control. Moreover, less time spent on the team should be associated with less
psychological identification with the collective (Paulus, 1980). Just as with task
interdependence, this should make contrasts such as teammates’ process control,
consistency of process control, and clarity of relative process control less important to the
team member.

Similarly, less time spent in the group means less time perceiving the
environment through the group lens (Hackman, 1992; Wegner, 1982). Thus factors which
are important to the group, like teammates’ process control, do not color perceptions to
the same degree in parallel teams. Lind et al.’s (1999) finding that time spent discussing
justice issues enhanced the impact of others’ justice on individual perceptions is also
relevant here. Parallel teams afford less time for the group to discuss issues, reducing the
importance of others’ treatment.

Thus arguments relating to collective identification, group influence, and group
discussion support a smaller relationship between the process control configurations and
outcomes in parallel teams. Another argument can be built using precisely the concept
that defines a parallel team -- the ability to exit. In the exit-voice-loyalty-neglect

framework, the four responses are alternatives for dealing with the same circumstance,

39

whether that circumstance is low job satisfaction or low procedural justice (Hirschman,
1970; Rusbult et al., 1988). In non-parallel teams exit is not an option, making voice,
loyalty, and neglect options more likely alternatives. Thus process control configurations
should relate more strongly to those three outcomes when exit is not a potential substitute
(Rusbult et al., 1988).

Another component of the exit-yoice-loyalty-neglect framework argues for a
stronger relationship between process control configurations and outcomes in non-
parallel teams. Rusbult et al. (1988) proposed that investment in one’s work role makes
exit a less likely response to dissatisfaction or unfairness. Non-parallel teams are
associated with more member role investment (Cohen & Bailey, 1997), thus members
should be more likely to opt for non-exit responses. I should note that, because exit does
not exist as a variable in non-parallel teams, it is not included as a dependent measure in
the below hypotheses.

Hymthesis 3: The relationship between the five process control configurations and
procedural fairness perceptions, conflict, leader evaluation, and mistakes during role
performance will be moderated by whether the team is parallel or non-parallel, such
that the process control configurations will explain less variance in the outcomes
when the team is parallel than when it is non-parallel.

Hypothesis 3a: The relationship between high vs. low individual process control

and procedural fairness perceptions, conflict, leader evaluations, and mistakes

during role performance will be moderated by whether the team is parallel or non-
parallel, such that individual process control will be less beneficial when the team

is parallel than when it is non-parallel.

40

Hypothesis 3b: The relationship between high vs. low teammates’ process control
and procedural fairness perceptions, conflict, leader evaluations, and mistakes
during role performance will be moderated by whether the team is parallel or non-
parallel, such that teammates’ process control will be less beneficial when the
team is parallel than when it is non-parallel.

Hypothesis 3c: The relationship between consistent vs. inconsistent process
control and procedural fairness perceptions, conflict, leader evaluations, and
mistakes during role performance will be moderated by whether the team is
parallel or non-parallel, such that the consistency of process control will be less
beneficial when the team is parallel than when it is non-parallel.

Hypothesis 3d: The relationship between unclear vs. clear relative process control
and procedural fairness perceptions, conflict, leader evaluations, and mistakes
during role performance will be moderated by whether the team is parallel or non-
parallel, such that the lack of clarity of process control will be less beneficial

when the team is parallel than when it is non-parallel.

Eguig Sensitivity: A Member-Based Moderator

Equity sensitivity is a relatively stable individual characteristic which assesses

one’s preferences for and sensitivity to various equity levels (Huseman et al., 1987). One

of the propositions in Adam’s (1965) conceptualization of equity theory was that higher

levels of inequity would be associated with higher levels of “equity distress” in

individuals. While Adams (1965) suggested that this proposition was generalizable across

most individuals, subsequent research has shown that everyone does not equate equity

with distributive justice. As reviewed earlier, other allocation norms can be used to fulfill

41

important goals in specific situations, such as equality or need allocations (Deutsch,
1975; Leventhal, 1976).

Past research has therefore examined individual characteristics which alter
responses to inequity. These include sex, age, ethnicity, and various specific personality
traits such as Protestant work ethic, machiavellianism, and moral maturity (Huseman et
al., 1987; Major & Deux, 1982; Vecchio, 1981). Major and Deux (1982) characterized
the work on individual differences in responses to inequity as “scattershot” and
unsystematic (p. 44). To attempt to fill this gap in the justice literature Huseman et al.
(1987) proposed the construct of equity sensitivity, which was later reconceptualized by
King, Miles, and Day (1993). While the construct is continuous conceptually, it an be
represented by three classes of people. Benevolents are more tolerant of situations where
their input/outcome ratios are less than those of the referent other (Huseman et al., 1987).
They are more focused on the giving of inputs than the other groups (King et al., 1993).
Equity sensitives need their input/outcome ratios to be equal than those of the referent
other (Huseman et al., 1987). Entitles are more tolerant of circumstances where their
ratios are higher than those of the referent other, and are especially intolerant of the
opposite circumstance. They are more focused on the receipt of outcomes than the other
groups (King et al., 1993).

Huseman et al. (1987) suggested that Benevolence is associated with altruistic
tendencies, where individuals give while expecting little in return. The Benevolent
behavior pattern could result from differences in cultural heritage, an empathic ability to
perceive others’ needs, or the desire for social approval and enhanced self-image. Equity

sensitives react to inequity as Adams (1965) suggested, feeling distressed whether they

42

are advantaged or disadvantaged. The Entitled behavior pattern can result from cultural
heritage, overly submissive childrearing practices, or generational tendencies.

Miles, Hatfield, and Huseman (1989) tested this conceptualization of equity
sensitivity in a sample of undergraduate business students. The students were presented
with two scenarios describing two jobs. One was a job coding questionnaires for a
professor -- a part time, low page, less skilled job. The other was a salaried position as a
staff member in a consulting firm. Students indicated what they perceived to be a fair
wage. They were then told to assume that they had been hired at that wage, and that their
abilities would allow them to perform at an above minimum standard level (which was
supplied in terms of questionnaires per hour coded and clients interviewed and reported
on). Finally, they were asked how much per hour more a co-worker would need to get for
them to feel angry, and how much less that co-worker would need to get for them to feel
guilty.

Miles et al. (1989) showed that Benevolents preferred that they themselves would
have higher inputs than either Equity Sensitives or Entitleds. They also showed the
expected differences in desired input/outcome ratios between Benevolents and either
Equity Sensitives or Entitleds. However, there was not the expected difference between
Equity Sensitives and Entitleds. Predictions for anger thresholds received marginal
support, with Entitleds having an especially low anger threshold for under-reward
conditions. In general these results confirm many of the propositions advanced in
Huseman et al. (1987), though less support was shown for a distinction between Equity

Sensitives and Entitleds.

43

As with Miles et al. (1989), King et al. (1993) tested the equity sensitivity
propositions using undergraduate business students and hypothetical vignettes. Subjects
were instructed that they and a referent other were very similar in their ability to perform
the tasks in the vignettes. Variation in inputs and outcomes across vignettes created
conditions where the subject felt either under or over-rewarded. Results were again
largely predictive of the equity sensitivity propositions. In cases where the participant
was underrewarded, Entitleds reported less satisfaction than both Equity Sensitives and
Benevolents. In cases where the participant was overrewarded, both Entitled and
Benevolents (who focused less on outcomes in general) were more satisfied than Equity
Sensitives.

King et al. (1993) then conducted a second study where employees in a large bank
completed a survey assessing equity sensitivity, the perceived importance of work and
pay, distributive justice, and job satisfaction. Benevolents focused significantly more on
work than Entitleds, while Entitleds focused significantly more on pay than Benevolents.
Equity sensitives were sandwiched directly between on both counts, though not
significantly different than either group. Also, a positive correlation between equity-
based distributive justice and job satisfaction was illustrated for all three groups. This
suggests that, while no group prefers inequity, both Benevolents and Entitleds are more
tolerant of certain types of it.

Given the implications of equity sensitivity for reactions to under or overreward
circumstances, it seems logical that equity sensitivity could alter reactions to certain
process control configurations. However, it should be noted that equity sensitivity has

only been used to assess sensitivity to overreward or underreward circumstances -- not to

44

differences in procedural justice operationalizations. Nonetheless, a circumstance where
an individual is over or underrewarded in comparison to a referent other seems similar to
a case where an individual is more or less fairly treated procedurally in comparison to
one’s teammates, particularly in terms of the monistic view (Cropanzano & Ambrose,
1996)

Specifically, it seems likely that Benevolents will be less affected by different
process control configurations, given their lack of focus on equity issues relative to other
groups (King et al., 1993). They are less likely to attend to justice issues, so should be
less likely to notice differences in own or other’s process control. One would therefore
expect that the four contrasts shown in Figures 2-4 will explain less variance in outcomes
for Benevolents.

Hypothesis 4: The relationship between the five process control configurations and
procedural fairness perceptions, exit from the team’s task, conflict, leader evaluation,
and mistakes during role performance will be moderated by the individual member’s
equity sensitivity, such that the process control configurations will explain less
variance in the outcomes when the member is high in Benevolence.

Hypothesis 4a: The relationship between high vs. low individual process control

and procedural fairness perceptions, exit from the team’s task, conflict, leader

evaluations, and mistakes during role performance will be moderated by the
individual member’s equity sensitivity, such that individual process control will
be less beneficial when the member is high in Benevolence.

Hypothesis 4b: The relationship between high vs. low teammates’ process control

and procedural fairness perceptions, exit from the team’s task, conflict, leader

45

evaluations, and mistakes during role performance will be moderated by the
individual member’s equity sensitivity, such that teammates’ process control will
be less beneficial when the member is high in Benevolence.

Hymthesis 4c: The relationship between consistent vs. inconsistent process
control and procedural fairness perceptions, exit from the team’s task, leader
evaluations, conflict, and mistakes during role performance will be moderated by
the individual member’s equity sensitivity, such that the consistency of process
control will be less beneficial when the member is high in Benevolence.
Hypothesis 4d: The relationship between unclear vs. clear relative process control
and procedural fairness perceptions, exit from the team’s task, conflict, leader
evaluations, and mistakes during role performance will be moderated by the
individual member’s equity sensitivity, such that the lack of clarity of process

control will be less beneficial when the member is high in Benevolence.

46

Chapter 3
METHOD

Choice of Setting

A laboratory setting was used to test the dissertation’s hypotheses. This setting
was chosen for several reasons. First, the laboratory allows the researcher to maximize
the variance in independent variables (Kerlinger, 1986). It also allows the researcher to
minimize error variance and control confounds and contaminants through randomization.
Second, the laboratory setting made it possible to assess the degree to which process
control, task interdependence, and parallel condition had a causal, not just correlational,
relationship with outcomes.
Participants

Participants were 300 undergraduates enrolled in an introductory management
course at a large Midwestern University. One hundred sixty-four participants were male,
the remaining 136 were female. The mean age for the participants was 21.35 (SD = 2.40),
and their average grade point average was 3.07 (SD = .42). All participants received
course credit in exchange for their participation. They also were given a chance to earn a
small cash incentive ($10) based on their performance on the task.
12$

Participants worked on a special version of the Team Interactive Decision
Exercise for Teams Incorporating Distributed Expertise (TIDEZ) computer simulation. An
extensive description of this task is given in Hollenbeck, Ilgen, Sego, Hedlund, Major,
and Phillips (1995). Four participants, termed Alpha, Bravo, Charlie, and Delta, served as

a team and were stationed at networked computer terminals. Participants were required to

47

classify aircraft as friendly or threatening based on certain information about the aircraft.
Bravo, Charlie, and Delta (referred to as staff members) provided classification
recommendations to Alpha, who was actually one of the experimenters serving as a
confederate.

Each team member was responsible for gathering specific pieces of information
about each aircraft (e.g., speed, altitude, angle, direction, radar type). Bravo, Charlie, and
Delta were each assigned four specific pieces of information to gather, and they were
given extensive training on how to transform information values into probable threat
levels. Team members shared information with one another via the computer network.
Bravo, Charlie, and Delta each needed at least one piece of information from a teammate
to complete the four pieces of information they were assigned to gather. In contrast,
Alpha was purportedly responsible for gathering all 9 pieces of information, but
supposedly did not have the depth of interpretation training that Bravo, Charlie, and Delta
had. Bravo, Charlie, and Delta were not able to ask for or share information with Alpha.
However, all team members could type text messages to one another through the
computer network. Messages from Alpha were sent to all three staff members
simultaneously, and were visible to all team members.

Once Bravo, Charlie, and Delta had obtained the four pieces of information they
were in charge of gathering, they made a recommendation to Alpha in terms of probable
aircraft threat level. This recommendation took the form of a course of action to take
regarding the aircraft, on a seven point continuum of aggressiveness (l = Ignore; 2 =
Review; 3 = Monitor; 4 = Warn; 5 = Ready; 6 = Lock-On; Z = Defend). Once Bravo,

Charlie, and Delta had made their recommendation to Alpha, he or she combined those

43

recommendations to form one final decision. This combination was made using a script
which manipulated the extent to which specific members’ views were considered.

Alpha’s final decision was then compared to the correct decision. Teams were
given feedback via the computer on the absolute difference between their decision and
the correct decision. No difference was termed a Hit, a one point difference a Near Miss,
two points difference a Miss, three points difference an Incident, and four points or more
difference a Disaster. Bravo, Charlie, and Delta’s recommendations were also listed on
this feedback screen, along with aggregate information on how the team had performed
on all trials to date. The same procedure was repeated for 36 trials (the first 3 of which
were practice only). Trials alternated between 150 and 120 seconds in length.

In addition to the decision accuracy feedback, team members were given bogus
feedback on the correlation between their recommendation and Alpha’s final decision
over the course of the trials. This feedback took the form of a green bar which ranged
from 0 (no correlation) to 1 (a perfect correlation). In actuality this green bar was scripted
to achieve specific levels depending on the experimental condition.

Manipulations

The manipulations formed a 5 (Process control conditions) x 2 (High or low task
interdependence) x 2 (Parallel vs. non-parallel team) design. The process control
conditions created the five situations shown in the circles in Figures 2-5: (1) uniformly
high; (2) uniformly low; (3) the individual member clearly has more control than his/her
teammates; (4) the individual member clearly has less control than his/her teammates; (5)

the individual member’s control falls in-between that of his/her teammates, making

49

relative process control unclear. From these five conditions, the four orthogonal contrasts
delineated in Figures 2-5 were created.

Process Control. Degree of process control was manipulated in three ways. First,
the script used by the confederate to combine Bravo, Charlie, and Delta’s
recommendations into one overall decision weighed members’ recommendations equally
in some conditions and to varying degrees in others. Where the weights varied the
equation was as follows:

Decision = .38(Low Person’s Rec) + 1.00(Medium Person’s Rec) + 1.63(High Person’s Rec)

Thus the individual in the high process control condition had his or her
recommendation weighed 4.28 (1 .63/ .3 8) times that of the person in the low process
control condition. The relative magnitude of these recommendation weights were based
on a past study which used TIDE2 (Colquitt, Hollenbeck, Sheppard, Ilgen, & LePine,
1999). In that study, the actual mean correlation between the leader’s decision and each
staff member’s recommendation was .40, with a standard deviation of .25. The weights of
1.63 to 1.00 to .38 were scaled to be consistent with those study correlations of .65, .40,
and .15.

While participants could, to some extent, naturally perceive how much they were
being weighed by the leader, the process control manipulation also provided feedback on
the size of that correlation via the green bars mentioned above. That feedback was
manipulated to insure that it was exactly the same within experimental conditions. Green
bars were a maximum of 10 centimeters long. In conditions where process control was
unifomrly high the bars centered around an 8 centimeter length. In conditions where
process control was uniformly low they centered around a 2 centimeter length. In

conditions where process control varied within teams the levels were 8 centimeters, 5

50

centimeters, and 2 centimeters, respectively. All bars started out at a full 10 centimeters
and gradually assumed their final positions during trials 4 - 9. The first three trials were
practice trials during which participants were told what to decide. Even after trial 10 bars
would periodically vary one centimeter above and beyond the assigned position to show
participants that they were still being updated.

Finally, process control was further manipulated using text messages transmitted
from the confederate to Bravo, Charlie, and Delta. These messages indicated the presence
of process control. Participants could read messages sent to them by Alpha, and could
also read messages sent to their teammates by Alpha. It was explained that this type of
“monitorable” communication was used by Air Force teams. Sample messages were
“BRAVO: I leaned toward your call on that one,” “CHARLIE: I leaned toward your
input there,” and “DELTA: I made my choice with your call in mind.” Two other
messages were given by Alpha, one as a greeting during training (“ALL: Heythere. . .”)
and one on trial 9 to underscore the importance of making recommendations in a timely
manner (“ALL: I need about 20-30 seconds to look at everything”).

In conditions where process control was unifome high, each participant received
six text messages over the course of trials 9 - 36. In conditions where process control was
uniformly low, participants received only the two introductory messages, not any of the
process control messages. In conditions where process control varied within teams, team
members were sent 6, 3, and 0 messages, respectively. These frequencies were all based
on a past study which showed that team leaders sent, on average, a total of 10 messages

per 36 trials, with a standard deviation of 3 messages (Colquitt et al., 1999).

51

Figure 6 summarizes the process control manipulations as they appeared in
conditions which varied control within a team. The figure also shows how the TIDE2
computer screen appeared to participants. The parts of Figure 6 which are in parentheses
did not actually appear on the game screen. In Figure 6, Bravo’s process control has been
manipulated to be high. Specifically, his or her recommendations received a weight of
1.63 in Alpha’s script, his or her green bars centered around the 8 centimeter mark, and
he or she received 6 text messages connoting process control. At the other end of the
spectrum was Charlie, whose recommendations received a weight of .38, whose green
bar centered around 2 centimeters, and who received no text messages connoting process
control.

Judgement Carrier

Hcesurc Rccc Ive Query Transui t

       

           

Charl 1e—

15 Time left: 36

(Script ueiglitaSS)
(0 Process Control Messages)

(Script weightflfiS)
(Sffipl weight = 1.00) (6 Process (‘ontrol Messages)
(3 Process Control Messages)

Figure 6 - Summary of Process Control Manipulations

52

flak Interdependence. Task interdependence was manipulated by varying the
degree to which the information a participant was in charge of gathering and interpreting
was measurable by the participant. In low task interdependence conditions 3 of the 4
pieces of information needed by the participants were measurable. This left 1 piece of
information that had to be delivered by a teammate, and it was only available from one
specific teammate. In high task interdependence conditions 1 of the 4 pieces of
information needed by the participants was measurable. This left 3 pieces of information
that had to be delivered by a teammate. One of those pieces of information could be
delivered by either teammate, the other two could only be delivered by one specific
teammate.

Par‘allel vs. Non-parallel. In the parallel condition, the version of TIDE2 also
included an individual decision making task that ran parallel to the team decision making
task. This task was similar to the team task, except that individuals had to classify an
aircraft based on six entirely different pieces of information (e.g., exhaust, acceleration,
and trajectory), all of which were available to them via their own computer. The game
was structured so that the team task and the individual task both ran simultaneously on
the participants’ computers. They could go back and forth between the two tasks by
clicking on the respective windows, which were arranged as in Figure 7. There were an
unlimited number of trials on the individual task, each of which could last a maximum of
15 minutes, despite the fact that trials could easily be completed in under one minute.
This allowed team members to leave the individual task and come back to it at their own
discretion. For example, some participants would initially measure their information on

the team task. Then, while waiting to receive information from their teammates, they

53

would go to the individual task and measure the information there. They would then
return to the team task, make their recommendation, before concluding the trial on the

individual task.

Trans-It Jud ,- enent. Carrier

{15.1

Pleasure Receive Quer

              

  

ft : 36
Due:- Trans-l

lll

[ ( l‘rlUlllIR
l

1‘3 Tim- Iv-[t‘ 16

Figure 7 - Screen with Individual Task

Procedure

Upon entering the research facility, four participants were brought into one of the
computer rooms where they filled out a consent form. This form is shown in Appendix A.
One of the participants was then led from that room to another room, ostensibly to fill the
role of Alpha. It was explained that Alpha would be separated from the other three team
members because actual Air Force teams were characterized by such geographic
dispersion. In reality, this person then took part in an entirely separate research study.

The three remaining subjects then received a booklet which provided an overview of the

54

team decision making task. This booklet is shown in Appendix B. Participants were given
10 minutes to read through these booklets.

After they had finished reading the booklets, the participants were given “hands-
on” training regarding how to perform the task. During the training they would make use
of individual instruction sheets (shown in Appendix C), which explained how to interpret
the four pieces of information members were in charge of gathering. The hands-on
training lasted for an average of 45 minutes and involved three practice trials, the
outcomes of which were not counted toward the final task scores. In parallel team
conditions, one practice trial was completed on the individual task. This occurred
between the second and third team task practice trials. Here members used an individual
task instruction sheet, shown in Appendix D.

At the conclusion of the practice trials, participants completed the a survey which
included the measure of equity sensitivity. The experimenter then asked whether there
were any remaining questions about the task before the real trials began.

The first trial which counted toward the participants’ scores was trial 4. Typically
the experimenter remained in the room for a few more trials, to insure that participants
understood the task and were getting their recommendations to Alpha on time. Upon
leaving the room, the experimenter took over the role of Alpha (filled up to now by
another member of the research laboratory’s staff). During these initial trials, the green
bars gradually drifted toward their manipulated level, arriving there on trial 10. At this
point the experimenter re-entered the room to note:

At this point the statistic that forms the Green Bars should be accurate. Remember

that the Green Bar is just the correlation between your recommendation and

Alpha’s final decision. The longer the Green Bar the more you’re influencing
Alpha. The shorter the Green Bar the less you’re influencing Alpha. So over the

55

remaining trials you will be able to see how much your recommendation
influences Alpha’s final decision.

The halfway point of the simulation was Trial 19, after which there was a short
intermission to fill out a survey which included the justice condition manipulation check
items and the measure of procedural fairness perceptions.

Participants then completed the remaining trials (trials 20-36). After the final trial
they were given a final survey, which included measures of conflict and leader
evaluation, along with the task interdependence manipulation check items. Participants
were then given a general debriefing, which was supplemented by a full debriefing at the
conclusion of the semester.

Measures

Equity Sensitivity. Equity sensitivity was assessed using the measure validated in
King and Miles (1994). This measure asks respondents to allocate 10 points between the
two choices for each of five items. One item reads “1 would be more concerned about:
(A) What I received from the organization; (B) What I contributed to the organization.”
Another item reads “It would be more important for me to: (A) Get from the
organization; (B) Give to the organization.” A third item reads “My personal philosophy
in dealing with the organization would be: (A) If I don’t look out for myself, nobody else
will; (B) It’s better for me to give than to receive.” Scores are created by summing the
number of points allocating to option (B) for the five items. Thus higher scores represent
higher levels of Benevolence.

Fairness Perceptions. Procedural fairness perceptions were assessed with five
items. Items followed Lind and Tyler’s (1988) recommendations for both direct measures

(i.e., those which simply ask “how fair” something is) and indirect measures (i.e., those

56

which assess one of Leventhal’s fairness criteria). Examples of the former were “The
method Alpha used to make the final decision was fair,” and “The procedure Alpha used
to make the final decision was fair.” Examples of the latter included “The procedure
Alpha used to make the final decision was free of bias,” and “The procedure Alpha used
to make the final decision was based on accurate information.”

ELit. Exit was assessed using the nrunber of trials completed on the individual
task during trials 20-36 on the team task. Exit was assessed during these trials so that the
manipulation check and fairness items temporally preceded the measure of exit.
Individual task decision making accuracy was explored as a potential control variable, in
case certain participants completed a high number of trials because they did not really try
hard on them. However, this control variable did not alter any exit results.

ppm The seven-item measure used by Saavedra, Barley, and Van Dyne
(1993) was used to assess conflict. Sample items included “There was a lot of tension
between myself and others in the group,” “I found myself unhappy and in conflict with
members of my group,” “I got along with most people in this group,” and “Given the way
group members performed their roles I often felt frustrated.” (I = Strongly Disagree to Z
= Strongly Agree).

Leader Evaluation. Leader evaluation was assessed with three items: “Alpha did a
good job,” “Alpha tried hard to do well,” and “Alpha helped our team perform well.” (1 =
Strongly Disagree to Z = Strongly Agree).

Mistakes duringflile Performance. Mistakes during role performance were
measured using the square of the absolute difference between what the participant Md

have recommended to Alpha (based on the four pieces of information they were to

57

interpret) and what the participants actually did recommend, aggregated across trials 20-
36. As with Exit, trials 20-36 were used to provide temporal precedence between the
manipulation check and fairness items and role performance.

Manipulation Checlg Perceived own process control was assessed with three
items (g = .85): “I have had a lot of input into Alpha’s final decision,” “I have had a lot
of influence on Alpha’s final decision,” and “My recommendations have influenced
Alpha’s final decision.”

Perceived teammates’ process control was assessed with three similar items (_ =
.82): “My teammates have had a lot of input into Alpha’s final decision,” “My teammates
have had a lot of influence on Alpha’s final decision,” and “My teammates’
recommendations have influenced Alpha’s final decision.”

Consistency of process control was assessed with three items (_ = .76): “Alpha
has considered each of our views and opinions equally,” “Alpha has been consistent in
the way s/he has acted on our opinions in making the final decision,” and “Alpha has
treated everyone the same way.”

Perceptions of task interdependence were assessed using Pearce and Gregerson’s
(1991) eight-item measure (_ = .80). Sample items included “My own performance on
the team task was dependent on receiving accurate information from others,” “I had to
work closely with others in doing my work on the team task,” “The way I performed my
job on the team task had a significant impact on others,” and “I frequently had to

coordinate my efforts with others on the team task.”

58

Data Analysis Strategy

Multiple regression was used to test all hypotheses, with moderated multiple
regression used to test interactions. Direct effects of justice comparisons were assessed
by regressing the five outcome variables on the four orthogonal contrast-coded variables:
(I) own process control high vs. low; (2) teammates’ process control high vs. low; (3)
process control consistent vs. not consistent; and (4) process control unclear vs. clear in
relation to one’s teammates. Hypotheses with interactions then entered the moderator
variable on the second step, followed by the four product terms on the third step.

Analyses indicated that the AR2 for the four interaction terms averaged .03, while
the Total R2 for the regression equations averaged .16. Using Cohen and Cohen’s (1983)
Equation 4.5.2 to assess the statistical power of sets of independent variables, these
values resulted in statistical power of .75 with N = 300. While this represents an
acceptable level of power to test the interaction terms, analyses using the Exit variable
are based on only N = 150, because the Exit variable is unavailable in non-parallel team
conditions. Thus, in reviewing results with the Exit variable, results reaching a .10 level
of statistical significance are discussed. Such results are n_ot discussed in the context of

any other outcome variable.

59

Chapter 4
RESULTS
_M_anipulation Checks

The four process control contrasts explained 38% of the variance in perceptions of
own process control (E = 45.29, p < .001 ), with the own process control contrast having
the largest effect (I; = .50, p < .001). Unexpectedly, the teammates’ process control
contrast had a negative effect on perceptions of own process control (11 = -.3 7, p < .001).

The relationship between the manipulation of own process control and
perceptions of own process control was especially strong in conditions in which process
control differed within the team. In those conditions, the own control contrast explained
48% of the variance in perceptions of own process control (E = 165.05, p < .001, I; =
.69).

The four process control contrasts explained 22% of the variance in perceptions of
teammates’ process control (E = 20.38, p < .001), with the teammates’ process control
contrast having the largest effect (13 = .41, p < .001). Again unexpectedly, the own
process control contrast had a negative effect on perceptions of teammates’ process
control (13 = -.19, p < .05).

The four process control contrasts explained 18% of the variance in perceptions of
the consistency of process control (E = 16.31, p < .001), with the consistency of process
control contrast having a significant effect (I; = .35, p < .001).

The interdependence dummy code explained 5% of the variance in perceptions of

task interdependence (_E = 15.68, p < .05).

60

Descriptive Statistics

The means, standard deviations, and zero-order intercorrelations for all variables
are shown in Table 1. Table 1 also includes participants’ perceptions of their own process
control. Coefficient alpha’s for self-report variables are shown on the diagonal.
Correlations among moderator and outcome variables were not the subject of any explicit
hypotheses, but several are notable. The positive correlation between parallel vs. non-
parallel team and conflict (; = .18) is consistent with the literature linking such teams to
higher levels of role conflict (Cohen & Bailey, 1997). Equity sensitivity was negatively
related to conflict (g = -.15) and positively related to leader evaluation (; = .19),
consistent with the existing nomological network for that construct (e. g., King & Miles,
1994). Procedural fairness perceptions were negatively related to conflict (g = -.35),
positively related to leader evaluation (; = .44), and negatively related to mistakes during
role performance (I = -. 12). This suggests that perceptions of fairness were not just
important to affective reactions, but also to objective levels of performance.

Tests of Process Control Direct Effects.

Hypothesis I predicted that the five process control configurations (expressed in
the form of the four orthogonal contrasts) would explain significant variance in
procedural fairness perceptions, exit from the team’s task, conflict, leader evaluation, and
mistakes during role performance. The regression results which directly test the
hypotheses are shown in Table 2. The means of the outcome variables across the different
process control configurations are shown in Figure 8. The five conditions on the
horizontal axis recreate the five circles shown in Figures 2-5. The effects of individual

contrasts in Figure 8 can be seen by constructing the contrast using the configuration

61

Table 1 - Descriptive Statistics for Dissertation Variables

 

M SD (1) (2) (3) (4) (5)

 

(1) Perceptions of Own PC 2.90 .93 ---

Process Control Conditions

(2) Own PC .00 .45 .50* ---

(3) Teammate PC .00 .45 -.37* .00 ---

(4) Consistency of PC .00 .45 .01 .00 .00 ---

(5) Unclear vs. clear PC .00 .50 -.04 .00 .00 .00 ---
Moderators

(6) Task Interdepend 1.50 .50 .00 .00 .00 .00 .00
(7) Parallel vs. Non 1.50 .50 .00 .00 .00 .00 .00
(8) Equity Sensitivity 27.22 5.70 -.05 -.05 .07 -.02 .03
Outcomes

(9) Fairness Perceptions 3.43 .62 .38* .32* -.11* .34* .03
(10) Exit 13.65 8.32 .05 -.10 -.10 -.01 -.21*
(11) Conflict 2.08 .58 -.10 -.19* -.07 -.03 -.10
(12) Leader Evaluation 3.78 .71 .18* .29* .06 .02 .13*
(13) Mistakes during 1.72 1.52 .03 .00 .07 -.18* -.20*

Role Performance

 

Note. N = 300, except for correlations with Exit. That variable only exists in parallel
teams, thus N = 150. * p < .05.

62

Table 1, continued - Descriptive Statistics for Dissertation Variables

 

(6) (7) (8) (9) (10) (11) (12)

 

(1) Perceptions of Own PC
Process Control Conditions

(2) Own PC

(3) Teammate PC

(4) Consistency of PC
(5) Unclear vs. clear PC

Moderators

(6) Task Interdepend ---

(7) Parallel vs. Non .00 «-

(8) Equity Sensitivity -.03 .05 (.79)

Outcomes

(9) Fairness Perceptions -.08 .01 .06 (.71)

(10) Exit -.14 --- -.06 .01 «-

(1 1) Conflict .04 .18* -.15* -.35* .02 (.78)

( 12) Leader Evaluation -.13* -.03 .19* .44* -.02 -.42* (.87)
(13) Mistakes during .07 .10 .05 -.12* -.01 .10 -.09

Role Performance

 

Note. N = 300, except for correlations with Exit. That variable only exists in parallel
teams, thus N = 150. * p < .05.

63

Table 2 - Beta-weights and Regression Results for Hypothesis 1

 

Fairness Exit Conflict Leader Mistakes

 

 

Eval
Own PC .32* -.10 -.18* .29* .00
Teammates’ PC -.11* -.10 -.07 .06 .07
Consistency of PC .34* -.01 -.03 .02 -.18*
Clarity of PC .03 -.21"‘ -.10t .13* -.20*
Total R2 23* .O6* .05* .11* .07*
Overall F 2206* 240* 3.76* 8.69* 588*

 

m Table contents are beta-weights. N = 300, except for correlations with Exit. That
variable only exists in parallel teams, thus N = 150. t p < .10; * p < .05.
labels. For example, the effects of own process control (PC) on fairness perceptions can
be seen by comparing the average of the “Low Own PC, Low Team PC” and “Low Own
PC, High Team PC” with the average of the “High Own PC, Low Team PC” and “High
Own PC, High Team PC.” That yields a contrast of approximately 3.2 for low own
process control and 3.65 for high own process control, suggesting a main effect of own
process control on fairness perceptions (a result verified by the regression results
discussed below).

As shown in Table 2, the four process control contrasts explained a significant
23% of the variance in procedural fairness perceptions, 6% of the variance in exit, 5% of
the variance in conflict, 11% of the variance in leader evaluation, and 7% of the variance

in mistakes during role performance. Nine individual contrasts exerted significant effects,

64

 

s.

 

Fair-cu Perception
l.) L.)
r; I“ \r
w LII V! VI

N
\n
u»

 

 

N
v.

Unclear Low Own Low Own Higi Own Hidi Own
Condition PC. Low PC. High PC. Low PC, Hiyi
Team PC Team PC Team PC Tenn PC

Protest Control Camille-

 

 

 

   

 

 

 

Unclear Low Own low Own Hiyi Own Higi Own Uncle! Low Own bow Own Hiyi Own Hidi Own
Condition PC. Low PC, Higi PC. Low PC. Hiyi Condition PC, bow PC. Hidi PC, Low PC, Hid!
Tenn PC Team PC Tenn PC Team PC Team PC Tech PC Ten PC Tm PC
Process Control Contitiol Pmm Control Coldfiol
4.25 2.5
4.05- 235

[nthr Meatlo-
s» s»
a 8
't'. N

P
&
‘JI
.—
N
ll.

  

 

 

 

   

Mistakes tiring Role Performance
3

3.25 ‘ . , l
Unclear [0W Own Low Own Higl 0W“ HI?! Own Unclear Low Own Low Own Higi Own Hiyi Own
Condltlon PC . Low PC, High PC. LOW PC. High Condition PC, Low PC. High PC. Low PC, High
Team PC Team PC Teln PC Team PC Team PC Team PC Tm PC Tm PC
Process Control Contition Process Control Condition

Figure 8 - Cell Means for Experimental Conditions

65

and all four contrasts had significant influences on at least one outcome variable. Taken
together, these results provide support for Hypothesis 1. Specific results for each of the
outcome variables are discussed below.

Fairness Perceptions. The significant effects on fairness perceptions were due
primarily to three of the four contrasts: own process control, teammates’ process control,
and consistency of process control. Table 2 shows the standardized regression
coefficients, which are identical to the zero-order correlations because the contrasts are
orthogonal and the group sizes are equal (Cohen & Cohen, 1983). Fairness perceptions
were higher when individual team members were granted high levels of process control
(13 = .32), partially supporting Hla. Contrary to Hlb, fairness perceptions were higher
when one’s teammates’ were given lo_w levels of process control (I; = -.11). Consistent
with ch, fairness perceptions were higher when levels of process control were
consistent within team (I; = .34). The effect of unclear vs. clear relative process control
was not significant.

E_xip The significant effects on exit were due solely to one contrast: unclear vs.
clear relative process control. Members were less likely to withdraw from the team’s task
when their relative process control was unclear than when it was clear (i.e., obviously
equal to, higher than, or less than teammates’ levels) (3 = -.21), partially supporting Hld.
This effect is very evident from Figure 8, where the mean levels of exit for the unclear
condition are markedly different from those of the other conditions.

Conflict. The significant effects on conflict were due primarily to one contrast:
own process control. Members perceived less conflict in their team when their levels of

process control were high (Q = -.18), partially supporting Hla.

66

Leader Evaluation. The significant effects on leader evaluation were due
primarily to two contrasts: own process control and the clarity of relative process control.
Individuals rated their leader more favorably when granted high levels of process control
(13 = .29), partially supporting Hla. Individuals also rated their leader more favorably
when their relative process control was unclear than when it was clear (i.e., obviously
equal to, higher than, or less than teammates’ levels) (fl = .13).

Mistakes during Role Performance. The significant effects on mistakes during
role performance were due primarily to two contrasts: consistency of process control and
the clarity of relative process control. Individual members made fewer mistakes in terms
of their recommendations to Alpha when the level of process control was consistent
within the team (fl = -.18), partially supporting ch. Members also made fewer mistakes
in terms of role performance when their relative process control was unclear than when it
was clear (i.e., obviously equal to, higher than, or less than teammates’ levels) (13 = -.20).
Tests of Process Control x Tag Interdependence Interaction Effects.

Hypothesis 2 predicted that the effects of the five process control configurations
(expressed in the form of the four orthogonal contrasts) would be moderated by task
interdependence, such that the effects would be weaker at low levels of interdependence.
The regression results used to test this hypothesis are shown in Table 3.

As shown in Table 3, process control x task interdependence interactions
explained an incremental 1% of the variance in procedural fairness perceptions, 3% of the
variance in exit, 2% of the variance in conflict, 1% of the variance in leader evaluation,
and 4% of the variance in mistakes during role performance. The exit effect approached

significance, while the role performance effect attained significance. In total, five

67

Table 3 - Beta-weights and Regression Results for Hypothesis 2

 

 

 

 

 

Step: Fairness Exit Conflict Leader Mistakes
Eval

(1) Own PC .32* -.10 -.18* .29* .00
Teammates’ PC -.11* -.10 -.07 .06 .07
Consistency of PC .34* -.01 -.03 .02 -.18*
Clarity of PC .03 -.21* -.10t .13* -.20"‘
R2 23* .06* .05* .1 1* .07*

(2) Task Interdep -.03 -.10' .04 -.12* .16*
AR2 .01 .01 ‘ .00 .02* .03*

(3) T1 x Own PC -.14 -.3ot .24 .20 -.12
TI x Teammates’ PC -.28t -.13 .11 -.21 .25
TI x Consistency PC .00 .38* -.36* .09 -.44*
TIxClarity PC .14 .12 .18 .04 .31*
AR2 .01 .03t .02 .01 .04*

Overall F 10.64* 2.34* 2.55* 4.83* 2.94*

 

Note. Table contents are beta-weights. N = 300, except for correlations with Exit. That
variable only exists in parallel teams, thus N = 150. t p < .10; * p < .05.

68

significant interactions were observed (counting Exit effects which approached
significance). The plot of each of the interactions was as predicted, with the process
control contrasts having weaker effects at high interdependence levels. A sample plot is
shown in Figure 9. Taken together, these results provide partial support for Hypothesis 2.

Specific results for each of the outcome variables are discussed below.

 

2.5 __ .. ______ , ,,_

 

 

 

....... Low Task Int

L

p.
I
I

 

 

Mistakes during Role Performance
N

 

 

fl

 

Inconsistent Consistent

Process Control Contrast

Figure 9 - Plot of Task Interdependence Moderation Effect

Fairness Perceptions. The interaction of task interdependence and the four process
control contrasts did not explain significant variance in fairness perceptions.

EL“; The interaction between task interdependence and the four process control
contrasts, which approached significance, was due primarily to interactions with own
process control and consistency of process control. The pattern was such that the negative

relationship between own process control, consistency of process control, and exit was

69

less negative when task interdependence was low than when task interdependence was
high. This offers partial support for Hypotheses 2a and 2c.

Conflict. The interaction of task interdependence and the four process control
contrasts did not explain Significant variance in conflict. While the task interdependence
x consistency of process control contrast had a significant t value, that interaction cannot
be interpreted because the overall step was not significant.

Leader Evaluation. The interaction of task interdependence and the four process
control contrasts did not explain significant variance in leader evaluation.

Mistakes duringRole Performance. The Significant interaction between task
interdependence and the four process control contrasts was due primarily to interactions
with consistency of process control and clarity of relative process control. The pattern of
the former interaction was such that the relationship between consistency of process
control and mistakes during role performance was less negative when task
interdependence was low than when it was high. Similarly, the pattern of the latter
interaction was such that the relationship between unclear vs. clear relative process
control and mistakes during role performance was less negative when task
interdependence was low than when it was high. These interactions offer partial support
for Hypotheses 2c and 2d.

Tests of Process Control x Parallel vs. Non-parallel Team Interaction Effects.

Hypothesis 3 predicted that the effects of the five process control configurations
(expressed in the form of the four orthogonal contrasts) would be moderated by whether
the team was parallel or non-parallel, such that the effects would be weaker in parallel

teams. The regression results used to test this hypothesis are shown in Table 4.

70

Table 4 - Beta-weights and Regression Results for Hypothesis 3

 

 

 

 

 

Step: Fairness Conflict Leader Mistakes
Eval
(1) Own PC .32* -.18"‘ .29* .00
Teammates’ PC -.11* -.07 .06 .07
Consistency of PC .34* -.03 .02 -.18*
Clarity of PC .03 -.10‘ 13* -.20*
R2 23* .05* .11* .07*
(2) Parallel vs. Non .01 .18* -.03 .10
AR2 .00 .03* .00 .01 ‘
(3) Para x Own PC .07 -.16 -.18 -.20
Para x Teammates’ PC -.40* .19 -.53* .24
Para x Consistency PC .07 .02 .20 .02
Para x Clarity PC -.10 .08 -.O6 .20
AR2 .02 .01 .04* .01
Overall F 10.65* 3.04 5.34* 3.47*

 

Note. Table contents are beta-weights. N = 300, except for correlations with Exit. That

variable only exists in parallel teams, thus N = 150. t p < .10; * p < .05.

71

As shown in Table 4, process control x parallel vs. non-parallel interactions
explained an incremental 2% of the variance in procedural fairness perceptions, 1% of the
variance in conflict, 4% of the variance in leader evaluation, and 1% of the variance in
mistakes during role performance. Exit is not shown as a dependent variable in Table 4
because that variable does not exist in non-parallel teams. Only the effect for leader
evaluation attained statistical significance. In total, only one significant interaction was
observed, with the plot as predicted (see Figure 10). The process control contrast had a
weaker effects in parallel teams. Taken together, these results provide only weak support

for Hypothesis 3. Specific results for each of the outcome variables are discussed below.

 

 

 

 

 

 

 

 

 

4
.-
3.75 ’\,; ___- _ -_..

5 \o
E 3.5 ______ 7 , 7 ParallelTeam
E . . ... . . Non-parallel team
.8 ---- __ --.- -
a
3 3.25 . _, ,- _ __ _ _

3 .

Teammates' Low Teammates' High

Process Control Contrast

Figure 10 - Plot of Parallel vs. Non-parallel Moderation Effect

Fairness Perceptions. The interaction of parallel vs. non-parallel team and the four

process control contrasts did not explain significant variance in fairness perceptions.

72

Conflict. The interaction of parallel vs. non-parallel team and the four process
control contrasts did not explain significant variance in conflict.

Leader Evaluation. The significant interaction between parallel vs. non-parallel
team and the four process control contrasts was due primarily to the interaction with
teammates’ process control. The pattern of the interaction was such that the relationship
between teammates’ process control and leader evaluation was less positive when the
team was parallel than when it was non-parallel. This represents partial support for
Hypothesis 3b.

Mistakes during Role Performance. The interaction of parallel vs. non-parallel
team and the four process control contrasts did not explain significant variance in
mistakes during role performance.

Tests of Process Control x Eguig Sensitivig Interaction Effects.

Hypothesis 4 predicted that the effects of the five process control configurations
(expressed in the form of the four orthogonal contrasts) would be moderated by
individual members’ equity sensitivity, such that the effects would be weaker for
members high in Benevolence. The regression results used to test this hypothesis are
shown in Table 5.

As shown in Table 5, process control x equity sensitivity interactions explained
3% of the variance in procedural fairness perceptions, none of the variance in exit, 4% of
the variance in conflict, 4% of the variance in leader evaluation, and 2% of the variance
in mistakes during role performance. The effects for procedural fairness, conflict, and
leader evaluation reached statistical significance. In total, six interaction effects were

significant. The plots of five of the six were as predicted, with the process control

73

Table 5 - Beta-weights and Regression Results for Hypothesis 4

 

 

 

 

 

Step: Fairness Exit Conflict Leader Mistakes
Eval

(1) Own PC .32* -.10 -.18* .29* .OO
Teammates’ PC -.1 1* -.10 -.07 .06 .07
Consistency of PC .34* -.01 -.03 .02 -.18*
Clarity of PC .03 -.21* -.10t .13* -.20*
R2 23* .O6* .05* .11* .07*

(2) Equity Sensitivity .01 t -.04 .04 .19* .16*
AR2 .01 ‘ .00 .02* .03* .00

(3) ES x Own PC -.23 .02 .36t -.40* .30
ES x Teammates’ PC -.51* .03 .59* -.68* -.09
ES x Consistency of PC -.41* .08 .53* -.13 .32
ES x Clarity PC .30t .13 -.21 -.17 -.40*
AR2 .03* .00 .04* .04* .02t

Overall F 11.82* 1.14 4.01 * 6.98* 3.57*

 

Note. Table contents are beta-weights. N = 300, except for correlations with Exit. That
variable only exists in parallel teams, thus N = 150. t p < .10; * p < .05.

74

contrasts having weaker effects for Benevolent individuals. A sample plot is shown in
Figure 11. Taken together, these results provide support for Hypothesis 4. Specific results

for each of the outcome variables are discussed below.

 

 

 

 

2.5
I, .
2.25 A -- -_ _ ,, L.
E r—e— Benevolents
s 2 — v
8 - - -I- - - Non-benevolents E
1.75 ——fi ,, .. ., -7-
1.5

 

 

 

Own Low Own High

Process Control Contrast

Figure 11 - Plot of Equity Sensitivity Moderation Effect

Mess Perceptions. The significant interaction between equity sensitivity and
the four process control contrasts was due primarily to interactions with teammates’
process control and the consistency of process control. The pattern of the former
interaction was contrary to predictions, as the relationship between teammates’ process
control and fairness perceptions was more negative for Benevolent members. The pattern
of the latter interaction was consistent with predictions, however. The relationship

between consistency of process control and fairness perceptions was less positive for

75

Benevolent members. These interactions therefore fail to support Hypothesis 4b, but do
support Hypothesis 4c.

ELit. The interaction of equity sensitivity and the four process control contrasts
did not explain significant variance in exit.

Conflict. The significant interaction between equity sensitivity and the four
process control contrasts was due primarily to interactions with teammates’ process
control and the consistency of process control. The pattern of both interactions was as
predicted, as the relationship between teammates’ process control, consistency of process
control, and conflict was less negative for Benevolent members. These interactions
therefore support Hypotheses 4b and 4c.

Leader Evalum The significant interaction between equity sensitivity and the
four process control contrasts was due primarily to interactions with own and teammates’
process control. The pattern of both interactions was as predicted, as the relationship
between own process control, teammates’ process control, and leader evaluation was less
positive for Benevolent members. These interactions therefore support Hypotheses 4a
and 4b.

Mistakes during Role Performance. The interaction of equity sensitivity and the
four process control contrasts did not explain significant variance in mistakes during role
performance. The interaction with the unclear vs. clear contrast was significant, but the
overall step only approached significance.

Supplementag Analyses
As mentioned in the discussion of the TIDE2 simulation in the Method section,

participants could send text messages to one another through the computer network. They

76

could also send text messages to Alpha, some of which may have dealt with reactions to
the process control manipulations. 1 content coded the text messages sent from
participants, focusing specifically on messages connoting conflict over the process
control configurations. Examples of such messages included “Listen a little more!”, “I
know what I’m doing!”, “Don’t ignore me.”, “Spread it around,” and “You should have
listened elsewhere, but oh well.”

Such messages were, in general, very rare, with a mean of .20 messages (SD =
.78) sent per individual, over the course of the entire experiment. They were most
prevalent in cases where an individual had lower process control than his or her
teammates (Mean = .36, SD = 1.09). They were least prevalent in cases where an
individual was in a team with uniformly high process control (Mean = .08, SD = .27).
The differences between these two conditions in terms of conflict text messages
approached significance (E(l, 117) = 3.49, p < .06). The other three conditions hovered
around the overall mean in terms of text message frequency, and were not statistically
different than the two aforementioned conditions. Thus, regressing conflict text messages
. on the four process control contrasts yielded non-Significant results (32 = .01, E = .96, p <
.43).

However, the effects of the process control contrasts on conflict text message
frequency were moderated by the nature of the team. The interaction of the four contrasts
and the parallel vs. non-parallel manipulation explained an incremental 4% of the
variance in text message frequency (E-Change = 2.64, p < .05). The bulk of this effect
was due to interactions with the other teammates’ contrast Q = 2.20, p < .05) and the

consistency contrast (t = -2.00, p < .05). The pattern of both of these effects was such that

77

the process control contrasts had a weaker effect on conflict text messages in parallel
teams. Presumably participants in parallel teams exited as a substitute for sending such

messages.

78

Chapter 5
DISCUSSION

As noted at the outset, the literature on team effectiveness has neglected the role
that the manager can play in fostering team performance, particularly in contexts where
the manager is incapable of leveraging task, team, or member factors. One variable which
could provide managers with a means of aiding their teams is procedural justice, which
has been shown to have many beneficial effects in individual contexts (Greenberg, 1990).
Such findings cannot be readily generalized to team settings, however, because collective
settings bring with them complexities that have not been addressed in the procedural
justice literature (James, 1993). Specifically, the collective nature of team settings begs
the question “is procedural justice a relative phenomenon or an absolute phenomenon?”

To attempt to capture the full complexity of procedural justice in team settings,
the effects of five possible configurations (shown in Figures 2-5) were modeled. Three of
these configurations modeled cases where justice levels vary within teams, particularly
likely given the leader-member exchange differences, cross-functionality, and diversity
that mark most teams. Differences among the five configurations were expressed in the
form of four orthogonal contrasts: own process control, teammates’ process control,
consistency of process control, and unclear vs. clear relative process control (see Figures
2-5) The effects of the process control configurations was examined using a theoretically
grounded set of outcomes based on the exit-voice-loyalty-neglect framework for

responses to dissatisfaction (Hirschman, 1970; Rusbult et al., 1988).

79

Discussion of Direct Effects of Process Control Configurations

The direct effects of the five process control configurations were examined for
fairness perceptions, exit, conflict, leader evaluation, and mistakes during role
performance. Fairness perceptions are a fundamental criterion in almost all justice
studies, and the positive relationship between process control and fairness perceptions is
among the most robust in the literature (Lind & Tyler, 1988). The results of this study are
consistent with past research, as high levels of individual process control increased
fairness perceptions. This provides further support for Thibaut and Walker’s seminal
theory of procedural justice (1975) -- that indirect control over a procedure enhances the
perceived fairness of it.

Interestingly, teammates’ process control was negatively related to an individual
member’s perceptions of procedural fairness. This initially counterintuitive result can be
explained by considering the results of the process control manipulation checks. Recall
that the teammates’ process control contrast was negatively related to own perceptions of
process control. Similarly, the own process control contrast was negatively related to a
member’s perceptions of his or her teammates’ process control. These results point out
that, in a team decision making context, process control is in many ways a “zero-sum”
situation. If a leader grants more voice to one member, that comes at the cost of less input
afforded to the other members.

This issue of process control in group settings has gained increased attention in
recent research on teams (Barker, 1993; Cohen & Bailey, 1997; Janz et al., 1997).
Process control takes the form of increased autonomy in teams marked by self-direction

and self-management. Cohen and Bailey’s (1997) review asks the question “Is team

80

autonomy a form of control?,” as norms of consensus and equality within teams minimize
the discretion of any one member (p. 231). Barker (1993) provided ethnographic
evidence that team settings can create “concertive” control that mimics the control
formerly afforded to direct supervisors. Similarly, Janz et al. (1997) provided empirical
evidence that the interdependence that characterizes teams can impinge on members’
process control.

Thus teammates’ process control was likely seen as hindering members’ own
process control. From the standpoint of the instrumental model of justice, such
hindrances compromise individuals’ ability to protect their long-term interests (Thibaut &
Walker, 1975). From the standpoint of the relational model of justice, such hindrances
reduce the status-signaling benefits of voice. The end result is a negative relationship
with procedural fairness perceptions. This suggests that the teammate contrast might
yield different results if procedural justice was operationalized as bias suppression or
interpersonal treatment rather than process control, though that remains an empirical
question which should be examined in future research.

In addition to the impact of own and others’ process control, the consistency of
process control contrast affected fairness perceptions. Specifically, procedures were
judged to be most fair when process control levels were consistent within the team.
Leventhal’s (1980) writings on procedural justice suggested that consistency was a
fundamental criterion of procedural justice. In practice, however, consistency has usually
been operationalized in terms of consistency across time, not persons. This is indicative
of the narrow, individualized focus of past research in the procedural justice domain.

Moreover, one of the few studies to examine across-persons consistency found that

81

 

 

im

re]
be
a r

162

the
the

Ire;

prc
jud
or i

Pyg

Pro
1‘ 6&5

P051

individuals preferred inconsistent treatment if it meant they were advantaged (Lind et al.,
1999). The benefits of consistency observed here were not in evidence in their study.

A different process control contrast predicted exit -- the clarity of a member’s
relative process control. Exit was lowest when a member’s process control fell in-
between that of his or her teammates. Exit was much more likely to occur in cases where
a member’s process control was clearly less than or greater than that of his or her
teammates, or where process control was uniformly low. In large teams it is likely that
most members will find themselves in this unclear circumstance, as some members have
more input but others have less. Thus this represents the first examination of a relatively
fundamental contrast.

It is likely that members in the unclear case focused their referent comparisons on
the member afforded less control. Thus they were able to feel that they were treated better
than someone else, but avoided being in the uncomfortable condition of being the “best
treated” team member. Moreover, the fact that their control was neither unexpected low
or unexpected high may have prevented them from even considering issues related to
procedural justice. Past research has shown that attribution seeking begins the justice
judgment process, and attribution seeking is most likely when confronted by unexpected
or negative circumstances (Brockner & Weisenfeld, 1996; Kelly & Michela, 1980;
Pyszczynski & Greenberg, 1981; Wong & Weiner, 1981). Brockner and Weisenfeld
(1996) suggest that the default assumption on the part of most individuals is fair
procedural treatment. Thus, because the person in the unclear circumstance had little
reason to examine the reasons behind his or her treatment, they likely were left with that

positive default assumption.

82

Unlike fairness perceptions and exit, conflict was only affected by members’ own
treatment. That is, the only contrast that affected conflict directly was individual
member’s own process control, with conflict levels being lower when members were
afforded more control. While the lack of effects for the other contrasts is disappointing,
the fact that the relationship between procedural justice and conflict generalizes to team
settings is important in-and-of-itself. As noted at the outset, conflict is a key feature of
many team failures (J ehn, 1995). To the extent that procedural justice can reduce conflict,
it becomes especially applicable to team settings.

Leader evaluations were affected by a member’s own process control, as with
conflict, but also by the clarity of relative process control, as with exit. Leader
evaluations have been one of the most commonly examined outcomes in the justice
literature (Greenberg, 1990), and the relationship between own process control and leader
evaluations is consistent with past research. However, this is the first study to Show that
individuals look to the procedural justice of others in making leader evaluations. As with
exit, the most favorable results for leadership evaluation were among the individuals who
fell in-between the treatment levels of their teammates.

Perhaps the most fundamental criterion examined here was mistakes during role
performance. Interestingly, mistakes were not affected by personal levels of treatment at
all, but rather by consistency and clarity of process control. Individual members made
fewer mistakes when their process control was consistent with that of their teammates,
and also when their levels fell in-between those of their teammates. Thus, ultimately,

performance was affected by the treatment of others, not of oneself. Linking procedural

83

jt

C(

justice to performance has been rare in any context (Greenberg, 1990), as researchers
focus more on affective outcomes (e.g., satisfaction, organizational commitment).

The results reviewed above make several contributions to the existing literature
on procedural justice. For example, this study showed that process control was capable of
predicting outcomes that were relevant in team contexts. Only one other study has
examined procedural justice constructs in the context of a group or team -- Korsgaard et
al. (1995). That study examined commitment to the team’s decision, but did not examine
any behavioral or performance variables. Thus this study echoes their contention that
procedural justice is important in teams, but goes beyond their study to show that it can
impact behaviors and performance, not just perceptions.

A second, and more significant contribution, of these results is the important role
played by other people’s treatment levels. Past research on procedural justice has been
egoistic, and has been driven by a focus on absolute, institutionalized norms of treatment
such as Leventhal’s (1980) justice rules (Cropanzano & Ambrose, 1986). As a result,
James (1993) suggested that current justice research has never adequately captured the
complexity of justice as it is experienced in collectives.

These results show that individuals are indeed affected by the procedural justice
that others experience -- they do not only attend to their own treatment levels. There were
a total of 10 significant direct effects for the process control contrasts. Of those 10, only
three were for the own process control contrast. Seven were significant effects for
contrasts requiring an individual team member to attend to the justice levels of others. If
interaction effects are considered, then there were 21 significant effects for the process

control contrasts, five of which were for the own process control contrast. Thus over

84

three-quarters of the effects demonstrated here show that individuals attend to others’
procedural treatment.

The fact that individuals look to others’ treatment is consistent with the relative
deprivation framework from which the justice literature developed. Nonetheless, past
research has shown little or no support for a relative perspective when it comes to
procedural justice in general, or process control in particular. Ambrose et al. (1989) found
no effects for other people’s process control on fairness perceptions, nor did own levels
interact with others’ levels. Ambrose and Kulik (1989) found similarly unsupportive
results -- no main or interactive effects of others’ process control. Grienberger et al.
(1993) found no main effect for others’ control, though they did find one significant
interaction effect. Finally, Lind et al. (1999) found that injustices to others mattered very
little to the participants in their study, when compared with injustices to the persons
themselves. The sole criterion in all these studies was fairness perceptions. It is likely that
the results would have been even less supportive of a relative perspective if other, more
distal outcomes had been examined.

Despite the dearth of support for effects of others’ procedural treatment, two
theories in the justice domain argue that such effects should occur. Folger’s (1986, 1987)
referent cognitions theory argues that fairness perceptions are derived from mental “what
ifs” regarding the treatment that has occurred. While the bulk of past research has used
Thibaut and Walker’s (1975) and Leventhal’s (1980) institutionalized procedural rules as
inputs to those “what ifs,” F olger did note that social comparisons could also be relevant.
Thus referent cognitions theory does provide a framework for examining process control

from a relative perspective.

85

While amenable to examining relative effects, referent cognitions theory has
spurred very little research. Cropanzano and Ambrose (1996) argue that this is because
the means vs. end distinction between procedural and distributive justice has created an
artificial distinction, wherein theories from the two domains are never integrated. This
distinction explains, in part, why distributive justice research has been built on the
concept of social comparison, whereas procedural justice has only examined such
comparisons in four published studies. Cropanzano and Ambrose (1996) proposed a
monistic theory of justice, which proposes an integration of procedural justice and equity
theory concepts. The effects reviewed above suggest that, to examine procedural justice
in teams, one should make use of the referent cognitions and monistic theory
frameworks. Relying on other theories which advance from an individual perspective will
not capture the full complexity of procedural justice as experienced in collective contexts.
Discussion of Interaction Effects with the Process Control Configmtions

The results reviewed in the previous section show that procedural justice does
have an impact in teams. Given this, the question becomes when is that impact especially
strong, and when is that impact especially weak. To begin to explore that question, this
dissertation examined three moderators of procedural justice’s effects, moderators based
on the team’s task, the team’s nature, and the team’s members. These moderators were
task interdependence, parallel vs. non-parallel team, and equity sensitivity, respectively.

Overall, eleven interactions between process control configurations and
moderator variables were observed. The most significant moderator was member equity

sensitivity, which yielded six significant interactions. Task interdependence also had

86

multiple significant interactions, yielding three effects. The parallel vs. non-parallel team
distinction resulted in only one significant interaction.

The four interactions with task interdependence Showed that the process control
contrasts had less significant effects on exit and mistakes during role performance at low
interdependence levels. At high levels of task interdependence, own process control and
consistency of process control reduced exit, but those effects were neutralized at low
levels of interdependence. Similarly, at high levels of task interdependence, consistency
and lack of clarity of process control reduced mistakes during role performance. Again,
those effects were neutralized at low levels of interdependence. As noted in the
introduction, high levels of interdependence increase members’ felt responsibility for
team outcomes, as well as their psychological identification with the team. Such effects
likely make own and teammate process control more critical, and make consistent
process control more appropriate (Kabanoff, 1991; Lind et al., 1999).

Less support was shown for parallel vs. non-parallel team as a moderator. It was
expected that parallel team environments would neutralize the effects of the process
control contrasts, for two reasons. First, parallel team demands prevent members from
spending all their time in the team role, limiting psychological involvement with the team
and group influences on the member. Second, the option of exit in parallel teams makes
alternative reactions such as loyalty and neglect less likely. Unfortunately, the results
showed little support for these predictions. Nonetheless, the results did support an
interaction with teammates’ process control, such that teammates’ control had a more
positive effect on leader evaluation when the team was non-parallel. Moreover, the

supplementary analyses of participants’ text messages showed that the sending of such

87

messages was driven less by process control conditions in parallel teams than it was in
non-parallel teams. Presumably the members of parallel teams exited rather than sending
such messages.

More support was shown for equity sensitivity as a moderator, with six significant
interaction effects. A stream of research on equity sensitivity has Shown that individuals
high in Benevolence attend less to outcome fairness issues (Huseman et al., 1987; King et
al., 1993; Miles et al., 1989). However, the construct had never been examined in
conjunction with procedural justice. The results indicated that, for five of the six effects,
process control configurations had weaker effects on outcome variables for Benevolent
individuals, as expected. For individuals low in Benevolence, consistency of process
control had positive effects on fairness perceptions; teammates’ and consistency of
process control had negative effects on conflict; and own and teammates’ process control
had positive effects on leader evaluation. Each of these effects were neutralized for
Benevolent individuals.

Taken together, these results suggest that justice configurations are less critical
where task interdependence is low, where team members are Benevolent, and (to a lesser
extent) where teams are parallel. Low or inconsistent levels of process control will not be
associated with negative outcomes to the same degree in such contexts. Alternatively, the
yield from beneficial justice configurations can be optimized in contexts where task
interdependence is high, members are low in Benevolence, and teams are non-parallel.

These results also make several contributions to the justice literature. Only two
other studies have examined moderators of procedural justice effects. Joy and Witt

(1992) examined delay of gratification as a moderator of the relationship between

88

procedural and distributive justice. Skarlicki, Folger, and Tesluk (1999) examined
negative affectivity and agreeableness as moderators of the relationship between fairness
perceptions and organizational retaliation behaviors. This neglect of moderation effects
constitutes an important gap in existing procedural justice theories. Whetten (1989)
argues that moderators, or boundary conditions, are a prerequisite for good theory,
because they can illustrate when the effects of an independent variable are especially
strong or weak.

Thus the interaction results presented above contribute to the justice literature by
identifying important moderator variables. Equity sensitivity was shown to be a
Significant moderator, adding to the personality moderators that have been examined by
Joy, Skarlicki, and colleagues (Joy & Witt, 1992; Skarlicki et al., 1999). Equity
sensitivity had already been shown to moderate distributive justice relationships, but had
never been examined in conjunction with procedural justice. Again, this is representative
of the artificial distinction between the two domains.

This study is the first to examine non-personality factors as moderators of
procedural justice effects. For example, task interdependence was a contextual factor
which increased the effects of process control. This moderation effect casts a new light
on the differences between the results of this study, and those of Ambrose, Grienberger,
Lind, and colleagues (Ambrose et al., 1989; Ambrose & Kulik, 1991; Grienberger et al.,
1993; Lind et al., 1999). As discussed above, those studies found much less support for
significant effects of others’ process control. One explanation for this may be that the
collectives in those studies lacked any interdependence. They were grouped together only

insofar as they participated in the experiment at the same time. They did not work on a

89

common task, nor did they have common goals or rewards. Their results may be an
example of low task interdependence neutralizing the effects of different process control
configurations.

While it had few significant interaction effects, the examination of the parallel vs.
non-parallel distinction also makes a contribution to the justice literature. Recent research
has seen an increased focus on how justice affects negative behaviors, such as theft or
organizational retaliation (e. g., Greenberg, 1990; Skarlicki et al., 1999). This study also
included negative outcomes in the form of conflict perceptions and text message
behavior. The results offer some support for the idea that injustice is less likely to result
in negative behaviors in parallel teams. Instead, such teams offer an avenue for
withdrawal from the injustice. It should be noted that none of the studies in Greenberg’s
and Skarlicki’s streams of research have involved settings in which exit could occur,
beyond forms of withdrawal such as absenteeism or turnover. As researchers continue to
study justice in the context of negative behaviors, they should pay close attention to
whether the context can create alternatives to such actions.

Suggestions for Future Resear_ch

This dissertation represents the first step in analyzing the full complexity of
procedural justice as manifested in team settings. Obviously much more research is
needed in this area. One definite need is to replicate the effects of the five justice
configurations using other operationalizations of procedural justice. As noted earlier,
process control in team settings is a zero-sum situation. Other operationalizations of
procedural justice, such as bias suppression or interpersonal treatment, can be supplied to

some members without detracting from others. It may be that the negative effects for the

90

teammate contrast would not be evident in such contexts, but that remains an empirical
question.

The above suggestion would extend our knowledge regarding justice
configurations, but further research is also needed to extend our knowledge regarding
moderator variables. Other team characteristics could be associated with stronger
configuration effects, as with task interdependence and non-parallel teams. In particular,
configuration differences are likely to be more critical in teams that are more
developmentally mature, as members may feel more comfortable acting on any perceived
injustices (Tuckman, 1965). Research should also examine other member characteristics
which could moderate configuration effects. Examples could include individualism-
collectivism (Wagner & Moch, 1986) or preference for autonomy (Wageman, 1995).

In addition, whereas this dissertation manipulated justice configurations, future
research should try to predict when specific configurations are more or less likely. For
example, a study could measure team member diversity and cross-functionality, or
directly assess leader member exchange differences. Those variables could then be linked
to specific configurations, such as inconsistent rather than consistent treatment. Such
research would also advance the status quo to a greater degree if it were conducted in a
field setting where the complexity of the teams exceeded those modeled in Figures 2-5.

Perhaps the most critical topic for future research is examining how to manage the
inevitable tradeoffs between individual treatment and collective performance. The results
of this study showed that, in many contexts, individual affective and performance
outcomes were maximized by high and consistent levels of process control. Oftentimes,

however, high and consistent levels of process control will not maximize the performance

91

of the collective. For example, leader member exchange theory suggests that aggregate
performance outcomes can be maximized when a leader tailors his or her treatment of
subordinates according to their skills and motivation (e. g., Gerstner & Day, 1997).
Similarly, the Multilevel theory of team decision making suggests that team decision
making accuracy is maximized when team leaders weigh staff members’
recommendations in proportion to staff members’ abilities to predict the correct answer
(e.g., Hollenbeck et al., 1995).

These two theories suggest that, in many contexts, high and consistent process
control will detract from collective performance. Given this, how should leaders
maximize collective performance without harming individuals’ fairness perceptions,
leader evaluations, and task behaviors? One possibility is by turning to a different
procedural justice operationalization in such circumstances. As mentioned in Chapter
Two, process control is not the only operationalization of procedural justice. Others have
focused on explanations and interpersonal sensitivity as markers of procedurally fair
treatment. Future research Should explore whether these different operationalizations
have compensatopy effects. That is, can the presence of explanations or interpersonal
sensitivity “make up for” low levels of process control? Or are the operationalizations
non-compensatory in their effects, meaning that individuals without process control are
unaffected by high levels of other forms of justice. If research supports a compensatory
model, then collective performance and individual treatment can both be maximized in
many contexts.

A second possibility for managing this tradeoff is exploring how the construct of

collective performance moderates the effects of individual treatment levels. As an

92

example, Elliott and Meeker (1986) conducted a study of distributive justice in which
they examined how reward allocators manage the tradeoff between promoting group
cohesion and enhancing individual productivity. They manipulated the contributions of
individual group members to collective performance, and also manipulated group success
and group morale. Their results showed that between-member differences in
contributions were much less relevant to perceptions of distributive justice when the
group achieved success and when the group had high morale. Such results suggest that, if
teams are successful, individual members will not care as much about their own treatment
levels. Rather it is when the group is struggling that personal treatment will become the
focus. Future research should investigate such a prediction in the context of the process
control manipulations employed in this study.
Limitations

This dissertation has some limitations which should be noted. While the
laboratory setting afforded the ability to test causality and control potential confounding
or contaminating influences, it resulted in a team environment that is quite different from
actual organizations. Specifically, the teams examined here were small and were
composed of members with no real past or future. The extent to which these findings will
generalize to more complex, long-standing teams is an empirical question that must be
the subject of future research. Nonetheless, it Should be noted that, traditionally,
laboratory settings have underestimated procedural justice effects (Lind & Tyler, 1988).
Indeed, it seems likely that the effects observed here would be stronger to the extent that

members have a future in the team and identified more strongly with it.

93

A second limitation is that the study examined justice differences along one
dimension only, that of process control. In organizational settings member treatment is
likely to vary along several justice dimensions simultaneously, including Thibaut and
Walker’s (1975) and Leventhal’s (1980) operationalizations. Future research should
model differences in justice configuration along multiple dimensions. Indeed, it may be
that the effects of justice configurations along one dimension compensate for any
negative effects of other dimensions.

Finally, the laboratory environment created a condition in which differences in
member performances and contributions were difficult to judge. AS noted earlier,
different process control levels may be deemed more fair when concomitant differences
in ability exist. Future research should model the potential effects of such ability

differences.

94

APPENDIX A

95

APPENDIX A

Consent Form

This study is designed to investigate team decision making effectiveness. If you choose to
participate in this study, you will be asked to learn a computer simulated target
identification task, operate the simulation task with other individuals, and complete a
series of questionnaire items. In addition, if you choose to participate in this study, you
authorize the researchers to have access to the questionnaires that you completed in your
Management 302 recitation section at the beginning of the semester.

Your participation in the simulation should take approximately three hours. In exchange
for your participation in this study, you will receive miscellaneous credit for your
Management 302 class requirement to participate in a research project. Other research
projects or alternatives are available from your instructor if you decide not to participate
in this study. If you do participate, you have the opportunity to earn a cash bonus. This
bonus will take the form of a check given to you in one of the final recitation sections.
This will be explained more fully during the study.

Your participation in this research is completely voluntary. You are free to decline to
answer any questions or to terminate your participation at any time. Your participation in
this study will be totally confidential. Your data will be included in a summary report
along with the data from others. The report will not include any information that will
allow anyone to identify any of your individual responses. If you have any questions or
concerns regarding this study, you may contact John R. Hollenbeck in the Management
Department at 355-2413, or David E. Wright of the University Committee on Research
Involving Human Subjects (UCRIHS) at 355-2180.

Participant’s Statement

I agree to participate in the Team Decision Making Study. I understand that I will learn to
operate a computer simulation, perform the simulation with other individuals, and fill out
a series of questionnaires. I authorize the researchers to use questionnaire that I already
completed in my Management 302 recitation section. It is my understanding that these
materials will be strictly confidential and will not be seen by anyone other than the
research team. I consent to having these materials used for research purposes. I
understand that the top performing teams will be eligible for cash prizes. I understand
that my participation is voluntary, that I may discontinue participation at any time
without penalty, that all of my individual responses will be kept strictly confidential, and
that I will not be identified in any report of this study.

 

Signature Date
Printed Name PID (MSU ID Number)

 

 

 

96

APPENDIX B

97

APPENDIX B

Training Materials

GENERAL OVERVIEW

INTRODUCTION: The year is 1999 and you are a part of a US. Air Force command and
control team stationed in the Middle East. A regional conflict between two nations in this
area has recently broken out, and your mission is to protect US. assets in the area from
accidental or intentional attacks. As history indicates, this is a highly sensitive task. For
example, in 1987, failure by a command and control team to quickly and accurately
identify a plane as threatening, allowed an Iraqi jet to accidentally fire two Exocet
missiles into the Frigate U.S.S. Stark, killing 37 American serviceman and crippling the
vessel. One year later, a command and control team error resulted in the USS. Cruiser
Vincennes accidentally shooting down an Iranian passenger plane killing 290 innocent
civilians. Another occurrence such as this will diminish public support for the current
mission and jeopardize peace in this region.

 

THE TASK FORCE: As a member of a four-person command and control team, you will be
linked to your fellow team members through an electronic data network that can supply bits
and pieces of critical information concerning possible enemy aircraft. Your mission is to
communicate and coordinate your information, so that the team commander ends up seeing
an accurate overall “big-picture”. Each team member is a specialist in interpreting particular
bits of information. Each person’s specialty must be considered as a whole in order to get
an estimate of the overall threat of a particular aircraft. This is necessary so that the
commander can make appropriate decisions concerning possible enemy aircraft.

TEAM MISSION: Monitoring Air Space: The team that you are a member of will monitor
the airspace surrounding an aircraft carrier group, making sure that your ships are not
attacked. In performing this role, you must make certain that you do not allow loss of life
resulting from enemy attacks on ally ships in the fleet. At the same time, it is also of
paramount importance that you do not inadvertently shoot down friendly military aircraft or
any civilian aircraft. Many passenger flights move in and out of the region, and fiiendly
military aircraft fiom nations not involved in the conflict also patrol the area.

OVERVIEW OF ROLES: There are four roles in this simulation, referred to as ALPHA,
BRAVO, CHARLIE, and DELTA. ALPHA is the team leader. BRAVO, CHARLIE, and
DELTA are staff members who specialize in interpreting aircraft information. The team’s
task is to decide what response should be made toward incoming aircraft. Team members
will assess the aircraft in accordance to their specific expertise, and make recommendations
to ALPHA, who will then make the final decision for the team. Team members base their
decisions on data they collect by measuring characteristics of aircraft that enter the task
force’s airspace. There are seven mssible judgments to make for each incoming aircraft.
These judgments are ordered in terms of their aggressiveness and there is one and only one
correct response for each aircraft.

98

SEVEN POSSIBLE JUDGMENTS

 

(l) IGNORE: This means that no further attention should be devoted to the aircraft.
Instead, focus should be directed on other possible aircraft in the area. Never ignore a
aircraft that might possibly attack.

(2) REVIEW: This means attention can be shifted away from this aircraft momentarily.
After a short period of time this aircraft should be returned to in order to update its status.

(3) MONITOR: This means that the aircraft should be continuously tracked.

(4) WARN: This means that a message is sent to the aircraft ordering it to turn away.
Warning aircraft that should be ignored detracts from the importance of legitimate warnings.
Warning aircraft that intend to attack is also bad, Since the warning makes it easier for an
attacker to locate aircraft the base.

(5) READY: This means to get into a defensive posture and to set defensive weapons on
automatic. A facility in a readied position is rarely vulnerable to attack. This stance should
not be taken to non-threatening aircraft since weapons set to automatic can fire mistakenly at
innocent aircraft that fly too close.

(6) LOCK-ON: This synchronizes radar and attack weapons so that the weapons fix
themselves on the aircraft. A ship at Lock-On position can take offensive action at a
moments notice. The capacity to track other aircraft is severely constrained once there is
Lock-On to a single aircraft, however. Thus, this should be reserved for aircraft that are
almost certain to be threatening.

(7) DEFEND: This is “weapons away” and means to attack the aircraft with missiles or
depth charges. A defend decision cannot be aborted once initiated and thus must only be
used when enemy attack is imminent.

NOTE: Do not worry about the exact definitions of these judgments.

All you really need to know is that they form a l - 7 scale of AGGRESSIVENESS,
where IGNORE is the least aggressive choice and DEFEND is the most aggressive choice.

99

CHARACTERISTICS OF AIRCRAFT

 

Aircraft can be measured according to nine characteristics. These are listed below along
with the range of possible values for each of the characteristics:

 

Aircraft cues

Definition

Range

 

RANGE

Distance from the base. Aircraft that are
closer are more threatening.

0 - 500 miles

 

ALTITUDE

Distance aircraft is above the ground.
Aircraft that are low in altitude are more
threatening.

10 - 10,000 feet

 

SIZE

Estimated size of the aircraft. Larger
aircraft are more threatening.

20 - 70 feet

 

CORRIDOR
STATUS

Distance from center of civilian
(friendly) corridor. Aircraft further
outside the civilian corridor are more
threatening.

0 - 1000 meters

 

FREQUENCY

Indicates type of communication signal.
Aircraft with high frequency signals are
more threatening.

10-60Mhz

 

RADAR

Indicates type of radar. Lower types are
civilian, higher types are threatening.

1 - 9 (type)

 

ANGLE

Indicates whether aircraft is ascending or
descending. The more negative the angle,
the more the aircraft is descending, and
the higher the threat.

-30 deg to 30
deg

 

DIRECTION

Indicates whether aircraft is flying
directly toward the base. Lower degrees
indicate more direct paths, and higher
threats.

0 - 90 deg

 

 

SPEED

 

Miles per hour. The faster the aircraft the
more threatening.

 

0 - 500 mph

 

100

 

DETERMINING THE LEVEL OF THREAT FOR YOUR SPECIALTIES

 

These nine characteristics combine according to three rules (or specialties) which
are used to determine the level of threat associated with any aircraft. Each staff member,
BRAVO, CHARLIE, and DELTA, has expertise for one of the three rules. They provide
a recommendation to the team leader, ALPHA, based on their specialty. ALPHA is then
responsible for combining the three recommendations of the team members into ONE
OVERALL team decision.

BRAVO’S SPECIALTY: Size, Radar, Frequency, Speed

These four values should be considered together to determine the threat level for
this aspect of the aircraft. If the aircraft is threatening on all of these variables, then the
threat level for this aspect is very high. However, if ANY ONE of these three values is
non-threatening, the threat level for this aspect is low. Thus, a aircraft can be very
threatening for three out of four of these values, and still be considered safe for this
aspect.

CHARLIE’S SPECIALTY: Speed, Direction, Angle, Range

As with Rule Bravo, these four values should be considered together to determine
the threat level for this aspect of the aircraft. If the aircraft is threatening on all of these
variables, then the threat level for this aspect is very high. However, if ANY ONE of
these three values is non-threatening, the threat level for this aspect is low. Thus, a
aircraft can be very threatening for three out of four of these values, and still be
considered safe for this aspect.

DELTA’S SPECIALTY: Range, Altitude, Corridor Status, Radar

As with Rules Bravo and Charlie, these four values should be considered together
to determine the threat level for this aspect of the aircraft. If the aircraft is threatening on
all of these variables, then the threat level for this aspect is very high. However, if ANY
ONE of these three values is non-threatening, the threat level for this aspect is low. Thus,
a aircraft can be very threatening for three out of four of these values, and still be
considered safe for this aspect.

101

THE ROLE OF THE TEAM LEADER (ALPHA)

The structure of your team will match that of an Air Force command and control team. In
those teams, the staff members are in close communication with one another while the
leader is separated a good distance away. In fact, your leader will be placed in a different
room to create that same level of separation. ALPHA’s role differs from that of BRAVO,
CHARLIE, and DELTA in 3 ways:

COMBINING RECOMMENDATIONS TO COME UP WITH ONE TEAM
DECISIONS

It is critical to remember that BRAVO, CHARLIE, and DELTA each see one-third of the
overall “big picture” concerning the aircraft. This is important because they could each
recommend something TOTALLY DIFFERENT and be doing the correct thing based on
their own specialty.

ALPHA’s responsibility is to take the three recommendations from BRAVO, CHARLIE,
and DELTA (along with information that he or she gathers independently) and combine
them into ONE OVERALL TEAM DECISION. For example, if BRAVO recommends a
DEFEND, and CHARLIE recommends a DEFEND, and DELTA recommends an
IGNORE (and these recommendation were correctly made), the overall judgement is
probably READY. Note that this DOES NOT mean that DELTA is “less right” than
BRAVO or CHARLIE. It is completely possible for each staff member’s information to
lead to a completely different recommendation.

SPECIALIZATION

As discussed on the previous page, BRAVO, CHARLIE, and DELTA are each in charge
of FOUR aircraft characteristics, and will be given very specific training about how to
interpret that information. In contrast, ALPHA will independently assess ALL NINE
characteristics. However, ALPHA’s will not be able to interpret that information with the
specificity, depth of understanding, and accuracy that BRAVO, CHARLIE, and DELTA
will have.

SIMULTANEOUS COMMUNICATION

In Air Force command and control teams the team leaders engage in “Simultaneous
Communication.” This means that they communicate with staff members as a whole, all
at once. This saves time, since team leaders are very busy and face many demands. Your
team will be set up this way. Specifically, each of you will be able to send messages

through the computer to another team member. However, messages from ALPHA will be
sent to BRAVO, CHARLIE, and DELTA all at once, simultaneously.

102

OUTCOMES OF DECISIONS

 

Once ALPHA makes his or her decision, there are five possible outcomes associated
with the accuracy or that decision (scoring is done automatically by the computer):

 

 

 

 

 

 

 

 

 

 

 

OUTCOME DEFINITION EXAMPLE POINTS
The decision was You said defend,
(1) HIT exactly correct. correct answer was +2
defend.
The decision was You said defend,
(2) NEAR MISS off by one level. correct answer was +1
lock-on.
The decision was You said defend,
(3) MISS off by two correct answer was 0
levels. ready.
The decision was You said defend,
(4) INCIDENT off by three correct answer was -1
levels. warn.
The decision was You said defend,
(5) DISASTER off by more than correct answer was -2
three levels. monitor, review, or
ignore.

 

103

 

APPENDIX C

104

APPENDIX C

Individual Instruction Sheets

INSTRUCTIONS FOR BRAVO

 

As BRAVO, you have unique knowledge that will allow you to gain a picture of one aspect of an aircraft’s
threat level. Specifically, you know how to interpret information regarding SIZE, RADAR, FREQUENCY,
and SPEED. You will be able to measure some of this information yourself, but you will also need to get
some of it from your teammates. Your responsibility is to make a recommendation to ALPHA based on
your interpretation of these four characteristics. ALPHA will then take your recommendation, those of
CHARLIE and DELTA, and the information that he or she has gathered individually, to make one overall
team decision.

YOUR SPECIALTY
As BRAVO, you are uniquely responsible for the following information:

 

 

 

 

AIRCRAFT MODERATE THREAT
CHARACTERISTIC
SIZE 20 - 36 37 - 53 54 - 70
RADAR l - 3 4 - 7 8 - 10
FREQUENCY 10 - 27 28 - 42 43 - 60
SPEED 0 - 170 171 - 329 330 - 500

 

 

 

 

REMEMBER: THESE FOUR CHARACTERISTICS INTERACT TO FORM YOUR RECOMMENDATION.
IF ONE OF THESE FOUR CHARACTERISTICS IS SAFE,
THEN THE “BRAVO” RECOMMENDATION MUST BE IGNORE.

o For example, an aircraft may have threatening Size, Frequency, and Speed, but if the Radar is safe, the
Bravo assessment for the aircraft is safe, and the decision should be IGNORE.

- As another example, the aircraft may have threatening Radar, Size, and Speed, but if the Frequency is
safe, the Bravo assessment for the aircraft is safe, and the decision should be IGNORE.

ALSO REMEMBER:
o If the aircraft is safe on one, two, three, or all four cues, the Bravo assessment is safe.
0 If the aircraft is moderate on all four cues, the Bravo assessment will also be moderate.

0 If the aircraft is threatening on all four cues, the Bravo assessment will be very threatening.

105

INSTRUCTIONS FOR CHARLIE

 

As CHARLIE, you have unique knowledge that will allow you to gain a picture of one aspect of an
aircraft’s threat level. Specifically, you know how to interpret information regarding SPEED,
DIRECTION, ANGLE, and RANGE. You will be able to measure some of this information yourself, but
you will also need to get some of it from your teammates. Your responsibility is to make a recommendation
to ALPHA based on your interpretation of these four characteristics. ALPHA will then take your
recommendation, those of BRAVO and DELTA, and the information that he or she has gathered
individually, to make one overall team decision.

YOUR SPECIALTY
As CHARLIE, you are uniquely responsible for the following information:

 

 

 

 

 

 

 

 

AIRCRAFT MODERATE
CHARACTERISTIC
SPEED 0 - 170 171 - 329 330 - 500
DIRECTION 9O - 65 64 - 36 35 - 0
ANGLE 30-15 l4to-l4 -15to -30
RANGE 500 - 340 339 - 161 160 - 0
_

REMEMBER: THESE FOUR CHARACTERISTICS INTERACT TO FORM YOUR RECOMMENDATION.
IF ONE OF THESE FOUR CHARACTERISTICS IS SAFE,
THEN THE “CHARLIE” RECOMMENDATION MUST BE IGNORE.

o For example, an aircraft may have threatening Direction, Angle, and Speed, but if the Range is safe,
the Charlie assessment for the aircraft is safe, and the decision should be IGNORE.

- As another example, the aircraft may have threatening Range, Speed, and Angle, but if the Direction is
safe, the Charlie assessment for the aircraft is safe, and the decision should be IGNORE.

ALSO REMEMBER:
o If the aircraft is safe on one, two, three, or all four cues, the Charlie assessment is safe.
0 If the aircraft is moderate on all four cues, the Charlie assessment will also be moderate.

0 If the aircraft is threatening on all four cues, the Charlie assessment will be very threatening.

106

INSTRUCTIONS FOR DELTA

 

As DELTA, you have unique knowledge that will allow you to gain a picture of one aspect of an aircraft’s
threat level. Specifically, you know how to interpret information regarding RADAR, RANGE,
ALTITUDE, and CORRIDOR STATUS. You will be able to measure some of this information yourself,
but you will also need to get some of it from your teammates. Your responsibility is to make a
recommendation to ALPHA based on your interpretation of these four characteristics. ALPHA will then
take your recommendation, those of BRAVO and CHARLIE, and the information that he or she has
gathered individually, to make one overall team decision.

YOUR SPECIALTY
As DELTA, you are uniquely responsible for the following information:

 

 

 

 

 

 

 

AIRCRAFT MODERATE
CHARACTERISTIC
RADAR l - 3 4 - 7 8 - 10
RANGE 500-340 339-161 160-0
ALTITUDE 10,000 - 6800 6799 - 3201 3200 - 100
CORRIDOR STATUS 0 - 320 321 - 679 680 - 1000
_

 

REMEMBER: THESE FOUR CHARACTERISTICS INTERACT TO FORM YOUR RECOMMENDATION.
IF ONE OF THESE FOUR CHARACTERISTICS IS SAFE,
THEN THE “DELTA” RECOMMENDATION MUST BE IGNORE.

o For example, an aircraft may have threatening Altitude, Corridor Status, and Radar, but if the Range is
safe, the Delta assessment for the aircraft is safe, and the decision should be IGNORE.

0 As another example, the aircraft may have threatening Range, Radar, and Altitude, but if the Corridor
Status is safe, the Delta assessment for the aircraft is safe, and the decision should be IGNORE.

ALSO REMEMBER:
o If the aircraft is safe on one, two, three, or all four cues, the Delta assessment is safe.
0 If the aircraft is moderate on all four cues, the Delta assessment will also be moderate.

0 If the aircraft is threatening on all four cues, the Delta assessment will be very threatening.

107

APPENDIX D

108

APPENDIX D

Instructions for Individual Task

INSTRUCTIONS FOR THE INDIVIDUAL TASK

On the individual task you will assume the role of an Aircraft Carrier. You will perform much the same
actions as on the team task. However, you and you alone will have the final say on what action to take
regarding the aircraft (e.g., Ignoring it, Monitoring it, etc.) As the Aircraft Carrier, you are uniguely

resmnsible for the following two rules:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

THE MOVEMENT SPECIALTY
AIRCRAFT Definition MOD THREAT
FACET
Length of the
EXHAUST exhaust trail left - 10 11 - 14 15 - 20
by the aircraft in
meters
Increase or
ACCELERATION decrease in speed -30 to - -9 to 9 10 to 30
in mph/second 10
Angle of aircraft’s
TRAJECTORY path since picked 90 to 30 29 to -29 -30 to -90
up by radar in
degrees
THE CATEGORY SPECIALTY
AIRCRAFT Definition SAFE MOD THREAT
FACET
Distance from
WIDTH wingtip to wingtip 20 - 30 31 - 39 40 - 50
in feet
Area of engine
ENGINES openings in 1 - 2 3 - 4 5 - 6
square feet
Frequency of
ESM “electronic security 40 - 60 61 - 79 80 - 100
measure” in Mhz

 

 

 

‘ AS WITH THE TEAM TASK, THE THREE CHARACTERISTICS WITHIN EACH SPECIALTY
INTERACT TO GENERATE A RECOMMENDATION. IF ONE OF THE THREE CHARACTERISTICS
WITHIN EACH SPECIALTY IS SAFE, THE RECOMMENDATION BASED ON THAT SPECIALTY

MUST BE IGNORE.

109

LIST OF REFERENCES

110

LIST OF REFERENCES

Adam, E. E. Jr. (1991). Quality circle performance. Journal of Management, 17,
25-39.

Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances
in experimental social psychology (V 01. 2, pp. 267-299). New York: Academic Press.

Ambrose, M. L., & Kulik, C. T. (1989). The influence of social comparisons on
perceptions of procedural fairness. Journ_al of Business and Psychology, 4, 129-138.

Ambrose, Harland, & Kulik, C. T. (1991). Influence of social comparisons on
perceptions of organizational fairness. Joumal of Applied Psychology, 76, 239-246.

Barker, J. R. (1993). Tightening the iron cage: Concertive control in self-

managing teams. Administrative Science Quarterly, 38, 408-437.

Barnard, C. I. (193 8). The functicps of the executive. Cambridge, MA: Harvard
University Press.

Brockner, J ., & Weisenfeld, B. M. An integrative framework for explaining
reactions to decisions: Interactive effects of outcomes and procedures. Psychological
Bulletin, 120, 189-208.

Campion, M. A., Medsker, G. J ., Higgs, A. C. (1993). Relations between work
group characteristics and effectiveness: Implications for designing effective work groups.

Personnel Psychology, 46, 823-850.

Campion, M. A. , Pappar, E. M. , Medsker, G. J. (1996). Relations between work
team characteristics and effectiveness: A replication and extension. Personnel

Psychology, 49, 429-452.

Cannon-Bowers, J. A. , Oser, R., & Flanagan, D. L. (1992). Work teams in
industry: A selected review and proposed framework. In Swezey RW, Salas E (Eds)
Teams: Their training and performance (pp. 355-3 77). Norwood, NJ: Ablex Publishing.

Cohen, S. G., & Bailey, D. E. (1997). What makes teams work: Group
effectiveness research from the shop floor to the executive suite. Journal of Management,
2_3_, 239-290

Colquitt, J. A., Conlon, D. E., Wesson, M. W., Porter, C., and Ng, K. Y. (1999).
A meta-analysis of the nomologic_al networkpof procedural justice. Research in progress.

Colquitt, J. A., Hollenbeck, J. R., Sheppard, L., Ilgen, D. R., & LePine, J. A.

(1999). Sex composition and communication technology in teams: Implications for
decision making accuracy. Research in progress.

 

111

Colquitt, J. A., Noe, R. A., & Janz, B. D. (1999). LeadinLWorkers to Help
Themselves: A Contingency Approach to Cooperation in Knowledge Wcflcer Tea_m_s
Manuscript under review.

Cropanzano, R., & Ambrose, M. L. (1996, April). Procedural and distributive

justice are more similar than you think: A monistic perspective and research agenda.
Paper presented at the Annual Meeting of the Society for Industrial and Organizational

Psychology. San Diego, CA.

Cropanzano, R., & Greenberg, J. (1997). Progress in organizational justice:
Tunneling through the maze. In C. L. Cooper & I. T. Robertson (Eds), International
review of industrial and organizational psychology; New York: John Wiley & Sons.

Cropanzano, R., & Randall, M. L. (1993). Injustice and work behavior: A
historical review. In R. Cropanzano (Ed), Jpatice in the workpl_ace: Approaching fairness
in human resource management (pp. 3-20). Hillsdale, NJ: Erlbaum.

Deutsch, M. (1975). Equity, equality, and need: What determines which value
will be used as the basis for distributive justice? Journal of Social Issues, 31, 137-149.

Deutsch, M. (1985). Distributive iustice. New Haven, CT: Yale University Press.
Dumaine, B. (1994) The trouble with teams. Fortune 130, 86-93.

 

Dunphy, D., & Bryant, B. (1996). Teams: Panaceas or prescriptions for improved
performance? Human Relations. 49, 677-699.

Elliott, G. C., & Meeker, B. F. (1986). Achieving fairness in the face of
competing concerns: The different effects of individual and group characteristics. Journal
of Persorglity and Socgl Psychology, 50, 754-760.

F olger, R. (1986). A referent cognitions theory of relative deprivation. In J. M.
Olson, C. P. Herman, & M. P. Zanna (Eds), R_elative deprivation and social comparison:
The On@ symposium (Vol. 4, pp. 33-55).

F olger, R. (1987). Reformulating the preconditions of resentment: A referent
cognitions model. In J. C. Masters & W. P. Smith (Eds), Social comparison, justice, and
relative deprivation: Theoretical, empirical, and policy perspectives (pp. 183-215).
Hillsdale, NJ: Erlbaum.

F olger, R., & Cropanzano, R. (1998). Organizational iustice and human resource
management. Thousand Oaks, CA: Sage.

F olger, R. & Greenberg, J. (1985). Procedural justice: An interpretive analysis of
personnel systems. In K. Rowland & G. Ferris (Eds), Research in personnel and human
resources management (Vol. 3, pp. 141-183). Greenwich, CT: JAI Press.

112

Goodman, P. S., Ravlin, E., & Schminke, M. (1987). Understanding groups in
organizations. In L. L. Cummings & B. M. Staw (Eds), Research in organizational
behavior (V 01. 9, pp. 121-173). Greenwich, CT: JAI Press.

 

Graen, G. B., & Scandura, T. (1987). Toward a psychology of dyadic organizing.
In L. L. Cummings & B. M. Staw (Eds), Research in organizational belmior (Vol. 9, pp.
175-208). Greenwich, CT: JAI Press.

Greenberg, J. (1990). Employee theft as a reaction to underpayment inequity: The
hidden cost of pay cuts. Journal of Applied Psychology, 75, 561-568.

Greenberg, J. (1990). Organizational justice: Yesterday, today, and tomorrow.
Journal of Management, 16, 399-432.

Greenberg, J ., & Folger, R. (1983). Procedural justice, participation, and the fair
process effect in groups and organizations. In P. B. Paulus (Ed.), Basic ggoup processes
(pp. 23 5-256). New York: Springer-Verlag.

Grienberger, I. V., Rutte, C. G., & van Knippenberg, A. F. M. (1997). Influence
of social comparisons of outcomes and procedures on fairness judgments. Journal of

Applied Psychology, 82, 913-919.

Guzzo, R. A. & Dickson (1996). Teams in organizations: Recent research on
performance and effectiveness. Annual Review of Psychology, 47, 307-338.

Guzzo, R. A., & Shea, G. P. (1992). Group performance and intergroup relations
in organizations. In M. D. Dunnette & L. M. Hough (Eds), H_andbool_< of Industrial and
Organizational Psychology (2nd ed., Vol. 3, pp. 269-313). Palo Alto: Consulting
Psychologists Press.

Hackman, J. R. (1992). Group influences on individuals in organizations. In M.
D. Dunnette & L. M. Hough (Eds), Handbook of Industrial and Organizational
Psychology, Second Edition, 3 (pp. 199-267). Palo-Alto, CA: Consulting Psychologists
Press, Inc.

Hirschman, A. O. (1970). Exit. voice, and loyalty: Responses to decline in firms,
organizations, and states. Cambridge, Mass: Harvard University Press.

Hollenbeck, J. R., Ilgen, D. R., Sego, D., Hedlund, J ., Major, D. A., & Phillips, J.
(1995). The multilevel theory of team decision making: Decision performance in teams
incorporating distributed expertise. Journal of Applied Psychology. 80. 292-316.

Homans, G. C. (1961). Social behaviour: Its elemenLary forms London:
Routledge and Kegan Paul.

Huseman, R. C., Hatfield, J. D., & Miles, E. W. (1987). A new perspective on
equity theory: The equity sensitivity construct. Academy of Management Review. 12,
222-234.

113

James, K. (1993). The social context of organizational justice: Cultural,
intergroup, and structural effects on justice behaviors and perceptions. In R. Cropanzano
(Ed), J_u_stice in the workplace: Approaching fairness in Human Resource Management
(pp. 21-50). Hillsdale, NJ: Erlbarun.

Janz, B. D., Colquitt, J. A., & Noe, R. A. (1997). Knowledge worker team
effectiveness: The role of autonomy, interdependence, team development, and contextual
support variables. Personnel Psychology, 50. 877-904.

Jehn, K. (1995). A multirnethod examination of the benefits and detriments of
intragroup conflict. Administrate Science Ou_arterly. 40, 245-282.

Joy, V. L., & Witt, L. A. (1992). Delay of gratification as a moderator of the
procedural justice-distributive justice relationship. Group and Organization Management,
fl, 297-308.

Kabanoff, B. (1991). Equity, equality, power, and conflict. Academy of
Management Review. 16. 416-441.

Kahn, R. L. (1981). Work and Health. New York: Wiley.

Kahn, R. L., & Byosiere, P. B. (1992). Stress in organizations. In M. D. Dunnette
& L. M. Hough (Eds), Handbook of Industrial and Organizaticmal Psycholgy (2nd ed.,
Vol. 3, pp. 571-650). Palo Alto: Consulting Psychologists Press.

Kelley, H. H., & Michela, J. L. (1980). Attribution theory and research. Annual
Review of Psychology, 31, 457-501.

Kerlinger, R. (1986). Foundations of beh_avioral research. New York: Holt,
Rinehart, & Winston.

Kiggundu, M. N. (1981). Task interdependence and the theory of job design.
Academy of Management Review, 6, 499-508.

King, W. C. Jr. (1993). A test and refinement of the equity sensitivity construct.
Journal of Organizational Behavior. 144 301-317.

King, W. C. Jr. (1994). The influence of sex and equity sensitivity on relationship
preferences, assessment of opponent, and outcomes in a negotiation experiment. Journal
of Management, 20, 605-624.

King, W. C. Jr., & Miles, E. W. (1994). The measurement of equity sensitivity.
Journal of Occupational and Organizational Psychology, 67, 133-142.

Korsgaard, M. A., Schweiger, D. M., Sapienza, H. J. (1995). Building
commitment, attachment, and trust in strategic decision-making teams: The role of
procedural justice. Academy of Management Journal, 38, 60-84.

114

Kozlowski, S. W. J., Gully, S. M., McHugh, P. P., Salas, E., & Cannon-Bowers,
J. A. (1996). A dynamic theory of leadership and team effectiveness: Developmental and
task contingent leader roles. In G. R. Ferris (Ed). Research in personnel and human
resources management, Vol. 14. (pp. 253-305). Greenwich, CT: JAI Press.

Lawler, E. E., & Cohen, S. (1992). Designing pay systems for teams. CEO
Publication T 92-12 (215).

LePine, J. A., Hollenbeck, J. R., Ilgen, D. R., & Hedlund, J. (1995). The effects of
individual differences on the performance of hierarchical decision making teams: Much

more than g. Journal of Applied Psychology.

Lerner, M. J. (1977). The justice motive: Some hypotheses as to its origins and
forms. Journal of Personality. 45. 1-52.

Leventhal, G. S. (1976). The distribution of rewards and resources in groups and
organizations. In L. Berkowitz & W. Walster (Eds), Advances in experimental social
psychology (V 01. 9, pp. 91-131). New York: Academic Press.

Leventhal, G. S. (1980). What should be done with equity theory? New
approaches to the study of fairness in social relationships. In K. Gergen, M. Greenberg, &

R. Willis (Eds), Social exchange: Advances in theog and research (pp. 27-55). New
York: Plenum Press.

Leventhal, G. S., Karuza, J ., & Fry, W. R. (1980). Beyond fairness: A theory of
allocation preferences. In G. Mikula (Ed.), Jistice and social interaction (pp. 167-218).
New York: Springer-Verlag.

Lind, E. A., Kray, L., & Thompson, L. (1999). The social construction of
injustice: Fairness judgments in response to own and others’ unfair treatment by
authorities. Organizational Behgior and Humap Decision Processes.

Lind, E. A., & Tyler, T. R. (1988). The social psychology of procedural justice.
New York: Plenum Press.

Magjuka, R. J ., & Baldwin, T. T. (1991). Team-based employee involvement
programs: Effects of design and administration. Personnel Psychology, 44, 793-812.

Major, B. & Deaux, K. (1982). Individual differences in justice behavior. In J.
Greenberg & R. L. Cohen (Eds), @uig and justice in social belgvior (pp. 43-76). New
York: Academic Press.

May, D. R., & Schwoerer, C. E. (1994). Employee health by design: Using
employee involvement teams in ergonomic job redesign. Personnel Psychology, 47, 861-
876.

McGrath, J. E. (1964). Social psychology: A brief introduction. New York: Holt.

115

Meindl, J. R. (1989). Managing to be fair: An exploration of values, motives, and
leadership. Administrative Science Quarterly, 34, 252-276.

Miles, E. W., Hatfield, J. D., & Huseman, R. C. (1989). The equity sensitivity
construct: Potential implications for worker performance. JouLnaI of Management, 15,
581-588.

Miles, E. W., Hatfield, J. D., & Huseman, R. C. (1994). Equity sensitivity and
outcome importance. Journal of Organizational Behavior. 15. 585-596.

Paulus, P. B. (1980). Psychology of group influence. Hillsdale, NJ: Erlbaum.

Pearce, J. L., & Gregersen, H. B. (1991) Task interdependence and extrarole
behavior: A test of the mediating effects of felt responsibility. Journal of Applied
Psychology. 76. 838-844.

Pyszczynski, T. A., & Greenberg, J. (1981). Role of disconfirrned expectancies in

the instigation of attributional processing. Journal of Persopality and Social Psychology.
59, 31-38.

Reis, H. T. (1984). The multidimensionality of justice. In R. F olger (Ed.), _T_’h_e
gase of injustice: Social psychologic_al perspectives. New York: Plenum.

Rusbult, C. E., Farrell, D., Rogers, G., & Mainous, A. G. 111. (1988). Impact of
exchange variables on exit, voice, loyalty, and neglect: An integrative model of responses
to declining job satisfaction. Academy of Management Journal, 31, 599-627.

Saavedra, R., Earley, P. C., & Van Dyne, L. V. (1993). Complex interdependence
in task-performing groups. Journal of Applied Psychology, 78, 61-72.

Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to
job attitudes and task design. Administrative Science Quarterly. 23, 224-253.

Scott, W. R. (1995). Ipstitutions_ and organizations. Thousand Oaks, CA: Sage.

Shaiken, H., Lopez, 8., & Makita, I. (1997). Two routes to team production:
Satin-n and Chrysler compared. Industrial Relationsg36. 17-45.

Sheppard, J. A. (1993). Productivity loss in performance groups: A motivational
analysis. Psychological Bulletin. 113. 67-81.

Simon, 3., Pantaleo, G., & Mummendey, A. (1995). Unique individual or
interchangeable group member? The accentuation of intragroup differences versus
similarities as an indicator of the individual self versus the collective self. Journal of
Personalig and Social Psychology. 69. 106-119.

Sisco, R. (1992). Put your money where your teams are. Training, 29, 41-45.

116

Skarlicki, D. P., Folger, R., & Tesluk, P. (1999). Personality as a moderator in the
relationship between fairness and retaliation. fldemv of Management Journal. 42. 100-
108.

Steel, R. P., Jennings, K. R., & Lindsey, J. T. (1990). Quality circle problem
solving and common cents: Evaluation study findings from a Unites States mint. Journal
of Applied Behavioral Science. 26. 365-382.

Steiner, I. D. (1972). Group process and productivity. New York: Academic
Press.

Stouffer, S. A., Suchman, E. A., DeVinney, L. C., Star, S. A., & Williams, R. M. R
Jr. (1949). The Americap soldier: Adiustrnent during army life (Vol. 1). Princeton, NJ:
Princeton University Press. Cited in Cropanzano, R., & Randall, M. J. (1993).

Sundstrom, E, DeMeuse, K. P., & Futrell, D. (1990). Work teams: Applications
and effectiveness. American Psychologist, 45, 120-133.

 

Thibaut, J., & Walker, L. (1975). Procedural justice: A psychological analysis.
Hillsdale, NJ: Erlbaum.

Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological
Bulletin. 63. 384-399.

Tyler, T. R. (1989). The psychology of procedural justice: A test of the group-
value model. Journal of Personality and Social Psychology. 57. 830-838.

Tyler, T. R., & Lind, E. A. (1992). A relational model of authority in groups. In
M. P. Zanna (Ed). Advances in experimental social psychology (V 01. 25, pp. 115-191).
San Diego, CA: Academic Press.

Wageman, R. (1995). Interdependence and group effectiveness. Administrative
Science Quarterly, 40, 145-180.

Wagner, J. A. III, & Moch, M. K. (1986). Individualism-collectivism: Concept
and measure. Group and Organization Studies. 11. 280-303.

Wegner, D. M. (1982). Justice and the awareness of social entities. In J.
Greenberg & R. L. Cohen (Eds), Equity and iustice in 80%] beh_avior (pp. 77-117). New
York: Academic Press.

Whetten, D. A. (1989). What constitutes a theoretical contribution? Academy of
Management Review. 14(4), 490-495.

Wong, P. T. P., & Weiner, B. (1981). When people ask “why” questions, and the
heuristics of attributional search. Journal of Personality apd Social Psychology, 40, 650-
663.

117

Vecchio, R. P. (1981). An individual-differences interpretation of the conflicting
predictions generated by equity theory and expectancy theory. Journal of Applied

Psychology, 66, 470-481.

Zuckerrnan, M. (1979). Attribution of success and failure revisited, or: The
motivational bias is alive and well in attribution theory. Journ_al of Psychology. 47. 245-
287.

118