'w".

um... «1...

u

.uvmm
m. -...

u
.':g&;§&
'1'. 3:19;:

”a...
My:
‘ ,. ‘

wand:
:a :w:."".
no“
”a . ~
v

”Wm."

. ,, . ‘

325$

2525,10
.

 

~

.
W... m
. .

”my!

I.

.. "’
fivq'éka

~32:-
;. ..
in?“

«w .
vw

:2“ ’f
’ 1-2:}:
’57 ‘1'?!»
5%"; .
“z .. m
Wm.

:33,

.M. ..
. way?
.r.

< r
v
:m» vain-v.-
aunv
rv
n: a.
1mm
.0
u‘.‘
v”

z?

..-.-.4- v...»
a... .v .. m
9., \~
h~v»{u . ‘ ‘4.-
-. ~¢nk
.
n.

5
§
t

i
t" .1
fish"
: ' i
3&3!

7

- 5%
.M. p A34~M
$5137;
as:
as?»
b 1'
w

of“.

.g..~.~.;:..q.:..;'
a ....,,.,...~..m.... .
, A 71‘

*2:-

.uvw
-o

~a

' L33.“
JW‘m‘B‘éf

~v

.

‘0 s

figii
I ‘ I

ran
x...»

k
l

 

".Ulz ’J'J'm- ' ‘
2 ‘ 'E‘ ' .A‘v

a.

L
_. . -
' a -
"fin-’u'.¢ -

.. ~—
:17
m-

51 "
Ms

.,A

rvfiss

MICHIGAN STATE UNIVERSITY Ui‘fiA‘RI‘ES“

\llllllllllllllll "llllli

‘1
ll
3 1293 01389 2801 r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l

 

 

This is to certify that the

dissertation entitled

THE ROLE OF INFORMATION IN THE INTERORGANIZATIONAL
COLLABORATIVE DECISION-MAKING PROCESS
presented by

Sandra M. Starnaman

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Communication

Major professor

Date mm // , l99é

MSU i: an Affirmative Action/Equal Opportunity Institution 0—12771

 

LIBRARY
M'Chlgan State
University

 

PLACE ll RETURN BOX to romanthb chockomttom your meat].
TO AVOID FINES Mum on or baton duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I. An Affirmative Action/Ema! Opportunlty lmtltulon
Walla-M

THE ROLE OF INFORMATION IN THE INTERORGANIZATIONAL
COLLABORATIVE DECISION -MAKIN G PROCESS

By

Sandra M. Starnaman

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Communication

1996

ABSTRACT

THE ROLE OF INFORMATION IN THE INTERORGANIZATIONAL
COLLABORATIVE DECISION -MAKING PROCESS

by

Sandra M. Stamaman

In increasing numbers, organizations have become involved in
interorganizational relationships (IORs) as a result of a rapidly changing
social and technical environment. Until recently, IOR research has focused
primarily on the antecedents and consequences of interorganizational
linkages. Little research has been devoted to the study of the
developmental processes in IORs. Research on the creation of IORs has
addressed the key dimensions of the environments in which organizations
exist and how the interactions among organizations within environments
have been conceptualized. Research on the contingencies that lead to
involvement in IORs confirms the idea that organizations enter IORs with
one or more agendas. Central to these agendas is the W that
unites the groups involved in creating an IOR.

The interorganizational decision making responsibilities associated
with these issues lies with the organizational representatives or boundary
spanners involved in the IOR. The problem addressed in this dissertation is
the role of information acquisition and exchange within that decision-
making process. A model was hypothesized with the exogenous variables
of information quality and quantity being positively related to participation
in decision making (PDM). PDM was viewed as antecedent to and
positively related to satisfaction. Finally, satisfaction was hypothesized as

antecedent to and positively related to commitment.

Survey data was collected from 129 professionals within a multi-
disciplinary, multi-organizational project on the constructs of information
quality and quantity, participation in decision making, IOR satisfaction, and
IOR commitment. Confirmatory factor analysis and path analysis were
used to assess the content of the constructs and the relationship among the
variable.

The path findings confirm the hypothesized model with an additional
link between information and perceptions of satisfaction. The factor
analysis findings suggest that for these respondents the constructs of
satisfaction and commitment may be more specifically linked to

professional identification versus organizational identification.

To Craig, with love i

iv

ACKNOWLEDGMENTS

I would like to express my sincere thanks to a few of the people who
helped make this work possible:
Thank you!

Dr. Katherine Miller - for nine years of guidance and support .

Dr. Rebecca Henry - for all that you've helped me learn in the last
four years and for your unwavering support.

Dr. Alicia Marshall and Dr. Sandi Smith - for your patience and
insights.

Craig - for your love and belief in me and lots of home cooked
meals.

Mom - for showing me that you can reach your dreams.
Barb - for listening to me and for believing in me.

Maureen - for understanding that I'm not that organized.

TABLE OF CONTENTS

Page
ListofTables.................vii
ListofFigures viii
Chapter

1. THEORY AND LITERATURE REVIEW . . . . 1
The Changing Environment as a Source of
Interorganizational Growth . . . . . . . 3
Interorganizational Relationships . . . . . . 7
Interorganizational Exchange . . . . . . . 17
Information Exchange as a Mechanism
for Collaborative Interaction . . . . . . 22
Organizational Representatives as the
Carriers of IOR Information . . . . . . 22
Boundary Spanner Commitment and
Satisfaction as Antecedents to IOR
Sustainability . . . . . . . . . . . 23
Decision Making within Collaboratives . . . 28
The Role of Emergent Communication in
the Decision Making Processes of IORs . . 36
Model of Participation in Decision Making within
Interorganizational Collaboratives . . . . . 44

1

vi

Chapter
2. METHODS .

The Kellogg Community Partnership Initiative
Respondents
Administration Procedure .

Instrumentation

Available Information (Quantity and Quality) .

Participation in Decision Making .
Project Satisfaction and Project Commitment
Analysis
Confirmatory Factor Analysis .
Path Analysis
3. RESULTS
Confirmatory Factor Analysis
Preliminary Path Analysis .
Path Analysis .
4. DISCUSSION .
Measurement Model .
Path Model
Future Research .

Bibliography

vii

46
46
47
49
49
50
50
51
53
53
54
55
55
60
6o
67
67
73
75
83

LIST OF TABLES

Table
1. Scale Items, Factor Loadings, and
Reliabilities .
2. Correlation Matrix.

3. Comparison of Models

viii

Page
58
63
66

LIST OF FIGURES

Figure
1. Cognitive Model of Participation in Decision Making

2. Affective Model of Participation in Decision Making .

3. Hypothesized Model of Information, Participation,
Satisfaction, and Commitment in IORs .

4. Hypothesized Model of Information, Participation in
Decision Making, Satisfaction, and Commitment in IORs
with Path Coefficients. .

5. Final Model of Information, Participation in Decision
Making, and Commitment in IORs .

ix

3O
3O

45

65

CHAPTER ONE

In discussing their vision of "organizations" in the year 2000,
Kanter, Stein and lick (1992) write that at the turn of the century there will
be:

"...more flexible organizations, adaptable to change, with
relatively few levels of formal hierarchy and loose boundaries
among functions and units, sensitive and responsive the
environment; concerned with stakeholders of all sorts -- employees,
communities, customers, suppliers, and shareholders....Overall,
these are global organizations characterized by internal and external

relationships, including joint ventures, alliances, consortia, and
partnerships" (p. 3).

In both practice and research, these characteristics of the
"organization of the year 2000" are being tested. In increasing numbers,
organizations are looking outside of their immediate boundaries to address
the broader, more open-ended, less divisible problems that arise within the
organizational environment (Aldrich, 1976; Gray, 1985; Ring & Van De
Ven, 1994). By the mid-19805 there were 6,000 partnerships among
businesses, public sector, and non-profit organizations in the United States
(W addock, 1988). Logsdon (1991) writes that the National Alliance of
Business reported that there were 140,000 operating linkages between
schools and businesses. As early as 1982, Fottler, Schermerhorn, Wong,
and Money, reported that multi-organizational linkages of some sort
existed among 35 percent of all nongovernmental, nonprofit hospitals in

the United States. In short, within all sectors of the organizational world,

2

interorganizational connections are becoming a prevalent part of "doing
business".

The research literature mirrors these societal trends. For
example, Smith, Carroll and Ashford (1995) indicate that there is a wide
body of literature supporting the idea that the ability and willingness of
organizations to establish cooperative ties across organizational boundaries
are essential to organizational growth, innovativeness, and environmental
responsiveness. Establishing and maintaining linkages that cross
organizational boundaries, however, present an enormous challenge.
Varying constraints such as national culture (e. g., Japan vs. the United
States), professional rivalry (e. g., medicine vs. nursing), and the vested
interests of different stakeholder groups are just a few of the challenges
faced by organizations and individuals interested in initiating
interrorganizational relationships (IORs) (Halpert,1982).

To date, IOR research has focused primarily on the antecedents
(organizational and environmental) and consequences of
interrorganizational linkages. Ring and Van de Ven (1994) note, however,
that "(r)elatively little scholarly attention has been devoted to studying
developmental processes of IORs" (p. 91). Process, as associated with
IORs, refers to the interactions of individuals involved in the development
of these linkages and how, according to Ring and Van de Ven (1994) "a
relationship might unfold over time" (p. 99). The consequences of IORs,
therefore, have their origins not only in the antecedents of these
interorganizational linkages, but also in the negotiation processes that occur
as organizational agents interact. These interactions, whether at the
organizational or individual level, are according to Miller (1995)

"inherently communicative" (p. 255).

3
The goal of this study is to more fully understand the

communicative challenges faced by organizational representatives when
they endeavor to create interorganizational linkages. In order to do this I

. will first examine the context or environment in which organizations exist.
This will be followed by a review of the literature on the antecedents to
IORs and the developmental processes through which IORs are created and
nurtured. Finally, arguments regarding a boundary-spanning approach to
and a model of participative decision making will be presented to address

the role of communication variables in creating and sustaining IORs.

[ ‘ Q .__n'I‘ -310! n 1- 1-0.- ‘ 0 9,; OI'_,'-._L_a. u... 10.

The environment in which IORs are created is defined by Miller
(1995) as "the larger world of institutions and individuals in which a given
organization must exist"(p. 252). In a general sense, the development and
maintenance of interorganizational linkages can be attributed to increased
external pressures, increased constraints on individual organizational
decision making (Oliver, 1990), increased levels of interdependence, and
the increased need to adapt to a changing environment (Emery & Trist,
1965; Hall, 1992).

Recognition of the impact of the environment on organizations
challenges many assumptions of early organizational research. Early
theories and the corresponding practices of organizational members
worked from the assumption that the environment was relatively stable
and, therefore, predictable. This approach fostered the idea that
organizations could operate as closed systems (Emery & Trist, 1965; Katz
& Kahn, 1978). Based on closed system assumptions, organizational

change could be planned and implemented with little responsiveness to

4

internal and external environments (Mink, Schultz, Mink, 1991). In closed
system organizations, the relationship between organizations and their
environment(s) has focused primarily on the connection between the
technical demands of the external environment and the activities within an
organization (Meyer & Scott, 1992).

This assumptive base had to be radically modified, however, to take
into consideration the reality of a rapidly changing technical and social
environment. As a consequence, changes have arisen in organizational
theory and practice that reflect an open systems approach. The open
systems approach is based on the assumption that systems (i.e.,
organizations) are not isolated units but associated in multiple ways to other
units within the environment(s) that link them (Bertalanffy, 1968; Buckley,
1967; Katz & Kahn, 1978; Laszlo, 1972). The open systems approach
- highlights the importance of conceptualizing organizational environment(s)
and developing an understanding of the key dimensions of environments
which may affect organizational relationships.

Emery and Trist (1965) first noted the increasing complexity of
environments in their discussion of the "causal texture of organizational
environments" or, in other words, "the processes through which parts of
the environment become related to each other" (p. 22). The environmental
contexts in which these relational processes operate were conceptualized by
Emery and Trist (1965) as falling into four categories: (a) "placid,
randomized environment" (p. 24) in which organizations can adapt as
single units within a relatively unchanging local environment; (b) "placid,
clustered" in which knowledge of the organizational environment and
increased coordination start to become important and in which " goals and

noxiants are not randomly distributed but hang together in certain ways"

5

(p. 25); (c) "disturbed, reactive" in which the existence of multiple
organizations of the same type create the need for strategic as well as
tactical planning within organizations (p. 25); and (d)"turbu1ent field" (p.
26). Within the turbulent field "the dynamic properties arise not simply
from the interaction of the component organizations but also from the field
itself. The 'ground' is in motion" (p. 26). In other words, the very way in
which the interorganizational field is conceptualized may be in flux.

More comprehensive and specific typologies describing the key
dimensions of the environment have been developed as interest in the
impact of the environment on organizations and organizational change has
grown. Child (1972) addressed the frequency and patterns of change in the
environment, the heterogeneity of activities, and competition within an
environment. Pfeffer and Salancik (1978) focused on the dispersion of
power, scarcity of resources, and the interconnectedness within an
environment. Perhaps the most complete typology was presented by
Aldrich (1979). Aldrich writes that previous research highlights six
environmental dimensions that affect an organization or a subset of an
organization. The first of these is W. Environmental
capacity refers to relative availability of resources for an organization
within its environment. Environments with a greater number of resources
are referred to as "rich" and those with less are referred to as "lean".
Resource availability affects the opportunities available to organizations
and helps structure their interactions within their environment. Aldrich
(1979) notes that "(r)ich environments also have been seen as a barrier to
the formation of interorganizational relations, at least of the more
formalized kind." (p 63).

6

The second dimension discussed is eevirenmentel hemegeneity-
heteregem or "the degree of similarity or differentiation between the

elements of the population dealt with, including organizations, individuals,
and any social forces affecting resources" (p. 66). In a homogeneous
environment, the creation of standardized procedures encourages the
development of products or services that are very similar to each other and
the simplification of organizational activities. Heterogeneity within an
environment, on the other hand, creates greater complexity because each
organization, individual or social force may have to be dealt with in a
unique way.

The extent to which there is turnover in the various components in
an environment can alter an organization's stability within that
environment. This Aldrich refers to as the vir nm nt 1 ili -
W. The formal organizational structures that are so
useful for dealing with the routines that emerge within a stable
environment may inhibit an organization's ability to deal with changes in
an environment.

The fourth dimension Aldrich considers is that of W
WWII. This entails the extent to which resources are
evenly spread out or consolidated in a specific location. When key
elements in an environment are concentrated in one area organizations can
create or adapt strategies to address those elements.

Another concern for an organization is the extent to which others
within the environment recognize the organization's domain claims within

the environment. W is of most concern to non-

profit organizations where domain conflict originates over "alleged

7

duplication of services and efforts by new agencies to encroach on the
domains of established organizations" (p. 68).

Finally, drawing on the work of Emery and Trist (1965) Aldrich
discusses W. A turbulent environment is one in
where the number of both potential and actual linkages and rate of
environmental connections are increasing rapidly. Turbulence obscures the
rules and definitions for interaction and often makes it difficult to plan for
the future.

These environmental dimensions act as challenges to organizational
actors by either constraining or enabling activities and interactions. The
concept of individual organizations existing within multiple, changing
environments is further complicated by the recognition that organizations
are not completely autonomous but are linked in multiple ways (i.e., by
type, product, or population) to other organizations or stakeholder groups
and may be "subsumed under broader structures" (Scott, 1992, p. 181) as
their environment becomes more complex or the number of organizations
increase (Scott, 1992).

These multiple links serve to create or emphasize the existence of
"bounded networks" (Scott, 1992, p. 181) or "domains" (Levine & White,
1961; Thompson, 1967) defined by Gray (1985) as "sets of actors
(individuals, groups, and/or organizations) that become joined by a
common problem or interest" (p. 912). These bounded networks are the

breeding ground for the development of interrorganizational relationships.

'2 ' n l ' n
In his review of the literature on organizational environments, Scott

(1992) discusses four ways in which organizational relationships within

8

environment(s) have been conceptualized . First, the role of a single
organization within an organizational set has been studied. The focus here
is on how the various organizations within an environment affect the
interests of a single organization (Whetten, 1981).

Second, a somewhat more comprehensive approach identifies
organizational populations as "aggregates of organizations that are alike in
some respect" (p. 127). At the core of this approach is the establishment of
boundaries by which organizations can be categorized. Hannan and
Freeman (1989) propose two ways of creating boundaries. Boundaries
may be created based on common-sense categories that employ either
theoretically-based or empirically-based rules to determine common
characteristics shared by groups of organizations. The second way in
which boundary creation may be viewed is as temporally bound, dynamic
forms. This implies that the characteristics used to categorize groups may
change over time or as an outcome to the rules being applied.

The next approach Scott considers is called the areal or area based
organizational field. This "focuses on the relations among a collection of
organizations (and perhaps other types of social units) within a geographic
area" (p. 128). It moves the emphasis away from the organization or
organizational category to a larger environmental field, and places,
according to Scott, two constraints on the study of interorganizational
linkages. First, this approach focuses on organizations within the same
geographic area. Second it emphasized horizontal relationships among
organizations and downplays vertical linkages.

Finally, Scott (1992) argues that in recent years the focus of

interorganizational scholars has been expanded to address the functional

9

similarities of organizations. In defining function, Scott (1992) draws on

the related concept of a "societal sector":

(1) a collection of organizations operating in the same domain,
as identified by the similarity of their services, products or
functions (2) together with those organizations that critically
influence the performance of the focal organizations; for
example, major suppliers and customers; owners and regulators;
funding sources and competitors (pg. 131)

Recognition of these varying approaches to IORs still doesn't answer
the question of why organizations are motivated to form
interorganizational linkages. Oliver (1990) raises the question..."For what
reasons and under what conditions do organizations establish linkages or
exchanges with one another?" (pg. 24). In order to link the causes of IOR
relationships across a number of different types of IORs, she proposes six
broad contingencies that serve as possible causal factors in the creation and
sustenance of IORs. Relationship development may, she notes, result from
any one of these contingencies alone, but is most often rooted in multiple
contingencies.

The first of the contingencies proposed by Oliver (1990) is necessity.
Organizations may be required to form IORs to address regulations or
legal requirements. This leads to the development of mandated linkages
rather than voluntary linkages. In this instance, failure to address these
issues may lead, for example, to loss of resources, or professional or legal
censure.

Asymmetry motivates IOR development when organizations want to

increase their control or power over other organizations. Much of the

10

research on the loss (or fear of loss) of organizational autonomy reflects
this contingency (Oliver, 1991; Pfeffer & Salancik, 1978).

An example of the development of an IOR motivated by both the
contingencies of necessity and asymmetry is presented by Scott (1992) in
his discussion of the convergence of public and professional control
systems in health care organizations in the 19803. In 1972 a federal statute
created professional standards review organizations (PSRO's) to monitor
the cost and quality of health care within institutions within the United
States. While these organizations are made up of both public and
professional stakeholders, physician involvement is such that "PSRO
legislation protects the fundamental concept that physicians are the most
appropriate individuals to evaluate the need and quality of medical service"
(Scott, 1992, p. 109)

When organizations are concerned with reductions in costs and
increased returns, they may be motivated to form IORs to increase their
effieieeey. Much of the research focusing on the efficiency contingency,
Oliver notes, has emphasized the negative financial consequences of IORs
without considering the relationship between concerns for efficiency and
the contingencies of necessity, power, and/or reciprocity. The United Way
is one example of multiple organizations maximizing their efficiency.
Through a single concerted drive a group of charitable organizations
reduces the effort necessary to collect donations for multiple causes and
increase visibility in the process.

The conception of the organizational environment as a highly
uncertain place may motivate organizations to form IORs to improve their

£13121!!! within an environment. The goal of these unions is to increase

11

organizations' "stability, predictability, and dependability in their relations
with others" (pg. 246).

The norms, expectations and beliefs of others within an environment
might also lead organizations to unite to increase their actual or perceived
legitimaey within the environment. Linkages with more established or
credible units within an environment can strengthen the role of new or less
visible organizations. Miller, Scott, Stage, and Birkholt (1995) report that
those who forge IOR's among agencies dealing with the needs of the
homeless view legitimacy -- along with the contingencies of reciprocity,
efficiency, and stability -- as an important motive for establishing
interorganizational linkages.

Finally, organizations may be motivated by the desire to pursue
common goals or issues. This Oliver (1990) refers to as the contingency of
reeipreeity. This approach, she indicates, is "theoretically rooted in
exchange theory" (p. 244).

While changes within the environment increase the probability of
IOR development, and one or more of the contingencies discussed by
Oliver (1990) may act as the catalyst for IOR formation, at the core of any
IOR is the issue around which the relationship develops. Organizations that
unite to form IORs are equally, and perhaps most specifically, tied to the
substantive Lssyes that caused the organizations to come together and the
activities and transactions that are developed to address those issues. For
example, Pasquero (1991) reports on what he refers to a
"supraorganizational collaboration" among the offices of the Canadian
government, industry, ecology groups, university, and groups of native
peoples. Each of these groups had one or more motivations for becoming

involved in this IOR, but all were united by the issue of environmental

12

protection within their country. The possibility or probability of
implementing change related to a specific issue, therefore, quite often
resides not with an individual organization or stakeholder but with a group
of organizations and stakeholders addressing a mutual problem.

When trying to find a way to categorize IORs in terms of the issues,
goals, or functions with which they are dealing, one broad distinction that
has been suggested is the extent to which there is agreement on the
definition of the issues and relevant activities by those involved (Boje &
Whetten, 1981; Hall, 1992; Levine & White, 1961; Mulford, 1984; Oliver,
1990; Scott, 1992; Van de Ven, 1976). Interorganizational activities or the
issues addressed by those activities fall to a certain extent along a
continuum. At one end are MW that arise
when there is a consensus among those within the domain about the
differentiation of role and task, and compatibility of ideological concerns
such as goals, problem definition, and process (Mulford, 1984; Boje &
Whetten, 1981; Hall, 1992; Scott, 1992). At the other end of the spectrum,
however, are W about which there is little consensus and a
high level of uncertainty about functions, roles and activities (Scott, 1992).

Routine interorganizational transactions illustrate the idea that
complexity in a domain in which there is a strong consensus can result in
"vast new forms of certainty, which organizations may obtain by the mere
process of conformity to environmental specifications" (Meyer, 1978, p.
361, 363). Conversely, those interorganizational activities or relationships
that are initiated without clearly defined roles, activities, or functions carry

a greater degree of uncertainty. Scott (1992) writes:

"Although new laws, administrative agencies, and professional
occupations are continually created, giving rise to new

l3

rationalized myths that provide a basis for organized action
where none existed before, we would not overemphasize the
amount of certainty that results. Laws are often ambiguous and
variously interpreted; state and federal bureaus represent an
increasingly vast and diverse collection of interests and programs
that are often contradictory or competing; and professional
occupations challenge one another's visions of truth." (p. 148)

Routine transactions may form the glue that helps preserve the
stability of systems in which there is domain consensus. However, within
the rapidly changing environments of many organizations and stakeholder
groups, it is the unique transactions that are of particular importance and
critical interest. It is on transactions at the unique end of the spectrum that
this dissertation will focus.

The more ambiguous the issues the greater the challenge to
organizations when they try to create mechanisms for interorganizational
relationships. In addressing how IORs meet these challenges researchers
have traditionally focused on the interorganizational meme chosen.

Whetten (1981), in his review of the interorganizational field,
describes three broad ways in which the participation of individual groups
involved in IORs is structured. The categories "vary in terms of intensity,
form of operational social power, formalization, and scope of coordination
activity" (p. 11).

W is the weakest form of coordination. In mutual
adjustment situations organizations maintain most of their own authority.
Involvement requires few costs or sanctions. Correspondingly, a mutual
adjustment arrangement provides, according to Whetten, a very narrow
range of benefits. Activities are more a reflection of the individual needs

of an organization or client rather than any overarching issue.

14

The strongest (i.e., most formalized) form of coordination is the
the. The corporate form most closely resembles a "single multiunit
organization" (Whetten, 1981, p. 13). There exists a single coordinating
body which is responsible for maintaining control, monitoring activities,
and sanctioning actions that don't reflect the collective goals of the system.
Whetten notes that this form is most often resisted by the individual
organizations within an interorganizational system.

The third form Whetten discusses, the Allieeee, lies between mutual
adjustment and corporate. Interorganizational relationships taking this
form adopt coordinating characteristics from both corporate and mutual
adjustment forms. Alliances attempt "to coordinate autonomous
organizations without the authority of a formal hierarchy" (Whetten, 1981,
p. 13). Federations, councils, and coalitions are examples of the alliance
approach, and organizations developing an IOR using this strategy
generally take one of two forms of power distribution. The first power
distribution form is the creation of a separate unit with limited decision
making authority to deal with the needs or wishes of the individual
organizations or groups. The second form involves the development of
coalitions or councils in which power rests with the individual members
and decision making requires some type of negotiation among members.

A central issue for each of these structure types is the extent to which
each of the participating organizations maintain their individual autonomy.
Loss of autonomy has been broadly cited as a principle reason
organizations fail to create or join existing IORs (Oliver, 1990; Pfeffer &
Salancik, 197 8). Research conducted by Oliver (1991), however, offers
empirical evidence to contradict this position. Her work suggests that even

given a potential loss of autonomy, organizations are motivated to work

15

together to address common issues. The question then becomes how
organizations that have chosen to become involved in IORs collaborate in
order to negotiate issues such as autonomy. Thus, investigations of the
process variables inherent in creating collaborative interactions take on
increasing importance.

The megs of developing a collaborative agreement relies on the
concept of a "negotiated order" among organizations (Day & Day, 1977;
Nathan & Mitroff, 1991; Strauss, Schatzman, Bucher, Erhlich, Sabshin,
1963; Trist, 1983). According to Nathan and Mitroff (1991), a
"negotiated order exists when organizations have jointly determined the
terms of their future interactions with one another" (p. 164). Negotiated
order, Trist (1983) notes "will need to be founded on collaboration rather
than competition, collaboration being the value base appropriate for
adaptive cultivation of interdependence" (p. 273).

Definitions of collaborative agreement highlight the process aspects
of the IOR interaction. Ring and Van de Ven (1994) write "....cooperative
IORs are socially contrived mechanisms for collective action, which are
continually shaped and restructured by actions and symbolic interpretations
of the parties involved" (p. 96). Wood and Gray (1991) define
collaboration as "occur(ing) when a group of autonomous stakeholders of a
problem domain engage in an interactive process, using shared rules,
norms, and structures, to act or decide on issues related to that domain" (p.

146). Organizations act collaboratively to:

- increase their individual control over or more efficient use of
resources such as funding, and political influence,

- increase their legitimacy within the domain or channels of
communication (Golich, 1991, Sharfman, 1991, Wood & Gray,
1991),

l6

- create the rules for accessing and using resources or for defining
a common problem (Wood & Gray, 1991).

Inherent in the process of collaboration are new roles within new
relationships. Therefore, organizations and/or stakeholder groups face
important challenges when they engage in collaborative IORs. Rather than
decreasing environmental and organizational complexity, collaborative
activities may actually increase uncertainty at all levels and create more
turbulence within the environment (Bresser, 1988). Wood and Gray
(1991) write that collaborative activities can increase uncertainty,
complexity and turbulence due to the creation of new levels of dependency

among organizations. These activities can (Wood & Gray, 1991):

- increase an organization's transactional costs

- involve them in new bilateral and multilateral relationships

- require the development of new skills and/or the abandonment
or reshaping of currently existing and much used skills

- lead to the creation of priorities (policy, economic, institutional)
that challenge what the organization may perceive as being its best
interest

- close off existing and create new avenues of "action,
interactions and relationships" (p. 158)

The potential for creating increased ambiguity in an already
ambiguous situation raises the question of how, in developing these new
relationships, organizations deal with the processes of collaborative
interaction that enable them to not only address the issues for which they
are created but also the challenges that arise from increased
interdependence. Ring and Van de Ven (1994) write that "...parties must
negotiate and commit to achieving congruency in presently expected
consequences by undertaking a line of behaviors regarding a (presently

uncertainty-filled) future"(p. 99). In order to more fully discuss the

17

process of IOR development it is important to understand some of the basic

ideas associated with the concept of m within IORs.

r ‘z ’ x han
A broad variety of theoretical perspectives (political,

microeconomic, resource dependency, strategic management) have been
applied to the study of interorganizational relationships (Whetten, 1981;
Ring & Van De Ven, 1994). Regardless of perspective, most approaches
are based on the concept of emhaege among organizations within a
domain. Exchange is designed to either benefit a specific organization or
stakeholder group (competition and control), or to create a new unit or
process or solve a problem involving all of those concerned (coordination
and cooperation). The outcomes of organizational exchange have been
characterized and analyzed in terms of resemeee (goods, services, or
personnel), mm (Aldrich, 1979; Weick, 1973) or am
W (Scott & Backrnan, 1990).

In their 1985 review of the literature, Eisenberg, Farace, Monge,
Bettinghaus, Kurchner-Hawkins, Miller and Rothman note that lack of
integration among the disciplines conducting IOR research (Gottfredson &
White, 1981) and conceptual differences among those taking a resource
exchange perspective and those taking an information exchange perspective
have limited and slowed down research in this area. They indicate that
there has also been a confusion in the literature between linkage types and
linkage levels. Linkage types consist of either material exchanges
associated with resource dependency or information exchanges which tend
to be more symbolic in nature and associated with reduction in

environmental uncertainty. Linkage levels, on the other hand, refer to

18
levels of analysis. At the MM level exchanges occur "without the

involvement of specific organizational roles or personalities (e. g., routine
data transfers between banks)" (p. 236). The next level is the representative
level. Exchange at this level "occurs when a role occupant who officially
represents an organization within the system has contact with a
representative of another organization (e. g., an interagency committee to
formulate joint policies)" (p. 236). Finally, the third level is the BET—3M1
level at which exchange "occurs when an individual from one organization
exchanges information or material with an individual in another
organization, but in a nonrepresentative or private capacity" (p. 237).

These authors credit Aldrich (pp. 106-135, 1979) with pointing out
that the resource and information perspectives are complementary and that
the properties of informational uncertainty and resource dependency are
related. Thus, one important issue in the study of IOR development is a
consideration of the relationship between information and resource
exchange.

The relationship between the exchange of information and resources
is addressed by McCann (1983). In addressing social problem-solving
(SPS) interventions, he proposes a broad three stage process by which
social problems are addressed by interorganizational groups. McCann
(1983) writes that "the dynamic, unbounded nature of social problems
creates many conceptual difficulties that limit a shared understanding of
their causes and effects" (p. 177). The difficulty with capturing the
definition, of the issues can impede the implementation of decisions that
relate to resource exchange, structural arrangements, and the renegotiation
of roles. The processes proposed by McCann are "three overlapping,

though not congruent processes that build upon each other" (McCann,

19
1983, p. 182). The first process is emblemgetg'eg. During the problem-

setting process, stakeholders must achieve some level of agreement on the
definition of the issue(s) under consideration and clarify the claims of
stakeholders through the exchange of information. McCann views this as a
critical stage because pressures to take action and/or differences in
stakeholder power or influence may constrain the process of problem-
setting. These pressures and differences may lead to what Kilrnan and
Mitroff (1979) refer to as effective solutions to the wrong problems. After
the problem has been clearly defined, and stakeholders share "a sense of a
common predicament" (p. 180), they move on to the W
process. It is during this process that the more desirable end state is
determined. Following this decision, the necessary actions of the
collaborative and the individual stakeholders to bring this state into being
are determined. These two decisions "ideally result in superordinate goals
that imply a more or less explicit direction for action by stakeholders"
(McCann, 1983, p. 180). It is through this process, McCann writes, that
the legitimacy and value of the end goal is established. This legitimacy
hinges on the quality of the problem-solving process through which the
definitions were created. The final process he refers to as the W
process. This is the stage in which the concerns of "functional viability" (p
180) are addressed. It is through the negotiation of a structural
arrangement that mechanisms are created to balance the benefits for each
stakeholder, implement policies and programs, facilitate the exchange of
resources, maintain the identity of the collaborative, and allow for
adaptation to change (p. 181). McCann makes an argument that is
intuitively appealing when addressing the development of collaboratives

dealing with unique transactions. The stages he proposes suggest a

20

movement from information exchange to resource exchange that can be
used to understand the interaction of individuals involved in the
development of IORs.

Before laying the groundwork for a decision making model within a
collaborative interaction, I will first summarize the literature discussed to
this point. In increasing numbers, organizations have become involved in
IORs as a result of a rapidly changing social and technical environment.
Until recently, IOR research has focused primarily on the antecedents and
consequences of interorganizational linkages. Little research has been
devoted to the study of the developmental processes in IORs. Research on
the creation of IORs has addressed the key dimensions of the environments
in which organizations exist and how the interactions among organizations
within environments have been conceptualized. Research on the
contingencies that lead to involvement in IORs confirms the idea that
organizations enter IORs with one or more agendas. Central to these
agendas is the eemmoe issue that unites the groups involved in creating an
IOR. These issues can be’ categorized on a continuum from M
(consensus as to role, task, goals, problem definition and process) to gene
(a high level of uncertainty about functions, roles and activities). The point
at which an issue falls along this continuum has an effect on the complexity
of interorganizational interactions.

In a review of the interorganizational field (Whetten, 1981) suggests
three forms of coordination by which the participation of individual groups
involved in IORs is structured. These include, at one end, Mutual
Adjustment in which there is minimal collaboration, few costs and a
narrow range of benefits for involved organizations to Cotperete at the
most formalized end. Between these two lies the Alliaece. IORs

21

(federations, councils, and coalitions) that adopt an Alliance approach
attempt to negotiate individual and organizational benefits while
minimizing the cost to the individual organizations. This middle ground is
where organizations dealing with unique issues most often operate.

The collaboration required for Alliance formation implies the
development of a negotiated order based on cooperation rather than
competition. In addition, however, to offering new opportunities this
collaborative process can create increased uncertainty due to the creation of
new roles, relationships, goals, necessary skills, and priorities. The move
from initiating an IOR with all of its uncertainty to the development of a
sustainable relationship relies on the concept of exchange. A review of the
literature on exchange within IORs suggests that within the process of
producing an IOR there are two forms of exchange (resource and
information) and these forms are interdependent. Consequently, to study
the process of IOR development requires investigating the relationship
between information and resource exchange. Finally, a three stage process
that addresses the development of the relationship between information
exchange and resource exchange in social-problem solving was reviewed.
This model indicates that stakeholders first define the problem. They then
determine the goals and activities necessary to solve the problem. Finally,
in order to sustain their collaboration, they create some form of
organizational arrangement. Drawing on the literature reviewed above,
the remainder of this chapter is organized into four arguments that form a
framework for considering participation and information exchange in the

IOR development process.

22

no .L:-'| E th'- , «.M' he. mfr 011'1001IV‘ Iner tin

0 91.9 .. n'nl ' - . - -n-.. 'vs. - - 'r oOfIR Infrrn 'on

Argument #1: While the goals and outcomes of IORs are
conceptualized primarily at the organizational or domain level, the
emergent processes or relationships needed to deal with the uncertainty of
IORs should be addressed at the representative level (Eisenberg, et a1,
1985). Exchanges at the representative level form the point of contact for
organizations attempting collaboration. Representatives are the assigned or
legitimate negotiators and, as a consequence, the carriers of information
for their individual organizations. They form leadership groups having
boundary spanning responsibilities for implementing interorganizational
activities and relationships (Ring & Van De Ven, 1994).

Organizational members holding boundary spanning roles within
organizations are defined by their activities. These activities are "those that
serve to functionally relate the organization to its environment" (Adams,
1980, p. 328). The boundary spanner's activities fall into five classes.
Specifically, the boundary spanner filters the inputs and outputs of the
organization, collects information, represents the organization to the
environment and protects the organization from threat or pressure from
the external environment (Adams, 1980).

Given these classes of activity, the boundary spanner role is complex
and fraught with uncertainty. Boundary spanners must represent their own
organization, balance the requirements of their own organization with
those external to the organization, interpret information acquired

externally and encode information for transmission. The expectation that a

boundary spanner will act as both an organizational buffer and an

23

informational conduit can create a high level of role conflict and stress
(Adams,l980). For instance, when the boundary spanning role conflicts
with other organizational, professional, or administrative roles, there may
be an increase in turnover (Eisenberg et a1, 1985; Ring & Van De Ven,
1994). In the case in which a professional is holding a boundary spanning
position, Eisenberg et al (1985) indicate that turnover

"can impact negatively on the stability of interorganizational
systems, since the turnover of key liaisons or boundary role
occupants can interrupt long-standing patterns of formal and
informal information exchange. Particularly in interorganizational
networks where linkages are voluntary, personal, and information-
oriented, exchange relations which have developed and strengthened
over time are vulnerable to the effects of professional turnover"
(pp. 250-251). '

An argument can be made, therefore, that if boundary spanners do not
display a certain level of commitment to and satisfaction with IOR
activities, the organizations in the IOR may withdraw from the relationship

before any substantive outcomes can be achieved.

:OJLeI our ouum'u 0." -_. 1'.|¢..5" '0': 0
ICE S . l .1.

Argument #2: Representative satisfaction with and commitment to
the boundary spanning activities are necessary for the continued existence
of the collaborative.

. What does commitment mean at the collaborative level? Because
collaboratives by their definition are concerned with bringing about change
related to issues of mutual concern, it is reasonable to presume that there is
a commitment to the collaborative relationship on the part of involved
individuals (Wood & Gray, 1991). In other words, it is likely that those

24

involved will continue their involvement in the IOR until the issues that
brought the IOR together have been dealt with. Wood and Gray recognize,
however, that those participating are autonomous and, therefore "retain
their independent decision-making powers even when they agree to abide
by shared rules within the collaborative alliance" (p. 147-48).

What does IOR commitment mean to the boundary spanner?
Organizational and/or stakeholder representatives play a linking role within
any IOR. This linking role brings with it: 1) expectations about how the
IOR will affect their organization or group, 2) expectations, in the case of
professionals, about how their IOR involvement will affect their
profession, and 3) expectations about how the IOR will affect them
personally in terms of their profession and organization. The interaction
among boundary spanners creates a context for negotiating, defining, and
legitimizing this new role and its accompanying expectations. If there is
little commitment to this process, then turnover of organizational
representatives may result. If a lack of boundary-spanner commitment
leads to turnover in this critical role, "levels of flexibility and efficiency
that may have existed in management of the cooperative IOR are likely to
be lost as the new 'agents' rely on the terms of the formal agreement and
their role designations in resolving matters that their predecessors had dealt
with based on psychological contracts and reliance on trust" (Ring & Van
De Ven, 1994, p. 104).

Organizational commitment at the individual organization level, as
defined by Mowday, Steers, and Porter (1979), refers to "a strong belief in
the organization's goals and values, a willingness to exert considerable
effort on behalf of the organization and a strong desire to remain a

member of the organization" (p. 226). One limitation to applying a

25

definition of "organizational" commitment to a professional in a boundary
spanning role is that it refers to only one object of commitment -- the
organization. Boundary spanners, however, have unique roles within an
organization. They may be "more distant psychologically,
organizationally, and, often, physically from other members of their
organization" (Adams, 1980, p. 329). Within the collaborative they must
rely more on referent and expert power versus the reward or coercive
power available to those who operate primarily within the organization
(Adams, 1980; French & Raven, 1959). Finally, "(o)ccupants of boundary
roles experience dynamic, dual conflicts with outsiders, on the one hand,
and with insiders, on the other hand" (Adams, 1980, p. 331). It is feasible
to assume, therefore, that a definition of collaborative commitment at the
boundary spanner level should reflect a balance -- though not necessarily
an equal balance -- between loyalty to the values and beliefs of the
organization and profession and those of the collaborative. To the extent
that boundary spanners feel that the goals and values of their collaborative
activities support those of their organization and profession, a
straightforward definition of organizational commitment can be applied to
the boundary spanner.

Boundary spanners' feelings about the extent to which the activities
of the collaborative support the values and goals of the individual
organizations are reflected in their level of satisfaction. Satisfaction has
been defined not only in terms of individual response to specific job tasks
(Glisson & Durick,l988; Locke, 1976), but also as a reflection of how the
job allows an individual to meet his/her physical (e. g., pay, working
conditions) and psychological needs (e. g., promotion, verbal recognition,

interaction with others) (Locke, 1976). The values and goals of

IEI

26

organizations and professions that lead to a sense of commitment have their
origins, at last partially, in the activities and interactions within
organizations and professions (Locke, 1976). These activities and
interactions allow those within an organization the opportunity to use their
skills and to be rewarded for their work (e. g., monetary, promotion, or
recognition) (Locke, 1976). Satisfaction with one or more aspects of these
variables has been repeatedly discussed in organizational research to as an
antecedent to commitment (Decotiis & Summers, 1987; Glisson & Durick,
1988; Locke, 1975; Marsh, & Mannari, 1972; Starnarnan & Miller, 1992;
Williams and Hazer, 1986).

The activities and interactions of boundary spanners within a
collaborative create a linking mechanism for the stakeholder organizations.
A certain level of boundary spanner satisfaction is required at each stage of
IOR development for the continued growth and/or evolution of the IOR.
Boundary spanner satisfaction should reflect both specific experience with
the work of the IOR and the psychological and physical needs linking work
within the IOR to their organizational positions. In other words, boundary
spanners have not only a need to feel that their work within the IOR uses
their skills, challenges them, and is worthwhile, but must also sense that
this work will bring both physical and psychological rewards within their
own organization.

In a general sense boundary spanning activities have been found to be
positively correlated with satisfaction with organizational co-workers,
promotion opportunities, and pay (Keller & Holland, 1975). If
organizational representatives feel that their interorganizational efforts
hinder other work goals by creating more difficult working conditions or

will not lead to rewards, they may report a lesser degree of satisfaction

27

with these activities. If they feel that there is little recognition by others
within the IOR of the values and goals associated with their organization
and/or profession, they may report less satisfaction with the IOR and its
activities. Conversely, if they feel that their activities within the IOR are
not recognized within their organization and/or profession and, as a
consequence, may not lead to the enhancement of their organizational or
professional roles, they may report less satisfaction with the IOR.
Therefore, those that report they are satisfied with their interorganizational
activities and that these activities reflect and reinforce their organizational
and professional expectations should also report commitment to the goals
and values of their collaborative group and a willingness to exert effort on

the behalf of the collaborative and remain a member of that

interorganizational group. 111W
in.-- ‘ -‘Q r -I- r I 1- 1-1 1' ‘ con 0 H nu; uu‘

In considering the ways in which satisfaction can be enhanced, early
research considered the ability of a job to meet the physical (e. g., pay,
good working conditions) and psychological needs (e. g., promotion,
recognition) of individuals. A growing body of organizational literature
has now, however, moved beyond these task characteristics to focus on the
ways in which organizational work is configured. More specifically,
researchers have recognized that the process of participating in the
decisions within an organization has an effect on individual satisfaction.
Because collaborative interorganizational relationships emphasize
cooperation over competition, the participation of boundary spanners in the

decision making process may be considered as a key mechanism in not only

28

achieving the outcomes of the collaborative, but also in enhancing the

satisfaction of boundary spanners.

' i n ' w' ' 01 r 'v

Argument #3: Decision making at the collaborative level represents
decision making in one of its most complex forms. The decision making
interactions of boundary spanners are hypothesized to reflect the needs,
expectations, values, and goals of all involved stakeholder groups. As a
consequence, decision making within a collaborative group requires the
participation of all boundary spanners.

At a most basic level, decision making centers around what decisions
need to be made, how they should be made, and who should make them.
Setting aside the question of "what" for a moment, I would like to briefly
discuss "how" and then in greater length consider "who".

The question of how decisions are made has been studied by
organizational scholars from a variety of perspectives. For instance, the
decision making process has been researched as a rational, normative
process in which a problem is defined, information gathered, options
evaluated, and ideal decisions made in an orderly fashion (Janis & Mann,
1977). The difficulty with taking the "rational" approach is, according to
March and Simon (1958), that it assumes that all alternative choices are
known, all the consequences of each choice are known, and that the
decisionmaker can rank all possible sets of consequences. The outcomes of
organizational decision-making are not always nearly as optimal (March &
Simon, 1958) or tidy (Cohen, March & Olsen, 1972) as the rational
approach might hope. March and Simon (1958) generated a "satisficing"

model of decision-making in which organizational actors search not for an

29

optimal solution, but for one that is "good enough" to deal with the issue
under consideration. Cohen, March, and Olsen's (1972) garbage can model
is predicated on the idea that problems, solutions and those given the
responsibility of decision making are "dumped" together. It is through
chance that solutions become matched to problems and the identification of
how the decision was made follows the identification of the outcome. So,
even in its most simplistic form, decision making is a complex process.

In addition to consideration of the decision making process itself,
research has also considered the scope of involvement and role of
participants in the decision making process. This interest is based
primarily on the assumption that the participation of individuals at all
levels within an organization would be good for the organization (i.e.,
increased productivity) and good for employees (i.e., increased
satisfaction) (Locke & Schweiger, 1979). Participation in decision making
(PDM) has been defined by Locke and Schweiger (1979) as "joint decision
making" (p. 274). This definition, the authors note, doesn't specify who is
participating (e.g., only subordinates), the equality of the participation, or
the content. It only specifies "participation in the process of reaching
decisions" (p. 274).

In a meta-analysis of 45 studies on participation in decision making,
Miller and Monge (1986) tested three models of participation suggested by
the theoretical literature -- a cognitive model, an affective model and a
contingency model. Research that addresses participation from a eegm'tiLe
mm emphasizes the relationship between PDM and the "flow and
use of important information in organizations" (Miller & Monge, 1986, p.
730). This model is based on the assumption that individuals possess high
quality information related to his/her specific job or position. The

30

opportunity to participate and influence decisions made within the
organization will facilitate the use of that information and lead to an
increase in productivity and an eventual increase in satisfaction.

I , Upward Information Flow\ ‘

\\
’/ \

, PDM: : Productivity---->Satisfaction
\ x /
‘ Downward Information FIOW’ ’

Figure 1. Cognitive Model of Participation in Decision Making

Those proposing an Meade] view the goal of PDM in terms
of the satisfaction brought about by the satisfaction of individual higher
order needs. Participation, in this model, is not necessarily related to
specific decision making issues but to the perception of a participative
climate..."for it is the act, not the informational content, of participation
that is the crucial mechanism" (p. 731). Increases in satisfaction must first

be achieved before there will be any increase in productivity.

Satisfaction Work
PDM---> of higher-order--—> satisfaction---> Motivation---> Productivity
needs

Figure 2. Affective Model of Participation in Decision Making

W focus on the differences in individuals and
situations that could have an affect on need for or use of participation in
decision making. For example, researchers taking this approach emphasize
that in dealing with certain situations or issues (i.e., complex versus simple)
participation may be more or less appropriate.

As a result of their meta-analysis, Miller and Monge found support
for both the cognitive and affective models of participation. There is
strong support for the idea that perceptions of participation affect

31

satisfaction. There is also weaker support for the notion that participation
in specific work-related decisions influenced productivity.

Using these models to discuss the participation of boundary spanners
within a collaborative is both useful and problematic. The models are
useful because they offer a sounding board against which ideas can be
tested. They are problematic because interorganizational collaboratives do
not necessarily have established structures, roles, values, or goals. As was
discussed when considering satisfaction and commitment at the
collaborative level, there may be multiple conflicting goals, roles, and
values. While these conflicting issues may create problems within
individual organizations, they can have even greater ramifications for those
involved in collaboratives.

For instance, in comparing participation in decision making in a
collaborative to these models, it is important to first consider the outcome
of interest. A great deal of previous research has focused on outcomes of
individual satisfaction or productivity. In the initial phases of a
collaborative, however, one of the key outcomes is the sustainability or
continuation of the IOR. This continuation allows for the possibility of
more substantial exchanges at a later date. As was proposed earlier, if we
consider this at the individual level, boundary spanners who are satisfied
with both their collaborative activities and feel that these activities meet
their physical and psychological needs, may potentially report a greater
level of commitment to the collaborative. This commitment may increase
the probability that the relationship will be sustained.

The question of interest for this dissertation, then, is whether the link
between participation in decision making and satisfaction can be expected
for those participating in the decision making of collaboratives. A growing

32

body of interorganizational literature highlights certain activities as being
essential to the development and sustainability of. collaborative
relationships. These activities include participation in decision making,
development of shared values, conflict negotiation mechanisms, role
definition and agreement (Gray, 1985; Ring & Van De Ven, 1994; Wood
& Gray, 1991). In discussing their definition of collaboration, Wood and
Gray (1991) highlight the significance of the decision making process.
Collaborative involvement presumes that "the participants must intend to

'act or decide (p. 148). Decisions about domain definition, outcome
identification, activities to facilitate outcome achievement, and level and
type of involvement must be dealt with in order to bring about
interorganizational mechanisms and structures to sustain the domain level
change. This definition of a collaborative hinges on those involved
engaging "in an interactive process using shared rules, norms, and
structures to act or decide on issues related to that domain." (Wood &
Gray, 1991, pg. 146).

Decision making within a collaborative, however, is conceptually
distinct from decision-making within a single organization. As indicated
earlier, decision making in its most basic form is a complex process.
When, however, there are multiple actors, representing multiple
organizations and/or stakeholder groups with varying cultures,
expectations, and needs, then the parameters for determining what
decisions to make and how to make them become even more obscured.
The negotiation and development of an agreed upon order for the
collaborative relies on boundary spanners reaching some level of

understanding regarding the underlying roles, rules and outcomes

associated with the issues that brought them together. Ring and Van De

33

Ven (1994) write that "Congruency is a cumulative product of numerous
interactions; through these interactions emerge trust in the good will of
others and an understanding of constraints on the relationship that may be
imposed by a person's organizational role" (p. 100). The decisions made
within those interactions will determine future relationships among
involved individuals and organizations.

Through the development of increased understanding and trust, a
new organizational arrangement or referent structure (Trist, 1983) can
emerge linking the stakeholders (Gray, 1985; McCann, 1983). This
organizational arrangement is based on members of an interorganizational
collaborative group creating at least a minimal set of agreed upon goals,
values, and rules related to the issue that forms the basis of their
interaction. As a part of this process issues of collaborative roles and the
extent to which each representative helps shape the emergent structure have
to be addressed. Because collaboratives are based on cooperation, the
extent of the influence of individual members may have to be negotiated.

An ideal collaborative situation would involve symmetrical
relationships in which each participant has equal opportunity to participate
and an equal chance of influencing decisions. Decision making at the
collaborative level, however, is very rarely ideal. For example, individual
influence based on professional or organizational affiliation could be
affected by status distinction or influence differences based on precedence
or tradition, expertise, resources, and/or law.

Participation, however, can and has been assessed through both the
opportunity to participate and the perceived or actual influence associated
with participation. These two dimensions (opportunity and influence) are

issues that are at the core of participation research. Early researchers on

34

participation report that while participation allows individuals to express
opinions or views (Argyris, 1955; Locke & Schweiger, 1979), it doesn't
necessarily guarantee influence in the final decisions that are made
(Hoffman, Burke, & Maier, 1965; Locke & Schweiger, 1979). More
current research by Marshall and Stohl (1993b) specifically considers the
issues of influence (empowerment) and opportunity (involvement) in a
participative structure. In order to obtain a clearer picture of the
relationship between these components of participation and the outcomes of
satisfaction and performance, Marshall and Stohl (1993b) examined both
communication network and individual indicators of empowerment and
involvement. They found that the extent that workers within a
participative organization are involved and have communication links that
would allow them influence in the decision making process is positively
related to their level of performance and satisfaction. Questions raised
about the differential effects of empowerment or involvement on these two
outcomes indicate that empowerment is more strongly linked to satisfaction
and performance appraisal than involvement.

While these concepts are discussed in the literature primarily in
terms of the participative opportunities and levels of influence of
subordinates within an organization, the broad definition used by Locke
and Schweiger (1979) (i.e., joint decision making) does not exclude
addressing the participation of individuals in collaborative groups. In fact,
using these dimensions of participation to discuss the collaborative
participation is most appropriate. Boundary spanners, as noted by Adams
(1980), collect information, filter organizational inputs and outputs and
represent and protect their organization within the external environment.

Based on their activities, decisions are made about their organization's role

35

within the environment and, as a result, decisions are made within their
organization. Adams (1980) notes that "(m)aladaptive decisions are a
consequence of inadequate information concerning external events"
(Adams, 1980, p. 348).

Because participation in decision making has been consistently linked
satisfaction (Jackson, 1983; Marshall & Stohl, 1993b; Miller & Monge,
1986, 1987; Schuler, 1980), it is reasonable to assume that this same link
should be considered for boundary spanners in collaboratives.
Organizational representatives in IORs have responsibility for decisions
made within the collaborative that affect their organizations. This
responsibility makes the issues of opportunity and influence in the decision
making process critical. It is feasible, therefore, to assume that the more
opportunities boundary spanners have for participation in decision making
and the greater their level of perceived influence, the more control they
will perceive that they have over the collaborative process. This control
should in all probability heighten the sense of satisfaction with
collaborative specific tasks and help create the perception that there could
be both physical rewards (e. g., pay increases) and psychological
recognition for their interorganizational activities. Conversely, those who
feel that they have limited opportunities for participation and/or limited

influence on collaborative activities should be more apt to report lower

levels of interrorganizational project satisfaction. Therefere, e pesits've
-_._n I_ hi ' AII ’lZ‘I I xi Ira. II I-JI lI-_III ' I‘ i in

l' 1 'SE .

36

The Role f Emer en omm '0 tion in e Decision Makin

Preeesses ef IQRs

Concerns that PDM research needs to be expanded to deal with the
critical role communication plays in facilitating participation have recently
been expressed. Monge and Miller (1988) write that "(a) critical
dimension of the participative process is the role of communication. To
date, however, little scholarly attention has been devoted to this process"
(213). These authors suggest that rather than just comparing the attitudes
and behaviors of individuals in participative and nonparticipative
structures, it would be more informative to study the communication
content of decision making. Participation in decision making implies
communication through the emphasis on the interaction of individuals
within an organization. Participation, however, is a complex process and,
as a consequence, so is the communication process required to initiate and
sustain it.

One example of this complexity is addressed by Marshall and Stohl
(1993b). These researchers indicate that a great deal of past research on
participative organizations has equated the participative smemse with the
participative meeess. Marshall and Stohl (1993b) suggest that the
participation literature so far contains a critical design bias. "Participation
traditionally has been viewed as a dichotomous variable, a static entity that
is either present or absent in an organization. Participation is assumed to
reside in the structure of the organization" (pg. 138). Changes, therefore,
in the roles, relationships and activities of individuals have been studied in
terms of the changes in structure. These authors propose that the emphasis

on participative structure does not allow researchers the opportunity to

3., »1-i

37

study the enactment of participation and limits their ability to address
performance and/or satisfaction variations within a participative structure.

A basic assumption throughout the research literature is that the
creation of a participative structure reflects the emergent participative
process. However, Marshall and Stohl (1993b), in their work on the extent
to which individuals actually participate within a participative structure,
suggest that the creation of a structure does not necessarily reflect the
process or even existence of individual participation. Presupposing that
participative structure equals participative process also limits the extent to
which emergent communication can be investigated.

Monge and Eisenberg (1987) in their review of the research on
emergent communication networks in organizations discuss the relationship
between structure and process. They write that in order to discuss the
rather fragmented literature on emergent communication it is necessary to
address the research on communication networks in terms of structural
inquiry. They define structure as "a collection of elements or parts and the
set of relationships that connect the parts together" (p. 305). Elements,
they note, can refer to organizations, people or even the language that
connects the parts of the structure.

Structural theory falls into the following three categories:

1. positional - in which the patterns of relationship within an organization
are linked to formally defined positions and roles, 2. relational - in
which the "structures lies in the emergent interactions between people" (p.
306). Structure, therefore, is determined at the individual level and
emerges through the dynamic processes of interaction, and 3. cultural -
which focuses on the transmission of symbols and their meanings

throughout a social system.

38

From the attempts to integrate these models, Monge and Eisenberg
(1987) draw the following three principles: First, both formal and
emergent networks exist and it is through their relationship to one another
that they are best understood. Second, each is valid. Third "the
predominance of either type of structure is to some degree a function of
where an organization is in it evolutionary life cycle" (p 309).

The work of Monge and Miller (1988), Marshall and Stohl (1993b),
and Monge and Eisenberg (1987) raise issues and suggest directions for
addressing emergent communication within individual participative
organizations. The issues raised are of even greater concern within the
often equivocal context of collaboratives. For example, collaboratives may
prove to be ideal for addressing Miller and Monge's (1988) suggestion that
researchers considering participation need to focus on the communication
content of decision making. Collaboratives may also allow researchers to
highlight the troublesome distinction between participative structure and
participative process proposed by Marshall and Stohl (1993).

The literature on emergent communication process and negotiated
order suggest a direction for research on decision making in collaboratives.
For example, initially a collaborative can best be viewed as a relationship
among boundary spanners. Its final structure is, as proposed by the
relational tradition discussed in Eisenberg and Monge (1987), determined
by the dynamic interactions of individuals. Within the development of a
collaborative, information must be shared on topics such as goals and
values, and joint definitions determined before decision making can take
place.

It is proposed in this dissertation, therefore, that another model of
participation in decision making needs to be considered which takes into

39

account the generative role of information exchange in collaboratives It is
based on this contention that the final argument is made and the model

proposed.

Argument #4: It is through boundary spanners' W
exchangeafinfmmatiau that the W of
collaboratives are based.

The large body of research that led to the development of the
cognitive and affective models of participation in decision making (Miller
& Monge, 1986) discussed earlier conceptualizes participation as the
exogenous variable. The communication process within "participation" is
only explicitly mentioned in one of the models. The cognitive model which
emphasizes the role of information is based on the idea that "participation
in decision making is a viable strategy because it enhances the flow and use
of important information in organizations" (Miller & Monge, 1986, p.
730). This presupposes an existing state or condition (e.g., participation)
that fosters information exchange. The emphasis is on the use of
information as a result of participation.

O'Reilly, Chatrnan, and Anderson (1987) in discussing the
distinction between the stem and use of information among decision

makers write:
a decision involves the use of information to assess the
results of a future course of action, given that some
future states of the world are more desirable than others.
This implies that a decision maker either has the requisite
information available to make the choice or will obtain
information through a process of choice (p. 605)

i The assumption that boundary spanners as they enter into an IOR

will have all of the requisite information for decision making about issues

40

at the unique end of the spectrum is debatable. They will, in all probability,
need to acquire information. The acquisition of information may be done
for a number of reasons. McCann (1983) suggests that early on in the
evolutionary cycle of groups dealing with social problems, stakeholders
need to clearly define the issues with which they are dealing and clarify the
claims of the multiple groups involved. As noted by McCann (1983),
difficulty with capturing the definition of the issues can impede the
implementation of decisions that relate to resource exchange, structural
arrangements, and the renegotiation of roles. The acquisition of
information may, therefore, be done to clarify the issue or problem under
consideration and in order to keep from coming up with effective solutions
to the wrong problem (i.e., Type 3 error) (Kilrnann and Mitroff, 1979).
What Kilrnann and Mitroff (1979) term as Type 3 error is not without its
critics. Nutt (1984) suggests that this is overly simplistic. He writes that
"there are few tests that can be applied to determine if the 'correct'
problem has been selected" (p. 447). Nutt (1984) does, however,
emphasize the importance of problem definition in decision making. He
indicates that issues such as the scope of the problem may affect the
definition chosen, and that greater breadth of inquiry will lead to the
perception of more available options. So, at a very basic level some
problem has to be defined, and solutions proposed. O'Reilly, Chatrnan and
Anderson (1987) note that "(i)f a decision maker requires information to
define a problem and to generate a list of alternative solutions, it logically
follows that the quantity and quality of information available will be
related to the alternatives considered, estimates of probabilities made, and

outcomes seen as desirable" (p. 610).

41

In addition to acquiring information for defining the issues, Adams
(1980) points out that boundary spanners are under pressure to filter
information based on organizational decision-making criteria.

There are 2 types of information that boundary spanners search for and
collect for their organizations. The first of these is operating information
which is "required for current decision making and policy formation
having short-and long-term effects" (p. 342). This type of information,
while not completely lacking in uncertainty is more directly accessed
because the source and content of necessary information is known. The
second type of information Adams discusses is related to situations or
events that are less predictable. It is this second type of information that is,
perhaps, most central and yet problematic to the activities of the
informational boundary spanner in situations in which there is limited
domain consensus. This information search is problematic first because the
boundary spanners' constituents "who require intelligence for decision
making and policy formulation, exercise pressure on them for information
that is at least proportional to their uncertainty and to the perceived
importance of the information" (Adams, 1980, p. 344). Secondly,
according to Adams, there is a tendency by constituents to overrely on the
information provided. This in turn may lead to informational boundary
spanners waiting until they have collected a requisite quantity of
information to corroborate their findings (Adams, 1980). So the boundary
spanner is also under pressure to acquire information of a quality and
quantity that will enhance or protect the organization's or group's stake in

the interorganizational collaboration.

42

Finally, at the individual level, boundary spanners acquire
information to maintain, advance, and/or protect their own role within
their parent organization or profession, and the collaborative.

O'Reilly, Chatrnan, and Anderson (1987) write that at the individual
organizational level, the structure of the organization, the incentive and/or
control systems, and the group norms may all affect the quantity and
quality of information available to the decision maker. In extending this
idea to boundary spanning decision makers, it could be hypothesized that
while the collaborative organization is participative by definition, the
organizations or professions that boundary spanners represent may vary in
structure, incentive systems, control systems, and group norms. These
variations may affect the quantity and quality of information that the
boundary spanner has available to share or feels comfortable sharing. As a
consequence, the quantity of information available to be exchanged as the
new collaborative relationship is being developed may initially be limited
due to reliance on traditional structures, systems, or norms. These may
serve to reinforce the differences among those involved in the IOR rather
than their capacity to work together to address a domain issue (McCann,
1983; O'Reilly, Chatrnan, & Anderson, 1987; Ring & Van De Ven, 1994).
If boundary spanners perceive that information is either not available in
sufficient quantity or not being shared by others, then they may also be less
inclined to provide information that could possibly further interaction.
The lack of available information at the representative level could
effectively block or, at least, seriously impede the development of joint
decision making. ’

In order to determine if the information being dealt with within the

collaborative is useful for defining the issue, protecting their individual

43

organizations or stakeholder groups, and advancing, maintaining or
protecting their individual role, boundary spanners must establish a context
for negotiated order. Putnam (1983) notes that "(n)egotiated order
assumes that multiple social realities makeup organizations. These
subcultures protect and advance their own interests through negotiating the
meaning of social events. Hence, individuals reach an agreement through
creating new meanings or by trading off interpretations of events" (p. 52).
In a collaborative, negotiated order is an expression, at its most basic level,
of the creation of a shared vocabulary. This vocabulary reflects individual
organizational and/or professional meanings or the trading off of
individual interpretations based to a large degree on organizational and/or
professional affiliations. For example, the same words can have very
different meanings depending on organizational, professional, or
community differences. Actions based on the, perhaps, false assumption of
shared meaning could quickly undermine a collaborative effort.

Heimer (1985) uses the term "negotiated information order" (p. 397)
to refer to the process of using information to create a negotiated order
among stakeholders. She defines negotiated information order as
occurring "(w)hen a system of criteria for the social sufficiency of
information is worked out by a group of interrelated organizations" (p.
397). Decision making within a collaborative, then, presupposes, relies on,
and reflects the gathering, negotiating, and acceptance of information that
meets the requirements of the members of the collaborative. This, Heimer

(1985) notes has an impact on information gathering.

When a single actor can implement a decision, then the question
is only which piece of information he or she will accept as
evidence of the fact that needs to be demonstrated. But when

44

several actors are required to carry out the decision, then the
problem is not so much to get evidence to answer the question,
but to get information that everyone concerned will agree is
evidence. That is, the information needs to be socially sufficient
as well as technically sufficient, and the two are to some degree
independent, so that socially sufficient information need not be
technically sufficient and vice versa. (p. 397)

This determination of how to define the "evidence" upon which
collaborative decisions are made reemphasizes the informational role of the
boundary spanner. Based on the idea that the definition and clarification of
issues and relationships within the negotiated order of collaboratives have
their origins in the quality and quantity of information acquired and used
by boundary spanners, the following link is hypothesized between
information (quality and quantity) and participation in decision making.

Aitiv 1‘ III'nhii hrII e'ZI I‘ ‘r‘nlanl'mgInI .' -.Id

 

MII‘l I P I iI .IIn iI D i in Man! i 'n I ‘I° , I° , IIIL.
011 'v

The model hypothesized in this dissertation assumes that sustaining
the interorganizational relationship at the representative level is an
important first outcome in creating a structure or situation which allows
for the exchange of more tangible organizational level resources (i.e.,
money, personnel, goods). In order to achieve these outcomes
(sustainability and resource exchange), it is necessary for representatives
to perceive that they have both the opportunity to participate in decision
making and exert influence on those decisions. As an antecedent to
participation in collaborative level decisions, boundary spanners must have

an adequate amount of what they perceive to be relevant information.

45

Therefore, I propose a model of PDM in IORs in which the generative

variable is information acquisition and exchange.

Information Quality --- + --->
Participation --- + --->Satisfaction--- + --->Commitrnent

Information Quantity --- + --->

Figure 3
Hypothesized Model of Information, Participation, Satisfaction and
Commitment in IORs

While commitment to the IOR is the outcome and participation is the
crux of this hypothesized relationship, the interaction originates with the
exchange of information. In summary, the model (Figure 1) hypothesized
in this study illustrates the decision making process for boundary spanners
as having its origins in the perceptions of an adequate quantity and quality
of available information. This model addresses the impact these variables
have on the outcome variable of interorganizational commitment. The next
chapter will describe how this model will be tested with groups of
boundary spanners involved in multiple inter-organizational/stakeholder,

multi-professional projects.

CHAPTER TWO
Methods

This chapter explains the methods used to assess the relationships among
the quality and quantity of information exchanged by boundary spanners in
a collaborative effort, their participation in decision making, satisfaction
and commitment. It includes: (1) a description of the project from which
the data were drawn, (2) a description of the population, (3) the process of
data collection that was used, (4) a description of the survey instrument
with a breakdown of the items, and (5) a discussion of techniques used to

analyze the data.

The Kellegg Communigt Partnership Initiative
The Kellogg Community Partnership Initiative is "an effort to

establish several models of academic, community-based, primary health
centers as a means for redirecting health professions education toward the
preparation of graduates more interested in and suited for practice of
primary health care in communities" (Richards & Bouhuijs, 1991, pg. 1).
In other words, the sites were charged with developing new models of
multi-disciplinary, community-based healthcare education for students
from medical, nursing, allied health, and social work schools. Based on
their proposals, seven sites within the United States received funding from
the Kellogg Foundation for five years to become demonstration sites for
this initiative. One of the unique characteristics of this project was the
mechanism for change. Those involved were expected to develop formal

organizational linkages between community and academe to produce out of

46

47

hospital, community-based health professions education In addition to
educational reform, these projects were also involved in informing state
and local policy, and the reallocation of resources. The number of
representatives in a given project site depended on the number of schools
and/or institutions involved and the number of communities involved. For
example, at one end of the spectrum, one project site involved an entire
state, including all of the medical and nursing programs. At the other end
of the continuum, one site was made up of the medical, nursing, and allied
health schools in one university and two small rural communities.

The activities of these groups -- which were in year four of the five
year project when these data were collected -- were monitored by a Cluster
Evaluation Team consisting of nine individuals with backgrounds in
Medical Education and Research, Political Science, Medical Economy,

Sociology, and Communication.

Respondents

This particular study was part of the larger research and evaluation
project. The population consisted of 308 individuals actively involved in
the governance and decision making processes of the seven projects.
Initially, individuals where chosen for this study based on their formal
assignment to governance and decision making groups (i.e., governing
boards, curriculum, or evaluation committees). This number was reduced
to those groups that were currently active in the implementation of these
projects. So, for example, members of the research committees were not
included due to their inactivity at the time data were collected. Thus, the
relevant population for this study consist of individuals from the medical,

nursing, allied health and social work professions representing 21 academic

48

institutions and individuals representing community health centers and the
involved communities at large who had membership on these projects.

Two hundred and seventy two individuals completed either the
telephone survey and/or the mail survey. This sample was composed of 198
individuals (73%) who were W (i.e., doctors, nurses,

allied health professionals), WW, health
effigals, or ptejeetsiteeuqtdiuam, and 64 individuals (24%) who
represent the involved communities (e. g.., business people, clergy,
government employees, K— 1 2 educators, homemakers and retirees). Ten
individuals (3%) responded W. Of this sample 145
were males (53.3%) and 127 females (46.7%). Education level reflected
the predominance of professionals. Two hundred and eight (84%) of the
respondents had a bachelor's degree or higher with 181 (66%) having post
graduate or professional degrees.

For the final path analysis, 129 cases were used in which individuals
responded to all items on all factors. These individuals differed
demographically from the broader group of respondents. This sample was
composed of 116 individuals (89.9%) who were healthpuu’essteuals (i.e.,
doctors, nurses, allied health professionals), We;
administram healtlufflcials, or W, and 12
individuals (9%) who represented the involved communities (ex., business
people, clergy, government employees, K-12 educators, homemakers and
retirees). One individual responded deujtkuewmpuueu. Of this
sample 63 were males (48.8%) and 66 females (51.2%). One hundred and
twenty-two (94.5%) of the respondents had a bachelor's degree or higher
with 103 (79.9%) having post graduate or professional degrees.

49

W

The data were collected through a combined computer assisted
telephone (CATI) survey and a follow-up mail survey by the Institute for
Public Policy and Social Research at Michigan State University. This
approach was used to maximize the response rate from the participant
group. Due to the focus on those boundary spanners actively involved in
the implementation of an interorganizational, multi-professional project,
subject anonymity was not possible. Subjects were given assurances,
however, that the data would be reported in aggregate format and
individual identity would be protected.

mammalian

Twenty-seven of the one hundred and fifty-five items included in the
larger measurement instruments were used to assess the factors in this
study. Initially, most factors were developed using items from reliable and
well-validated instruments. During the process of adapting items to
address the evaluation needs of the larger study, however, item content
changed or new items were developed. These revisions or new items were
developed by the members of Cluster Evaluation Team who had been
conducting evaluation data collection at the seven sites for three years prior
to the administration of this survey. Every effort was made, therefore, to
insure the conceptual and face validity of the items. This face validity had
its foundation not only in the evaluation needs of the project, but more
basically in the literature on IORs, organizational change, community
change, and boundary spanner relationships. The factors used in this study
related to individual perceptions of information quality and quantity,

SO

participation in decision making, boundary spanner satisfaction, and
boundary spanner commitment. Specific items in these factors are
discussed below. The items in the Satisfaction and Commitment factors
were measured using four-point Likert-type scales ("to a great extent" to
"not at all"). The items in the Information, and Participation factors were
measured using five-point Likert-type scales ("strongly disagree" to
"strongly agree").

A ' l rm ' ' li
These seven items were developed by the Kellogg Community
Partnership Cluster Evaluation Team. They were based on findings from
the literature (i.e., IOR, organizational change, community change, and
boundary spanning) and site visit evaluations made on this project.
f rm ' n ' I

- Far too little information on important topics is shared among
Partnership members.

- Information on the partnership is widely shared among the
Partnership participants.

- I receive too much information related to the Partnership.
I E . Q 1. I _

- The information I receive about the Partnership gives me a clear
understanding of the Partnership.

- The information I receive about the partnership is accurate.
- I receive information about the partnership in a timely fashion.

- The information I receive about the partnership is relevant to my
needs.

51

P 'i 'ninD iinM'

The individual's perceptions of opportunities for participation and
influence in the decision making process were assessed using items
developed from the Survey of Organizations developed at the Institute for
Social Research at the University of Michigan (Taylor & Bowers, 1972)
and Wandersman and Goodman's Community Participation Survey (1991).

i i 'n m '
- It is easy to get my ideas across to the project leadership ifI have a
suggestion.

- I feel I have many opportunities for participation in the partnership.

- When decision are being made in the partnership, the persons
affected are asked for their ideas.

- Participation by community representatives on partnerships boards
and committees is high.

- Participation by university representatives on partnership boards and
committees is high.

- Participation by project staff on partnership boards and committees
is high.

- I feel that it is useless to make suggestions about the partnership
because decisions are made regardless of my attempts to influence
them.

- Decision making in the partnership is broad-based.

- Decisions are made only by a small group of leaders.

E'S'E' lE'C'

The items used to assess these two factors were developed by the

Kellogg Community Partnership Cluster Evaluation Team. They were

52

based on findings from the literature (i.e., IOR, organizational change,
community change, and boundary spanning) and site visit evaluations made
on this project.

MW

- To what extent do you feel that others on the Partnership respect
your profession?

- To what extent do you feel that other on the Partnership respect the
organization for which your work?

- To what extent do you think people in your profession or the
organization for which you work positively recognize your work on
the Community Partnership?

- To what extend do you find work in your profession to be very
satisfying?

- To what extent do you find your work on the Community
partnership to be very satisfying?

- To what extent do you expect to receive career benefits from
working on the Partnership, such as in terms of promotion or merit
pay increases.

E'E' I'

- To what extent would you say you really care about the future of the
Community Partnership organization?

- To what extent would you say you are proud to tell others that you
are part of the Partnership organization?

- To what extent are the Partnership values similar to the values of
your profession or the organization for which you work?

- To what extent does your work on the Partnership contribute to your
professional or career development?

- To what extent will your work on the Partnership be detrimental to
your professional or career development?

53

M31313

This section addresses the types of analysis used in the examination
of this model. The analysis proceeded in two steps. The first was the
evaluation of the measurement models developed to operationalize the
concepts that comprise the proposed path model. Secondly, the structural
equation model was estimated.

W

The measurement model was analyzed using the confirmatory factor
analysis subroutine of the PACKAGE computer program (Hunter & Lim,
1987). The model was defined by an a priori analysis of the item content.
The fit of the specified measurement model to the data was then evaluated
by comparing the observed correlations between the variables with the
correlations predicted by the measurement model. The items should not
only "share a common meaning, the observed correlations must conform to
the product rules of internal and external consistency" (Hunter & Gerbing,
1982, p. 276). The criteria, therefore, for confirming the factor structure
of each scale were as follows:

1. The content of the items should be htuuegeueeus. This criterion
was assessed by checking the face validity of the items to assure that the
items fit the underlying construct being measured.

2. The scale should be internal): eensistent. The items should satisfy
the Spearman Product Rule. "That is, the correlation between two items in
the same cluster should be the product of their correlations with the

underlying uait" (Hunter & Gerbing, 1982).

54

3. It should satisfy the petflglism requirement. Parallelism refers
to the extent to which items in a factor are related in a similar way to other
factors.

An i

The analysis of the path model was performed using the LISREL
computer program. Because LISREL uses full information, maximum
likelihood techniques, the overall fit of the structural equation model can
be evaluated. The output used to assess the fit of the model is the Chi-
square estimation of the " goodness of fit" of the model. An insignificant
Chi-square indicates a good fitting model. Since Chi-square is sensitive to
sample size, the critical N statistic (Hoelter, 1983) was also considered,
along with the Adjusted and Unadjusted Goodness of Fit Indices. At a
more micro level of analysis, LISREL provides information that helps the
researcher assess whether the links specified in the model should be
retained or links not specified should be added. This is assessed by looking
at the T-values and the Modification Indices. The T-values evaluate the
significance of the path coefficients and the Modification Indices assess
unspecified parameters and indicates the degree to which Chi-square would
drop if a parameter was estimated. The amount of variance accounted for
by the model is determined by examining the R-squared values and the

Coefficient of Determination.

CHAPTER THREE

Results

This chapter includes the results of the confirmatory factor analysis
and the path analysis. Of the 272 respondents, 140 confirmed cases were
used for the confirmatory factor analysis. The factors addressed in this
dissertation were included in either the mail survey or the telephone
interview. These 140 cases represented individuals who returned the mail
survey and participated in the telephone interview and, therefore,
responded to all factors. For the path analysis, factors were computed
using only those cases in which individuals had responded to all items

within each factor. These 129 cases were used for the path analysis.

W

The confirmatory factor analysis was used to determine the
dimensionality of the measurement scales used to assess information quality
and quantity, participation in decision making, satisfaction, and
commitment. Two separate information scales were initially predicted--
tufeuuauouguauty and W. These, however, were found
to be strongly correlated (.94) and, therefore, a one factor solution was
evaluated. The combined factor had four items from the information
quality scale and one item from the information quantity scale. The two
items dropped emphasized the extremes of information sharing ("Far too
little information on important topics is shared among partnership

members" and "I receive too much information related to the

55

56

partnership"). This factor focused on the extent to which information that
is widely shared, relevant, timely, accurate, and leads to a clear
understanding of the IOR activities. The resulting scale was internally
consistent, parallel with outside factors and had a reliability of .89. The
items and factor loadings for this factor are presented in Table 1.

A 9 item scale was initially analyzed to assess W
deeisieutuakurg. Due to a lack of internal consistency, low factor
loadings, and fact that one item loaded more strongly on another factor,
five items were dropped. The resulting scale had a reliability of .89 and
focused on broad-based member participation. The items eliminated all
addressed high levels of involvement by judiyidualmps (e. g.,
community representatives, university representatives, project staff). The
remaining items emphasized in a more broad-based fashion the opportunity
to participate, access to project leadership, and the extent to which persons
affected are asked for their ideas. Scale items, factor loadings and
reliabilities are presented in Table 1.

Three items were dropped from the mum scale leaving a 3
item scale. Examination of the remaining items indicate broad contextual
similarity. They all imply that satisfaction with IOR involvement relies on
being meguizeu for one's individual professional identity as well as one's
work on the project and being rewarded (i.e., promotion, merit pay
increase) for one's work on the project. This factor was internally
consistent with a reliability of .57. Scale items and factor loadings
presented in Table 1.

Of the five items on the original cummiuueut scale, two were
dropped. One was dropped because it loaded more strongly on two other

factors, and one due to a lack of internal consistency. Commitment to the

57

project results from the similarity in the values espoused within the project
and those espoused within an individual's profession and organization; a
concern about the future of the project, and, finally, unlike the reward
concerns associated with satisfaction, a concern that project involvement
should not be detrimental to an individual's professional or career
development. The resulting three item Scale was internally consistent with a
reliability of .56. The items, factor loadings and reliabilities are presented
on Table 1.

58

Table 1
Scale Items, Factor Loadings, and Reliabilities

LII m 5mm
WW1!) (Alpha = -89)
1. The information I receive about the partnership .82

gives me a clear understanding of the partnership.
2. The information I received about the partnership

is accurate. .81
3. I receive information about the partnership

in a timely fashion. .74
4. The information I receive about the partnership

is relevant to my needs .79
5. Information on the partnership is widely shared

among the partnership participants. .77
WW (Alpha = -89)
Item Was
1. It is easy to get my ideas across to the project

leadership if I have a suggestion. .82
2. I feel I have many opportunities for participation

in the partnership. .81
3. When decision are being made in the partnership,

the persons affected are asked for their ideas. .82
4. Decision making in the partnership is broad-based .82

W (Alpha = .57)

1. To what extent do you feel that others on the Partnership
respect your profession? .59

59

Table 1 (cont'd.)
To what extent to do you think people in your profession
or the organization for which you work positively recognize
your work on the Community Partnership? .53

To what extent do you expect to receive career
benefits from working on the Partnership, such
as in terms of promotion or merit pay increases .5 3

Cammltment (Alpha = 56)

1.

To what extent would you say you really care
about the future of the Community Partnership
organization? .5 1

To what extent are the Partnership's values similar
to the values of your profession or the organization
for which you work? .58

To what extent will your work on the Partnership
be detrimental to your professional or career
development? .53

60

li'nP li

After the factor structure of the measurement model was confirmed,
a correlation matrix was run using the PEARSON CORR subroutine of
SPSS-PC. This correlation table was then corrected for attenuation due to
measurement error using the reliabilities from the confirmatory factor
analysis. These corrected correlations were the basis for the path analysis.
The initial correlations along with the means and standard deviations and
the corrected correlations used in assessing the fit of the model are I
included in Table 2.

EathAnalxsis

Analysis of the original model (Figure 2) using LISREL showed the
following. The goodness-of-fit index was .965 (Adjusted goodness-of-fit -
.883). Twenty percent of the standardized residuals were above 2.0. The
coefficient of determination indicates that 8% of the variance is explained
by the interaction of the variables. The R-squared values indicate that a
majority of the variance is explained by boundary spanner satisfaction (R-
squared = .563), with lesser amounts explained by participation in decision
making (R-squared = .078) and commitment (R-squared = .102). While
the goodness-of-fit index points to a model that is a good reflection of the
data, the chi-square analysis of the model (chi-square = 9.67; required chi-
square = 7.82/df = 3; p. <.05) suggests that the model does not fit
sufficiently well. Examination of the modification indices shows a link
between the information factor and the satisfaction factor that is
unaccounted for in the model (5.323). Conceptually, adding this link

61

makes sense as information is hypothesized to play a critical role in the
activities of boundary spanners. Since boundary spanner satisfaction has
been defined as a reflection of individual recognition and reward both
within the collaborative and an individual's organization, a direct link
emphasizes the centrality of information in assessing both the psychological
and physical rewards necessary for boundary spanner satisfaction.
As a result, the model was modified and reassessed.

In model 2 (figure 3), the chi-square analysis indicated a good fit of
the model (chi-square = 4.24; required chi-square = 5.99/df = 2, p < .05).
The goodness-of-fit index indicated an even better fit of the data to the
model (goodness-of-fit = .984; adjusted goodness-of-fit = .919). The
standardized residuals above 2 dropped to 10%. The total coefficient of
determination rose to 12%. The R-squared values indicate that even more
of the variance is explained by boundary spanner satisfaction (R-squared =
.581) with participation in decision making (R-squared = .078) and
commitment (R-squared = .102) remaining the same. The chi-square of
the difference between models 1 and 2 was computed (chi-square = 5.43;
required chi-square = 3.84/df = l, p < .05) The difference was significant
with one degree of freedom, therefore, adding the link makes sense
mathematically as well. See Table 3 for a comparison of the models.

Finally, the critical N (CN) for Model 2 was computed. Sample size
can play a decisive role in analysis of covariance structures. Testing models
using a small sample size increases the probability of failing to reject the
null and, therefore, accepting the model. Conversely, using a large sample
increases the probability of rejecting what might be a useful and
informative model. Hoelter (1983) suggests that "(w)hile the CN provides

a straightforward method for estimating the sample size for which a model

62

is statistically acceptable, there are no firm guidelines for assessing the
magnitude of CN in relation to deciding whether or not a model is
generally acceptable and reasonably reproduces the observed covariance"
(p 331). He does, however, provide a rule of thumb for using CN. "CN
values exceeding 200(G) indicate that a particular model adequately
reproduces an observed covariance structure" (p. 331). For Model 2 the
CN value would have to exceed 200. The CN for Model 2 is 206 and,

therefore, adds support to the contention that Model 2 is acceptable.

63

 

 

Table 2

rrl' Amn 'l fInfrminP 'ia'on
inDe'II .._.,I' III _II. I °uen '
Items 1 2 3 4
l -- .75 .25 .28
2 .85 -- .32 .34
3 .35 .45 -- .27
.4. .39 .48 .48 --
Means 12.78 9.83 6.85 4.35
SD. 4.49 4.05 1.78 1.55
Key:

1. Information Received

2. Participation in Decision Making

3. Satisfaction

4. Commitment

@ Correlations in the lower half of the matrix are uncorrected for

attenuation; corrected correlations appear in the upper half.

64

.28 .75 .32
Information—-> Participation in-—-—> Satisfaction—a Commitment
Decision Making
Figure 2

Hypothesized Model of Information, Participation in Decision
Making, Satisfaction, and Commitment in IORs with Path Coefficients

All coefficients significant, p < .05

65

(.28 ) (.71) (.32)
Information—I Participation in——* Satisfaction—r Commitment

Vision Making

(.14)

Figure 3

Final Model of Information, Participation in Decision making,
Satisfaction, and Commitment in IORs

All coefficients significant, p < .05

66

Table 3

Comparison of Models

 

 

 

Chi-square df Coefficient R-squared
of Determination
Model #1 9.67 3 .078
PDM .078
SAT .563
COM .102
Model #2 4.24 2 .1 17
PDM .078
SAT .581
COM .102
T-Values
INFOR PDM SAT COM
Model #1
INFO
PDM 3.274
SAT 12.728
COM 3.791
INFOR PDM SAT COM
Model #2
INFO
PDM 3.274
SAT 2.348 11.826

COM 3.791

CHAPTER 4

Discussion

This chapter includes a discussion of the findings from the
measurement model and the path model. Given the extension of the
constructs (i.e., information, participation in decision making, satisfaction,
and commitment) from the individual organizational context to the
interorganizational context, each scale will be discussed individually to
assess the degree to which the items reflect or differ from the review of
these areas in the literature. These will be discussed in the measurement
section. In the path analysis section the total interaction of the final path
model will be discussed. Finally, potential research questions and practical
issues raised by this dissertation will be addressed.

MeasurementMedel

In order to facilitate the discussion of each of the factors in the
measurement model, the items that define each factor are presented prior

to the discussion.

mm

1. The information I receive about the partnership
gives me a clear understanding of the partnership.

2. The information I received about the partnership
is accurate.

3. I receive information about the partnership
in a timely fashion.

4. The information I receive about the partnership
is relevant to my needs

5. Information on the partnership is widely shared among the
partnership participants.

67

68

In the theoretical discussion of interorganizational exchange,
information acquisition and exchange is conceived of as establishing a
context for a negotiated order in which future decision making can take
place. Three uses of information were discussed in Chapter 1. Boundary
spanners must have access to information that is useful for defining the
issues of collaborative interaction; protecting their individual organizations
or stakeholder groups; and advancing, maintaining, or protecting their
individual roles. The items within this factor reflect these three categories
of information acquisition.

Three of the five items in the information scale focus primarily on
the quality of information acquired. Information received was assessed in
terms of being accurate, timely, and providing a clear understanding of the
project. This adds support to the contention of O'Reilly, Chatman, and
Anderson (1987) that in order to define a problem or develop a list of
possible solutions, decision makers must have information of a quality that
allows them to consider alternatives, and determine the most desirable
outcomes.

In addition to items tapping information acquisition, one of the five
items in the final scale emphasizes that information must be relevant to the
needs of the boundary spanner. A final item in the information factor
focuses on the idea that information should be widely shared among
participants.

To summarize, high quality information is defined in this factor as
that which gives a clear understanding of the Partnership. Such
information is widely shared, accurate information that is received in a
timely fashion. Finally, high quality information is information that is

relevant to individual boundary spanner needs.

69
Partici ation in Decision Makin cale:

1. It is easy to get my ideas across to the project
leadership if I have a suggestion.

2. I feel I have many opportunities for participation
in the partnership.

3. When decision are being made in the partnership,
the persons affected are asked for their ideas

4. Decision making in the partnership is broad-based

Participation in decision making has been cited as essential in the
development and sustainability of collaborative relationships (Gray, 1985;
Ring & Van De Ven, 1994; Wood & Gray, 1991). In the measurement
model confirmed in this research, both opportunity and influence emerged
as defining characteristics of the participation construct.

Within the participation in decision making factor, three items that
specified a high level of participation by individual groups (i.e.,
community representatives, university representatives, or project staff)
were nOt confirmed as part of the final scale. A fourth item that said that
decisions were made by a small group of leaders also failed tests of
unidimensionality. What remained after confirmatory factor analyses were
two items that emphasized the ideas that representatives should have many
opportunities for participation in decision making, and that that decision
making should be broad-based. Additionally, items reflecting ease of
communicating ideas to project leadership, and the expectation that those
affected should be asked for their ideas where also confirmed as a part of
the final scale. Thus, for these interorganizational participants the notion
of broadbased, and targeted decision making was associated with

effectiveness in participation.

70

Satisfaetion Scale:

1. To what extent do you feel that others on the Partnership
respect your profession?
2. To what extent to do you think people in your profession
or the organization for which you work positively recognize
your work on the Community Partnership?
3. To what extent do you expect to receive career benefits from
working on the Partnership, such as in terms of promotion or merit pay inc;

Building on the reward and recognition aspects of previous
satisfaction research, this construct was defined as follows in Chapter One.
Boundary spanner satisfaction should reflect both the more specific
experience with the w_ork of the IOR and WW
linking work within the IOR to their organizational positions. However, in
empirically validating a measurement scale based on these ideas, two items
dropped out of the satisfaction factor. The first is "To what extent do you
find working in your profession to be very satisfying?" The second is "To
what extent do you find your work on the Community partnership to be
very satisfying?" The elimination of these two items suggests that general
satisfaction with the m of either one's profession or that of a
collaborative relationship were not related to other characteristics of
satisfaction considered in this research. Instead, for these boundary
spanners the items that proved unidimensional centered around respect for
one's ptofessien by others within the Partnership. This suggests that for
professionals within a collaborative, the primary focus of identification --
and perhaps a primary source of satisfaction -- may be with their
profession and not their organization.

A second characteristic of satisfaction confirmed in this study is
the extent to which people in their profession or organization positively

recognize a boundary spanner's work within an IOR. This may indicate a

71

need for external validation of boundary spanner activities in
collaboratives. Because both "profession" and "organization" were used in
the item, it is not possible to parse out the influence of each.

The final aspect of satisfaction confirmed in this scale is the
expectation that boundary spanners will receive career benefits based on
their work in an IOR, such as promotion or merit pay increases. In
summary, it appears based on the items confirmed in this scale that
boundary spanner satisfaction is less focused on the work associated with
the endeavor itself and more specifically focused on the individual

recognition and reward for that work.

1. To what extent would you say you really care
about the future of the Community Partnership
organization?

2. To what extent are the Partnership's values similar
to the values of your profession or the organization
for which you work?

3. To what extent will your work on the Partnership
be detrimental to your professional or career
development?

Commitment was hypothesized in Chapter One to reflect a balance
between the boundary spanner's loyalty to his/her organization and/or
profession and the collaborative. The extent of this loyalty or commitment
is determined by boundary spanners' expectations of the effect of
collaborative activities on their organization or profession and the effect on
them individually.

In the commitment factor, an item assessing the extent of boundary
spanner concern about the future of the Community Partnership

organization was confirmed. An item that dropped out, however, dealt

72

with the extent to which respondents were proud to tell others that they
were part of the Partnership organization.

A second item that supports the hypothesized definition of
commitment centers around the similarity of the underlying values of the
IOR to those of boundary spanners' professions or organizations. At the
core of Mowday, Steers, and Porter's (1979) definition of commitment is
the idea that an individual must have "a strong belief in the organization's
goals and values" (p 226). It was hypothesized in this dissertation that due
to the fact that boundary spanner roles may distance individuals both
psychologically and physically for their organizations (Adams, 1980), that
boundary spanner commitment to a collaborative hinges on the extent to
which the values of the collaborative reflect those of the organization
and/or profession. This item, however, also refers to both profession and
organization so it is not possible to sort out which is the primary referent
or whether the values of the academic organizations and the various
professions overlap.

The third item making up this factor focuses on the extent to which
individuals' work on the Partnership will be detrimental to their
professional or career development. It suggests that to the extent that
collaborative activities do net hinder professional or career growth,

boundary spanners will continue to be committed to the work of an IOR.

Path M l
The revised model supports the hypothesis that there is a significant
relationship between information exchange and participative decision
making (.28) and expands on this hypothesis to add a direct link to

satisfaction (.14). Given the discussion of the information scale, these

73

linkages make sense conceptually. In order for boundary spanners to feel
that they have both the opportunity to participate in collaborative decision
making and influence over decisions, they must first perceive that
information is widely shared, leads to increased understanding of the
collaborative, and is relevant.

The strength of the path between participation in decision making
and satisfaction (.75) might seem to reconfirm the relationship between
these variables that has been consistently found in participation research
within the intraorganizational arena (Miller & Monge, 1986). As noted in
Chapter 1, in an ideal collaborative situation each participant would have
an equal opportunity to participate and an equal chance of influencing
decisions. One of the items confirmed in the participation in decision
making scale in this study is the extent to which decision making is broadly
shared. This may suggest that for satisfaction to occur, the impact of status
distinction or influence differences based on precedence, tradition,
expertise, resources, and/or law must be dealt with.

As the discussion of the PDM scale suggests, these individuals may
have entered the Community Partnership Initiative expecting to have
extensive influence on the decisions of the Partnership. When the
organizational and/or professional positions of the respondents are
considered, this makes sense. Boundary spanners, no matter what their
level in an organization, are by definition decision makers for their
organization. Boundary spanner activities, as discussed in Chapter 1, fall
along a continuum based on the ambiguity of the activities, roles, and
norms of interaction. At the more formalized end of the spectrum are
boundary spanners whose activities and decision making capacity are

specifically defined (e.g., bank tellers, receptionists). As boundary

74

spanning activities become less defined, either by a lack of definition of
roles, outcomes, or norms (e.g., collaborative development) or by design
(e.g., research), then the expectations of boundary spanners and the
potential for organizational or individual professional loss increase.

As a consequence, it is feasible to suggest that, when creating
interorganizational collaborations around issues about which there are not
established goals, or norms of interaction and for which the organizational
risk potential is relatively high, those chosen to represent an organization
or profession would be individuals who have substantial decision making
capacity within the organization.

The final relationship in the model between satisfaction and
commitment (.32) is the second strongest link. It indicates that reward and
recognition for boundary spanner activities leads to increaSed commitment.
Commitment, however, hinges on not only the similarity of values between
the collaborative and the organization and/or profession but also on the
negative impact of collaborative involvement on individual career or

professional development.

Future Research

Due to the limitations of this study more questions were raised than
answered. For example, the examination of the measurement model
suggests that further psychometric evaluation is warranted to address the
complexity of the information, participation in decision making,
satisfaction and commitment factors when used in the collaborative arena.

Additionally, the limited representation of community members in
the initial respondent pool and their almost non-existent representation

(i.e., 116 professionals vs. 12 community members) in the group used for

75

the final analysis allows only half of the picture of the collaborative process
within the Community Partnership to be presented. The data provide a
picture of the collaborative process among the professionals who were
involved in this relationship, but not the community members. The criteria
for satisfaction and commitment, the expectations for participation in the
decision making process, and the role of information may differ radically
for representatives of organizations or stakeholder groups that fall outside
of institutionalized organizations.

The limitations of the sample, however, also provide an opportunity
to explore the ways in which the professional boundary spanners studied
constitute themselves as distinct from the organizations they represent. In
order to more fully understand the "separate identity" of professionals, two
related themes that were touched on in setting up the model for this
dissertation, but were not explicitly incorporated into the model or the data
collection need to be examined. Their omission limits the findings of this
research. These are the role definition of boundary spanners and the

organizational forms in which these boundary spanners work.

The Rule Defiuiu'on of Beundaty Spanners
The first of these deals with the primary role definition of the

respondents. The assumption was made that the predominant linkages of
interest were among etga_nizatienal representatives. Because multiple
organizations and communities were involved, it was assumed that the goals
and outcomes most beneficial to these groups would be the key issues under
discussion and negotiation and/or of concern to boundary spanning
representatives. The findings of the current study suggest that the

dominant source of identification for these boundary spanners is their

76

W (i.e., physician, nurse, public health professional).
It appears, then, that a distinction should be made between those
collaboratives in which the representatives solely represent the organization
and those in which the representatives identify with both their professions
and their organizations. Abbott (1988) writes that collaboration among
professionals threatens their jurisdiction over knowledge and the use of
knowledge. In studying the similarities and differences among definitions
of "profession", Abbott notes that the definitions are similar in the
following ways: "Certainly all agreed that a profession was an
occupational group with some special skill, Usually this was an abstract
skill, one that required extensive training ...... In addition, professions were
more or less exclusive" (p. 7). Projects such as the Community
Partnership Initiative may threaten that exclusivity to the extent that
individual representative satisfaction and commitment rely on the impact of
iudividual iuvelvemeut on recognition, reward and the extent to which that
involvement will be detrimental to professional or career development. If
professionals are involved in a collaborative that involves the potential
redefinition of their professions, the role of information exchange as a
mechanism for clarification of roles, expectations, and outcomes may
become even more crucial.

The area of role clarification is one that warrants future research.
For example, through their identification with their organizations,
professions or stakeholder groups, boundary spanners may have certain
expectations about the scope and content of their responsibilities. These
expectations are associated with the roles they fill within these groups.
Roles have been defined by Katz and Kahn (1978) as "standardized patterns
of behavior required of all persons playing a part in a given functional

77

relationship" (p. 43). No matter what the source, these role definitions
form the basis for initial interaction among boundary spanners. If, due to
a lack of consensus about the domain, its definition, and individual rights
and responsibilities, role definitions are challenged or unclear, then
attempts to redefine roles can lead to increased uncertainty.

Role uncertainty has been shown to fall into two categories: role
conflict and role ambiguity. Adams (1980) writes that one negative
outcome of boundary spanner role conflict is an increase in distrust among
representatives which tends to inhibit cooperative interaction, interfere
with conflict negotiations and resolution among boundary spanners and, as
a consequence, increase the negative response to others involved in these
activities. Van Sell, Brief, and Schuler (1981) suggest that engaging in
boundary spanning activities sensitizes individuals to the conflicting
messages they receive and it is as a result of this conflict that they
experience increased uncertainty. At the representative level, then,
information of importance may center around role explication and stress
reduction because role conflict and stress have been cited as causes of
boundary spanner turnover (Adams, 1980; Eisenberg et al, 1985; Ring &
Van De Ven, 1994). Information acquired about the collaborative, its
process or outcomes, may, for example, confirm for the boundary spanner
that he or she wants to build or maintain a high profile association with the
collaborative or, conversely, distance him or herself from the
collaborative.

Researchers addressing the place of shared information in the
emergence of participative structure may also want to consider the role of
shared information in the distribution of power and the extent to which

information exchange builds and fosters trust among participants. For

78

instance, when organizations come together around issues that are
ambiguous and may, as a consequence, lead to goal conflict, a legitimate
concern is the extent to which a single group or a coalition may seek to
control the overall outcomes. One mechanism for control or power within
an organization that is widely recognized is the control of critical
information. Power, as defined by Kanter (1977), is "the ability to get
things done, to mobilize resources, to get and use whatever it is that a
person needs for the goals he or she is attempting to meet" (p. 275). If
information is shared among all members of the collaborative, then it can
be supposed that power may also be shared.

Addressing the issue of trust, Ring and Van de Ven (1994) suggest
that through the numerous interactions of boundary spanners "trust in the
good will of others" (p. 100) emerges. Trust is defined by McAlister
(1995) as "the extent to which a person is confident in, and willing to act
on the basis of, the words, actions, and decisions of another" (p. 25). Two
categories of trust have emerged from the literature -- cognitive and
affective. Cognitive trust is based on evidence of trustworthiness, while
affective trust is identified with the emotional bonds that are created
between individuals based on interaction. McAlister indicates that
cognitive trust must exist prior to the development of affective trust. The
broad exchange of information that meets the quality standards indicated in

this factor may be antecedents to the development of cognitive trust.

Institutional anizational Form
The second general theme that emerges from this research involves
the organizational models or forms (e.g., university professional schools)

within which professionals work. Organizations may be said to fall along a

79

continuum. At one extreme are the technical or production organizations
that are evaluated on output. At the other end are what are referred to as
institutionalized organizations "whose success depends on the confidence
and stability achieved by isomorphism with institutional rules" (Meyer &
Rowan, 1992, p. 36). Some organizations operate within and/or as a result
of what Scott (1992) refers to as institutional environments. "Institutional
environments are broadly defined as including the rules, and belief systems
as well as the relations networks that arise in the broader societal context"
(p. 14). Meyer and Rowan (1992) note that "(n)ew and extant domains of
activities are codified in institutionalized programs, professions or
techniques and organizations incorporate the packaged codes" (p. 26). One
outcome of organizational institutionalization, according to Meyer and:
Rowan (1992), is the fact that organizational activities and structure may be
predominantly defined by the rules, regulations and norms associated with
the key professions that make up the organization. Examples of domains
around which institutionalized organizations develop are government,
professions, union and trade associations (Scott, 1992). Scott and Backman
(1990) write that researchers have noted that professionals who work in

health care organizations:

did not behave as conventional employees. They exhibited
considerable independence. They were more likely to be
guided by the norms and standards of collegial groups
beyond organizational boundaries; they were more oriented
to the development and exercise of their distinctive skills
than to advancement in the organization hierarchy; they were
likely to consider career moves between organizations

rather than developing commitment to a specific
organization. (pp. 23 - 24)

The above goes a long way in explaining the findings of both the

measurement and path model. For example, the satisfaction and

80

commitment of these respondents appears to be oriented to a set of external
professional norms. These individual and organizational distinctions,
however, raise questions about and create unique challenges to the
development of IORs among institutionalized organizations. For example,
when applying the concept of institutionalization to organizations, Meyer
and Rowan (1992) note that "organizations tend to disappear as distinct,
and bounded units" (p. 28). In essence then, institutionalized organizations
are mechanisms for sustaining definitions of specific social realities, such as
those that determine professional boundaries (Scott, 1992). The process of
developing a collaborative such as the Community Partnership calls,
therefore, for the revision of the social reality that defines and legitimizes
the professional groups involved and their interaction with each other. The
recreation of a social reality suggests a more complex process than the
exchange of resources. At the organizational level this one central issue
suggests that those doing research in this area need to carefully rethink the
underlying assumptions about IORs at least to the extent that they are
associated with institutionalized organizations. In addition to continuing to
probe the process variables, researchers need to examine and possibly
redefine the antecedents and consequences of IORs for these types of
organizations.

If sustainability is a legitimate outcome for IORs, research
addressing the emergent structure of IORs would also be valuable. An
underlying argument of this research is that participation in a collaborative
relationship leads to the emergence of participative structures. One of the
features of the model tested in this dissertation is its emphasis on the role of
communication in the emergence of the participative structure. That is, it

was not assumed that involvement in a structure defined as participative is

81

isomorphic with the process of actual participation. As suggested by
Marshall and Stohl (1993b) the creation of a structure (e.g., collaborative)
does not necessarily reflect the process or even the existence of individual
participation. This argument needs to be tested.

In practical terms, this concept of institutionalized organizations and
the dominating role of professions means a shift in perspective is necessary
to fully deal with the process of initiating, developing and/or sustaining
collaborative activities and structures. If the professional "structures" are
the "organizations" that are forming the collaborative, then issues such as
collaborative leadership, transactional costs, new relationships,
collaborative priorities, and avenues for collaborative action need to be
addressed from this perspective.

In summary, the findings of this dissertation add support to both the
central and critical role of communication for professionals within the
collaborative process. The measurement model suggests that the
collaborative process itself may be more complex than initially thought
with distinctions among organizational types creating different needs within
the collaborative. Continued research in this area would benefit those who
have responsibility for coordinating collaboratives that call for
representation from multiple types of organizations and those who evaluate

the success of those collaboratives.

82

BIBLIOGRAPHY

83

BIBLIOGRAPHY

Abbott, A. (1988). The system of professions: An essay en the division of
expert labor. Chicago: The University of Chicago Press.

Adams, J. S. (1980). Interorganizational processes and organization

boundary activities. Research in Organizational Behavior, 2, 321-
355. _

Argyris, C. (1955). Organizational leadership and participative
management. Jemal ef Business, 28, 1-7.

Aldrich, H. E. (1976) Resource dependence and interorganizational
relations. Administratieu yd Seeiety, 2, 419-454.

Aldrich, H. E. (1979). Organizations and environments, Englewood
Cliffs: Prentice Hall.

Bertalanffy, L. von. (1968). Gepetal systems Qeog. New York: George
Brazillier.

Boje, D. M., & Whetter, D. A. (1981). Effects of organizational strategies
and contextual constraints on centrality and attributions of influence

in interorganizational networks. Administrative Science Quarterly,
26, pp. 378-395.

Bresser, R. K. (1988). Matching collective and competitive strategies.
Strategie Mgagement Jeumal, 2, 375-385.

Buckley, W. (1967). i 10 d Modern stems The
Englewood Cliffs: Prentice Hall.

Child, J. (1972). Organizational structure, environment and performance:
The role of strategic choice. Sociolegy, 6, 1-22.

Cohen, M. D., March, J. G., & Olson, J. P. (1972). A garbage can model

of organizational choice. Adminisuative Science ngterly, 11, l-
25.

84

Corbin, J. M. & Strauss, A. L. (1993). The articulation of work through
interaction. The Sociological Quarterly, fi, 71-83.

Day, R., & Day, J. V. (1977). A review of the current state of negotiated
order theory: An appreciation and a critique. Sociological

Quarterly, 1S, 126-142.

DeCotiis, T. A., & Summer, T. P. (1987). A path analysis model of the
antecedents and consequences of organizational commitment. Human
Relatiens, dd, 445-470.

Dimaggio, P. J ., & Powell, W. W. (1983). The iron cage revisited:
Institutional isomorphism and collective rationality in organizational

fields. The Socielogical Review, 48, 147-60.

Eisenberg, E. M., Farace, R. V., Monge, P. R., Bettinghaus, E. P.,
Kurchner-Hawkins, R., Miller, K. I., and Rothman, L. (1985).
Communication linkages in interorganizational systems. In B.

Dervin & M. Voigt (Eds), Progress in eommunieatien seiences
(Vol. 6, pp. 231-261. Norwood: Ablex.

Emery, F. E., & Trist, E. L. (1965). The causal texture of organizational
environments. Human Relations, IS, 1, 21-32.

Fottler, M. D., Schermerhom, J. R., Wong, J ., & Money, W. H. (1982).
Multi-institutional arrangements in health care: review, analysis and

a proposal for future research. Academy of Management Review, 1,
67-79.

French, J ., & Raven, B. H. (1959). The bases of social power. In D.
Cartwright (Ed.), Studies ef social power. Ann Arbor, MI: Institute
for Social Research.

Glisson, C., & Durick, M., (1988). Predictors of job satisfaction and
organizational commitment in human service organizations.

Admnistrative Quarterly, 33, 61-81.

Gottfredson, L. S., & White, P. E. (1981). Interorganizational
agreements. In P. C. Nystrom and W. H. Starbuck (Eds),

Handboek of orggizational design (Vol 1, pp. 471-486). New

York: Oxford University Press.

85

Gray, B. (1985). Conditions facilitating interorganizational collaboration.
Human Relations, SS, 10, 911-936.

Hall, R. H. (1992). Qrganizations: Structures, Processes and Qutcomes.
Englewood Cliffs, NJ: Prentice Hall.

Hall, R. H., Clark, J. P., Giordano, P., Johnson, P., & Van Roekel, M.
(1977). Patterns of interorganizational relationships.

Administrative Science Quarterly, 22, 3, 457-474.

Hannan, M. T. & Freeman, J. (1989). Organizational ecology.
Cambridge, MA: Harvard University Press.

Heimer, C. A., (1985). Allocating information costs in a negotiated
information order: Interorganizational constraints on decision
making in norwegian oil insurance. Administrau've Science
QuaneLly, 3_Q, 395-417.

Hoelter, J. W. (1983). The analysis of covariance structures. Smielegical
Memeds and Researeh. L1, 3, 325-344.

Hoffman, L. R., Burke, R. J ., & Maier, R. F. (1965). Participation,
influence, and satisfaction among members of problem-solving

groups. Esyeholegieal Reperts, L6, 661-667.

Hunter, J. E., & Gerbing,, D. W. (1982). Unidimensional measurement
second order factor analysis, and causal models. Research t'u

Qrganizatioual Behavier, 4, 374-392.

Hunter, J. E., & Lim, T. S. (1987) LIMSTAT. Unpublished manuscript.
Department of Communication, Michigan State University, East
Lansing, MI.

Jackson, S. E. (1983). participation in decision making as a strategy for

reducing job-related strain. oumal of A lied Ps ch , fl,
3-19.

Janis, I. L., & Mann, L (1977). Decisi maki : A s chol 'cal an
of cenflict, ehoice and commiment. New York: Free Press,

Kanter, R. M.& Stein, B. A., & Jick, T. D. (1992). The challeuge of
erganizatienal change. New York: The Free Press.

86

Katz, D., & Kahn, R. L. (1978). The social psychology of organizations.
New York: Wiley.

Keller, R. T., & Holland, W. E. (1975). Boundary-spanning roles in a
research and development organization: An empirical investigation.

Academy of Management Review, LS, (2), 388-393.

Kilman, R. H., & Mitroff, I. I. (1979). Problem defining and the

consulting/intervention process. Qalifomia Management Review, 2_l_,
(3), PP. 26-33.

Laszlo, E. (1972). Introduction to systems philosophy. New York:
Gordon & Breach Science Publishers

Levine, S. & White, P. E. (1961) Exchange as a conceptual framework for
the study of interorganizational relationships. Administrative

Seience Quarterly, S, 1, 583-610.

Locke, E. A. (1976). The nature and causes of job satisfaction. In M. D.

Dunnette (Ed.), Handbook of Industrial and Qrganizatienal
Psychelogy. (pp. 1297-1349). New York: McGraw-Hill.

Locke, E. A., & Schweiger, D. M. (1979). Participation in decision
making: One more look. In B. M. Straw (Ed.), Research in

organizational behavier (Vol 1, pp. 265-339). Greenwich:
JAI.

Logsdon, J. M. (1991). Interests and interdependence in the formation
of social problem-solving collaborations. Journal ef Applied

Behavioral Science, 27, 1, 23-37.
March, J. G., & Simon, H. A. (1958). Organizations, New York: Wiley.

Marsh, R. M., & Mannari, H. (1977). Organizational commitment: A
predictive study. Administrative Science Quarterly, 22, 57-75.

Marshall, A., & Stohl, C. (1993a). Being "in the know" in a participative

management system. Mana ement ommunication uart r1 , 6,
372-404.

Marshall, A., & Stohl, C. (1993b). Participating as participation: A
network approach. Qemmunication Monogtaphs, 69, 137-157.

87

McAlister, D. J. (1995). Affect- and cognition-based trust as foundations
for interpersonal cooperation in organizations. Academy of

Management Journal, 38, 24-59.

McCann, J. E., (1983). Design guidelines for social problem-solving

interventions. Journal of Applied Behavioral Science, 29, 177-
189.

Meyer, J. W. (1978). Strategies for further research: Varieties of
Environmental Variation. In M. W. Meyer (Ed.), Environments and

Qrganizations. (pp. 353-368). San Francisco: Jossey-Bass.

Meyer, J. W., & Rowan, B. (1992). Institutionalized organizations:
Formal structure as myth and ceremony. In J. W. Meyer & W. R.

Scott (Eds), thanizau'onal Envirenments: Ritual and Ratienality.
Newbury Park: Sage.

Meyer, J. W., & Scott, W. R. (1992). Qrganizational Environments:
Ritual and Rationality. Newbury Park: Sage.

Miller, K. I. (1995). anizatio om 'c tion: A roac e
fiocesses. Belmont: Wadsworth Publishing.

Miller, K., Scott, C., Stage, C., & Birkholt, M. (1995). Communication
and coordination in an interorganizational system: Service provision
for the urban homeless. Unpublished manuscript.

Department of Communication, Arizona State University, Tempe,
Arizona.

Miller, K. 1., & Monge, P. R. (1986). Participation, productivity and

satisfaction: A meta-analytic review. Academy ef Management
Jeurnal, 22, 727-753.

Miller, K. I., & Monge, P. R. (1987). The development and test of a
system of organizational participation and allocation. In M. L.

McLaughlin (Ed.), Qommunication Yearbook, (pp. 431-455),
Newbury Park: Sage.

Mink, O. G., Schultz, J. M, & Mink, B. P. (1991). Developing md
Man anaging Qpen thmizations. Austin: Learning Concepts.

88

Monge, P. R., & Eisenberg, E. M. (1987). Emergent communication
networks. In E. F. Jablin, L. L. Putnam, K. H. Roberts, & L. W.

Porter (Eds.), Handbook of Qrganizational Qommunication, (pp.
304- 342), Newbury Park, CA: Sage.

Monge, P. R., & Miller, K. I. (1988). Participative processes in
organizations. In. G. G. Goldhaber & G. A. Barnett (Eds.),

Handbook of Qrganizational Communication, (pp. 213-229),
Newbury Park, CA: Sage.

Mowday, R., Porter, L., & Steers, R. (1982). Qrganizational linkages:

The psycholegy pf comitment, absenteeism, and turnover. Academic
Press: New York: NY.

Mulford, C. L. (1984). Jnterotganizational relations: Implications for
community develepment. Human Sciences Press, New York: NY.

Nathan, M. L., & Mitroff, I. I. (1991) The use of negotiated order theory
as a tool for the analysis and development of an interorganizational

field. Journal of Applied Behavieral Science. _2_7, (2), 163-180.

Neal, M. A., & Northcraft, G. B. (1991). Behavioral negotiation theory:
A framework for conceptualizing dyadic bargaining. In L. L.

Cummings & B. M. Straw (Eds.), Research in organizatienaJ
behavier, vol 13: 147-190. Greenwich, CT: JAI Press.

Nutt, P. C. (1984). Types of organizational decison processes.

Administrative Seienee Quarterly, 22, 414-450.

Oliver, C. (1990). Determinants of interorganizational relationships:

integration and future directions. Academy pf Management Review.
25, (2). pp. 241-265.

Oliver, C. (1991). Net work relations and loss of organizational autonomy.
Human Relations, fl, 943-961.

O'Reilly, C. A., Chatrnan, J. A., & Anderson, J. C. (1987). Message flow
and decision making. In F. M. Jablin, L. L. Putnam, K. H. Roberts,

& L. W. Porter (Eds.), Handbook ef Qrganizational Cemmum'cation:

An Interdisciplinary Perspective. 600-623. Newbury Park:
Sage Publications.

89

Pasquero, J. (1991) Supraorganizatinal collaboration: The Canadian

environmental experiment. Journal of Applied Behavioral Science,
2.7.. (1). pP- 38-64-

Pfeffer, J ., & Nowak, P. (1976). Joint ventures and interorganizational

interdependence. Administrative Science Quarterly, 21, pp. 398-418.

Pfeffer, J ., & Salancik, G. R. (1978). The External Qontrol of
Qrganizatiens. New York: Harper & Row.

Putnam, L. L.(1983). The interpretive perspective: An alternative to
functionalism. In L. L. Putnam, & M. E. Pacanowsky (Eds.),

Communieation md organizatiens; m intemretive approach. (pp.
31-54). Beverly Hills: Sage.

Richards, R. W., & Bouhuijs, P. A. J. (1991) Leadership for change:
Closing the gap between health professions and the community.
Kellogg Foundation, Unpublished manuscript.

Ring & Van De Ven (1994). Developmental processes of cooperative

interorganizational relationships. Academy of Management Review.
_2. (1) pp. 90-118.

Schuler, R. (1980). A role and expectancy perception model of

participation in decision making. Academy of Management JeumaJ,
23, (2), 331-340.

Scott, W. R. (1992) r aniza' S' Rati nal atur an
Englewood Cliffs: Prentice Hall.

Scott, W. R., & Backman, E. V. (1990). Institutional theory and the
medical care sector, in S. S. Mick (Ed.), Jnnevauens in Health care

delivem: Insights fer erganizatienal thegtx (PP. 20-52). San
Francisco: J ossey-Bass.

Scott, W. R., & Meyer, J. W. (1983). The organizational of societal
sectors. In J. W. Meyer & W. R. Scott (Eds.), thanizational
envi entS' ritual an ati n ' . vol, pp Beverly Hills: Sage.

Smith, K. G., Carroll, 8. J ., & Ashford, S. J (1995). Intra- and
interorganizational copperation: Toward a research age. Academy

ef Management Joumal. 3, 7-23.

90

Stamaman, S. & Miller, K. I. (1992). A test of a causal model of
communication and burnout in the teaching profession.

Communication Educatien, At, 40-53.

Strauss, A. (1988). The articulation of work: an organizational process.

The Sociological Quarterly, 3, 163-178.

Strauss, A., Schatzman, L., Bucher, R., Ehrlich, D., & Sabshin, M. (1963).
The hospital and its negotiated order. In E. Friedson (ed.) The

Hospital and Modern Society (PP. 149-169). New York: Free

Press.

Taylor, J .C., & Bowers, D. G. (1972). Survey pf organizations: a

machine-seored standardized guestiennaire instmment. Ann Arbor:
Institute for Social Research.

Thompson, J. D., (1967). Qrganizations in Action. New York:
McGraw-Hill.

Trist, E. L. (1983). Referent Organizations and the Development of Inter-
Organizational Domains. Human Relations, &, 3, 269-284.

Van de Ven, A. H. (1976) On the nature, formation and maintenance of

relations among organizations. Academ of Mana e e t eview, 1_,
24-36.

Van de Ven, A. H., & Walker, G. (1984). The dynamics of

interorganizational coordination. Administrative Scienee Quagerly,
22, 598-621.

Van Sell, M., Brief, A. P., & Schuler, R. S. (1981). Role conflict and role
ambiguity: Integration of the literature and directions for future

research. Human Relan'ens, SA, (1), 43-71.

Waddock, S. (1988). Building successful social partnerships. Slean
Management Review, 22, 17-23.

Wandersman, A., & Goodman, R. M. (1991). Community partnerships for
alcohol and other drug abuse prevention. Family Reseurce

Cealition, R2, 8-9.

Weick, K. (1979). The Socim Psycholegy ef Qrganizing, 2nd ed. Reading:
Addison-Wesley.

91

Whetten, D. A. (1981). Interorganizational relations: A review of the
field. Journal of Higher Education, 2, 1-28.

Williams, L. J ., & Hazer, J. T. (1986). Antecedents and consequences of
satisfaction and commitment in turnover models: A reanalysis using
latent variable structural equation methods. Journal of Applied

Psychology, fl, (2), pp. 219-231.

Wood, D. J. & Gray, B. (1991). Toward a comprehensive theory of

collaboration. Journal of Applied Behavioral Science, 2_7, 2, pp. 139-
162.

HICHIGAN sran UNIV. LIBRARIES
ll11111111“11111111111111111111111
31293013892801