I)ate

l\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

3 1293 0140

This is to certify that the

thesis entitled

Formalization as a Mediator of
Hierarchical Level,
Role Conflict, and Role Ambiguity

presented by
Betty Helen La France

has been accepted towards fulﬁllment
of the requirements for

Master of Arts degree in Comunication

//£//+

Major ﬁffssor

 

 

“4% ﬂ

 

0-7639

MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

 

 

LIBRARY

Michigan State
University

 

 

 

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

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU le An Nﬂnnetlve Mattel Opportunity lnetltmlon
7 , WM!

FORMALIZATION AS A MEDIATOR OF HIERARCHICAL LEVEL,
ROLE CONFLICT, AND ROLE AMBIGUITY

By
Betty Helen La France

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF ARTS
Department of Communication

1 996

ABSTRACT

FORMALIZATION AS A MEDIATOR OF HIERARCHICAL LEVEL,
ROLE CONFLICT, AND ROLE AMBIGUITY

by
Betty H. La France

A model that posits a mediating effect of formalization on hierarchical level and
role conﬂict and role ambiguity is tested. In addition, adaptations of the Rizzo, House,
and Lirtzman (1970) role ambiguity and role conﬂict scales are provided and statistically
validated using the a Mi, theoretically grounded assumptions of conﬁrmatory factor
analysis. Data were gathered ﬁom self-report questionnaires and completed by
organizational members (3 = 81) within the National Cancer Institute’s Cancer
Information Service, a geographically-dispersed federal government health information
program. The model provided a poor overall ﬁt to the data ()6 = 4.34, with d_f = 3),
p>.05 with two nonsigniﬁcant paths. These results are discussed with speciﬁc concern for
the validation of the role conﬂict and ambiguity measures, as well as the implication that

the model holds for future research.

ACKNOWLEDGMENTS

I would like to thank the Ofﬁce of Cancer Communications as well as the following CIS
oﬁices which participated in the August 1995 data collection:

Region 1: (CT, ME, MA, NH, RI, VT)

Region 2: (NYC, Long Island,
Westchester County, NY)

Region 3: (NY State, WestPA)

Region 4: (DE, NI, EastPA)

Region 5: (DC, MD, NorthVA)

Region 6: (GA, NC, SC)

Region 7: (FL, PR)

Region 8: (AL, LA, MS)

Region 9: (AR, KY, TN)

Region 10: (OH, WV, SouthVA)

Region 11: (W1, IA, MN, ND, SD)

Region 12: (MI, IN)

Region 13: (IL, KS, MO, NE)
Region 14: (OK, TX)
Region 15: (AK, NorthID, MT, OR WA)

Region 16: (AZ, CO, SouthID, NM, VT, WY)
Region 17: (NorthCA, NV)

Region 18: (SouthCA)

iii

CIS at Yale Comprehensive Cancer
Center

CIS at Memorial Sloan-Kettering
Cancer Center

CIS at Roswell Park Cancer Institute
CIS at Fox Chase Cancer Center
CIS at Johns Hopkins

CIS at Duke Comprehensive Cancer
Center

CIS at Sylvester Comprehensive
Cancer Center/University of Miami
CIS at University of Alabama at
Birmingham

CIS at Kentucky Cancer Information
Service

CIS at West Virginia University
Health Sciences Center

CIS at University of Wisconsin
Comprehensive Cancer Center

CIS at Meyer L. Prentis
Comprehensive Cancer Center of
Metropolitan Detroit

CIS at University of Kansas Medical
Center

CIS at MD. Anderson Cancer
Center

CIS at Fred Hutchinson Cancer
Research Center

CIS at Penrose Cancer Hospital

CIS at Northern California Cancer
Center

CIS at Jonsson Comprehensive
Cancer Center/UCLA

Region 19: (HI) CIS at Cancer Research Center of
Hawaii

Research supported by: Subcontract No. 73 7-4241 ﬁom the AMC Cancer Research
Center for P0] CA5 7586-01Al grant from the National Cancer Institute. The conclusions
reached in this paper are solely those of the authors and do not necessarily reﬂect the
views of the National Cancer Institute or the AMC Cancer Center.

Special Thanks:

I would like to thank Deb Tigner for her help in preparing and mailing the questionnaires.
I would also like to extend thanks to the members of the Network Analysis Advisory
Board for their help throughout the many phases of the research process: Donna Cox, Jo
Beth Speyer, William Stengle, Marsha Woodworth, Maureen McClatchey, and Diane
Ruesch. In addition, four excellent researchers must also be thanked for their extensive
help with gathering the data as well as working through all the revisions of this paper:
Marcy Mayer, Carolyn Ethington, Susan Pobocik, and Hui-Jung Chang.

iv

IN ADDITION...

I would like to thank my advisor, J. David Johnson, who I am grateful to for his
many stimulating comments as well as much needed guidance not only through this thesis
but through my graduate education as well. I would also like to thank my committee
members, Dr. Frank Boster and Dr. John Wagner, III, for their expert input and direction.

Additional thanks must also go to Marcy Meyer who helped me survive and
provided me with profound theoretical grounding about graduate school, life, happiness,
priorities, family, as well as role conﬂict and role ambiguity.

Many additional, heart-felt thanks to my partner, William E. Donner without
whom the journey through life would not have been as fun or made easier and to Melissa
M. Emmett, a soul ﬁiend and true believer when no one else was.

Finally, to my Mom and Dad who have a lot to do with the person I was, I am, and

forever will be.

TABLE OF CONTENTS

Page Number
List of Tables ......................................................................................... vi
List of Figures ........................................................................................ vii
Chapter 1. Introduction ......................................................................... 1
Chapter 2. Literature Review ................................................................. 3
Chapter 3. Methods ............................................................................... 17
Chapter 4. Results ................................................................................. 26
Chapter 5. Discussion ........................................................................... 30
Appendix ............................................................................................... 37
List of References .................................................................................. 40

LIST OF TABLES

Page Number
Table 1. The Expected and Error Correlation Matrix .......................................... 24
Table 2. The Intercorrelation and Factor Matrix ................................................ 25
Table 3. Correlations Among Constructs ........................................................... 27

LIST OF FIGURES

Page Number
Figure l. Formalization as a Mediator of Hierarchical Level,
Role Conﬂict, and Role Ambiguity ................................................... 12
Figure 2. Path Model with Path Coefﬁcients .................................................... 28

viii

Chapter 1

INTRODUCTION

The constructs role conﬂict and role ambiguity have received much attention from
sociological, psychologicaL communication, as well as managerial scholars (see Fisher &
Gitelson, 1983 and Jackson & Schuler, 1985 for a complete list). From Kahn, Wolfe,
Quinn, Snoek, & Rosenthal’s (1964) role explication decades ago, the literature on role
conﬂict and role ambiguity has grown considerably.

Despite some of the equivocal results of various empirical studies, both role
constructs are described as role stressors (Kahn et al. 1964; Schaubroeck, Cotton, &
Jennings, 1989; Coverman, 1989; Siegall, 1992). The literature has identiﬁed several
correlates of role conﬂict and ambiguity, including many structural variables such as:
formalization, participation in decision making, span of subordination/supervision, size,
and organizational level (Kahn et al., 1964; Greene, 1978; Morris, Steers, & Koch, 1979;
Organ & Greene, 1981; Nicholson & Goh, 1983). Many negative psychological,

emotional, and behavioral outcomes have also been associated with both role stressors.

These include: tension, anxiety, dissatisfaction, absenteeism, as well as lack of
commitment, low performance, low involvement, and reduced levels of autonomy (Brief
& Aldag, 1976; Rizzo, House, & Lirtzrnan, 1970; Van Sell, Brief, & Schuler, 1981; Fisher

& Gitelson, 1983; Jackson & Schuler, 1985).

Chapter 2
LITERATURE REVIEW

Role Theory

Kahn, Wolfe, Quinn, Snoek, and Rosenthal (1964) have usually been credited for
introducing and explicating the concepts of roles and role processes as well as
popularizing both constructs in the organizational literature. However, before these
researchers, other works predated Kahn et al. (1964) with the initial concept of role. In
193 6, Linton brought the idea of m to the social science forefront. Others followed in
relating the construct of role to their respective ﬁelds (Newcomb, 1950; Merton, 1957,
Shibutani, 1961). These authors proposed the notion of a role as the primary means of
linking the person with her/his environment. In a historical context, the concept of role
linked psychology to sociology. It is where structural and personal forces meet that a role
is created, altered, and maintained. These early researchers formed the beginning of role
theory. A role, then, can be deﬁned as, “a set of expectations about behavior for a
position in a social structure” (Rizzo et al. 1970; p. 155)

In 1964, Kahn et al. (1964) extended these ideas to organizational settings. They
fully explicated what a role was and how it was created through a theory that incorporated
the Role Episode Model. The theoretical framework of the role episode consists of ﬁve

major elements: organizational factors, attributes of the person or personal factors,

interpersonal factors, the role sender (including her/his role expectations and sent role),
and the focal person (including her/his received role and role behavior). The role sender is
the individual who conveys expectations of behavior to another person. In most cases, the
person who fulﬁlls this position is a supervisor. The focal person, on the other hand, is the
individual who receives the role expectations sent from his/her superior.

These ideas can be captured in Burt’s (1991) concept of structural equivalence
where, “similar actors lmve similar patterns of relations with others. Pattern similarity
deﬁnes social boundaries around reference groups and feelings... creating homogeneous
beliefs and behavioral tendencies among equivalent actors” (p. 8). Thus, organizational
members are inﬂuenced by their role sets. As a speciﬁc extension of Burt’s assumptions,
it could be argued that those individuals who occupy a similar position within the
hierarchy will also, through similar communication patterns, feel and perhaps act similar to
those who hold the same position within the organization. The same could be said for
those individuals who hold the same roles or who interact with the same role sets
(Johnson, 1993).

A role is the summation of the necessary requirements deﬁning what the behaviors
of a member within a certain position should be. As a way of operatioﬂﬁliZing systems,
Katz and Kahn ( 1966) refer to roles as, ”the building block[s] of social systems and the
summation of the requirements with which such systems confront their members as
individuals" (p. 197). Similarly, role expectations are the set of activities that a person is
responsible for engaging in as part of her/his position within the organization. These

expectations are task-related; however, this is not the limit of their content. They also

provide norms of both professional and social behavior. Expectations can be explicit,
implicit, direct, or vague; they can also compete with each other. This is where the
concepts of role conﬂict and ambiguity develop.

Kahn et al. (1964) originally conceptualized both role conﬂict and role ambiguity
as multi-dimensional constructs. In addition, Kahn et al. (1964) proposed two different
types of role conﬂict and ambiguity: subjective and objective. In this sense, objective
means quantiﬁably "real,” veriﬁable conditions in the environment in which contradictory
(conﬂict) or unclear (ambiguity) messages are sent by the role sender(s). Subjective role
conﬂict and ambiguity is the internal, psychological state of the focal person; it is the
individual's perceptions of such messages. Kahn et al. (1964) suggested that the existence
of one type does not necessarily include the existence of the other type. In other words, if
one experiences objective role conﬂict, she/he may or may not experience subjective role
conﬂict and vice versa.

Role Conﬂict. Kahn et al. (1964) also created a typology of role conﬂict. Because
individuals receive role expectations at any given time, these expectations can compete
with one another. The target individual is said to have psychological conﬂict because of
the pressure ﬁom these competing expectations. Thus, the ﬁrst, and most popularized

type of role conﬂict is, sent role conﬂict. and is deﬁned as:

 

the simultaneous occurrence of two (or more) sets of pressures such that
compliance with one would make more difﬁcult compliance with the other. In the

extreme case, compliance with one set of pressures excludes completely the

possibility of compliance with another set; the two sets of pressures are mutually

contradictory. (p. 19)

 

Other types of role conﬂict include: intra-sender incompatible role expectations from a
member of the role set; inter-sender, incompatible role expectations from two or more
members of a role set; inter-role, incompatible expectations ﬁom individuals of two or
more different role sets; mrson—role (others refer to this as intra-role), role requirements
violate a person's morals, beliefs, etc; role overload, is the combination of person-role and
inter-sender conﬂict.

The underlying assumption with these different types of role conﬂict is that a
person is able to experience multiple contexts of conﬂict depending upon the individual’s
referent organizational network. This possibility is explicated through role theory which
speciﬁes that organizational members belong to various role sets, or networks. Because
of the vast membership in these networks, conﬂicts are able to more readily develop.
Kahn et al. (1964) gave labels to different types of role conﬂict to emphasize the many
nuances of the construct itself.

The importance of one type of conﬂict over another may be contingent upon which
role set is most salient to organizational members. For instance, duties that a priest may
feel contradict his religious morals will probably be more important (at least perhaps the
need to alleviate the conﬂict) than diﬁ'erent expectations that come ﬁ'om a fellow priest.

That is, it would be expected that the priest may view person-role conﬂict as a more

important consequence than intra-sender conﬂict. The signiﬁcance of one facet of role
conﬂict, then, is contingent upon the context embedded within an individual’s role set(s).

It could be argued that intra-sender, inter-sender, and inter-role conﬂict are
common forms of conﬂict within organizations and thus need to be investigated; therefore,
they will be the focus of this paper.

Role Ambiguity' . Kahn et al. (1964) deﬁne role ambiguity as, ”the degree to which
required information is [un]available to a given organizational position" (p. 26). Role
holders fulﬁlling that organizational position can experience four types of ambiguity. The
ﬁrst is the uncertainty about the focal person's requirements or some of responsibilities.
The second refers to the uncertainty about the set of behaviors that are necessary to fulﬁll
those responsibilities. Third, the focal person may be uncertain about whose expectations
must be met, role sender A or role sender B? Forth, and last, refers to the uncertain
consequences of role behaviors on the focal person herself/himself or the organization as a
whole.

Together, these four types can be partitioned into two major dimensions of
ambiguity: task and socioemotional. The ﬁrst three types fall under the task dimension
where Kahn et al. (1964) suggested that task ambiguity, ”results from lack of information
concerning the proper deﬁnition of the job, its goals and the permissible means for
implementing them,” (p. 94). The last type, uncertain consequences, falls under the rubric
of socioemotional uncertainty where this dimension of ambiguity, "manifests itselfin a
person’s concern about his [her] standing in the eyes of others and about the consequences

of his [her] action for the attainment of his [her] personal goals,” (p. 94).

Kahn et al. (1964) recognize that not receiving adequate information to perform
one's duties can stem from various causes: inadequate communication, availability of
information, nonexistent information, etc. All of these factors contribute to the focal
person experiencing role ambiguity.

Role conﬂict and role ambiguity have been explicated in the preceding paragraphs.
Kahn et al. (1964) asserted that these two role constructs were expected to be
independent of each other and that the simultaneous occurrence of both was unlikely. In
addition, ﬁom their explication it seems as if role conﬂict could result from role ambiguity
and vice versa, as the deﬁnitions of each allow for crossover points (King & King, 1990).
Therefore, the relationship between these two role constructs is vague, even as
conceptualized by their creators.

Since Kahn et al. (1964), there have been a host of researchers that have used
these role concepts to theoretically ground their empirical or conceptual research (Rizzo,
House, & Lirtzman, 1970; Lichtman & Hunt, 1971; Schuler, Aldag, & Brief, 1977; Batlis,
1980; Kemery, 1991; O'Driscoll, Ilgen, & Hildreth, 1992; Sawyer, 1992). Each have
supplemented or slightly altered Kahn et al.’s (1964) initial explication of role constructs;
however, little theoretical work has been developed since the original inception of role
conﬂict and role ambiguity.

Many relationships have been hypothesized between and among role conﬂict and
ambiguity and other variables. Brief and Aldag (1976) reported signiﬁcant, negative
correlations between role conﬂict and task identity, autonomy, work satisfaction,

supervisory satisfaction, and perceived work quality; signiﬁcant positive correlations were

found for anxiety-stress, tension, and propensity to leave the organization. For role
ambiguity, the same study found signiﬁcant, negative correlations for feedback from job
and performance ratings. Likewise, signiﬁcant, positive correlations were found for
anxiety-stress, tension, propensity to leave, and turnover.

Meta-analyses performed since then have corroborated these results. The ﬁrst
meta-analysis of the correlates of role conﬂict and ambiguity was done by Fisher and
Gitelson (1983). Using 42 studies, these researchers found signiﬁcant, negative mean
correlations between role conﬂict and organizational comnritment, job involvement,
satisfaction with job (e.g., pay, co-workers, and supervision), and participation in decision
making. Role conﬂict was also found to be signiﬁcantly related to boundary spanning:
those who reported higher levels of role conﬂict also reported higher levels of boundary
spanning activity. Role ambiguity was found to be signiﬁcantly and negatively related to
the same constructs including boundary spanning. In addition, Fisher and Gitelson report
that both tenure and age is also negatively correlated with role ambiguity while education
was positively related to the same role construct.

A second meta-analysis of these two variables was offered by Jackson and Schuler
(1985). Negative average weighted correlations, which were within a 90 percent
conﬁdence interval not containing zero, were reported between role conﬂict and task
identity, feedback from others and task, leader initiating structure, participation, job
satisfaction (various aspects), commitment, involvement, and organizational level.
Positive average weighted correlations, within the same conﬁdence interval, were found

between role conﬂict and locus of control, task/skill variety, tension/anxiety, and the

lO

propensity of an individual to leave her/his organization. These same relationships were
found for role ambiguity, except for three cases. Task/skill variety dﬁ correlate with role
ambiguity; however, it was n_egativ_e but not within the conﬁdence interval speciﬁed. The
second exception was with regard to the locus of control construct, while the correlation
was in the positive direction, it was in a conﬁdence interval that did include zero. Most
importantly, at least for this study, formalization was negatively correlated with role
ambiguity within the 90 percent conﬁdence interval. Note that there was no signiﬁcant
relationship found between formalization and role conﬂict.

While some of these hypothesized relationships above have been supported or
partially supported by research, others have not. The negative relationship between
formalization and role ambiguity has repeatedly been shown to exist; however, research on
the relationship between formalization and role conﬂict has demonstrated inconclusive
results. Organ and Greene (1981), while ﬁnding the traditional strong (r= -.40)
correlation between formalization and role ambiguity, found that formalization and role
conﬂict were positively correlated (r = .24). They attributed this latter relationship to
discrepancies between organizational requirements and professional norms. To the
contrary, Morris, Steers, and Koch (1979) found a signiﬁcant, negative relationship
between formalization and role conﬂict. As can be ascertained ﬁom above, the nature and
strength of the relationship between formalization and role conﬂict is unclear. Some
researchers hypothesize and empirically support a positive relationship, while others

anticipate and demonstrate a negative correlation.

11

The following study was undertaken to offer insight into the relationship between
role conﬂict and formalization by explaining the function of organizational level. More
speciﬁcally, two avenues will be investigated in an attempt to reconcile these ﬁndings.
First, a deﬁnite, negative relationship between formalization and role conﬂict will be
hypothesized. Secondly, a direct, causal relationship will be applied ﬁ_orr_r a structural
variable (formalization) t_o_ role conﬂict. This same model will also examine the
relationship between formalization and role ambiguity. Finally, this study will extend
previous research by testing the model of the impact of hierarchical level on role conﬂict

and role ambiguity as mediated by formalization. This model is presented in Figure 1.

12

Figure 1:
Forrnalization as a Mediator of Hierarchical Level, Role Conﬂict, and Role Ambiguity

 

Role Conﬂict

 

 

 

 

 

 

Hierarchical F orrnalization
Level

 

 

 

 

 

/\

 

Role Ambiguity

 

 

 

13

The Model

HieLarchﬁl Level. Past researchers have described the construct of hierarchical level in
various ways. Organizational level (Schuler, 1977; Mossholder, Bedeian, & Armenakis,
1981), functional role (Johnson, Meyer, Chang, Ethington, Pobocik, & LaFrance, 1995),
and position have also been offered to describe the same conceptualization. For the
purposes of this investigation, hierarchical level can be deﬁned as a person’s position on a
scalar chain within an organization (Jablin, 1987). Graphically, it can easily be seen on an
organizational chart where a chief executive ofﬁcer’s position, for example, is ﬁxed at the
top of the hierarchical chain and where unsupervisory workers are displayed at the lower
level of the chart.

Much research regarding role conﬂict and role ambiguity have considered
hierarchical level as a moderating variable of role stressors and behavioral or psychological
outcomes. Schuler (197 7) found that organizational level and participation in decision
making moderated the relationships among role perceptions, satisfaction, and
performance. Likewise, Mossholder, Bedeian, and Armenakis (1981) found that
organizational level and self-esteem were moderators of role stressors, satisfaction, and
performance. While these moderating relationships were found to be signiﬁcant, they
accounted for a small proportion of the variance in the dependent variables.

Hierarchical level has also been posited as an antecedent to role conﬂict and role
ambiguity. In their meta analysis of both role stressors, Jackson and Schuler (1985)

reported that previous research had found no relationship between organizational level

14

and role conﬂict nor role ambiguity. This relationship directly contrasts the data that Kahn
et al. (1964) initially reported where they hypothesized that higher level employees would
experience more role ambiguity because they had more direct contact with the uncertain
environment. The task, then, seems to be in determining the position of hierarchical or
organizational level within a causal model.

The higher an individual is on the organizational chart, the higher that person's
hierarchical level within the organization. Again consider the chief executive ofﬁcer, an
individual who is at one of the highest levels within an institution, who will have fewer
written rules, policies, or procedures than an individual at lower levels of the
organizational hierarchy. The chief executive oﬁcer is often required to ﬁrlﬁll different
roles that mandate her/him to perform unique duties at different times; because of this
ﬂexibility required of the job, there can be virtually no standardization of operating
procedures. In stark contrast, an individual who is stationed at the lower level within an
institution's hierarchy, a retail cashier for example, will have very speciﬁc, written policies
and procedures to follow while performing her/his duties. With this example in mind, it is
hypothesized that a higher hierarchical position (the exogenous variable in the model)
within an organization will lead to a decrease in the level of formalization an individual
will experience.

Forrnalization. Pierce and Delbecq (197 7) conceptualized formalization as, "a form of
control employed by bureaucratic organizations, [and] refers to the degree to which a
codiﬁed body of rules, procedures or behavior prescriptions is developed to handle

decisions and work processing" (p. 31). Jablin (1987) notes the explicitness of

15

formalization. As he states, it is, "the degree to which the behaviors and requirements of
jobs are explicit..." (p. 404). F orrnalization may be achieved through the utilization of
procedures manuals, standardization of operating procedures, or the presence of job
descriptions. As a result of its importance as an index of organizational structure,
formalization has recently been included in the National Organization Study (Marsden et
al.1994). For the purpose of this investigation, formalization (the mediating variable) can
be seen as the degree to which organizational members perceive the existence of policies,
procedures, and rules that precisely deﬁne their daily activities. Therefore, the ﬁrst
hypothesis is generated,

H1: A person residing at a higher position within an organization’s hierarchy
will experience a direct m in the level of formalization related to
her/his job.

Role Conﬂict. Role conﬂict, for this study will be deﬁned as the, "incongruity of the
expectations associated with a role" (Van Sell, Brief, & Schuler, 1981). The source or
sender of these expectations may vary. Perhaps an organizational member receives
contradictory expectations from the same sender or she/he may receive conﬂicting
expectations from more than one supervisor. Kahn et al. (1964) proposed that the
communication "style" between the focal person and role sender was an important variable
to be considered in the role process. When expectations of the subordinate are
contradictory, the quality of communication between the supervisor and subordinate

suffer.

16

In their seminal work, Kahn et al. (1964) theorized that the relationship between
formalization and role conﬂict would be negative. Following this lead, a negative
relationship is hypothesized for this study. Ifan organization is highly formalized (e.g.,
established, written policies and procedures exist), then the expectations of an individual
working under those policies are known. Therefore, the chance that conﬂict may arise
between such expectations decreases considerably. Thus,

H2: A person who has a highly formalized position within an organization will

experience a direct decrease in her/his level of role conﬂict.

 

Role Ambigu_rty' . As previously stated, role ambiguity has often been conceptualized as
the degree to which information regarding a person's role expectations, means of
perfomring organizational duties, or the outcomes of organizational behavior associated
with her/his role, is either vague or lacking (Van Sell et al.1981).

Also in keeping with past literature, it is proposed that formalization will also be
negatively correlated with role ambiguity. This relationship is intuitive. If there are clear
policies and procedures speciﬁed for organization members‘ duties, then their uncertainty
on how to perform those duties is greatly diminished. In addition, it is predicted that there
is a direct causal relationship between formalization and role ambiguity. Therefore,

H3: A person who has a more formalized position within her/his institution will

experience a direct decrease in her/his level of role ambiguity.

Chapter 3
METHODS

This research was conducted in the National Cancer Institute’s (NCI) Cancer
Information Service (CI S), a geographically-dispersed federal government health
information program that disseminates cancer information to the public. The CIS was
established in 1975 by the NCI to disseminate accurate, up-to-date information about
cancer to cancer patients, the relatives and friends of cancer patients, health care
professionals, and to the general public. Over the past 20 years, the CIS has compiled a
remarkable record of achievement in ﬁrlﬁlling the critically important function for the NCI
(Morra et al. 1993a). The public health mandate of the CIS is grounded in the National
Cancer Act of 1971 and the amendments to that act made over the past 20 years (Morra,
et al.1993b). The core element of the 1971 National Cancer Act that led to the formation
of the CIS stipulates that the NCI, “Provide a program to disseminate and interpret... for
practitioners and other health professionals, scientists, and the general public, scientiﬁc
and other information regarding the causes, prevention, detection and treatment of
cancer.” In response to this mandate, the CIS currently maintains a network of 19
regional oﬂices that are typically linked to NCI-ﬁmded regional cancer centers. The
activities of the CIS network are coordinated and supervised by the Oﬁice of Cancer

Communications at the NCI. These activities fall into two broad categories:

17

l8

1) responding to requests for information over the telephone (the CIS operates a toll-free
telephone number, 1-800-4-CANCER, in which callers are automatically triaged to their
regional ofﬁce from response from a professional cancer information specialist, and 2)
conducting community outreach activities. The communication outreach programs of the
CIS can be ﬁrrther subdivided into mass media campaigns that promote use of the CIS
toll-ﬂee telephone number and/or encourage speciﬁc cancer prevention and control
behaviors (e. g., smoking cessation or screening mammography), and more interpersonal
community outreach efforts that typically involve working with community intermediaries.
The data used for this investigation were part of the eighth recurring quarterly data
collection period of a larger study that took place in August 1995. Each individual
participating in the larger project within the CIS was mailed a series of questions relating
to various organizational issues as well as a network questionnaire. These questionnaires
were sent with a stamped, self-addressed return envelope to the respondents
approximately ten days prior to the sample time period. In addition, a personalized letter
was included to explain the various issues that would be examined and urge participation
in the data collection. Concurrently, an e-mail was sent to all participants to notify them
that they would be receiving the questionnaire. In addition, an e-mail was sent the day
after the sample time period had concluded and reminded respondents to return their
questionnaires. Through these extensive follow-up efforts, a very satisfactory response

rate (90 percent) was achieved. A total N = 81 was the size of the sample.

l9

Qperationalization. Hierarchical level was assessed using a directory and organizational
chart of the CIS where the following positions were ranked ﬁ'om a higher to a lower level
within the hierarchy: National Cancer Institute, Ofﬁce of Cancer Communications and
Public Inquiries (level 1), Principal Investigators (level 2), Project Directors (level 3),
Telephone Service Managers and Outreach Coordinators (level 4).

The remainder of the constructs were measured using scales comprised of 11
point, Likert-type items. Formalization, a perceptual self-report instrument, was assessed
using ﬁve items. These items were used and validated in a previous study investigating
perceived organizational innovativeness (Johnson, Meyer, Berkowitz, Ethington, & Miller,
1995). Examples of formalization items are: “The policy and procedures manual for the
CIS covers what happens in a typical day;” “Policies and procedures are strictly enforced
at the CIS;” and, “I follow established procedures exactly.” Standardized alpha was or:
.80 for this measure.

Role conﬂict was operationalized using all eight items ﬁom the Rizzo, House, and
Lirtzman's (1970) role conﬂict scale. Six items, addressing different aspects of role
conﬂict, were also added, making the total number of items equal to 14. Examples of
such items include: "I have to do things that should be done diﬂ‘erently;" and, "I receive
an assignment without the manpower to complete it;" ”I have to buck a rule or policy in
order to carry out an assignment.”

Role ambiguity was assessed using all six items from the Rizzo et al. (1970) scale
and again an additional ﬁve items were included for a total of 11 items measuring the role

construct. Items such as the following were included: ”I feel certain about how much

20

authority I have;" "Clear, planned goals and objectives exist for my job;" and, ”I know
what my responsibilities are. " All items were reversed coded so that the higher the
number, the more role ambiguity a person would report experiencing. For both scales,
Rizzo et al. (1970) had originally used seven-point Likert-type items; as stated earlier, CIS
members responded utilizing an 11 point scale.

While there have been criticisms of the Rizzo et al. (1970) scales and
conceptualizations (King & King, 1990; Kelloway & Barling, 1990) there has also been
support (Schuler, Aldag, & Brief, 1977). Based on these comments and criticisms,
additional items were added to these scales to solidify a good measure of both constructs.

In addition, organizational members were speciﬁcally asked to respond to each
item on two levels: within her/his ofﬁce and across the CIS network. However, for the
purposes of this study, since macro-structural variables were being assessed, so too would
the role constructs at a macro level: across the network. Data gathered on role conﬂict
and role ambiguity within individuals’ ofﬁces are not part of these analyses because these
perceptions are held at a more micro level.

Individual item means and standard deviations for the formalization, role conﬂict,
and role ambiguity scales are contained in the Appendix. The basic statistics for the scales
averaged across items were as follows: formalization, M = 3.20, SD = 2.00; role conﬂict,

M = 4.42, SQ = 2.28; and role ambiguity, M = 6.94, S_D = 1.61.

21

Analysis

Conﬁrlnatory Factor Analysﬁ. A conﬁrmatory factor analysis was performed on the data
with the initial theory that there were a total of three factors to which 30 items belonged.
These factors were role ambiguity, role conﬂict, and formalization. As a result, a total of
eight items were deleted from the matrix. Item number ten from the role ambiguity scale
as well as items number 13, 14, 15, 18, 23, 24, and 25 from the role conﬂict scale were
removed (see the Appendix). Another conﬁrmatory factor analysis was performed on the
remaining 22 by 22 matrix.

While the standardized item alpha coefﬁcient is an indicator of the reliability of the
measure, a more advanced statistical technique will provide two different tests gauging the
validity of a scale. Conﬁrmatory factor analysis is such a technique (Hunter & Gerbing,
1982). Three criteria utilized in this type of analysis includes establishing face validity or
homogeneity of item content, internal consistency and parallelism.

The 22 items were found to be face valid based on item content; each item appears
to be tapping its theoretically-based factor. Internal consistency is a test employed
through conﬁrmatory factor analysis to determine if the correlations among the items are
as predicted for the factor model. This is ultimately achieved through comparing the
obtained versus predicted correlations between items. The larger the difference, the larger

the error. The predicted correlations are generated among items one through ten, 11

22

through 17, and 18 through 22 using the internal consistency theorem. This theorem
calculates a predicted correlation by multiplying the factor loadings.

The expected and error correlation matrix, shown in Table 1, displays the residuals
for all 22 items in the upper diagonal. While there are some larger errors (e. g., error =
.23), the percentage of items for which the errors are greater than sampling error is 4.44
percent for role conﬂict and 4.76 percent for the role ambiguity scale. For formalization,
none of the errors were greater than sampling error.

Table 2 shows the intercorrelation and factor matrix. There seems to be no
substantial cross-loadings on any of the factors. For role ambiguity, the factor loadings
ranged from .59 to .94, with the highest cross-loading at -.62. The range of factor
loadings for the role conﬂict measurement was ﬁom .53 to .86, while the greatest cross-
loading value was .33. Last, the range of the factor loadings for formalization was from
.43 to .86; the highest cross-loading was -.42 for this factor. Thus, it appears as if all
three measurements have achieved internal consistency.

Parallelism, the second test within conﬁrmatory factor analysis, is generated by
taking the product of the factor loadings and multiplying that product by the correlation
between the constructs themselves. These predicted correlations between items one and
ten, 11 and 17, aswellas 18 and 22 are alsoinTable 1. Ascanbeseen, someofthe
residuals are higher than one would expect or ﬁnd optimal (e.g., error = .26). The
percentage of items for which these errors are larger than sampling error is 3 .23 percent.
With a sample size of N=8 l , the size of the residuals, as well as the percentage or errors in

proportion to sampling error, was acceptable.

23

Using conﬁrmatory factor analysis, scale reliabilities were calculated. For role
ambiguity, standardized or = .94; while the role conﬂict scale received a standardized or =
.86. Standardized or = .79 was achieved for the formalization scale.

From these four criteria, face validity, internal consistency, parallelism, and
standardized reliabilities, one can be conﬁdent in all three scales: role ambiguity, role
conﬂict, and formalization.

Path Analysis. To test the above hypotheses regarding the predicted path model, shown in
Figure 1, a path analysis was performed on the data (Hunter & Hamilton, 1995). The
construct hierarchical level varied in range from one to four, with one being hierarchically
higher on the organizational chart. Note that this would then create a positiye correlation
or path coefficient between hierarchical level and formalization, so that the higher the level
the higher the perception of formalization. This coding of the hierarchy was taken directly

ﬁom the CIS’ organizational chart and conﬁrmed by experts within the network.

24

Heart. A

.25 mxwooﬂoa go m3: mono—lune: 331x...

t

\OWQONU'I-bDJNv—t

n—
nn

3
a
.3
3
mo
.3
S
8
3
a
a
3
S
3
a
G
-8
-8
-8
-8
-2

n

An
mm
mm
A
An
3
Am
mA
An
3
no
:
3
Am
5
-nn
Lo
-no
-nm
LA

u

A
on

3
me
on
mw
3
oo
on
Am
5
no
_A
no
3
nu
-nq
-nu
-u m
-w A
.. A q

A

Oh)

mA
mn
GA
3
an
an
no
no
nA
3
nm
nu
nm
-uA
-no
-Au
.3
-n_

m

93me

mo
mm
m—
.3
no
no
nA
no
5
no
nA
no
-u u
-no
-Ao
Ao
-nn

o

r—IANWO

am
no
am
3
no
nA
nm
5
no
nu
no
-um
-no
.Am
.3
-nn

q

r—le—I
Nr—rmOONOO

on
we
mo
3
6
no
_A
no
3
nu
-nq
no
.3
A3
-3

m

onva—A

.2
no
3
nn
no
A q
nA
n A
no
-u w
-nq
-An
-w g
.nn

0

UtxlNNAu-dm-h

am
3
nA
nn
5
nn
no
no
.3
-no
Ao
.3
.5

ﬂ
NHN—w—awww O

Antbnnn—nnr—
uuounowoxoomAxo

-uo
-nn

:

b-i t—t
OAHQAROOONUI

H

AwAwAw
oooowo—xo

Lo
-m
Ln
.2
-o

n

l—d
\JN—‘HMONQOOQON-b-

GALI’IWM
ooowxr.‘

Ln
Lo
.3
LA

-w

on

NU’IQOF—N

no
n
no
no
u
3

no
mm
mo
on
Lu
.2
-3
LA
.m

H
A

i—d
wNAQONHﬂﬂbN—‘O

H

Ln
Lo
-o

3

\0

_n

A0
3

OOWQOOOOO

Mn
3
Ln
-2
.3
.3
-m

3

mum—to

mm
Ln
Lo
Lu
Lu

-u

3

pap—.3

O\O\~AA\IAwnww——-N.—~

5
o
_n
u
q
.o
3
u
n
w
A

nu
w
G
3
n
m
5

om
mm
mo
no

no no n_
u q n
G I no
o w A
o m n
n A m
o 3 on
: u —o
n o m
m A w
n w A
0 o no
.V o A
w A m
n 3 c
m n o
m n o
o w _o
S A n
An A
A0 n
Au 3
nm .3 wn

nn

no
H:
0
AA
3
x

00000

s .26 ESQ ammoaa conga :6 oxeoonegcaomoaa cone—macaw $399: amass—mo woa w: moan :8 €qu mimosa £52m :5
Emma—cam.

25

.320 n

MHz-1238:8530: Ba 3888 253x...

n n n w a u 8 a m o 8 : B 8 z 8 8 S 8 8 8 2 8. P» we m
_ B a 8 an
n .5 B a. no .8
u t 3 n» q... no a...
a .3 8 8 8 8 3 -8
-.. 8n 8 8 3 Mn 2 8 .3
8 8 3 8e 8 8 mm 3 a -3
q 8 8 3 we 3 2 uln- an 8 E
m 8 an a 8 8 8 S n» 8 3 .3
e .8 a .3 3 .8 3 8 8 8.8. an no .3
8 8.. 3 8 2 8 8 8w 8 n... Mn 3 8 -8
: 8 8 q 8 8 8 -n S n. e B- 8 .8 L
s. s B .n S 8 .8 8 .3 8 B n ale 8 8 .8
a 8 S : S no 3 o B 8 e 8 a M 8 q... a
8 w 8 8 8 5 S 8 8 8 S 8 8 8 .8- 8 a -8
8 8 8 8 3 no 8 .n B 3 8 2 8 on .8 Mm 8 8 -8
8 8 8 8 an 8 8 no 8 ns 8 8 .3 8 8 8.. An 8 a. -8
S no 8 B 8 no an 8 no 8 8 8 8 8 3 8 8.. n» n 8 -8
8 an .8 -8 r: .8 -8 -8 .2 .8 -8 a -8 n .8 -8 an n An -8 a 3
8 .8 -8 .3 -8 -8 -8 -5 -8 -8 -2 .8 -3 .8 .8 -8 -8 -8 8 Mn -8 -8 .3
no .8 .8 an -8 an -8 an -8 -8 an -n -8 -8 -8 -8 -3 -8 we 2 .8 an -8 8
B -8 -2 -3 .8 -8 -3 -8 an -8 .8 8 -8 .8 -8 a -8 h 3 t 8 M an -5 8
S -8 .8 a .8 -8 -8 -8 -3 -8 an .: -8 a -8 -8 -8 -8 8 8 3 on 8 -3 -8 a

P» 3 an an on P. on an no on on no no no G .3 no nn .8 -no .An An -3 3o 3 .3
wn nn no no 3 nu An S 3 no n8 we no nu ma 8 8 no .8. .5 .3 Ln -na 3 So -nq
T. .An .on .3 -3 .An -3 -3 An An -3 L -nn .o -8 no no -3 an an «a am 3 -3 no no:
_.. .38 enema—Son 85:88 Beam :8 gnomes— Emug. :8 85358 we one: 83. .28 $92 835% an» 358 388. an 88 ”.86me? won 38

8:33. go mu mouse—Enos

Chapter 4
RESULTS

Comtions. The correlations for the four constructs are shown in Table 3. There
was a signiﬁcant, negative correlation found between hierarchical level and role conﬂict (p
= .02). This surprising ﬁnding is examined in the discussion section of this paper.

As can be seen, there seems to be no relationship between organizational level and
role ambiguity. Again, this may be attributable to the CIS speciﬁcally. As a government
agency, where the policies and procedures of the daily routine may be ﬁrmly established,
other environmental elements (e. g., the budget, contract renewals, etc.) contributing to
role ambiguity may be more salient to organizational members.

A signiﬁcant correlation was found between role conﬂict and role ambiguity.
While past research as shown such a correlation, when partialing out their shared
antecedent variable, formalization, as well as organizational level, the correlation between
the role constructs drops to a non-signiﬁcant pr = .22.

Figure 4 shows the results of the path run. The model provided a poor ﬁt to the
data. None of the indirect links were signiﬁcant (at p<.05) and the overall chi-square for
the entire model was )6 = 4.34, with 1f= 3 (p>.05). Despite this ﬁnding, two of the three
predicted paths had zeroes within their 95 percent conﬁdence intervals. The reproduced
correlations for the indirect links were relatively low; between hierarchical level and role

conﬂict, r=.04

26

Table 3

C orrelgtions Among Constructs

27

 

 

Variable Level F onnalization Conﬂict
Level

Formalization . 1 5

Conﬂict -.29* -21

Ambiguity -.08 -49" .34”

 

 

*p<.05. **p<,01

 

 

Hierarchical
Level

 

 

.17

28

Figure 2
Path Model with Path Coefﬁcients

-.25

\

 

Formalization

 

 

 

/

-.57

 

 

Role Conﬂict

 

 

 

Role Ambiguity

 

 

 

29

with r=.10 between level and role ambiguity. \Vithin this model, role conﬂict and role
ambiguity were shown to have a correlation of r=. l4.

Hypothesis one was not supported; organizational level was not a signiﬁcant
predictor of the level of formalization experienced by individuals whereby those
individuals at the lower level of the hierarchy, overall, did not seem to experience higher
levels of formalization (Pathmwrm = .17). This path had a 95 percent conﬁdence interval
that included zero (-.07 g p S .41).

Hypothesis two was not supported by the data in that formalization failed to be an
adequate predictor of role conﬂict (Patthlemm= -.25). The 95 percent conﬁdence
interval around this point (-.51 5 p 5 .01) also included zero.

Hypothesis three was strongly supported by the model. Higher levels of
formalization did cause lower levels of role ambiguity (Pathkmyml, mum = -.57). The 95

percent conﬁdence interval around this point did not included zero (-.77 5 p 5 -.37).

Chapter 5
DISCUSSION

It is probably most ﬁtting to begin the discussion section with some brief words
regarding Rizzo, et al.’s (1970) role conﬂict and role ambiguity scales. While the scales’
inception was achieved through exploratory factor analysis, their continued growth has
come from their repeated usage within the social sciences; this has been despite some
opposition. King and King (1990) reported problems with the measures and a seemingly
lack of convergent and discriminant validity. These authors also criticize the research
community whose attention has shifted away ﬁ'om the nomological network of
hypothesized relationships ﬁrst posited by Kahn et al. (1964).

This study attempted to somewhat bridge the gaps between the conceptualizations
of role conﬂict and role ambiguity as well as improving their measurements. Additional
items were added to the role conﬂict and role ambiguity measurements based upon the
conceptualizations and theoretical framework initially introduced by Kahn et al. (1964).
Using conﬁrmatory factor analysis, halfof the items that had to be discarded were Rizzo
et al.’s (1970) items. Using this type of statistical validation technique, which is based
upon a priorj assumptions between and among items and factors, internal consistency and

parallelism were successﬁrlly achieved. Thus, supporting the new scales purported here.

30

31

There has also been support for the Rizzo et al.(1970) scales; however, that
support has usually come through the implementation of exploratory, rather than
conﬁrmatory, factor analysis (Schuler, Aldag & Brief, 1977). This study offers two new
scales that have been scrutinized under four validating elements: face validity, internal
consistency, parallelism, and reliability. Through a statistical technique that requires a
“theory ﬁrst” approach, conﬁrmatory factor analysis, along with much theoretical
consideration, provides a bridge between the nomological ﬁamework and the
measurements of role conﬂict and role ambiguity.

“frth that stated, additional conceptual work needs to be done with items for each
scale. Future research that attempts to measure these role concepts ultimately needs to
address the various types of role conﬂict and ambiguity ﬁrst proposed by Kahn et al.
(1964) to determine their actual existence in the workplace. While Kahn et al.’s (1964)
conceptualizations are well-thought out and profound in providing an understanding of
role theory, the intricacies of such concepts may be far ﬁom distinguishable for
organizational members. It may be asking individuals to differentiate between and among
the types of role conﬂict and role ambiguity that contributes to the variation of responses
to many of the items.

While this possible dilemma poses no initial threat for the framework of the role
constructs, the way researchers measure both role conﬂict and role ambiguity would have
to be altered. Thus, in the “ﬁnal” analysis, ﬁxture research must ﬁrst try to establish the
existence of all the types of role conﬂict and role ambiguity; it is only with this information
that future researchers can create measurements that reﬂect experiences shared by the

organizational members who suffer from role conﬂict and role ambiguity. While this is

32

proposed for ﬁxture work and will take time (and perhaps more importantly money), the
current study offers a path regarding the salience of certain items. Knowing which items
seem to work and those that do not, provides an effective springboard ﬁ'om which other,
hopefully better items, can be generated.

This study choose to investigate the relationships between structural variables and
perceptions of role conﬂict and role ambiguity. While it was shown that formalization
mediates the relationship between hierarchical level and role conﬂict/role ambiguity, the
path coefﬁcients, not equal to 1.00, seem to indicate that the model should be expanded to
include additional variables. Future studies that include informal network measures, such
as prominence or power measures, would no doubt contribute invaluable information to
the understanding of how role conﬂict and role ambiguity function within organizations.
Which is more important, formal or informal structures on the perceptions of role conﬂict
and role ambiguity as experienced by organizational members? This is an important
research question that must be empirically addressed if the research on such topics is to
broaden its ﬁ'amework.

Another interesting ﬁnding to be ﬁrrther investigated, needs to be addressed in
more detail here. The relationship found between hierarchical level and role conﬂict was a
bit of a shock although not shocking. Kahn et al. (1964) had suggested a negative
relationship between the two variables. This would seem plausible since those individuals
at the lower strata of the hierarchy have more supervisors, who have more messages
instructing them what to do and thus increases the chances that a higher frequency of

messages are contradictory.

33

Equally as plausible is an argument that reverses such a relationship.
Organizational members at the higher echelons experience different types of conﬂict
pressures today. As organizations compete for ever-dwindling resources, top decision
makers must face enormous pressure and demands that often times conﬂict with their
personal beliefs or executive board wants. Take an example where a chief executive
oﬂicer must make a decision where solution A means laying off thousands of workers or
solution B that entails placing her/his organization at risk of bankruptcy (which the board
is ardently against). Sohrtion A evokes a moral dilemma (intra-role conﬂict); solution B
could create intra-sender conﬂict. Taken together, both solutions may embody sent role
conﬂict. Thus, the evidence presented in this study, that moderately support the idea of
those members at higher levels within the organization experience more role conﬂict than
those at lower levels, seems to be consistent within organizations today.

Focusing on the organization studied in this investigation may also contribute to
this ﬁnding. Contractual organizations are often embedded in other, perhaps larger
organizations. Such is the case with the CIS, whose 19 regional ofﬁces are “housed”
within community organizations. This situation may be especially salient for those
members who reside in higher positions and who must abide by the sometimes conﬂicting
rules or policies of b9t_h the CIS and its host institution. Thus, the negative correlation
found between hierarchical level and role conﬂict is reasonable given the substantiative
changes witnessed in organizations today; in addition, this relationship may be especially
salient to the CIS speciﬁcally.

The conﬁdence intervals around the paths speciﬁed in the ﬁrst two hypotheses

included zero (Pumas/{0m= -.07 _<_ p 5 .41; Pathrommkmm= -.51 5 p 5 .01). This

34

nonsigniﬁcant ﬁnding should lead future researchers to apply different models; speciﬁcally,
it may be the case that formalization is a moderator, rather than mediator as posed in the
current study. Formalization and hierarchical level may both directly contribute to role
conﬂict and role ambiguity.

While offering many insights, this like all other studies, also has a few limitations.
The ﬁrst of which is the static nature of the investigation. As was stated in the previous
paragraphs, this study is part of a larger project where future investigations will look at
such relationships over time. However, this speciﬁc study concentrates on the ﬁrst point
in time when role construct data was collected, therefore creating a strong baseline
measurement for the CIS. While it provides insights not already established, many
questions regarding time emerge. Do role conﬂict and role ambiguity have a time factor?
Are they static? Are they processes rather than outcomes? Are they cyclical in that
everyone experiences either or both at different times in their organizational lives? How
long does each last? These important questions can only be answered through
longitudinal analysis.

A second limitation that many may ﬁnd with this research is that it investigates
macro- and micro-level variables simultaneously. This “fear” can be somewhat alleviated
because the formalization measure is an individual perception measure that rests ﬁrmly
upon the reality experienced by most CIS members who live with binders and binders of
procedures manuals on a daily basis. In other words, the measurement, while on the micro
level, closely reﬂects the macro level (if it were to be indexed). The focus was shifted to
only include perceptions of role conﬂict and role ambiguity across members’ omces in an

attempt to limit suspicion of this data.

35

A third limitation to this study possibly includes the former. In doing research,
individuals must limit the scope of their topic and considerations. In doing so, they may
also eliminate data that possibly hold keys to the locked doors of unanswered questions.
While the focus across the CIS ofﬁces was necessary, ﬁxture analysis including within
members’ oﬂices will hopefully provide ﬁrrther insight to both role constructs. In addition
to this type of analysis, other analyses would be fruitful to perform. This would include,
but would not be limited to, comparing the means between organizational level and role
conﬂict/role ambiguity. Because the standard deviations for the role conﬂict items were,
in general, larger, this may provide additional information needed to pose new research
questions and hypotheses. In addition, regression formulas may also provide detailed
information regarding the speciﬁc contributions of each variable.

Lastly, a fourth limitation is the measurement of hierarchical level. While formal
organizational charts and experts were consulted, the hierarchy within the CIS is complex.
As is consistent with the above discussion on new organizational forms, it is diﬁcult to
draw perfect hierarchical levels within institutions today. This is the case with the CIS
who, as stated earlier, has oﬂices residing in larger organizations. Due to this
environment, members have sometimes parallel hierarchies, one for the host institution and
one for the larger organization. Individuals’ formal reporting relationships are not as clear
as they once were; with the increasing types of new organizational structures, the
construct of organizational level may have to be measured differently.

In summary, the model in Figure 1 provided a poor ﬁt to the data. Hierarchical
level was not a predictor of individuals’ perceptions of formalization whereby the lower

the organizational level, the higher the perception of formalization. In addition, no

36

signiﬁcant relationship was found between formalization and role conﬂict for
organizational members. The strongest, and perhaps most intuitive relationship was found
in the negative path between formalization and role ambiguity where the higher one
perceived her/his level of formalization within the organization, the lower the perception

of role ambiguity.

APPENDIX

APPENDIX

INDIVIDUAL ITEM STATISTICS FOR THE
ROLE PERCEPTIONS AND FORMALIZATION QUESTIONNAIRES

Directions: The following statements ask your opinion about the job you perform in the CIS. Each item
is designed to tapvarious attitudes you may have toward speciﬁc responsibilities that are associated with
your job. We would like you to respond by indicating how much you agree or disagree with each
statement on a scale of 0 to 10 where 0 indicates "total disagreement" and 10 indicates "total
agreement.” Plaseanswereachquestionashonestlyaspossiblemsalways, allresponseswill remain

 

 

 

 

 

 

 

 

 

 

 

 

strictly conﬁdential.
Role Ambiguity“ Within your Across the
ofﬁce CIS network

1. lfeelcertainabouthowmuch authoritylhave. g'lD:;-;
2. Clear. planned goals and objectives exist for my job. SAD==3275
3. 1 know that I have divided my time properly. 1:113:23
4. I know what my responsibilities are. SAD==2254
5. I know exactly what is expected of me. SAD==3213
6. Explanation is clear ofwhat has to be done. 240:3;
7. I am conﬁdent in performing my duties. §4D==2i39
s. I feel certain how I will be evaluated SAD-3276
9. The duties that come with my job are clear to me. 1:13:295
10 M ’ ' ' ' u M = 2.2

. yjobdutleschangewrthtlme. SD=28
ll. Itypieallyknowwhatisexpectedofmeaspartofmyjob. SJD==2283

 

 

 

 

*

Items 1-6 are from Rizzo et al. (1970).

** Item was removed from analysis.

37

 

Directions: The following statements ask your opinion about the job you perform in the CIS. Each item
is designed to tap various attitudes you may have toward speciﬁc responsibilities that are associated with
your job. We would like you to respond by indicating how much you agree or disagree with each
statement on a scale of 0 to 10 where 0 indicates "total disagreement" and 10 indicates "total
agreement.” Please answer each question as honestly as possible; as always, all responses will remain

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

strictly conﬁdential.
’ *
Role Conﬂict Within your Across the
ofﬁce CIS network
12. 1 have to do tasks that should be done differently. 1;;5381
13. I receive an assignment without the manpower to complete it” 241;}?!
14. Ihavetobuckaruleorpolicyinordertoearryoutan M=3-0
assignment." SD = 2.7
15. I work with two or more groups who operate quite differently." ISJD=-6392
16. I receive incompatible requests from two or more people. ISdD=-4330
l7. Idothingsthatareapttobeacceptedbyonepersonandnot M=_5-3
accepted by others. SD - 3.1
18. I meive an assignment without adequate resources and M = 4-9
materialstoexecuteit" 39:11
M = 3 9
19. I k . ‘
wor onunnecessarytasks SD= 3.0
20. I often ﬁnd conﬂicting rules and procedures for my job duties. $1241
21. My supervisor and colleagues have contradictory expectations of gdszfo
me. .
22. There are important differences between what I expect as part of ngi‘ézo
my job responsibilities and what others expect — .
23. I personally disagree with some of the tasks that are expected of SAD—:3340
me as a result of my position.“ - .
24. My job description is not compatible with what I feel should be M =__2.9
doing.” SD - 2.5
25. I do not have enough resources to complete my job gapiséoz
responsibilities." - -
* Items 12-19 are from Rizzo et al. (1970).
u

Items were removed from analysis.

38

 

CIS INNOVATION SURVEY

Directions: The following statements deal with your opinions about the climate for
intervention strategies within the CIS nationally. Intervention strategies are initiatives
that relate to the development or implementation of new methods for reaching target
audiences such as counseling protocols for special target populations, targeted outreach
activities using the telephone, responses to calls associated with communication
campaigns, and other national initiatives like those developed by this program project.

We would like you to respond by indicating how much you agree or disagree with each
statement on a scale of O to 10 where 0 indicates “total disagreement” and 10 indicates
“total agreement.” Please answer these questions for the CIS as a whole, rather than
focusing on purely local concerns.

 

 

 

 

 

 

 

 

F ormalization

l. The policy and procedures manual for the CIS covers what happens in §4D==6£07
a typical day.

2. Policies and procedures are strictly enforced at CIS. ISdD==72(.)(‘)52

3. I follow established procedures exactly. 2413:63‘0

4. My job requires that I comply with set procedures 2413:7533

5. My job description is similar to what I actually do when I am at work. 213:2?39

 

39

 

LIST OF REFERENCES

REFERENCES

Batlis, N. C. (1980). Dimensions of role conﬂict and relationships with individual
outcomes. Perceptual and Motor Skills, 51, 179-185.

Brief, A P., & Aldag, R. J. (1976). Correlates of role indices. Journal of Applied
Paychology, Q1, 468-472.

Burt, R. S. ( 1991). STRUCTURE reference manual.

Coverman, S. (1989). Role overload, role conﬂict, and stress: Addressing
consequences of multiple role demands. Social F orcea, 92, 965-983.

Fisher, C. D., & Gitelson, R (1983). A meta-analysis of the correlates of role
conﬂict and ambiguity. Jouanal of Applied Psychology, Q8, 320-333.

Greene, C. N. (1978). Identiﬁcation modes of professionals: Relationships with
formalization, role strain, and alienation.” Academy of Management Jomﬂ, 2_1, 486-492.

House, R. J ., & Rizzo, J. R. (1972). Role conﬂict and ambiguity as critical
variables in a model of organizational behavior. 913m tional Behavior and Him
Performance, 1, 467-505.

Hunter, 1., & Gerbing, D. (1982). Unidirnensional measurement, second order
factor analysis, and causal models. Research in Organizatignal Behavior, 5, 267-320.

Hunter & Hamilton, 1995. Path Analysis computer program. Michigan State
University: version, August 24, 1995.

Jablin, F. (1987). Formal organizational structure. In F. M. Jablin, L. L. Putnam,
K. H. Roberts, and L. W. Porter (Eds), Handbook of prgan_lza' tional communication: An

W (pp. 389-419)- Newbury Park, CA: Sage.

Jackson, S. E., & Schuler, R S. (1985). A meta-analysis and conceptual critique of
research on role ambiguity and role conﬂict in work settings. Organizational Bahavior and
Human Decis_ion Processea, 3_6_, 16-78.

Johnson, J. D. (1993). Organizational communication structure. New Jersey:

Ablex Publishing.

40

41

Johnson, J. D., Meyer, M, Chang, H-J., Ethington, C., Pobocik, S., & LaFrance,
B. H. (1995). Technical rapgrt #2. Analysis of the CIS/Program Project Network for the
period April 1994 to September 1994. East Lansing, MI: Department of Communication,
Michigan State University.

Johnson, J. D., Meyer, M., Berkowitz, J., Ethington, C. & Miller, V. (1995). A
test of two contrasting models of perceived organizational innovgm'veness in j contractual
network of health services organizations. Manuscript submitted for possible publication.

Kahn, R L., Wolfe, D. M., Quinn, R. P., Snoek, J. D., & Rosenthal, R. A (1964).
Organizational stregs; Studies in role conﬂict and ambiggy’ . New York: Wiley.

Katz, D., & Kahn, R. L. (1966). The sw'al paychology of organizations. New
York: “Wiley

Kelloway, E. K., & Barling, J. (1990). Item content versus item wording:

Disentangling role conﬂict and role ambiguity. Journal of Applied ngchology, 2;, 73 8-
742.

Kemery, E. R. (1991). Aﬂ‘ective disposition, role stress, and job withdrawal.
Perrewe, P. L. (Ed). Handbook of job stress [Special Issue]. Journal of Social Bahavior
Lndl’eriomlini, 6, 331-347.

King, L. A., & King, D. W. (1990). Role conﬂict and role ambiguity: A critical
assessment of construct validity. Pychological Bulletin. 107. 48-64.

Lichtrnan, C. M., & Hunt, R G. (1971). Personality and organization theory: A
review of some conceptual literature. ngchological Bulletin, 1g, 271-294.

Linton, R. (1936). The study of man_. New York: Appleton-Century.

Marsden, P. V., Cook, C. R, & Knoke, D. (1994). Measuring organizational
structures and environments. American Behavioral Scientiat, 37, 891-911.

Merton, R K. (1957). Social theogz and social structure. Glencoe, IL: Free Press
of Glencoe.

Morra, M. E., Bettinghaus, E. P., & Marcus, A C. (1993a). The ﬁrst 15 years:
What has been learned about the Cancer Information Service and the implications for the

future. Monographs: National C_a_ncer Institute, 15, 177-185.

Morra, M. E., Van Nevel, J. P., Nealon, E. O., et al. (1993b). History of the
Cancer Information Service. Monoggaphs: National Cancer Institute. L4, 7-33.

42

Morris, J. H., Steers, & Koch, J. L. (1979). Impacts of role perceptions on
organizational commitment, job involvement, and psychosomatic illness among three
vocational groupings. Journal of Vocational Behavior 141, 88-101.

 

Mossholder, K. W., Bedeian, A. G., & Armenakis, A A. (1981). Role
perceptions, satisfaction, and performance: Moderating effects of self-esteem and level.
Organizational Behavior and Human Performance A, 224-234.

 

Nicholson, Jr., P. J., & Goh, S. C. (1983). The relationship of organization
structure and interpersonal attitudes to role conﬂict and ambiguity in different work
environments. Agdemy of Management Journal, g6, 148-155.

Newcomb, T. S. (1950). Sgial pgchology. New York: Dryden.

O'Driscoll, M. P., Ilgen, D. R, & Hildreth, K. (1992). Time devoted to job and
off-job activities, interrole conﬂict, and affective experiences. Journal pf Applig
Psychology, 21, 272-279.

Organ, D. W., & Greene, C. N. (1981). The effects of formalization on
professional involvement: A compensatory process approach. Administrative Science

Qlaarterly, L6,, 237-252.

Pierce, J. L., & Delbecq, A. L. (1977). Organization structure, individual attitudes
and innovation. Academy of Management Review. _2_, 27-3 7.

Rizzo, J. R, House, R. J., & Lirtzman, S. I. (1970). Role conﬂict and ambiguity in
complex organizations. Administrative Sciapg Merly, Q, 150-163.

Sawyer, J. E. (1992). Goal and process clarity: Speciﬁcation of multiple constructs
of role ambiguity and a structural equation model of their antecedents and consequences.

Journal of Applied P§ychology, ]_7_, 130-142.

Schuler, R. S. (1977). Role perceptions, satisfaction and performance moderated
by organization level and participation in decision making. Academy of Myagemmt
Jgumal, 20, 159-165.

Schuler, R S., Aldag, R J., & Brief, A P. (1977). Role conﬂict and ambiguity: A
scale analysis. Organizational Behavior and Human Performance, _2_Q, 111-128.

Shlbutani, T. Spaie_ty aad pmonality. Englewood Cliﬂ’s, N. J .: Prentice-Hall.

Siegall, M. (1992). Some effects of role conﬂict source on the experience of role
conﬂict. Jgumal of Qplied Social Paychology, 22, 628-637.

 

43

Van Sell, M. Brief, A D., & Schuler, R S. (1981). Role conﬂict and role
ambiguity: An integration of the literature and directions for future research. Human
Relations 141, 43-72.

 

.445. -

"’l'lilllllllllllilllllllf