IIILTILLILILILL

 

LIL LID:

             

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

                    

 

MSU

LIBRARIES

 

RETURNING MATERIALS:
PIace in book drop to
remove this checkout from

 

 

your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

4%Qrfi- HUI-06v
' QM igip
Wives x W293 5

 

m4? ‘ DUO LL “-1)

356%"

' 4‘2““; .916
@3361; rm t».

 

 

 

AN INVESTIGATION FOR A MULTIDISCIPLINARY SOCIAL NETWORK
WITHIN AN ACADEMIC SETTING

By

Ronald J. Stump

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Educational Administration

1985

ABSTRACT

AN INVESTIGATION FOR A MULTIDISCIPLINARY SOCIAL NETWORK
WITHIN AN ACADEMIC SETTING

By
Ronald J. Stump

Blau, in The Organization of Academic Work (l973), postulated that

 

the formation of integrative Inultidisciplinary social networks \Nithin

University communities is highly unlikely. He suggests that the high

degree of departmental specialization makes communication between

different academic fields difficult. Friedkin (1978), however, identified
a multidisciplinary social network within the physical science

departments of a major research institution and concluded that

"...settings may exist in science where a fairly high degree of

multidisciplinary integration is systematically fostered."

The present study examined Blau's postulation against Friedkin's
conclusion by investigating for a multidisciplinary social network of the
personnel involved with the energy education projects at Michigan State
University in l981. The investigation attempted to determine if
multidisciplinary integration can exist in this research setting. The
questions for the investigation were developed after a selective review
of the literature concerning communication network concepts and
procedures, as well as organizational structure and multidisciplinary

interaction within the academic setting. The literature supported the

Ronald ll. Stump
appropriateness and the value of using communication network concepts and
procedures to identify multidisciplinary integration.

The participants were asked in this study to indicate their
frequency of energy education related communication with other energy
education personnel in terms of Berlo's three functions of communication:
production, innovation and maintenance. The data collected was then
analyzed by NEGOPY, a computer program developed by Richards (T975) which
generates the communication structural properties of a given population's
communication patterns. The analysis (Hi the communication structural
properties iin‘ the energy education personnel's communication patterns
identified nmltidisciplinary social networks. Network structures were
found to exist related to the content of the energy education research
communication shared by the personnel involved with the projects. The
membership in the communication network linkages and cliques were
determined not to be restricted by formal departmental structure.

The finding of’ multidisciplinary integration supports Friedkin's
conclusion that Blau's postulation cannot be generalized to all-academic
institutions. The implication is that through additional similar
studies, a generalization may be built which states multidisciplinary
integration is likely to exist at research-oriented institutions, a

conclusion which substantially limits Blau's postulation.

This dissertation is dedicated to my wife, Janet Stump, for her
love and support of' myself and our children, Michelle and Heidi.
The dissertation is also dedicated to my parents, Robert and Gladys
Stump, 'hi acknowledgment (H: the love, dedication, and guidance they

have provided for their children.

ii

ACKNOWLEDGMENTS

The successful completion of a doctoral program requires the
support of many people. The following comments are intended to
recognize the individuals who played such a role in my program.

First, Dr. Eldon Nonnamaker served as chairman of my committee and
program advisor. Eldon gave his time and attention to me whenever it
was needed.

Dr. Michael Moore, committee member and major professor in my
cognate area, challenged and encouraged me academically and was an
on-going source of assistance. His commitment and belief in me will
always be appreciated.

The time and assistance provided by my other committee members,
Dr. Howard Hickey and Dr. Kay White, is also sincerely appreciated.

Cathy Helms provided me with significant assistance in focusing my
study and developing the methodology. I thank Cathy as well for her
patience and good humor in working with me.

The typing of the drafts and the final copy of the dissertation
was done by Sharon Kukuk. Sharon's skills, commitment, and interest
for the endeavor were an important contribution to the dissertation's
completion.

I wish to express gratitude as well for the encouragement,

assistance, and understanding by my family, friends, colleagues, and

supervisors during this period of study. Each in their own way made a
contribution to my success.

Finally, special recognition is given to the sacrifices and
patience provided to me by my wife, Jan, and children, Michelle and
Heidi. The completion of the degree is an acknowledgement of their

support, as much as, any personal success.

iv

TABLE OF CONTENTS

Chapter Page
I. PURPOSE AND ORGANIZATION . . . . . . . . . 1

Study Overview

Introduction .

Statement of the Problem . .

Significance and Purpose of the Study.

Limitations of the Study. . .

Background of Energy Education as a Research Problem
at Michigan State University . . . . . . 12

Organization of the Study. . . . . . . . 15

t—Il—a
I—‘OCDr—‘t-J

II. REVIEW OF RELATED LITERATURE . . . . . . . . 16

Introduction . . . . . . . . 16
Multidisciplinary Integration. . . . . . . 17
Networks . . . . . . . . . . . 22
Clarifications . . . . . . . . . 24
Historical Background . . . . . . . 27
Network Analysis . . . . . . . 35
Communication Structure and Content . . . . . 37
Communication Content . . . . . . . 38
Communication Structure . . . . . . . 40

Problem Hypotheses . . . . . . . . . 45

III. METHODOLOGY. . . . . . . . . . . . . 50

Introduction . . . . . . . . . . . 50
Assumptions . . . . . . . . . . . 51
Research Setting . . . . . . . 51
Network Analysis Methodology . . . . . . . 52
Sampling. . . . . . . . . . 52
Measurement . . . . . . . . . 53
Boundary Specification . . . . . . . 57
Analyzing the Data . . . . . . . 6O
Methodology in Relation to Purpose . . . . . 64
Study Population . . . . . . . 66
Data Collection . . . . . . . . . . 7O
Questionnaire . . . . . . . . . 7O

IV. ANALYSIS OF THE DATA AND RESULTS
Introduction .
Description of the Study Population .
Analysis of the Communication Relationships
Hypotheses .

V. SUMMARY, CONCLUSIONS AND RECOMMENDED RESEARCH .
Introduction .
Conclusions . .
Recommended Research .

APPENDICES

Appendix A: Glossary of Terms

Appendix B: Instructions and Network Analysis
Questionnaire

REFERENCES .

vi

73
73
85
93
98
98
101

104
107

117

Table

2-1
4-1
4-2
4-3

4-5

LIST OF TABLES

Who to Whom Communication Matrix.
Members of the Energy Education Study Population.
Departments Represented in the Study Population .

Colleges, Divisions, and Programs Represented
in the Study Population .

Study Population' 5 Reported Type of Involvement
in Energy Education Activities .

Comparison of Network Properties by Communication
Function. . . . . . . . .

vii

Page
33
74
79

80

80

82

Figure

2—1

2-2

2-3

4-1

4-3

LIST OF FIGURES

Sociogram of three communication cliques joined
by a liaison. . . . .

Illustration of a network and selected communication
roles . . . . . . . . . .

An example of high and low degrees of density
within cliques . .

Figure representing the energy education network

for production content and for all-function content .

Figure representing the energy education network for
innovation content

Figure representing the energy education network for
maintenance content . . . .

viii

Page

29

34

43

86

88

90

CHAPTER I
PURPOSE AND ORGANIZATION

Study Overview

 

This study investigates the existence of a multidisciplinary
social network of the energy education personnel at Michigan State
University in l98l. The study's investigation tests Blau's (l973)
postulation that; a multidisciplinary network is unlikely against
Friedkin's (T978) finding of such a network. Communication network
concepts and analysis techniques are utilized in conducting the
investigation. The results of the study are intended to add to the
knowledge regarding the relationship of college and university

structures with academic endeavors.

Introduction

 

An organization is intended to bring people together in a coor-
dinated way to accomplish a defined purpose over a period of time
(Porter, Lawler and Hackman, l975). The organization's policies,
procedures, and administrative structure provide order and direction
for the members to contribute their particular skill and/or knowledge
toward the realization of a desired purpose. Academic institutions
like other public and private enterprises are organizations which
require a complex administrative apparatus to organize and coordinate
the academic work of faculty, staff, and students.

Although the organizational structure may appear similar, a

college or university has only slight resemblance to the general

1

conceptions of business or governmental organizations (Millet, l962).
Contrary to the typical structures of authority and communication
found in most business or governmental agencies, the academic
structure» has significant. authority' and autonomy invested in its
academic departments. The departments are characterized by their
loose connection, separateness, and decision-making power (Woodburne,
l958).

The unique nature of the organizational structure of academic
institutions has been studied and discussed by numerous authors
(Dressel and Reichard, l970; Blau, l973; Perkins, T973; and
Mintzberg, l979). However, Beyer and Lodahl (1976) in their review
of the literature could not find a consensus on how universities
should be viewed as formal organizations. The characteristics of a
bureaucracy are found in the university administrative structure, but
the level of decentralized control and authority found in the
academic department; is not consistent. with the traditional
bureaucratic model.

Mintzberg (l979), however, offers an enlightening view of the
university structure. In his discussion of organizational
structures, Mintzberg describes the decentralized bureaucracy of a
university as a "professional bureaucracy." Such organizational
structures are found by Mintzberg to exist where the environment is
both complex and stable. An environment in which extensively
educated professionals are required for complex work problems, but a
stable enough environment to allow for an expected level of

performance. The organizations which have this particular

environment are law firms, hospitals, social work agencies as well as
universities.

The structure of the organizations is decentralized because the
work is too complex to be directly supervised (Beyer and Lodahl,
l976). The actual coordination of the work is based on the standard
results expected of a professional in a given situation. The profes-
sional is given autonomy to carry out the work and a certain level of
results are expected of the professional by the organization
(Mintzberg, 1979).

The autonomy is supported as well by the fact that the training
and indoctrination of the professional is actually more of a factor
in setting the performance level than is the organization (Mintzberg,
l979). The professional's education process begins with formal
schooling and is followed by internships and certification
examinations to enter the field. The professional association then
expects the professional to continually upgrade his/her skills, to.
stay abreast of new concepts, and to generate new knowledge. As a
result, the professional is more often directed by his/her
educational background and colleagues than by the organization within
which the professional works.

Although the professional bureaucracy is an effective
organizational response to a complex work environment, Mintzberg
(l979) identifies major problems which arise within the professional
bureaucracy caused by the professional's autonomy. The professional
bureaucracy has difficulty in contending with professionals who are
either incompetent or not loyal to the organization. Supervision and

modification of a professional's skills are opposed by professions as

an unwarranted interference. Regarding the latter problem, some
professionals see the organization as a place to work within one's
expertise and not a place to be involved with broader organizational
concerns. Professionals often resist cooperating with and depending
on other professionals as it may interfere with their own work
success.

The strong and insular academic department as a professional
bureaucracy is a relatively new and unique phenomenon to higher
education in the United States. Until early in this century, the
college faculty as a whole regularly met and decided upon academic
policy and personnel matters for the school. The traditional orga-
nizational structure was slowly eroded, however, by the growth in
size of universities, the rapid increase in knowledge, the growth of
graduate study and research activities, and the gradual
departmentalization of colleges and universities by distinctive
disciplines (Dressel and Reichard, l970; and Perkins, T973).

This departmentalization was not pre-conceived as a well-defined
system to manage the growing university (Dressel and Reichard, l970).
Rather, it was caused by the need for flexibility to carry out the
diverse and changing functions of the university, and in particular,
the increasing dimensions of research and knowledge which affected a
reciprocal specialization by faculty members (Veysey, 1965). A
scholar was no longer a botanist, but a botanist specializing in such
areas as plant genetics, cellular structure, or taxonomy. The
advancing knowledge and specialization prohibited a unified faculty
reasonably aware of the university's major disciplines and the

activities of their colleagues. Decisions on curricula and research,

as well as, employment and promotion, could now only be made by those
competent in the field. As a result, the department system brought
together faculty of common disciplines in order-for them to pursue
their interests and to make decisions which a total faculty was
incapable of making (Millet, l962).

Although the authority and autonomy of the department has come
as a logical result of size, specialization, and the necessity to
delegate decision-making, Dressel and Reichard (l970) cite several
criticisms of departmentalization in their review of selected
literature on the advantages and disadvantages of the departmental
organization. They find departmentalization criticized because it
erodes university unity, central planning, and shared communication
and because departments have become political and social forces for
their own means. These criticisms are supported by Mintzberg's
(l979) citation of the problems found within the professional
bureaucracy.

Even in light of the disadvantages of departmentalization, the
dynamics of higher education are seen by Tucker (198T) to require an
academic structure of departmentalization by discipline. From his
study of the literature and his analysis of the Florida state
university system, Tucker feels that there is "...no better way to
organize colleges and universities than through academic departments
based on recognized disciplines. Almost all efforts to organize
colleges and universities differently have failed, particularly those
efforts in public institutions...."

The actual effect of the departmental structure on the overall

operation of the university or college, however, has not been

significantly researched. In fact, writers in the 1960's (Moran,
1968; Bolman, T965; and McConnell, 1963) were calling for' more
research in the administration of higher education. McConnell (1963)
cited so little research being done on how colleges and universities
were organized and administered that the conceptual framework was
lacking for the formulation of an investigative search. Bolman
(1965) recommended investigations of the rigidity and communication
problems of the department and college organization. Moran (1968)
suggested studies on the effect of new scientific problems on
academic structure.

The call for this research was answered by Peter Blau's seminal

study in 1973. His book, The Organization of Academic Work (1973),

 

provided the conceptual framework for further research of the
university organization. In his study, Blau examined how academic
performance at colleges and universities is affected by the structur-
ing of human resources. In particular, he studied the influence of
administrative structure on several factors which have indirect
bearing on the teaching and research orientation of faculty. These
factors included the size of the institution, decentralization of
authority, faculty social integration, and patterns of faculty
communication.

Blau's study is based on his investigation of the administrative
structures and academic environments of 115 American universities and
colleges. He examined their operations through surveys, on-campus
interviews, and review of their organizational materials. One of
Blau's findings from his investigation was concerned with the

relationship of academic departmentalization and multidisciplinary

integration. Because of the independence in which faculty can pursue
their academic professions and because the highly specialized nature
of the disciplines makes communication difficult between the fields,
Blau concluded that. academic departments are centers of" academic
networks, but multidisciplinary integration is unlikely in the
academic setting.

Blau perceived the unlikely existence of multidisciplinary inte-
gration in the academic setting as an important issue for colleges
and universities. The collegial atmosphere is intended to be an open
and sharing environment of knowledge and ideas. If
departmentalization and specialization have a deleterious effect on
the academic environment, there is cause for concern for the
effective advancement of academic pursuits.

Nearly all studies related to academic integration have been
supportive of Blau's conclusion that integration exists within the
academic department, but nmltidisciplinary integration is Luflikely
(Breiger, 1976; Crane, 1969 and 1972; Mulkay, Gilbert, and Wolgar,
1975; and Mullins, 1972). However, the studies supportive of Blau
focused on single disciplines and not on research being conducted by
faculty in different disciplines. Freidkin (1978) studied the
research communication of 128 faculty in six physical science
departments of an American university, which he described as an elite
research institution.- Acquiring data gathered by a questionnaire on
their communication behavior, Friedkin found the faculty to be linked
in a single network of research communication. Each of the faculty
had one or more communication paths to each of the other faculty

members. In addition, interdepartmental cliques were found to exist

which had densities higher than that which existed within the
individual departments. Friedkin did not find specialization to
prohibit cohesion between the departments. Rather, he found linkages
and cliques which were independent of departments. Friedkin
suggested that multidisciplinary integration may exist in higher

education, particularly in research oriented institutions.

Statement of the Problem

 

Friedkin's study suggests possible limitations of Blau's
postulation that integrative multidisciplinary social networks
within the academic setting is unlikely. The conclusion of
Friedkin's study is that integrative multidisciplinary social
networks can exist at least within the physical sciences faculty
of a research institution.

The intent of this study is to further assess Blau's
postulation by conducting an investigation for a multidisciplinary
social network within the departments of a large land-grant
university. The personnel composition of the social network
investigated, as well as the setting, will differ from the
previous studies which have tested Blau's postulation. Whereas
the other studies sought social networks within physical science
departments by asking only the physical science faculty who they
communicated with related to academic research, this study' will
seek to identify a nmltidisciplinary social network of all faculty
and staff involved in a recent research problem area, energy
education. The study also alters the focus of previous studies

from the social network of general research within selected

academic departments to £1 social network of’ a specific research
problem.

If the social network indeed exists, the study will also
investigate the content and the structure of the network members'
communication. The intent of this investigation is to analyse the
members' communication patterns in relation to the network itself
and with their department. The level of integration of the
departments can be determined by this comparison. The
communication content to be investigated fbcuses (n1 Berlo's (1970)
three types of communication--production, maintenance, and
innovation. The level of involvement will be neasured by
identifying each participant's network position and amount of
communication within the network.

The investigation for the social network. will use a
communication network analysis. This method utilizes a survey
process and quantitative analysis to identify members of a network
and the structure of their communication. The results of the
communication network analysis identifies networks by communica-
tion patterns and the amount and type of involvement individuals
have within the network.

The hypotheses for the study assume the existence of the
social network(s) including communication content and cliques
within the energy education personnel. The communication network
analysis investigates for the validity' of 'these hypotheses. The
specific problem hypotheses are presented in the conclusion of the
second chapter. The hypotheses are written in the second chapter

because they are based on the literature review. A better

10

understanding of the hypotheses can be gained by examining them in
the context of their supporting literature.

In summary, this study uses 51 communication network analysis
to investigate the existence of a social network of all faculty
and staff involved in energy education projects at a major
land-grant institution. Provided the network exists, the study
further seeks to identify the communication function and the
communication structure of the network. The intent of this
investigation is to test Blau's (l973) postulation that
multidisciplinary integration is unlikely within academic settings
against Friedkin's (1978) finding that multidisciplinary inte-

gration is possible under certain circumstances.

Significance and Purpose of Study

 

The significance of the present study is that it attempts to add
to the research on the administration of higher education. The need
for this research was cited earlier in the chapter. Bolman (1965),
McConnell (1963), and Moran (1968) decried the lack of investigative
research on the administration of colleges and universities. Of
particular importance to them was research on communication problems
between departments and the relation of new scientific problems with
academic structure. The concern about communication between academic
departments was expressed also by Dressel and Reichard (1970). They
criticize the lack of interdependence between academic departments.
Blau (l973) responded to these concerns in his investigation of the
effect of college and university structures on academic pursuits.

One of his areas of study was interdepartmental communication in an

academic setting. Blau described the existence of a
multidisciplinary social network to be unlikely in a college setting.
His postulation was generalized from a study of 115 American
colleges. Blau's generalization is challenged, however, by
Friedkin's (1978) finding of multidisciplinary integration within the
physical science departments at a major research institution.

The present study specifically seeks to identify and analyse the
existence of a multidisciplinary social network at a major land-grant
institution. The study is intended to test the validity of Blau's
claim and extend the» meaning of Friedkin's finding by examining
whether multidisciplinary integration can be found within the
academic personnel responding to a research problem at a single

academic setting. In writing The Organization of Academic Work, Blau

 

h0ped that his generalizations would be treated as a frame of
reference for further research on the administration of colleges and
universities. This study responds to his work and is intended to
further explain how the structure of universities affects academic

performance.

Limitations of the Study

 

The individual characteristics of the study may not be
applicable to other campuses or to future similar studies of the same
population. Such characteristics as network properties and type and
level of faculty involvement may not be consistent in other research
settings because of the dynamics inherent to college campuses and to
the field of research. The forces influencing the dynamics of this

type of study include the continual change in research activities,

12

personnel, and client needs. It is also quite possible because of
these continual changes and the relative autonomy existing within
academe, that not all of the potential study population were
identified and/or analyzed in the current study.

The study conclusions, however, have general applicability for
future research and for consideration in higher education
administration. The strength of the study's general application is
in the conventional setting in which the study was conducted and in
the utility' of’ the' methodology; The study site and the faculty
research behavior studied are not idiosyncratic, but common at many
other campuses. The methodology is a standard process which uses a
computer program for data analysis. Because the computer program is
based on algorithms the methodology, as well as, the overall study is
replicable for future similar studies.

Background of Energy Education as a Research Problem
at Michigan State University

 

 

Beginning with the 1973-74 international oil crisis, the
economic and social life style of the American society was disrupted
and drastically affected by petroleun1 scarcity' and rising costs.
Americans faced energy constraints which affected their living
standards, particularly for heating, cooling, and transportation.
Moreover, these constraints were seen by scientists not to be
temporary but expected to last into the next century and beyond.

The energy crisis was caused by several factors not the least of
which included the high energy consumption which had become a way of
life in the United States, the dependence by the United States on

importation of over fifty percent of its oil supply, and the regular

13

petroleum price raises by the Organization of Petroleum Exporting
Countries (OPEC). Regarding this latter factor, the United States
paid $90 billion for petroleum in 1980, up from $60 billion in 1979 -
even though Americans used substantially less oil (Bernstein, 1980).
Exacerbating the situation was the unsettled political climate of the
Mideast, the major supplier of United States imported oil. The flow
of oil for export was often impeded due to wars, changes in political
leaders, and violations of agreed upon oil exportation policies.

Because it was doubtful that the United States would ever again
enjoy an era of cheap and abundant energy, the solution to the energy
crisis was felt to be through both the development of alternative
fuel supplies and energy conservation practices. These solutions
have not, however, been easily and quickly developed and implemented.
A new American energy ethic, as well as energy technology, has been
difficult to develop after years of excessive use of inexpensively
priced fuel supplies.

In the mid-1970's, institutions of higher education began
fostering a new energy value system and technology. Colleges and
universities gathered energy data and information, researched new
energy concepts, and began training today's and tomorrow's leaders
and technicians in energy concepts. Michigan State University
responded to the energy education problenl in character' with its
land-grant mission. The faculty and staff initiated a large number
of energy research and energy education projects both off-campus and
on-campus in such areas as agriculture, education, administration,

and community development. These projects were in response to

14

research ideas of personal and institutional origin, as well as, from
research grants from Federal, State, and private funding sources.

Hagstrom's description (1965) of how scientists choose research
problems provides additional insight to the manner in which Michigan
State University and other universities responded to the energy
crisis. Typically, problems for investigation are given to
scientists or identified by the scientists themselves. The
motivation for working on the problems includes the challenge of
discovery, the potential for recognition, and the funding
availability. If the problem is new and the solution(s) will be
highly valued, an important dimension is added to the research
problem. In these cases, many scientists including the best ones are
attracted to the problem. The scientific effort during World War II
and the "race for space" after the Russian satellite Sputnik orbited
the Earth are examples of these special problems. The energy crisis
offered these same factors of major national importance, a new area
for investigation, and significant discoveries to be made. The large
number of energy education projects at Michigan State University and
at other universities is very understandable in light of the impor-
tance of the energy crisis problem.

Typically, applied research into problems like the energy crisis
tends to be interdisciplinary (Hagstrom, 1965). Collaboration
between scientists from different disciplines is often necessary to
solve practical problems. Temporary interdisciplinary groups form to
work on the problems, or projects related to the problem, and
dissolve as the projects are completed, or the problem is solved.

This study will investigate the existence of such interdisciplinary

integration within the energy education personnel working on the
energy crisis problem at Michigan State University in 1981. Blau has

postulated that the existence of such integration is unlikely.

Organization of the Study

 

Chapter I consists of an overview, an introduction concerning
studies of academic departments, the problem statement, the purpose
and significance of the study, limitations of the study, the
background of the problem setting, and the organization of the study.

Chapter II is a review of selected literature related to the
purpose of the study. The review expands on the Blau and Friedkin
studies and overviews the history, types, and functions of
communication networks, as well as, network analysis. A statement of
the problem hypotheses is given at the conclusion of the review.

The methodology of the study is described in Chapter III. The
identificathm1 of the population, the method of investigation, the
research setting, and the treatment of the data is explained in this
chapter.

Chapter IV contains the results of the data analysis, as well
as, the evaluation of the hypotheses.

Chapter V briefly summarizes the study, offers conclusions with
a discussion of their implication for theory and practice in relation
to Blau's postulation and Friedkin's finding, and completes the

report of the study with suggestions for additional research.

CHAPTER II
REVIEW OF RELATED LITERATURE

Introduction

 

The present study is a network analysis of the communication
patterns of the 1981 energy education personnel at Michigan State
University. Based on the results of the analysis, the study seeks to
identify the existence of a multidisciplinary social network within
this setting. The presence of such a social network would be
unlikely according to Blau's (l973) postulation on the
multidisciplinary integration of faculty. A postulation which was
challenged by Friedkin's (1978) study which found a relatively
integrated social network within the physical science departments at
a major research institution. The present study tests Blau's
postulation as it applies to a selected all-campus research problem
at a major land-grant institution. In this manner, the study seeks
to enhance the understanding of the level of integration of academic
specialization within university environments.

A review of selected literature related to the study is
presented in this chapter. The background, methodology, and findings
of Blau and Friedkin, as well as, the concepts and research on
networks and communication network structures, content, and analysis
are presented in order to provide a foundation for the study's
purpose and unethodology. The chapter concludes with the problem

hypotheses.

16

17

Multidisciplinary Integration

 

In The Organization of Academic Work (1973), Peter Blau

 

attempted to answer’ questions about the influence of the
administrative structure of colleges and universities on academic
pursuits. He looked at such issues as the conditions necessary for
academic innovation, the dependence of faculty research on collegial
climate, the relationship of faculty authority to organizational
structure, and the characteristics of the academic stratification
system. Blau's work on these issues was intended to provide insight
and suggestions which would be expanded upon and used to enhance the
pursuit of knowledge in the higher education setting.

Blau's study began in 1968 when he collected data on 115 univer-
sities and colleges within the United States. These schools were a
sample of all four year institutions granting liberal arts degrees in
the mid-1960's. Specialty colleges such as junior colleges and
teachers colleges were excluded from the study. The data were
collected from three sources. First, the top administrators at all
115 schools were visited and interviewed regarding their
administrative structures. In these visits, the researchers met with
the president and/or the chief academic officers and discussed such
topics as organizational structure, how authority was shared, and the
creation or elimination of academic departments. The second source
was publications and internal records. The academic institutions'
own qualitative information was reviewed, as well as the tenth

edition of the American Council on Education's American Universities

 

and Colleges and the 1968 survey information of’ college faculty

 

18

conducted by the American Association of University Professors. The
final source was a survey of 2,577 of the investigated schools'
faculty members which was conducted by the Survey Research Center at
the University of Michigan in 1967. All but one school participated
in the survey for an average of 23 participants per institution.

Blau utilized regression analysis techniques on his data to find
the conditions which influenced selected characteristics of academic
institutions. The procedure decomposed simple correlations between
independent and dependent variables to show which associations were
spurious or directly effected. From the data analysis, Blau
discussed and offered generalizations about a variety of topics
related to the effect of organizational structure on the academic
process. For example, he presented postulations on the effect of
institutional size on research, the process by which an academic
specialty area becomes a department, and the relation of
differentiated academic specialties with the overall academic en-
vironment. Blau described these research conclusions however as
"highly tenuous and tentative." He felt that not enough research had
taken place in academic settings to "...declare any causal precedence
for the existing social conditions." He offered analyses from his
study, however, as building blocks to encourage further research and
testing toward a theory of the academic organization.

Blau's study of the relation of differentiated academic speci-
alities with the overall academic environment found a positive
relationship of increasing institutional size with increasing
institutional differentiation at declining rates. Blau suggested and

discussed several hypotheses to explain this finding. One of his

19

hypotheses offered the idea that "...differentiation creates the
interdependence necessary’ to integrate large numbers of' employees
into a common enterprise." However, Blau's discussion of this
hypothesis arrived at a negative conclusion on the positive
relationship of differentiation and integration within an academic
institution.

In his discussion, Blau considered as ea dilemma the
specialization required of faculty for scientific progress and the
social relationships faculty have with the colleagues in their own
university. Because academic scholars and scientists have a strong
orientation toward their individual discipline and seek support and
stimulation from the specialists in their discipline, social
integration and networking with other campus faculty essential for an
academic community is lost. In other organizations, specialization
causes an interdependence between departments for either parts,
tools, or resource support related to the product. In an academic
institution, however, specialties are> not. directly interdependent
(Hagstrom, 1965). The members of each academic department can
research and teach without interacting with faculty from other
departments. In addition, the specialization itself causes
difficulty in communication between the faculty of different disci-
plines. Because' of’ the lack of familiarity' with the speciality
discipline, the academic research conducted by the faculty of’ a
department is not readily understood by faculty outside of the
discipline. This apparent inability to exchange work-related
information is exacerbated by Bess' (1982) opinion that faculty

generally lack interest in interpersonal work-relations with their

20

academic colleagues ir1 other departments because (If status
distinctions. Bess contends that faculty tend not to give full
attention to other departments with fewer members and/or with less of
a scholarly reputation than their own.

For these reasons, Blau postulated the formation of academic
social networks to be essentially' within the department and the
formation of nmltidisciplinary social networks 1x1 be difficult and
highly unlikely. The differentiation of universities into
departments based (Ml academic disciplines has been essential for
progress in the specialty disciplines. The differentiation has,
however, caused obstacles to nmltidisciplinary communication about
scholarly matters.

Since Blau's study, most research on the social networks of
faculty focused on the activities of the scientists within a single
discipline (Breiger, 1976; Crane 1969 and 1972; Mulkay, Gilbert, and
Wolgar, 1975; and Mullins, 1972). These studies found scientists to
be connected by a discipline network to researchers at other campuses
and project centers. Price (1963) described this network as an
"invisible college". Through these networks, scientists received
research information and respond to and support the work of other
scientists which is critical to the progress of research. In effect,
social structures have developed around the communication of research
information (Mullins, 1968).

More recently, Friedkin (1978) examined the social network of
scientists in different disciplines at a single institution and found
an integrative social network existing within the multidisciplinary

population. Friedkin studied the pattern of research communication

I‘D
pa

for the faculty of six physical science departments at an American
mid-west university, which he described as an elite research
institution. Using a survey questionnaire listing the names of all
of the faculty in the six physical science departments, Friedkin
asked the faculty to indicate with whom on the list did they have at
least three conversations about research problems during the academic
year. Based on a 58% response, Friedkin found that 128 of the 133
faculty were linked by one or more communication paths into a single
network.

Friedkin (1978) broke the network into sub-networks of direct
one to one contacts or linkages and found the six networks with the
highest degree of direct linking to be multidisciplinary. He also
found 39% of the linkages to be multidisciplinary. Departmental
boundaries were of no effect on the social network patterns for these
physical science disciplines. Friedkin concluded that academic
specialization by departments does not appear to automatically
constrain multidisciplinary social networks. Friedkin suggested that
research oriented universities may be settings where
multidisciplinary integration is fostered.

The present study extends Friedkin's finding that
multidisciplinary integration may be fostered in research oriented
institutions against Blau's postulation that such integration is
unlikely. The study is conducted utilizing network concepts and
communication network analysis techniques. The nature of these
concepts and the basis for their application in this study is

presented in the following sections.

22
Networks

As evidenced by Barnes' (1979) observations and Freeman's (1976)
bibliography of social networks, the concept of network is becoming
increasingly popular among social scientists to describe and
understand social relations. The rising popularity of the network
concept is attributed by Whitten and Wolfe (1973) to its
effectiveness in identifying and studying clusters of informal
relations within a system; an approach which has only recently been
formalized and supported by computer technology.

The term network is similar to the idea of a group; except
"network is distinct from 'group' in that it (network) refers to a
number of individuals (or other units) who persistently interact with
one another in accordance with established patterns" (Rogers and
Agarwala-Rogers, 1976). Thus, the regular interacthm1 of network
members differs from group members who may not consistently interact
with one another in a continual pattern. The established patterns of
a network are based on the participants' communication relationships
and common concerns or interests. For example, a network may ferm
around a common problem or a common attraction to the arts, sports,
or hobbies. Whatever the reason for the network's existence, the
patterns of the members' communication provide definition and
structure to the network.

Sociologists and communication specialists have found networks
to be a rich source for studying behavior.Researchers like Boissevain
(1974), Rogers and Agarwala-Rogers (1976), and Farace (1978) have
found that the behavioral patterns of the network members can be

understood by analysing the characteristics of the network structure

23

and the network members. These behavioral studies are possible
because network analysis identifies the level and type of
relationship each network member has with other network members.
This relationship data is then utilized in investigating whether the
network's members have an affect on other individual network member's
behavior. With an analysis of each participant's involvement level
and personal characteristics, a selected trait or knowledge can be
traced through the network channels and investigated in relation to
the individual's behavior. Indeed, research tn/ Boissevain (1974),
Lin (1973), Rogers (1983), and others have well documented the
influence communication networks can have on the acceptance of new
ideas, practices, and innovations by people in various cultural and
organizational settings.

In addition to their effect on member behavior, communication
networks have been studied for their relationship to the formal orga-
nization in which they exist (Jacobson and Seashore, 1951; and Weiss
and Jacobson, 1955). In this approach, the informal network is iden-
tified within the organization and its structural properties (i.e.
cliques, liaisons, isolates, etc.) are analysed against the formal
structure and its desired communication patterns. These studies not
only establish the existence of an informal network within the formal
organizational structure, they also can be beneficial for identifying
how communication flow can be enhanced within the organization's
system and what role the network has on the organization's goals

(Taylor, 1976).

24

Clarifications

 

Before proceeding into the history and application of network
concepts, it is important to clarify the term network as it is used
in this study. The contributors to the network literature are so
varied in their background and perspective of social behavior that
their discussions on networks can be confusing. Indeed, Barnes
(1972) describes network literature as '%1 terminological jungle in
which any newcomer may plant a tree." Because of their relation to
the purpose and approach of this study, three particular issues are
presented at this point to assist the reader in understanding
networks and the writing perspective.

The first issue to clarify is that the concept of network is not
a theory (Barnes, 1972; Hammer, 1979; and Mitchell, 1974). Rather, as
a concept, network means that certain people are informally
interacting with each other in a regular patterned way. As mentioned
earlier, not all aggregates of people are a network, unless there are
consistent communication patterns. The actual existence of a group
of people as a network must be established empirically through
network analysis, a research approach for investigating and labeling
a particular set of social reactions. Until the communication
relationships are established between an organization's members, it
cannot be assumed that a network exists solely because people and
communication channels are present within an organization
(Boissevain, 1974). In addition, knowledge of' the .structure and
function of communication in one system's network does not

necessari1y predict the behavior and relations in another system's

25

network. Networks are a concept for labeling and describing
behavior, but not for predicting behavior (Burt, 1980).

Second, the term network is used to refer to an aggregate of
people who are regularly interacting with each other as defined
above. However, the term often describes in the literature an
approach for identifying a network and its structural properties and
content by studying and explaining a study population's social
relations. For this paper, network means an aggregate of persons who
by their pattern of’ regular interaction are a network. Network
analysis is the term which will be used to signify a research method
for identifying a network and its communication structure and
function.

The final clarification to be made concerns networks, social
networks, and communication networks. Networks are referred to in
the literature as social networks by sociologists and social
anthropologists, and as communication networks by communication
specialists. Descriptions by Rogers and Kincaid (1981) for
communication networks and by Mitchell (1974) and Boissevain (1974)
for social networks will be presented to explain a subtle difference.
In concept, the terms are very similar; each network description
refers to an aggregrate of people with patterned relationships. The
communication network concept however emphasizes more the
identification and interpretation of communication patterns in
determining a communication network and its properties. The
communication network analysist will begin researching a selected
population of people by identifying communication relationships and

the nature of the relationships. The structure and function of

26

communication will be established from this data and the communica-
tion patterns and content will be studied for their affect on
behavior.

The sociologist defines a social network from a more macro-level
approach and views the social network as slightly apart from the
communication network. For this reason, the sociologist identifies a
social network through focusing on the communication content and the
personal relationships rather ‘than (N1 the communication patterns.
Choosing a behavior to study within a given population, the
sociologist will identify the existence of the behavior in the
network and how the behavior influences and is influenced by the
personal relationships. Using dominance as an example of a behavior
which a sociologist might study within a given population for its
affect on relationships, the sociologist's attention would
concentrate more on the dominance behavior itself than on the
communication roles played by the network members. As a result, a
social network could include individuals who rarely or never share
interpersonal communication (Rogers and Agarwala-Rogers, 1976). In
comparison, the communication network would include only those
individuals who have consistent communication patterns.

A. discussion of the differences between social networks and
communication networks is akin to splitting hairs. Nonetheless, the
presentation establishes a frame of reference from which to
understand the nuances of the network literature, as well as, the
perspective utilized in approaching the study's problem. This study
utilizes the terms network, social network, and communication network

interchangeably. There is no intent to signify any substantial

27

difference between them. In that the purpose of this study is to
investigate the existence of interdisciplinary' communication, the
communication concept of network is the chosen methodological
approach.

To assist the reader' in further- understanding this study, a
glossary of terms is provided in Appendix A. The glossary defines
such terms as cliques, reciprocation, density, and networks; words

which are used often in this report.

Historical Background

 

The foundations for network concepts and methodology originate
in sociology, particularly sociometry, and in the social
psychological small group studies of the 1950's which were performed
in laboratory settings (Farace, Monge, and Russell, 1977). Georg
Simmel is generally recognized as one of the first sociologists to
recognize the effect of network relationships on individual behavior.

In 1922, Simmel published The Web of Group-Affiliations which

 

introduced the concept of informal networks (Rogers and Kincaid,
1981). To understand human behavior, Simmel studied the system in
which the individual participated for interpersonal relationships
that influenced the individual's behavior. Simmel's approach
differed from the predominant concepts of the time which focused on
the individual for behavioral causes. Rogers and Bhowmik (1971)
describe the prevalent methodological approach during this era to be
the study of intrapersonal characteristics, leaving interactive
relationships largely neglected. Although his ideas offered a

significant new insight into behavior, Simmel's work was limited by

28

the lack of £1 satisfactory means to "wasure network relationships
(Rogers and Kincaid, 1981). The measurement technique to collect
quantitative data on these network relationship's was provided in the
mid-1930's by Jacob Moreno's sociometry (Moreno, 1953).

Sociometry is described by Chapin (1940) as a generic term for
the measurement of societal and interpersonal relationships. The
sociometry method determines these relationships by directly
questioning respondents about their social relations (Mouton, Blake,
and Fruchter, 1955). For instance, a respondent may be asked
questions regarding with whom they speak with the most, or with whom
they discuss certain issues. The sociometry method then portrays the
collected data in a sociogram (see Figure 2-1) as a graphic of the
communication patterns or social relationships (Moreno, 1953).

In reviewing the history of sociometry, Nehnevajsa (1955) found
the sociometry method to have had a significant impact on
contemporary behavioral research and theory. Sociometry deeply
influenced the theoretical and conceptual development of group
dynamics, role playing, psychodrama, socialization, and spontaneity.
In addition, Rogers and Kincaid (1981) cite Moreno's research on
communication patterns as a primary source for much of present day
network analysis. Of particular importance 1x1 the development of
these theories and concepts was sociometry's methodology of
systematic data reports in contrast to intuitive observation (White,

Boorman, and Brieger, 1976).

29

Clique A Clique B

   

Clique C

 

Key to Symbols
- Clique member
L - Liaison
-—-——» - Clique link indicating who is communicating with whom
Figure 2-1

Sociogram of three communication cliques joined by a liaison.

Although sociometry has proven to be a useful investigative tool
and has been a foundation for modern measurement techniques, a
drawback to the technique is that it can be applied to a maximum size
network of only 80-100 people. As the study size increases, it
becomes impossibly difficult to draw sociograms. In addition to

being an arduous task, the drawing of the sociogram cannot

3O

necessarily be replicated due to the lack of systematic rules
(Farace, Monge, and Russell, 1977). For these reasons, the
sociometry approach did not maintain significant scholarly interest
beyond the 1940's, although interest in communication networks and
network analysis did not perish (Rogers and Kincaid, 1981).

The interest was maintained by three centers of research which
are often cited in the literature for their role in carrying forward
the studies of' communication networks in the 1950's and 1960's.
Anthropologists in Europe led by Bott (1955), Mitchell (1969), and
Boissevain (1974), studied the effect of communities on individual
behavior. In their studies, which were concentrated on very small
networks often of families and neighborhoods, they found significant
community influence on the behavior of individuals. At approximately
the same time in the early 1950's, research on network analysis of
organizational communication was occurring at the University of
Michigan. Led by Jacobson, Seashore, and Weiss, these researchers
investigated the organization communication patterns of the regular
work-related interpersonal communication between organization members
(Schwartz and Jacobson, 1977). Using the sociometry research
approach, Jacobson and Seashore (1951) and Weiss and Jacobson (1955)
identified the communication cliques and the liaisons linking the
cliques as an informal structure within the organization and compared
the informal structure with the formal structure. From these
comparisons, generalizations were developed on the communication
efficiency of the organization and the relationship between the

informal and formal structures.

31

In addition to advancing the studies in communication networks,
the research conducted at the University of Michigan was significant
for continuing the methodological shift from focusing on individuals
to an analysis of communication relationships (Rogers and
Agarwala-Rogers, 1976). Rather than studying the functions and
characteristics of individuals in organizations, Jacobson, et a1.
studied the whole organizational system for communication patterns of
relationships.

While the studies by Jacobson, et al. were being conducted
within organizations, studies of communication patterns were also
being conducted by Bavelas (1948), Leavitt (1950), and Shaw (1964)
within laboratory small groups. These researchers were concerned
about the relationshna of fixed communication patterns upon group
process. In their laboratory studies, they experimented with various
structural patterns of communication networks on such dependent
variables as group efficiency, member satisfaction, perception of
leadership, and member adaptation (Shaw, 1964). The results of their
experimentations included the conclusion that the greater the
interconnectedness of a network, the greater the member satisfaction,
as well as, the better the group's problem-solving efficiency on less
routine tasks. (Shaw, 1964).

The overall laboratory methodology and results are criticized
however by Rogers and Agarwala-Rogers (1976), Burgess (1968), and
other social scientists. The primary criticism is that studies
replicating their work often concluded with confusing and
contradictory results (Freeman, 1978). In addition, the experimental

results were found to lack reliability outside of, as well as, within

32

the laboratory. This reliability problem may be related to a second
criticism that the laboratory "conditions did not reflect the real
life situations of large organizations" (Rogers and Agarwala-Rogers,
1976). Unaccounted for in the experiments was the influence on
individuals and groups which established integrated networks have
within an organization.

Although the conclusions of the small group laboratory studies
may be limited in value, the contribution of these small-group
studies to network analysis is significant in that they generated
information and interest regarding the structural effects of
communication networks on network members. In particular, Bavelas'
idea of structural centrality (the network participants' degree of
connectedness) provided impetus for many follow-up experiments
(Freeman, 1978).

The results of these network studies conducted at the three re-
search centers did not have an immediate and significant impact on
the study of behavior. The primary focus of the field continued to
be on the analysis of the individual for explanations of behavior
(Rogers and Kincaid, 1981). The traditional approach was supported
by the advent in the 1950's of computers capable of manipulating
large quantities of data on individuals. Similar computer technology
to analyse communication relationships as the unit of study was not
available until the 1970's. Consequently, the design of a sociogram
displaying communication relationships for a network of 100 could
take weeks to design and then not necessarily represent an accurate
picture of the communication patterns. The availability of computer

technology for individual-based data and the lack of the same for

33

relationship-based data resulted in a dormant period for network
studies in the late 1950's and the 1960's (Rogers and Kincaid, 1981).
In the early 1970's, however, several computer methods became
available for the network analysis of large systems. The computer
analysis methods include NEGOPY, CONCOR, SOCK/COMPLT, and CLIQUE
(Rogers and Kincaid, 1981). The common procedure for all of these
computer methods is described by Rogers and Kincaid (1981) as the use
of a Inatrix to order the data of' who is communicating with whom

Table 2-1

Who-to-whom communication matrix.

 

ll Whom"

A] #10 #11 #12 #13 #14 #15 #16 #17

#10 - 2 3 O O O 4 O

#11 2 - 1 0 3 1 2 1

#12 3 1 - 3 2 3 1 0

"Who“ #13 O 0 3 - 4 0 2 1
#14 0 3 2 4 - 4 2 2

#15 O 1 3 0 4 - 1 1

#16 4 2 1 2 2 1 - O

#17 0 1 O 1 2 1 0 -

Note: These data represent who is communicating

with whom and for how many hours per month. For

example, #10 is communicating two hours per month with

#11, three hours per month with #12, four hours per month

with #16 and not at all with #13, #14, #15, and #17.
(see Table 2-1). In the matrix, each participant is listed on the
"who" and "with whom" dimensions. The reported frequencies of dyadic
interaction are then entered into the N x N matrix set and applying
an algorithm, a specific mathematical and procedural criteria, the
computer generates an analysis of the data. The results of each

computer method, however, emphasizes a different set of network

structural properties.

   

34

 

Clique 2
Clique 1
.1
X
y‘ . / x ‘
JO Ta: 1
X 3
Cliques: Group Linkers
Clique 1 - 4,5,6,7,8 Bridges - 5,9
Clique 2 - 9,10,11,12 Liaison - 13
Clique 3 - l4,15,l6,l7,l8 Other - 19
Key to Symbols Isolates
- Clique Isolated Dyad - 2,3
Jf'- Clique Member Isolate Type One - 1
X -Isolate Isolate Type Two - 20
---- - Isolate Link Tree Node - 21
_- Group Link
7 - Tree Node
L - Liaison
o - Other
Figure 2-2.

Illustration of a network and selected communication roles.
[Based on a similar figure by Farace, Monge, and Russell (1980)]

For instance, Farace and Mabee (1978) describe CONCOR as a

group-detection devise and NEGOPY as a means for identifying cliques
and network property measures.

Of most importance is that all of the methods cope with the
primary problem of analysing the large-volumes of data which are
A second advantage of the

found in communication network studies.

computer methods is their standardization of analysis. Because

strict algorithms are assigned to the analysis procedures, the
subjective nature of determining communication structures is removed.
With these methodological breakthroughs, social scientists could now
conduct studies on topics which had previously been unapproachable.
The advent of computer technology in the 1970's sparked a new surge

of interest in network analysis (Rogers and Kincaid, 1981).

Network Analysis

 

Network analysis is a research method for identifying a system's
communication structure and function based on the study of
interpersonal relationships (Rogers and Agarwala-Rogers, 1977). From
the analysis, a systematic picture is composed of the interpersonal
relationships' social structure (Knoke land Kuklinski, 1982). The
social structure's picture is called a network existing within the
system.

A glossary of terms is provided in Appendix A to assist the
reader in better understanding a network and its properties. In
addition, figure 2-2 provides a visual illustration of a network.
Such terms as cliques, liaisons, and isolates are presented in the
figure and in the glossary.

The network analysis may be accomplished at several levels and
from two different frames of reference or approaches (Burt, 1980).
The level of analysis may concentrate on the individual, the system's
sub-groups or cliques, or the total system level (Knoke and
Kuklinski, 1982). The frames of reference for analysing these levels
are described by Burt (1980) as the relational approach and the
positional approach. In a relational approach, the content and the

intensity of each relationship is described for the units analysed.

36

The intensity of the relationship is measured by the frequency of
exchange, the greater the reported frequency' of interaction, the
stronger the relationship. The relationship's content is the
function (Hi the communication shared. The communication functions
could be for work information, new ideas and practices, or
interpersonal support (Berlo, 1970).

The status and role of the units of analysis are described in a
positional approach, the second frame of reference. Again, as in the
relational approach, the position's strength is measured by the frequen-
ey of the exchanges for the unit. The positional approach differs
however from the relational approach in that the unit of analysis is
referenced against all other units of analysis within the system,
whether a relationship exists or not, to determine the unit's role or
status in the system. This approach allows for measures of social
integration and configuration which reveals a specific network struc-
ture. For example, a unit of analysis can be isolated or be a central
figure (e.g. bridge or liaison) in the network based on the frequency
and type of relationships it has with other units.

This dual analysis provides a description of the unit's communica-
tion structural properties and content. Networks can now be generated
based on overall communication or selected functional content to illus-
trate how and where certain types of information are flowing through the
system. The roles and the degrees of relationship can also be defined
for the units of analysis indicating: (1) how connected or integrated
the units are with each other; (2) the units who are bridges, isolates
or liaisons; and (3) the cliques within the total system (Rogers and

Agarwala-Rogers, 1976). With this information, a social scientist can

37

study the relationship of the network's content and structural charac-
teristics on individual and group behavior. The network can also be
related to the system's formal structure for comparisons of their
structural properties.

This latter property of network analysis is the reason for network
analysis' selection as the methodology in the present study. Because
the study attempts to identify a multidisciplinary network within a
formal organizational structure, the network members would necessarily
have communication relationships which connect across the formal struc-
tural lines (if the network exists). Network analysis can identify a
network from the communication relationships within the system and
describe the network's communication structure. The network can then be
compared with the formal organizational structure to test fOr differ-

entiation between the two structures.

Communication Structure and Content

 

The products of the communication network analysis in the
present study are the communication structure and content of the
investigated population. These results are necessary to determine
the level of multidisciplinary integration within the energy
education personnel, the study problem. Although communication
structure and content have been introduced in reviewing the selected
literature related to networks and network analysis, they are
presented as a separate section to provide a clear and distinct
understanding of their concepts, particularIy as they concern this

study.

38

Farace, Monge, and Russell (1977) define communication structure
as repetitive patterns of interaction which occur within members of a
system. The patterns of interaction are based on communication
content. Monge, Edwards, and Kirste (1978) in reviewing the
literature assert that the repetitive relationships of a system may
serve several different purposes and have a variety of content. For
this reason, communication networks based on the different content of
the relationships may differ in structure. For example within the
same system, the network of work-related communication could be quite
different from the friendship network of informal social relations,
even though some relationships may be the same for both types of

communication content.

Communication Content

 

A meaningful way to look at the content of communication within
an organization is through Berlo's (1970) three functions of
communication: production, maintenance, and innovation. Berlo's
categorization can be used to examine any communication relationship
for content by determining the function the communication serves in
the relationship. The relationship itself may be based on one or
more of these three major communication functions (MacDonald, 1970).

Production is described by Berlo (1970) as communication related
to getting the job done. It is a nmchaniyn of control. Messages
related to specifications and amount and type of work are included in
the production category.

While production communication is serving to complete the

day-to-day work, innovation is communication related to new ideas and

39

practices. Innovation communication provides the sensitivity to the
changes taking place which the formal work-related network may be
missing. The innovation network is critical to the long term survival
of the system (Farace, 1978).

Maintenance refers to that communication which establishes and
improves self-image, inter-personal relations, and attitude toward
the organization. The maintenance function is referred to by Farace
and Connelly (1970) as the communication which serves to encourage
the smooth operation of the organization and its members. This
function differs from the other two functions in that system outputs
are not directly concerned (Albrecht, 1978). Often in studying the
maintenance function in a system, researchers will further define it
into subfunctions of' maintenance of self-concept, maintenance of
interpersonal relations, and maintenance of the production and
innovation functions (Farace and Connelly, l970).

Monge, Edwards, and Kirste (1978) in their review of research
related to communication content have found two difficulties as well
as emerging trends within the studies. The first difficulty is that
the results of the studies are difficult to compare because of
terminology and operational differences. The second concern is for
the capability of the study populations to discriminate reliably
between Berlo's fair1y similar content categories. If the survey
instrument does not present clear and understandable descriptions of
the functional behavior sought by the researcher, the survey
respondents may not provide valid data (Erickson et al., 1981).

The results of the communication content studies, however, have

been summarized by Monge, Russell, and Kirste (1978) into interesting

40

trends. One conclusion is £1 possible relationship between certain
types of content and the role expectations of network members. The
communication content shared by subordinates with superiors and by
superiors with subordinates can vary by amount and type in terms of
quantity. The channels of communication can also vary with the type
of content. For example, face to face communication was found to be
related to production content. The content channels can be
influenced, however, by the system's social atmOSphere.

The benefit of studying the content of communication in terms of
networks is to understand the system's communication patterns of
production, maintenance, and innovation. For the present study, the
stated content networks as well as the overall communication network
related to energy education can be compared with the existing
departmental structure for multidisciplinary integration. The
content network which exhibits the greatest multidisciplinary
connectedness or density is important to determine because it offers
insight into the possible causes for the existing level of

integration.

Communication Structure

 

Farace and Mabee (1978) describe three broad categories of
communication structure properties which can be revealed by network
analysis procedures. The terms used in describing the structural
properties may be uncommon to the general reader. For this reason,
the reader may again be aided in comprehending this material by a

glossary of terms which is available in Appendix A.

41

The first category of communication structure are the properties
of the structure as a whole, including structure size and density and
the number and size of detectable cliques. The second category are
the properties of the cliques, including the number of members and
the density within each clique, and the relationship of each clique
with other structural components. The last category is the
individual member properties, including the number of dyads,
communication roles(s), and density.

The benefit of looking at the communication structure as a whole
is to determine the size of the network and its level of density.
This perspective also displays how the network is divided into
cliques and individual members (Farace and Johnson, 1974). The level
of density of the whole network is measured in terms of the "percent-
age of existing within network relationships out of the total number
of such relationships which would have been theoretically possible"
(Monge, Edwards, and Kirste, 1978). This measure indicates the
relative integration of the respondents in the network (Albrecht,
1978). The density measure is influenced however by system size,
with large systems having a lower percentage of density because of
the difficulty in communicating with all members (Friedkin, 1981).

Boissevain (1974) describes £1 clique, the second category of
structural properties, as clusters of network members who have a
relatively high density. These network members are speaking more
regularly' with each other than they are with the other network
members. The existence of cliques in the present study would be an

indication of multidisciplinary integration.

42

Farace and Johnson (1974) and Rogers and Agarwala-Rogers (1976)
describe several variables which can be measured at the clique level.
The first measure is group size. The size of a clique can be three
or more members depending on the parameter used for the level of
interaction. For example the computer method of analysis, NEGOPY,
can set the number of linkages for a clique member to be 50% or more
of his/her total network linkages. This parameter can be raised to a
higher percentage and thus affect the clique's size and inner
connectedness.

A second measure at the clique level is clique dominance or the
degree to which the pattern of communication within a clique is
dominated by one or nmre individuals. This measure may illustrate
the importance of an individual's communication role in the clique.

The degree of density of a clique can be assessed of the members
within the clique and of the clique with other cliques. As in the
previous measure for density of the whole system, the measure of
connectedness of the individuals within the clique and between the
cliques of the network is based on the percentage of actual linkages
against the possible number of linkages (Rogers and Kincaid, 1981).
A high degree of density within a clique would indicate the members
have extensive and strong links. The existence of strong density
also suggests a similarity of knowledge and attitudes within the

clique members (Farace, Monge, and Russell, 1977).

43

LOW DENSITY HIGH DENSITY

Actual Linkages _8 . . Actual Linkages 4 .4
Possible Linkages: 10 “ Possible Linkages: 10 '

 

 

Key to Symbols
— " Link
*- Clique Member
Figure 2-3

An example of high and low degrees of density within cliques.

The final category of communication structural properties is the
individual. The important structural properties considered for
individuals are communication roles, links, and density (Farace and
Johnson, 1974). Communication roles can be broadly divided into
participants and isolates, with participants further defined by
clique members, bridges, liaisons, tree nodes, and others (Farace and
Johnson, 1974). An isolate is defined as an individual in the
network who has less than two communication links (Richards, 1975).
A non-participant or an isolate in the network then is a person who
has one or less communication interactions with other network
participants. Richards (1975) describes a person with no
communication links as an isolate type one. A person with only one
link he describes as an isolate type two.

In reviewing the descriptions of participants, clique members

have been defined previously as individuals who are more frequently

44

in communication with other clique members than with non-clique
members (Larkin, 1978). Rogers and Agarwala-Rogers (1976) describe
liaisons as persons who interpersonally connect two or more cliques,
but who do not belong to any clique. Persons who interpersonally
connect two or more cliques, but who are members of a clique are
labeled as bridges. A tree node is a person who is not a clique
member, but connects a clique with an isolate type two. The
communication roles of liaison, bridges, and tree nodes, which are
called group linkers, are critical for network cohesion and
information exchange because they control the rate and nature of
information flow 'h1 the network (Albrecht, 1978). Ihi the present
study, the identification of the existence of group linkers can be an
indication of the integration of a multidisciplinary network.

An individual who cannot be identified as an isolate, a group
linker, or a clique member is simply referred to as an other
(Richards, 1975). This individual may have two or more linkages
which do not meet the requirements for clique membership or group
linking.

The frequency of these roles in a communication network may vary
according to Monge, Edwards, and Kirste (1978) with the communication
content of the network. For example, Farace and Johnson (1974) found
that the number of group linkers and isolates varied within the same
system for networks of production, innovation, and maintenance.

The second structural property of individuals for review is the
link, the most fundamental of structural properties (Rogers and
Kincaid, 1981). A link indicates the communication relationship

between network members (Larkin, 1978). Farace and Johnson (1978)

45

report three characteristics of links often studied are the number of
links for each member, the percentage of reciprocated dyads, and the
strength of the links. Together, these characteristics provide an
indication of the important linkages within the network.

Individual density, the last individual structural property pre-
sented, is the degree to which a member is linked to the network
members (Rogers and Kincaid, 1981). This measure is defined as the
actual number of links the individual has with other network members
divided by the possible number of system links. The level of
connectedness for an individual can provide a perspective of an indi-
vidual member's integration with the network.

The concepts of network content and structural properties
presented in this section provide an understanding of the
communication relationships which can be identified in a
communication network and how they can be applied to an investigation
of behavior in a system. These concepts will provide the basis for
the determination of multidisciplinary integration within the energy
education personnel, the problem of the present study. The
determination is possible because the existence of multidisciplinary
integration can be graphically illustrated by placing content

networks and cliques against the formal departmental structure.

Problem Hypotheses

 

This study seeks to determine multidisciplinary integration
within the faculty involved in energy education at Michigan State
University in 1981. The foundation for the study is from the work by

Blau (1973) and Friedkin (1978), as well as, from network and network

46

analysis concepts. Friedkin suggested that multidisciplinary
integration may be fostered in a research setting. A suggestion
which is contrary to Blau's postulation that multidisciplinary
integration is unlikely. Friedkin's finding was from a study which
focused on research communication in a research institution. Blau's
conclusion was drawn from a study of general communication of faculty
at a sample of American universities.

As the present study more nearly replicates Friedkin's problem
and setting, the hypotheses for this study imply that
multidisciplinary integration will be identified within the
investigated population. To prove the existence of multidisciplinary
integration, the first three hypotheses are related to the structural
properties which will be sought by the communication network analysis
of the personal involved in energy education. These hypotheses are
intended to establish that the communication patterns existing
between the energy education personnel comprise a network. The
fourth and final hypothesis determines the multidisciplinary
integration, the problem of the study.

The chapter's review of the literature presented selected
research and writings regarding networks and network analysis.
Networks were described as aggregates of individuals who have regular
patterns of communication interaction. Network analysis was defined
as a methodology for determining networks and their characteristics
within a system.

Based on the network description, if Inultidisciplinary
integration exists within a given set of faculty at a college

setting, the faculty participants would in effect form a network

47

(Friedkin, 1978; Mulkay, Gilbert and Woolgar, 1975; and Mullins,
1972). For this reason, the present study's problem, the
identification of multidisciplinary integration within the faculty
and staff involved with energy education projects, can be
investigated utilizing network concepts and network analysis
procedures. The network analysis methodology which will be followed
includes identifying the possible existence of a network of the
energy' education personnel. If' the network exists, according to
Richards (1975) and Rogers and Kincaid (1981), the participants in
the network will have direct and indirect linkages with each other
and each relationship will be based on regular communication
concerning energy education. In addition, the network may have
groupings of linkages composed of members who communicate more with
each other than with other network members. These groupings are
called cliques, if strict conditions are met by the members.

The first hypothesis is intended to establish the existence of
regular communication patterns within the energy education personnel

as a network.

H]: The communication behavior of the involved
personnel in energy education will be defined
by the data analysis to have regular communication
linkages.
The first hypothesis will be accepted if two conditions are met.
First, the data analysis completed by the NEGOPY program must
determine the individual communication relationships. In effect, the

participants must indicate communication interaction on energy

48

education with the other involved personnel. Second, the
relationships must have at least the strength (or frequency of
communication) of one hour per month (Richards, 1975).

In addition 1x1 the overall communication network, the
sub-networks of specific communication content of production,
innovation, and maintenance can be determined. These content
networks will be noteworthy as to whether they exhibit
multidisciplinary integration.

H2: The data analysis of the communication content

of the participants based on production,
innovation, and maintenance will identify
regular communication linkages for each content
topic.

The acceptance of the second hypothesis is based on the same
criteria used for the first hypothesis. The test of the second
hypothesis is for the existence of the linkages. The second
hypothesis will be accepted if communication linkages of at least one
hour per month are identified within the communication behavior of
the energy education personnel.

Sub-groups or cliques of relationships can be identified which
have direct linkages to each other. These cliques can offer a more
micro inspection (Hi the, composite relationships existing between
members of a network.

H3: The data analysis will identify cliques as a

structural property of the network linkages.
The acceptance of the third hypothesis is based on Richards'

(1975) definition of a clique. A grouping of linkages is considered

49

a clique, according to Richards, if it is composed of three or more
persons who have direct linkages to each other and have more than 50%
of their links within the clique. In addition, no single link nor
individual can be removed and cause the dissolution of the group.

The NEGOPY program's algorithm can be set to determine cliques
according to this definition.

The network linkages and cliques identified may be centered in
academic departments or display communication patterns across depart-
mental lines. In order to verify the existence of multidisciplinary
integration, the linkages and the cliques must be constituted by
persons from different academic departments.

H4: Participation in the network linkages and the cliques

will not be restricted by departmental membership.

Friedkin (1978) concluded multidisciplinary integration to exist
in his study population based on the finding of 39% of the research
relationships to be interdepartmental. The acceptance of the fourth
hypothesis will be based on Friedkin's analysis. The fourth
hypothesis will be accepted if at least 39% of the linkages are
interdepartmental and at least 39% of the membership of any cliques
is interdepartmental.

The network analysis procedures for identifying the
communication content and structure were brief1y mentioned in this
chapter without conclusions made as to which techniques would be
proper for the present study. The discussion of these procedures in
the research methodology of this study is presented in Chapter III.
The results of the data analysis and their implications for the

study's hypotheses are reported in Chapter IV.

CHAPTER III
METHODOLOGY

Introduction

 

The purpose of this study is to investigate for the existence of
a multidisciplinary social network within the energy education
personnel at Michigan State University in 1981. The presence of such
a network would be unlikely according to Blau (1973), although a
study by Friedkin (1978) has found a multidisciplinary network within
a research setting. If the social network exists, the communication
structure, and content of the network members will be analysed and
discussed 'h1 relation 1x1 the previousLy noted research on
multidisciplinary integration.

The existence of the social network and the nature of the
members' communication content and structure was investigated by a
communication network analysis. The analysis began with a survey of
the study population's overall communication patterns related to
energy education. From the survey data, the study population was
identified and communication networks were constructed using NEGOPY
(a statistical network analysis package). The networks determined by
NEGOPY provided an identification of the (xmmmnication content and
structure, including the specific communication roles, linkages, and
groupings existing within the projects' personnel. The results of
the analysis are discussed in the context of Blau's (l973)
postulation and Friedkin's (1978) finding on multidisciplinary social

networks within academic settings.

50

51

Assumptions

 

In developing the~ methodology for this study, certain
assumptions had to be made which were integral to the methodology's
success. It was assumed that:

1. The Michigan State University energy education projects and

their directors could be located and identified;

2. The project directors would consent to cooperate with the

study;

3. The respondents would give honest replies on the question-

naire; and

4. The social structure of the study population would be

stable enough to allow measurement.

Research Setting

 

The data were collected on the campus of Michigan State Univer-
sity between February 19 and March 24, 1981. The campus was active
in the energy field with energy education projects primarily related
to its land-grant heritage. Energy education projects which were
reported included improving energy information for community
decision-making, alternative energy techniques for grain drying,
biomass from livestock waste, and Michigan tourism and energy.

It is important to note the campus climate at this time was
unsettled, tense, and directed away from typical academic research
and teaching due to current budget restrictions and further imposed

program constrictions. Numerous respondents and non-respondents

52

indicated difficulty' in completing the study because of' pressing

budget matters.

Network Analysis Methodology

 

The objective of network analysis is to describe a system's
communication structure (Rogers and Kincaid, 1981). This section
presents the methodological procedures for collecting and analysing
the data to describe the structure. Presented as well are the
problems and strengths of these procedures and the reasons for
selecting the procedures used in the present study.

Knoke and Kuklinski (1982) state that there are three matters
which must be considered in any empirical research: boundary
specifications, sampling, and measurement. The consideration of
these three matters in a network analysis is influenced however by
the nature of networks and the network analysis methodology. A
description of these matters as they relate to network analysis is
presented in the order of sampling, measurement, and boundary
specification. Following these descriptions, a presentation of
NEGOPY, the computer data analysis method, completes this section on

network analysis methodology.

Sampling

Because network analysis studies interpersonal relationships and
not individuals, sampling in a network analysis is of the entire
study population whenever possible (Rogers and Kincaid, 1981). Each
case of the study population is considered, otherwise the opportunity

is lost for identifying relationships with other cases. Without a

53

total population study, the network and its properties can be severly
misrepresented by the absence of relationships which may play
significant network roles. For this reason, random sampling is not
appropriate for network analysis (Knoke and Kuklinski, 1982).
Likewise, the application of conventional statistical procedures
based on random sampling measures is not suitable for network

analysis (Knoke and Kuklinski, 1982).

Measurement

 

Rogers and Kincaid (1981) cite the three main methods of
measuring communication relationships to be observation, unobtrusive
techniques, and survey sociometry. All of these methods have
particular strengths which make each appropriate for different types
of studies. In the observation method, the researcher monitors the
behavior of a system and records the nature and frequency of the
communication relationships as they occur. The strength of the
observation method is that if the communication relationships are
observed there can be little doubt about their validity. With survey
methods, the population studied is asked to report their
communication relationships. The respondents may jeopardize the
validity of the data by not accurately reporting their behavior due
to fatigue, misunderstanding, or other problems related to the survey
process. The researcher however is trained to observe and to
interpret behavior accurately and thus ensure a higher validity of
measurement (Rogers and Kincaid, 1981).

One disadvantage of observathm1 as a method of measurement is

its practicality. To be successful, the observation method must be

54

implemented in a small closed system. For example, Bernard and
Killworth (1973 and 1978) have effectively used the method in prisons
and on ships. In research problems such as the present study, with a
study population widely separated, relatively large, and at least
initially unknown, the technique is impractical and probably
ineffective.

A second disadvantage of the observation method cited by Rogers
and Kincaid (1981) is the obtrusiveness of the researcher in
collecting the data. Because of the technique's nature and that it
is most effective with a small closed study, the observer can become
a part of the study (Bernard and Killworth, 1978). In cases where
this involvement occurs, the obtrusiveness may well influence the
behavior investigated.

The second method of measurement cited by Rogers and Kincaid
(1981) is the unobtrusive method. In this approach, the observer is
not directly involved with the behavior studied. Rather, records of
events are studied in a lfistorical fashion for relationships. The
records can be tape recordings (Killworth and Bernard, 1976),
legislative roll-calls (Stokman, 1977), and correspondence (Stern,
1979). From these records, relationships and roles can be identified
and networks determined. Use of records has the advantage of
providing the researcher with a view of the relationships over time
for an analysis of change, as well as, greater data validity because
of the unobtrusive approach. The problem with the approach however
is that the data must be recorded and available for the researcher.
These conditions are not always possible for most studies, including

the current study.

55

The final method of measuring communication relationships is
survey sociometry. Of the three methods cited by Rogers and Kincaid
(1981), this method offers the most effective approach for collecting
data in the present study. A researcher using the survey sociometry
method gathers information about communication patterns of the
system's individuals by asking them with whom they interact about a
given topic. Depending on the nature of the study, the respondents
may also be asked for the frequency of the contact and/or to indicate
who are the most important contacts.

To encourage a greater identification of all relationships, a
roster of the study population is frequently provided with the
survey. The respondent can use the roster rather than memory to
recall with whom they talk about a given topic. For this reason, the
roster technique is beneficial for identifying the relationships
which Granovetter (1973) calls "weak ties." Although these
relationships are characterized by less frequent interaction,
Granovetter has found weak ties to be critical for bridging
communication flow between cliques of the networks. Without their
inclusion in a network, the network properties could be
misrepresented.

Because the roster technique facilitates the identification of
weak ties, it is also one means of counteracting a criticism of
sociometric surveys, the validity of self-reporting. Due to that the
source of information is self-reports based on recall, serious
questions have) been raised about the sociometric data collection
method. Based on their research, Bernard, Killworth, and Sailer

(1981) feel that it is unrealistic to expect respondents to have

56

perfect recall of their communications and to handle the amount of
cognitive data requested in a sociometric survey. Burt and Bittner
(1981) state, however, the conclusion of Bernard et al. is
unwarranted. Burt and Bittner's counter' argument. is based on a
re-analysis of the data used by Bernard et al. to reach their
conclusion. Their assessment is that measurement by the sociometric
procedure can be valid through a well conceived and conducted survey
approach.

Erickson, Nosanchilk, and Lee (1981) plus Knoke and Kuklinski
(1982) suggest that there are ways to minimize the respondent
measurement error in a sociometric survey. The most effective way is
to be as precise as possible with the respondents when defining the
network content being solicited in the survey. The respondents
should readily understand the survey and what is being asked of them
for information. The roster technique as previously described is
likewise an important means for aiding the respondents in completing
the survey. Fatigue in reviewing the roster however can be a factor
in respondent measurement error and must be considered when composing
the survey instrument. Erickson et al. (1981) suggest that a roster
of 130 can be safely used but as many as 200 may be too much for even
the most willing respondent. Other suggestions for enhancing the
survey are to pretest the survey, include the respondent's name in
the roster list which he/she receives, and make a brief introductory
statement in the respondents' own terms about the purpose of the

study.

57

Boundary Specification

 

Knoke and Kuklinski (1982) regard boundary specifications as a
difficult consideration for a communication network analysis.
Boundary specification refers to the parameter limits set for
collecting data. These limits are usually defined by the members of
the studied entity. Consequently, only those persons who are
relevant to the study are included within the boundaries.

Normally, the boundary specifications for many communication
network analysis studies can readily be set if the problem is to
study an intact unit's communication patterns. However, this
approach is not practical in a large organization if the network
members are unknown and the network to be identified has the
potential to spread across the organization and include relatively
few organization members. This situation exists in the present study
where the energy education personnel and the extent of their spread
across a major institution are not known at the time of the
investigation.

In cases of unknown boundary specifications, Burt (1980)
suggests identifying the boundary while sampling for data-collection
using the "snowball sampling" technique. With this technique,
persons who by their prominence are known to be members of the study
population are used as an initial set of respondents. In a manner
similar to the sociometric data collection procedure, these persons
are surveyed for information not only on their communication patterns
as related to the study topic but also for other persons with whom
they have significant relationships within the study's criteria. The

persons named in the data collection sampling are added to the list

58

of respondents and they alone are surveyed in a second phase of data
gathering. This technique can be repeated until a satisfactory study
population is determined.

Burt (1980) indicates that the technique has been successfully
used to identify networks of "elites" or specific types of persons
within business corporations, academic disciplines, and neighborhood
communities. Rogers and Kincaid (1981) also cite the technique's
success in diffusion studies within neighborhoods and academic
settings.

Rogers and Kincaid (1981) describe two disadvantages' of the
snowball sampling technique if the sampling phases are not continued
to a final phase where no new members are identified. First,
individuals and small cliques who are not well connected to the
prominent core set of' respondents may' be Inissed in the sampling
process if the phases are not repeated enough. It is assumed that
unless the clique or person is an isolate, the snowball sampling will
eventually identify them. However, some less integrated network
members may not be identified if the sampling phases terminate too
soon to reach these members.

A second disadvantage is that the rate of reciprocation is
affected. The study population added to the roster at the last phase
will not have an opportunity to respond to the survey. For this
reason, they cannot affinn the communication relationships between
themselves and other study population members.

Related to Rogers and Kincaid's (1981) concern with the problem
of sufficient number of sampling phases is the effect of time on a

network (Knoke and Kuklinski, 1982). The question Knoke and

59

Kuklinski (1982) raise is how many phases are sufficient to reach all
the appropriate study population members in light of any sociological
considerations for the network under investigation? By conducting
the survey over an extended period of time, there is a risk that the
actual network itself’ may change by the final phase. Networks
typically are not composed of a static set of communication
relationships; network members do change as well as their
relationships. Knoke and Kuklinski, as well as the literature, do
not present any clear answer to this problem. The proper balance of
the number of sampling phases and identifying all the possible
population members is considered to be relative to each individual
study situation.

Because the population of the present study is large, unknown,
and spreads across a large system, the snowball sampling technique is
the proper technique for data collection and for identifying the
study sample. In sampling the population, however, there is not a
need to identify all energy education personnel who are members of
the network. If persons or small cliques are not located, the
structural properties of the network will be affected but not the
overall trend of the network relationships. Using the snowball
sampling technique, the predominant members are identified and their
relationships can be examined for multidisciplinary integration.
Whether integration is found to exist or not, the persons and cliques
not located by the sampling phases are either weak ties or isolates
within the network because of their apparent low level of involvement

with the predominant network members.

60

As all network members are not required for this study, the
sampling phases was arbitrarily limited to two phases and set to not
extend over a four-five week time span because of the potential
change in projects and personnel on the energy education problem. If
the sampling was to continue for a longer time period, the population
sampled in the later phases would not be the same network which

existed in the beginning of the study.

Analysing the Data

 

The purpose of the network analysis in the current study is to
identify the communication content and structural properties of the
population under investigation. These results are identified in the
final research step, the data analysis. For the present study, the
analysis of the data can be accomplished by any one of the previously
indicated computer methods; however, NEGOPY will be the computer
method used for the analysis.

The selection of NEGOPY as the computer method is based on its
capability of providing the desired analysis for the research
problem. Rogers and Kincaid (1981) describe NEGOPY as being
preferable for research studies seeking to identify a large size
study population's communication structure, particularly for
networks, cliques, individual communication roles, and indices of
connectedness. These dimensions of NEGOPY are essential for the
successful implementation of the present study's methodology.

NEGOPY was invented by Richards (1975) in the mid-1970's in re-
sponse to the need for a computer method to analyse large quantities

of communication network data. The NEGOPY method analyses linkages

61

among persons in large organizations for the purpose of identifying
the communication structure at the individual, group, and network
level. The properties of the communication structure which can be
identified include the individual communication linkages and roles,
the cliques and their linkages, and the degree of connectedness or
density of the linkages (Rogers and Kincaid, 1981). In addition to
the general communication structure, selected content of the
communication shared by the network members can be analysed as
separate communication structures. Each of these structures can then
be compared to the formal organizational structure for similarities
and differences (Farace, Monge, and Russell, 1977).

NEGOPY determines these communication structural properties from
a data base of reported dyadic interaction frequencies. The study
population is asked to indicate with whom they communicate about
specific topics (i.e. identify ‘their ‘linkages) over a: given time
period. This data is entered into the computer and arrayed on an N x
N matrix of respondents and their reported contacts. Using cluster
analytic techniques, the data are decomposed and, based on each
respondent's linkages, a communication position and role is assigned
to each individual. From this analysis, cliques and networks can be
constructed, as well as, the degree of density measured at the
individual level and the clique level (Richards, 1975).

Because networks are based on the dyadic interaction
frequencies, determining what constitutes a dyadic relationship or a
link is an important consideration (Farace and Johnson, 1974) . In
their review of research, Farace and Johnson (1974) have found that

the determination of a dyad can vary between one person reporting the

62

relationship, to a more stringent requirement of both persons
reporting the relationship at the same frequency for the same content
topics. However, intermediate definitions are found to be often
used. For this study, reciprocated linkages are identified, but the
network analysis is based on all relationships reported by the
respondents, reciprocated (n: not. Reciprocated relationships were
not required in the analysis because of the study's sampling
technique and because valuable information would be lost otherwise.
In this study, the snowball sampling was terminated after two phases.
Those study population members identified in the second phase were
not surveyed and given an opportunity to provide reciprocation
responses. To not consider these unreciprocated relationships as
well as others, from the earlier phase would be a loss of relevant
network information. Support for this action is found in discussions
on the topic by Alba and Kadushin (1976) and Rogers and Kincaid
(1981). They feel that "forcing reciprocation" or considering all
relationships as reciprocated is appropriate because persons do not
always remember their relationships while completing the survey. In
addition, relationships are sometimes not reported because of the
status of the respondent or the respondent feels that the frequency
of the relationship is too minimal to report.

Richards' (1975) NEGOPY method is distinctive in that it
details an algorithm of mathematical and operational rules for
identifying network properties like cliques, group linking roles, and
connectedness or density. A brief description of how NEGOPY
determines these properties will provide understanding of this

analytical procedure. The NEGOPY method identifies a clique based on

63

three or more individuals who have direct linkages to each other and
have more than 50% of their links within the group. In addition,
NEGOPY applies what Richards (1975) calls a cnjtical criterion in
determining a cflique. The criterion is no single link must exist
within the clique which if removed from the group, would cause the
remaining clique members to not be able to meet the other criteria.

The group linking roles are identified by NEGOPY (Richards,
1975) as those persons who have linkages which connect two or more
cliques. Their role is to facilitate communication flow between the
cliques and for this reason are important to the functioning of the
communication network (Guetzhow, 1965). Indeed, Schwartz and
Jacobson (1977) have found group linkers to be the informational
leaders for the network and to be considered by the network members
as highly influential.

Connectedness or density refers to the degree to which
individuals or cliques are related to each other respectively (Rogers
and Kincaid, 1981). Friedkin (1981) describes connectedness as a
measure of density or cohesion within the system. Such a measure
indicates the amount of interaction which is taking place within the
level of analysis. The degree of density is indexed by NEGOPY to the
actual number of links which exist for the level of analysis, divided
by the number of possible links for the level of analysis.

By using an algorithm for identifying these properties, NEGOPY
has standardized network property meanings from study to study. In
this way, researcher error and subjectivity are removed in
identifying the communication structure's properties and content.

For the present study, the standardization provides validity to any

64

conclusions reached (”1 the degree of nmltidisciplinary integration
existing within the study population. NEGOPY's standardization also
provides generalization of the study results, as well as, the data
analysis to other similar campuses and studies.

Methodology in Relation to Purpose

 

The intent of this study was to investigate the existence of a
social network of the energy education personnel at Michigan State
University, how the network members were linked through
communication, and the function of their communication. The
investigation was accomplished using the procedures of communication
network analysis. Communication network analysis is a research
method developed out of sociometric techniques (Rogers and Kincaid,
1981). It takes a set of paired data which specify the interactions
between individuals within a social environment and displays the
interaction on a N by N person matrix (Taylor, 1976). NEGOPY, a
computer based technique, was used to analyze these interactions
among persons and within groups, and identify the communication
roles, groupings, and functions in the system (Richards, 1975).

For this study, the participants of the survey population were
asked to review a list of the names of people active in energy
education at Michigan State University and record the number of hours
they communicated with them in the functional areas of work
(production), ideas (innovation), and personal (maintenance) in an
average month during a normal academic term. The p0pulation was
directed to leave blank the appropriate space next to any person
listed with whom they did not normally communicate in any or all of

the functional areas. If the participants did not find a person on

65

the roster with whom they normally communicated, they were asked to
include the person's name on the blank spaces at the end of the
survey roster and to indicate the frequency of communication (see
Appendix B).

The survey data were then entered in the NEGOPY program. The
frequency parameters of the NEGOPY program were set to consider only
those communication relationships with a minimum frequency of one
hour and a nwximum frequeney of 255 hours per month. The reported
contacts within the frequency parameters were compared against each
other in a matrix to determine reciprocation of contact among
respondents. To prepare the data for analysis, NEGOPY's statistical
parameter forced reciprocathm1 for any unreciprocated respondents.
The data were then reduced through multivariate analytic techniques
into network properties. The cliques (groups) were separated out and
members were classified into one of the following role types: group
member, isolate, or group linker (Albrecht, 1978).

The properties revealed by the network analysis procedures can
be broken into three broad categories (Farace and Mabee, 1978): (1)
properties of the network as a whole, including the size and, the
connectedness of the structure, and the number and size of detectable
cliques; (2) properties of groups or (fliques, including number of
members in each clique and the amount of structure binding the clique
members; and (3) properties of the sub-networks associated with each
individual node (person), including the role of the node in the
network.

In addition to identifying the communication structure, a goal

of the study was to identify the function of communication existing

66

between the energy education personnel. This goal is important to
the study because roles and groups may vary across "content
networks" and provide an entirely different communication structure.
A look at content networks may show that communication related to the
energy education project directors' need for work or idea information
may be flowing across the formal department lines. The analysis of
the content networks may also indicate the importance of one or more
of the communication functions to the social network's existence.

The content of communication was solicited from the participants
by asking them to report their amount of communication in the
functional areas of' work, ideas, and personal with each of the
individuals listed (M1 the survey. The fhnctional areas replicate
Berlo's (1970) production, maintenance, and innovation categories.
The titles were changed in the survey for the purpose of clarity and
understanding. Appr0priate clustering was completed on the content

data in the same manner as the all-function communication network.

Study Population

 

The population studied was those full-time persons at Michigan
State University who were involved with energy education projects in
1981. This population was selected because the purpose of the study
was to investigate and identify the existence of'ai social network
based on their communication structure and content. Only full-time
faculty-staff were considered because of the temporary nature and
lack of full campus involvement normally associated with part-time

faculty-staff and graduate assistants.

67

An initial problem in studying the population was to identify
the respective members. The members were not readily known because
little knowledge existed in 1981 as to what was being done in energy
education and by whom. In fact, Schriver (1979) described a problem
which existed in coordinating the energy related projects and in
avoiding a duplication of effort. The difficulty in coordinating the
energy projects was due to the constant change of the energy
projects, the lack of a formal means to identify who was doing what
in the energy field at M.S.U., and the general autonomy of the
departments at Michigan State University in conducting the projects.

This lack of knowledge of the energy projects conducted by the
various University departments caused instances of duplication of
effort. An example of this duplication was found "in Genesee County,
at one point in time in late 1978 and early 1979, (where) there were
no less than six different MSU energy projects conducting research or
other activities" (Schriver, 1979). These activities were in some
cases attempting to accomplish the same goals and to utilize the same
resources. Eventually, the people in Genesee County affected by the
projects raised a concern about MSU's lack of coordination and lack
of knowledge of the projects.

A project in 1979 to identify all of the energy information ser-
vices at MSU was a response to this problem. The project, conducted
by the College of Education, attempted to contact all colleges and
departments at Nfichigan State University which were active in the
field of energy for descriptions of their energy programs. These
descriptions were summarized and published in a directory entitled

"Energy and the University: A Guide to Energy Programs and Resources

68

at Michigan State University". However, the value of this directory
in enhancing communication and integration of the departments'
efforts is not known because a follow—up study did not occur.

The initial source to identify the population for this study was
Schriver's 1979 directory. The second source was Mr. John Sarver,
Michigan Department of Energy, who was contacted to ascertain the
Michigan State University faculty and staff who were leaders of
energy' education projects funded by the Energy Department. Mr.
Sarver served a coordinating role for the Energy Department with the
Michigan State University energy projects which were funded by his
department.

The names of the energy project directors were composited into a
preliminary list which was reviewed by selected campus energy
education personnel for additions, suggestions, and correction. The
reviewers were Dr. Herman Koenig, Research and Graduate Studies; Dr.
Jon Bartholic, Agricultural Experiment Station; Dr. William Cooper,
Department of Zoology; Dr. Wally Piper, Student Activities Office;
Mr. Tommy Mc Peak, Cooperative Extension; Dr. Thomas Edens,
Agriculture Economics; and Dr. Bill Stout, Agriculture Engineering.
These resource people~ were asked to review the preliminary list
because of their high visibility and contacts in the energy education
field at Michigan State University. The revised list was further
refined through telephone calls to selected people on the list and to
academic departments having energy projects. These latter contacts
were also intended to build their commitment to complete the survey.

At this point, 126 persons had been identified as population

study members. Although the list of energy education project

69

directors was not considered to be complete, the communication
network analysis (the survey) was initiated on this population. The
reason for this step was that one of the first exercises in the
analysis is to have the population being studied identify those
energy education persons at Michigan State University with whom they
are in frequent contact regarding energy education projects. Those
persons identified who were not listed were added to the study
population. From this first sampling phase, nineteen new members were
included in the study for a total of 145. The revised survey was
sent to these new population members in a second sampling phase and
they were asked to participate with the analysis. In the second
phase, seven new members were identified for a total of 152 in the
study population. A third phase was not conducted.

The procedure used in identifying the population was based on
the snowball sampling technique (Burt, 1980), as well as,
Granovetter's (1976) work with social network sampling techniques.
For this reason, it was assumed that by following their procedures a
high percentage of the study population had been identified. Those
persons not identified may either have been recent entries into the
system or have had little visibility across campus. Further time and
effort was not put into locating other possible energy education
personnel because carrying the survey out over an extended period of
time would have distorted the results. The communication network
analysis is intended to be a tool for describing communication as it
exists at a particular static time. If the study is over a prolonged
period of time, persons will move in and out of the network and the

structure and functions of communication will change.

70

Data Collection

 

Questionnaire

 

A questionnaire (see Appendix B) was administered to obtain the
data on the communication patterns of those persons who were involved
in an energy education project(s) at Michigan State University. The
study population was limited to fhll-time personnel located on the
East Lansing campus. The purpose of the questionnaire was to seek
data which would identify:

1. the energy education projects at Michigan State University;

2. the goals and activities of the projects;

3. the population served by the projects;

4. who the participants communicate with on-campus regarding

energy education;

5. the frequency of the communication contacts regarding

energy education; and

6. the function served by the communication contacts.

The questionnaire was in two sections, a personal data sheet and
the communication network analysis survey. The personal data sheet
requested the study population members to provide information related
to their campus position and energy education responsibilities. The
data sheet was based on the 1979 questionnaire administered for the
data found in "Energy and the University". Mac Donald's (1970)
"personal contact checklist" and other selected communication network
analysis surveys were the basis for the communication analysis.
Selected on-campus energy education program directors were also

contacted for their assistance in developing the questionnaire.

71

Before the questionnaire was administered, it was pretested with the
same selected university energy education personnel who were asked to
help identify the campus energy education personnel. This pretest
was intended to: (1) identify any aspects of the questionnaire which
were misleading, ambiguous, or unnecessary; and (2) evaluate data
collection procedures.

The questionnaire and a return envelope was mailed to each of
the project directors identified in the population. A cover letter
(see Appendix B) explaining the purpose of the study was also
included in the mailing. The participants were given a designated
period of time by which to return the completed questionnaire.
Follow-up written messages and telephone calls were made to those
participants who did not return the questionnaire by the designated
date. Of the 145 persons surveyed in the study, 102 usable surveys
were returned for a response rate of 70%. The seven persons
identified in the second phase were not surveyed.

The survey procedure was completed between February 19 and March
24, 1981. Approximately five weeks was given to the survey because
new names for the study population were generated by the initial
communication network analysis and time was needed to contact and
survey them with a second sampling questionnaire. As previously
mentioned, the survey process was not extended beyond five weeks and
two phases because of the changing dynamics of communication networks
and the high percentage of energy education members felt to have been

identified in the two phases.

72

The results of the analysis of the data collected through the
survey questionnaire and their relationship to the study hypotheses

are presented in Chapter IV.

CHAPTER IV
ANALYSIS OF THE DATA AND RESULTS

Introduction

 

The intent of the present study is to test Blau's (1973) postu-
lation that multidisciplinary integration is unlikely within univer-
sity and college institutions against Friedkin's (1978) finding that
multidisciplinary integration may be possible in certain academic
research settings. The study is being investigated by co.ducting a
communication network analysis of the personnel involved with energy
education projects at Michigan State University in 1981. The
analysis will identify the possible existence of a communication
network within the study population. The presence of such a network
will indicate multidisciplinary integration in this academic setting.

The first section of Chapter IV describes the study
population.The results of the communication network analysis for the
study is presented in the second section of the chapter. The
analysis was accomplished by using ‘the NEGOPY computer program
developed by William Richards (1975). The analysis identified
communication networks from the content of the surveyed population's
communication as related to energy education. The specific content
of energy education communication sought was based on Berlo's (1970)
three functions of communication: production, maintenance, and
innovation. The final section of the chapter reports the evaluations

for the hypotheses based on the data analysis.

73

 

4
7

mmogaommm
mmogaommm
mmugzommm
mmogzommm
mmoc30mmm
mmuczommm
mmoL30mmm

mmugzommm

mmoL=Ommm
mmucsommm

mmogaommm

Fmgzumz
_meaumz
pagapmz
_mgzpmz
_mgzpmz
”meaumz
_mgzumz

Pmesumz

_mtapmz
_mgaumz

_meapmz

use mgappzowcm<
wucm_om _mczpmz
new exam-:uwcm<
mucmwum _mgzumz
ucm mgzapzuwgm<
vcm we:u_:uwgm<

amo_oum scan:
mucmwum _mwuom
was menu—:u_em<

mmwuaum mumaumgw

new weappaowem<
ecu mL:u_30wcm<
mucwwum _mgzpmz
coeumoaum

use mL=HF=uwcm<
mucmwum Pagapmz
amopoum sass:
ecu mL:p_=uwem<
mucmwum _mcapmz
new we:p_:uwem<
mucm_om _mgzpmz
vcm mgzp_:u_gm<
mewcmmcwmcm
mucmwum _mcaumz
cowumoauu
xmo_oum cuss:
mucmwum _m-oom

mmogzommm _mczpmz ucm mg:u_:uwgm<

 

EnemoL¢\cowmw>wo\mmm_—ou

newgmmcpmcm _mgaupzuwgm<
muwmzcm

moveocoom pagap.:u_gm<

xmo_oow

muwsocoum Pmczp_:u_cm<
xgpmmgom

:mwmmo cam “cmacogw>:m cuss:
somezm :ucmmmmm mucmwum _Mwuom
copmcmwxm m>wpmemqoou

mmwuaum magnumcw new gogmmmmm
m:_cwmcwmcm _mc:p_=UwLm<
mewcmmcwmcm _mg:p_:u_cm<
wowewo m.=mmo

cowpmusum ewzmwz use cowpmeumwcwEuq
mocwwum wood

meFooN

smopoom v.2;u ecu »_w5md
mowsocoou _mc:u_:uPcm<
Amo_o50pcm

cowumum pcmswemaxm _mczu_=uwcm<
Ame-ocpma pcmpa cam Acmpom
m:_cmm:wmcm pansy-:u_em<
mcwgmmcwmcm —mu_guum_m

nm4 :ucmmmmm u:m_m

cowgmuzum mcwzcwpcou

smo_oum u_wsu use »,wsma
mucwwom _mwoom

mu_socoum _mgzpF38wLm<

 

pamEumewo

.cowpmpzqoa xuzum cowpmuzum amemcm mg“ to mcmnEmz

.Fiv mpnmp

coco
xm_zmgu
cgotzmcu
Lagoou
Loccou
m—_mqamnu
mmmu
FFoLLmu .»
F_ogcoo .<
coFacmu
axogm
xoogm
coucmcm
amazom
accummgom
mc__om
meom
xumpm
cgwm
ow_o;pgmm
mecaucmm
msmxg<igmxxmm
:mmm=Em<
cm~gce<
comgmuc<
mzmguc<
:m__<
:wxn<

mEmz

75

mmussOmmm

measuommm
mmossommm

mm0s30mmm
mmussommm

meussommm
mmussommm

meussomem
mmussommm

moussommm

_essuez

_essuez
Feszuez

Pesspez
_esspez

Fesspez
_esspez

_esseez
Pesspez

_esspez

use mssupsu_sm<
eusmwum _e_uom
musm_om _esspez
use mssppsowsm<
use es:u_:uwsm<
musmwom Fessuez
mosmwum _esspez
mswsmms_msm
amo_oum seas:
use mssp_suwsm<
musewom _ewoom
use ms:p_:uwsm<
muswwum _ewuom
use ms:w_:uwsm<
use muse—sowsm<
msesemswmsm
mswsmmswmsm
mose_om _essuez
AmOFOUm seas:
use assquUPLm<
use essu_:uwsm<
so_peo:um
mosmwum _essuez
use weep—so_sm<
emopoum seas:
mswsmesemsm
musmwum Pesspez
emo_oom seas:

mmusm_om use mus< sowpeo_ssEEou

mmUL:Ommm
mmussommm
mmussommm
m¢8L3Omma

mmossommm

Pessuez
_esspez
_esspez
_esspez

Fessuez

use msuupuu_sm<
use ms:u_:uwsm<
use ms:u_:uwsm<
use mssppsuwsm<

ms_smmswmsm
mosmwum Fewuom

msmpumu use mus<

use essppsuwsm<

mmwuspm euesueso

ems—sum seas:

mmossommm so_pemsumm use gees
weapomu_sus< useUmuse4 use ms_sse_s sens:
sepsmu ms_suemh mu_ue5mspez use mosewum
musmwum Pwom use sosu

musmwum uoos

muwmxss

emoFoEopsm

mswsmmswmsu _equmsu

eso_oee eswse ese e_wses

asymmsos

amopowoom

Asymmsos

mssuuepwsus< mseumuses use ms_sse_s sees:
mswsmmswmsm —es:p_=o_sm<

psmsmmsez uses

mswseeswmsm _eowamsu

mswsmms_msm _euwspum_m

emo_osues useps use Aseuom

seepage e__;u ece e__ses

mswsmmswmsm Fes=u_:owsm<

so_mseuxm m>wuesesooo

sepswu ms_suemp mo_uesspez use musmwum
emo_osopsm

meussommm sowpemsumm use gees

sm_mmo use usmEsosw>sm seas:

musm_um mFewsmuez use muwsesumz .xmss__eumz
espmwsmsuowm

emosoem e__ss ece e_sses

sowpeo_ssssou

mswsmmswmsm _essupuu?sm<

mosm_um Fwom use soso

mswsmmswmsm _essp_=uPLm<

Amo_oEousm

mswsmms_msm F_>Pu

emu—oaossys<

mu_weo m.semo

asymmsos

seems:

musw_om uoos

xmum_o:
emOhosw:
soumswsmsum:
_mmFmI
seEu—m:
mesxe: .m
mmsxez .o
xm_3e:
xezesue:
use:
Wessex
sm>osez
sepsez
__e:
smsee:
mx_:su
sesuoow
uoou
psesue_u
smessmw
seFmemu
smsme__ew
mmew
seeeees .e
seeeess .u
xs_s
smspws
u_e_s
mueses
xesmm
semxu_sm
musezum
msmuw
sespmem
Lexie
exmzoxesoo
sesxuwo
sowsexuwo
Pseo

76

mswsmmswmsm
musewum _ewuom
emoPOQM seas:

maesmoss use mm_u:pm _esowpessmpsH

mmussomea

maus30mmm
meULsOmwm
m¢ug=0mmm
mmus30mmm

mmussommm
mmossommm
mmussommm
mm0s30mmm

mmussommm
mmussommm
mmussomms
mmos30mmm

_essuez use mssppsowsm<
emo_oum seas:
sowpeosum

mosewom _esspez
musmwom _ewuom

emopoom seas:

mmm__ou msoso:
mspsmmswmsm

Fessuez use ms:p_:o_sm<
_esspez use mssu_sowsm<
_essuez use ecsp_suwsm<
Pessuez use mssupso_sm<
mosewum _essuez
mosmwom _esspez

_essuez use mszppsuwsm<
Fecuuez use mssp_:u_sm<
_essuez use msspp:u_sm<
_essuez use mssu_:owsm<
museeum _ewuom
sowpeosum

mosmwum Pesspez
psmsso_m>mo sens:
mosmwom _eeuom
mswsems_msm

emOFQUM seas:

mmwuspm euesuesw
mswsmmswmsm

emo_oum seas:

Amopoom seas:

AmOFOUM seas:

_esswez use es:p_:uwsm<
Pessuez use es:u_sowsm<
_esspez use msspP=uwsm<
_esupez use mgsppauwsm<
msmuueq use mus<

ms_gmmswmsm _eoespow_m

esspumuwsus< mseumuses use mswsseFQ sens:
sm_mmo use psmssosw>sm seas:

maesmoss use mmwuapm _esowaesswusH
sowmsmpxm m>muesmsoou

sowmsmpxm m>wuesmsoou

sowueusum sesmw: use sowuesumwswsu<
emo_ospes aseFQ use Asepom

amopo_uom

emosooe e_wsu ece e_.ses

mumppou msoso:

mswsmms_msm _euPsesomz

mmossommm soweemsumm use xses

soweeum usmswsmsxm Fessp~:u_gm<
mmussommm sowpemsuem use xses

seemsmpxm e>wpesmsoou

smpsmu mswsoemh muwuesmsuez use mosewom
mosmwum _esspez

mmussOmmm sowpemsoem use sees
mswsemswmsm Pes:u_30_sm<

mswsmmswmsm Pessupsu_sm<

mquOsoom —es:pesowsm<

meFOsuems

sepsmu mswsueme muwpesmspez use musmwum
see susemmmm usefis

mswe$w< sees:

esspomuwsus< mseomuses use mswsseps sens:
sosemmmm mswsmms_msm

musewum uoos

mmwuspm muesuesw use sosemmmm

susemmmm mswsmmswmsm

emo_oee e_wse use aseses

eeoeoem e_sss ese e_wses

eeosoee e__su ese e__ses

usmEso_m>mo mossommm

mswmsem museum so» smusmu

mosm_um _eews<

espmesos

mo_weo m.seeo

sees
xuw_so
smeaXz
mmosm_z
sse52mz
eemz

spa:
Asses:
sumwssoz .o
somwssoz .m
eesooz
xsmsompsoz
wmwsosoz
Le302
_sz
xemsoz
uomuoz
xsesuz
smzosoouz
sue:
sousoe
eee_s
ms_>m4
mmuses
“Losses
messes
sewage;
mmsesx
eme_o¥
mesmox
sensmx
soussmx
spwmx
mwmmx
emeex
somusx
Less: .u
sens: .:
susozmmswPFo:

77

mmus20mmm
mmossommm

maussOmmm
mmossomem

mmussomem
maussomem
mmussomem

mmussommm
mmussomem

mmussomem
mwussommm
mmus30mmm
mmussommm

mmussommm
mmussommm
mmus30mmm
mmosaomem
mmossommm
mmossommm

mmussommm

meussommm

_essuez
Fesspez

Fessuez
Fesspez

_eszuez
Pecspez
_essuez

pessuez
~esspez

Fesspez
Fesspez
Fessuez
Fessuez

_essuez
_esspez
_essuez
_essuez
Pessuez
_essuez

_esspez

_essuez

meow>sem use

mues30mem
mmussomes

_esspez
_essuez

mswsmmswmsm
musmwom _ewuom
use mssupsuPsm<
mswswmswmsm

use mssupsuwsm<
musmwum _esspez
use mssu_suwsm<
use mssupsu?sm<
mswsmmswmsm

use mssppsuwsm<
use mssppsowso<
use ms=u_30_sm<
ms_smms_msm

use mssppsupsm<
use msuppzuwsm<
Amo_oum seas:
use mssu_suwsm<
use mssp~so_sm<
use mesa—sowsm<
use mssu_:uwsm<
mosmwum Pewoom
mmmwsm see»;
use mssp_suwsm<
use mssu_=owsm<
use weep—sowsm<
use mssupsuPsm<
use mssu_:uwsm<
use mssupau_sm<
mswsmms_msm
mosmwum _esuuez
use ms:p_:owsm<
xmo_oom seas:
musmwum _essuez
mosmwum Fessuez
use menuF36wsm<
mswe$$< useusem
musewum _esspez
use ms:p_38wsm<
use menu—sowsm<

mswsmmswmsm _euwsmso
msspumuwsus< mseumusee use mswsse_s sees:
musmwom _esws<

muwwmo m.sewo

wusewum Fwom use sosu
amo_00N

mquosoum _es:u_:owsm<
ms_smmswmsm —es:ppsowsm<
mswswms_msm FeuwsuomFN
asymmsos

asum_5msoo_m

musewum xsweo

mosmwum mpewsmpez use mowsesomz .xms=F_epmz
asymmsom

mmussOmmm suppemsomm use xses
sawmmo use usessosw>sm seas:
mswsmms_msm Fes:p_:o_sm<
usesso_m>ea museummm
mswsmmswmsm _essppzopsm<
ms_smms_msw _essu_:uwsm<
essesmomo

mmmwsm sesxu

morsesoum _essppsuwsm<
muwsosoom _essp_:o_sm<
psmsso_e>wo mus30mmm
usessopm>mo museummm
usesmmsemsm _es:u_:uwsm<
so_msmuxm m>wuecmsoou
mswsmms_msm _euesesumz
muwmxss

mquosoum _es:u_:uwsm<
mow>smm uoos

xsumwemsu

Amopo_soso_z

xsummsos

mmwpw>wpu< psmusum

essesmomw

ms_seeswmsm _essppso_sm<
morassoum _es:u_:u_sm<

somswxpwz
xoensez
smp_e3
somsmpso>
smou_>
esumsmw>
xezmsm>emse>
mmse>
e_esssh
smsensoe
usmn_oe
mesosh
smpsm
usepmsusm
museum
mzoum
usopm
sw__essswmpm
memmum
e>eume>wsm
msmEEom
zosm
mewesm
nezsom
xo_u_:sum
xuwsum
Npom
usesuom
msmssmmom
msmmom
semwsom
smears
somsmm
euumm
EEem
sears
wNNQQws
somsms
somees

78

musmwom _ewuom
mswsmmswmsm

me0s30mem Pesspez use mssppuuwsm<
mmmwsm see»;

musm_om Fesspez

meossommm Pessuez use mssp_sowsm<
mmussommm _essuez use ms:p_su_sm<

meFowuom

ms_smmswmsm Peu_spom~m

musmwom Feaws<

mmmwsm seEXu

mowmxsa

muwaosoom msup_:uwsm<

so_ueum useswsmsxm _essppsuwsm<

mesu_sN
sseN
eEeAOxo>
“gees:
wwssuooz
use:
smzppwz

a. we; CCwfl;
“arisimts

 

79
Table 4-2.

Departments Represented in the Study Population.

Department # 0f Members

 

 

Administration and Higher Education
Agricultural Economics

Agricultural Engineering
Agricultural Experiement Station
Animal Science

Anthropology

Biochemistry

Botany and Plant Pathology

Center for Remote Sensing

Chemical Engineering

Chemistry

Civil and Sanitary Engineering
Communication

Cooperative Extension Service
Continuing Education

Crop and Soil Sciences

Dairy Science

Deans Office (Arts and Letters)
Deans Office (Engineering)

Deans Office (Natural Science)
Electrical Engineering and Systems Science
Engineering Research

Entomology

Family and Child Ecology

Food Science and Human Nutrition
Forestry

Geography

Honors College

Human Environment and Design
International Studies and Programs
Land Management

Mechanical Engineering

Metallurgy, Mechanics and Materials Science
Microbiology and Public Health
Museum

Natural Science

Park and Recreation Resources

Plant Research Lab

Physics

Psychology

Research Development and Graduate Studies
Resource Development

Science and Mathematics Teaching Center
Social Science

Social Science Research Bureau
Sociology

Student Activities

Urban Affairs

Urban Planning and Landscape Architecture
Zoology

Lyman Briggs

dfl

mwm—n—Iw-a—I-b-bm-A-p-Nm—a—a—amm—amwdmwmmbmmdam-awam-a-a-Aw-uwN—ubwmam

80

Table 4-3.

Colleges, Divisions, and Programs Represented in the Study Population.

 

 

College/Division/Program # of Members
Agriculture and Natural Resources 64
Arts and Letters 3
Communication Arts and Sciences 1
Education 6
Engineering 16
Honors 1
Human Ecology 17
International Studies and Programs 2
Lyman Briggs 2
Life Long Education 1
Natural Science 22
Research and Graduate Studies 3
Social Science 12
Student Affairs and Services 1
Urban Development 1
Table 4-4.

Study Population's Reported Type of Involvement
in Energy Education Activities.

 

 

Involvement Type # of Members
Research Only 31
Research and Public Service 16
Research and Teaching 15
Public Service 13
Research, Public Service, and Teaching 12
Teaching Only 6
Teaching and Public Service 2

(Note: Seven of the respondents did not indicate their type of energy
education involvement activity.)

81

Description of the Study Population

 

The snowball sampling technique was used to identify the study
population of full time faculty and staff who were involved with
energy education projects at Michigan State University in the spring
of 1981. Through this technique, 152 participants were found in the
energy education population (see Table 4-1). The participants were
from 51 University departments representing 15 colleges, divisions
and programs on campus (see Tables 4-2 and 4-3). The Departments of
Agricultural Engineering and Agricultural Economics had the largest
number of members with 16 and 11 respectively. Nineteen other
departments had only one member in the study population. Of the
colleges, the College of Agriculture and Natural Resources and the
College of Natural Sciences had the greatest participation with 64
and 22 members respectively. The Colleges of Communication Arts and
Sciences, Urban Development, and Honors had only one member in the
study.

Research was reported to be the primary involvement of the study
population in energy education (see Table 4-4). Nearly thirty
percent of the respondents (31) indicated their involvement to be
solely research. An additional forty-two percent of the respondents
(43) indicated their involvement to be research with teaching and/or
public service. The next greatest activity was in public service
with 12 percent of the respondents (13) indicating their only
involvement in energy education to be public service.

The energy education projects described by the participants were
numerous and varied. The projects were intended to service such

publics as the general citizenry, farmers, other scientific

82

.uousomossos ompe use
maesmoes use msowm .uousomessos .ueosom
-.>_u .mooe__oo some ompe use maesmoss use -ogsos om_e use msesmocs
.musosucesou is_sp .soosm mwsu s? msoemw>Pu .mouoppoo sop use mso_mv>eu .mouoppoo

 

 

 

 

usosoww_u Loom «senses o>es ausum .soosm «_su s, «Lessee so» .saosm m.su s. memo
eosw stooges es» s. usaoe muses o>es auaum as» s_ .502 o>es xuoum as» s_
mo uomoseoo losesou osoiauw.» usso» musoEusesou oso usoom mososusesou use
up saocm use use mo xwmia»sos team,» as» oo ozuieusos ixuu_m use mo o3u-»usom so_uvmossou
m._ i i u mxsvo ssosu souzuom wo sumsocum omeco><
“.5 i i . mxs.4 ssosu swsupz mo sensuoum omeso><
e.¢ . i . ass.4 noose ——< so suasosum oneso><
ue.~m . i . sovueoosspuos mo usoogos
n. i i i xupmsoo
e - i . mes.4 _eusoEQLesouesus_
m 1 i i asses PeusoEHLesougous_
up i u . uxs.4 mo Logan:
e _m_ —N— _N— o~.m
_o noose
moaavpu
osos osos pm pm mouuvsm
e osos osos osos msom.e_4
~ osos osos osos mouoz ouch
«N o_ up up or» ooze age—on“
—N a m o . use waxy one—om~
osos _ osos osos muexc
33s 533.2238
um.o_ u..m~ um.m~ um.m~ «use; uoueooss_oos so «soosos
e.m m.~ mm.~ mm.~ mxspe __< so suasosum sew:
uni. om-_ cal. omi— someosum so oases
mm_ mm mm_ mmp «sense: mo sensaz
New _em Pmm _mm asses FeosoEpsesouesus_
emu mo_._ No~._ ~m~.P mess; peusosuseoousous_
emu ueo.~ Mme.— mmo.p mxs_4 mo songsz
mos—e> xooxuoz
osmosou moseseusrez «seesou sowue>oss~ usousou sowuosuoss usousou sowuosssipp<

.so_uos=s so_oeo_s=esou xm mowugososs egoxuoz yo sompseseou

.miv m4m<h

{33

.seusou msPsueop
moeuesosuez use
ooso.um as» s.suez
eupsuem mo uomos
-eoo e. ssosu as»

.xmo—oum
seas: so one—poo
as» swsups muses
lusenou oxu sosm

xu_=uem o3» use
mouso'om _e.oom so
umeppou osu spsu.:
musoEusesmu moss»
eosm appsues Laos
"moosaomos pessuez
use os:u_:owsm<
so mam—_ou use
spsu.3 musmEusesou
o>we sosm xu—soee
soouswsu so ummos
-eoo m_ ssosm esp

m.p

s.m

N.N

um_.m—

m_.

em

we

we

a,

.usauoou_sos< useom
tuses use usesseps
sens: so usoEusesoo
es» Eoss m_ sesame
saw.» use .meossom
to: so_»eesoom use
sees so usosusesoo es»
soss use «sesame
mace—o use so Laos
es.
e.o_
_.e
amm.mm
on.
N—
N
e—
m

sowu_moasou

mxs_4 ssosm soozuom so seososum omeso><
mess; saosu spsupz so suososum woeso><
mxsps goose pp< mo sumsosum maesm><
sowueuoss.uo¢ so osmosos

xuvmsoo

A_eoseseseseoecoss __<v mes.o so guess:

oewm

me ssoso

.mssuuouesus< oseom
-usee use ms_sse_s
seas: so usoEusesoo
as» soss m_ sesame
see?» us» .moossom
tom sovueusoos use
sees so usososesoo oso
Eoss use meooEee

encepo as» so Laos so_u_mossou

us. mxs_4 ssoso smmxuom mo sumsmsum mmesm><

m.o_ mxs_4 noose swsu_3 so somsosum woeso><

_.s mxs_o ssoso __< so somsosom omese>s

emm.mm sopueoosseoos so useoses

es. eo_mees

up asses _eosososeseuesess
m asses —eusoeuseseusousH
e, asses so guess:
m m~_m

we ssoso

84-

.msssmmssm

ism so woos—cu
ms» sssosz muses
lusesou usososssu
moss» Eoss use
«senses moss» ass

.usuEusesoo muse
isum uoos use so
so senses use use
usmEusesoo msssoos
ismsm pesos—soss
im< as» so «sesame
soousoos so uousss
isoo ms ssosm ass
N.
s.
o.
.a.
so
oms
os
oss
mm

a d

p
p
—
mm

sosusmossou

mess; noose soozuom so sumsosum omeso><
mess; saoso sssusz so sumsosum omese><
mess; noose __< so sumsosum mueso><
sosueoosssoom so usoosos

sosmsos

Aseososoceseesooss sssv exsso so ceases

sosusmoseou
mess; ssosm soozuom so sumsosum oueso><
mess; sooso sssusz so sumsosum ooeso><
mess; noose _s< so sumsosum omeso><
sosueuosssoos so osmoses
xusmsoo

exsss PeasoEusesouesoss

mess; seusosuseoousouss
messe so sense:
oesm

vs azosw

85

researchers, public agencies, and private corporations. The projects
mentioned included solar water heating systems, wind energy, methane
gas production from manure, energy education curriculum development,
utilization of' waste heat, the relationship of construction and
landscape material to energy use, improving energy information for
community decision-making, and energy audit systems for homes and

buildings.

Analysis of the Communication Relationships

 

In analysing the communication relationships described by the
study population respondents, the NEGOPY program found communication
networks for each of the three separate communication functions, as
well as, for all—function communication (the three communication
functions considered together). The analysis identified the network
values, linkages, and the cliques including size, properties, and
composition for each of the four network structures (see Table 4-5).

The reader is again reminded of the glossary of terms which is
located in Appendix A. The glossary can be of assistance in better
understanding the structural values determined by the analysis.
Definitions are also given within the text where appropriate.

The all-function content network and the production content
network were found by the network analysis to have the same
structural values (see Figure 4-1). Both networks were composed of
152 members with 1,683 links or relationships. Of the 1,683 links,
1,292 (77%) were of interdepartmental relationships and 391 (23%)
were of intradepartmental relationships. The percent of reciprocated

relationships was 25.3% for each network. This property means that

86
approximately a fourth of the reported relationships were
agreement by each party.

in
The range of the strength or frequency for
all of the relationship's within each network was from one hour to 90

‘

  

"---x
x .

x - Isolate

--- at - Isolate Link
0 - Other

Key to Symbols
- Clique
r- Clique Member
_ - Clique Link

Figure 4-1.

Figure representing the energy education networks for production
content and for all-function content.

hours per month with a mean strength of 2.95 hours per month.

include dyads,

The communication roles for the members of each network did not
liaisons,

or tree nodes. There were, however,
thirty-one bridge links, nine isolate type ones, fourteen isolate

87

type two's, and four others. 0f the 31 bridge links, 22 of them
involve individuals Lfrom the College of Agriculture and Natural
Resources. The nine isolate type ones (individuals with no
communication links) are all from different academic departments.
However, five of the nine departments are within the College of
Natural Science. The majority of the isolate type two's have links
to group members from Agricultural Economics and Botany and Plant
Pathology. Two of the four others are from the Dean's Office,
College of Arts and Letters.

The network analysis identified two distinct cliques within each
of the two networks. As noted before regarding the two networks
having the same structural values, the cliques in the all-function
network are identical to the cliques in the production network.
Because of the size of the first clique, the NEGOPY program could not
perform an analysis for density, percent of reciprocation, and number
of links. Although the clique is unusually large, the group met the
criteria for a clique in that each group member has direct linkages
to each other member, the clique is composed of more than two
members, and 50% or more of each member's links are within the group.
For this reason, the NEGOPY program could not break the group down
into smaller cliques for further analysis of relationships and meet
the given criteria for being a clique. The group is composed of 121
members from 42 of the 51 departments represented by the study
population. The College of Agricultural and Natural Resources has
the largest number of members with 60, fellowed by the College of
Natural Science with sixteen and the College of Engineering with

fourteen.

88

The second clique within the two networks is composed of five
members, four from the Department of Park and Recreation Resources
and one from the Department of Urban Planning and Landscape
Architecture. Each group has 14 links with a density of .70. Of the
fourteen links within the clique, two of the links were
interdepartmental and twelve were intradepartmental links. The
percent of reciprocation is 33.33%. The average strength is 10.6 for
within group links and 4.7 for links between group members and
network members outside the clique. The average strength for all

group links is 4.1.

o X ’1! X
x \a /
1' l x
’1
X-..
a!
r “*~x x
‘\
I”’ ‘\\\ X
1!” X ‘ \\
x x X
Key to Symbols
0 - Clique 0—0 - Isolated Dyad
* - Clique Members s--x - Isolate Link
x - Isolate
Figure 4-2.

Figure representing the energy education network for innovation
content.

The third network, the innovation content network (see Table 5
and Figure 4-2), is composed of 152 members with 1,446 links. The

percent of reciprocated links is 23.1%. Interdepartmental

89

relationships comprised 1,105 (76%) of the 1,446 links and
intradepartmental relationships comprised 341 (24%) of 'the ‘links.
The range of strength for the links is from 1 hour to 50 hours per
month with a mean strength for all links of 2.3.

The communication roles for the innovation content network do
not include group linkers as there is just one clique in the network.
There are, however, within the network one dyad, nine isolate type
ones, and ten isolate type two's with links to group members from ten
different departments. Seven of the network members who are isolate
type two's were also isolate type two's in the two previously
described networks. However, three of the seven's group links are
different in the innovation network.

The dyad is composed of two members of the Urban Planning and
Landscape Architecture Department. The nine isolate type ones are
from eight different departments with the College of Natural Science
having five of the nine isolates. The Physics Department has two of
the isolates. In comparing the nine isolate type ones in this
network to the nine isolate type ones in the all-function content
network and the production content network, it is found that six of
these individuals are isolate type ones in all three networks.

The only clique for the network is composed of 131 members from
forty-six of the fifty-one departments represented by the study popu-
lation. The College of Agricultural and Natural Resources has 62
members in the clique. In addition, the College of Natural Sciences
has 16 members; the College of Human Ecology has 15 members; and the
College of Engineering 14 members. Again, because of the large clique

size, the NEGOPY program could not perform an analysis for density,

9O

 

Key to Symbols

9 - Clique I -

f - Clique Member --- y -

I. - Liaison f' -
Figure 4-3.

Figure representing the energy education
content.

Isolate
Isolate Link
Other

Tree Node

network for maintenance

91

number of links, and percent of reciprocation. As with the first
clique in the all-function network and the production network, the
group met the criteria for being a clique and could not be broken
down into smaller cliques which could have been further analysed.

The final network, the maintenance content network (see Table
4-5 and Figure 4-3), presents a wide variety of communication network
properties. The network is composed of 152 members with 754 links.
Of the 754 links, 457 (61%) are interdepartmental relationships and
297 (39%) are intradepartmental relationships. The range of strength
for the links is from 1 hour to 30 hours per month with a mean
strength for all links of 2.4 hours per month. The percent of
reciprocated links is 16.8%. The communication roles identified
within the network include 27 isolate type two's, 21 isolate type
ones, two tree nodes, 50 others, and four liaisons from four
different departments.

The 27 isolate type two's represent 25 different departments.
The Departments of Agricultural Economics and Family and Child
Ecology have two members each who are isolate type two's in this
network. In addition, three of these isolate type two's are found in
all four networks. Only one of the isolates maintained the single
link to the same group member in all four networks.

The twenty-one isolate type ones are from eighteen different
departments. The Department of Physics, Botany and Plant Pathology,
and Family and Child Ecology have two members each who are isolate
type ones in this network. The College of Natural Science has eight
of the twenty-one isolate type ones. Four persons were isolate type

ones in all four networks.

92

One of the tree nodes, a member of the College of Agriculture
and Natural Resources, links with a member of his department to
another member from a different department in the same college who is
an isolate type two. The second tree node, a member of the College
of Human Ecology, links with a member of the College of Agriculture
and Natural Resources to another College of Human Ecology member (in
the same department as the tree node) who is in isolate type two.

As evidenced by the high number of linking roles, the network
has five cliques. The first clique has seven members with 14 links
and a density of .33. Interdepartmental relationships comprise 8
(57%) of the links and 6 (43%) are intradepartmental relationships.
The percent of reciprocation is 32.4%. The average strength of links
is 4.6 hours per month, with 7.7 as the average strength of links
within the group, and 1.9 as the average strength for links between
group members and outside group members. The clique is composed of
one member from the Office of Student Activities, two from the
Department of Zoology, three from the Department of Entomology, and
one from the Department of Higher Education Administration.

The second clique is the largest with nineteen members and six-
ty-six links, but it has the lowest density of .19 and the lowest
percent of reciprocation at 15.18%. The sixty-six links are split
between 42 (64%) interdepartmental relationships and 24 (36%)
interdepartmental relationships. The average strength of all group
links for the clique is 2.7 hours per month, with 3.7 as the average
strength of links within the group, and 1.3 the average strength of
links between group members and outside group members. The clique is

composed of thirteen faculty from five departments within the College

93

of Agricultural and Natural Resources; four faculty from three
departments within the College of Social Science; and two faculty
from two departments within the College of Human Ecology.

The third clique has the second lowest number of members with
four and a perfect density of 1.0. The group has 12 links and a
percent of reciprocation of 20.0. The average strength of all group
links is 6.1 hours per month, with an average strength of 9.1 for the
within group links, and .9 for the links between group members and
outside group members. The group is composed entirely of faculty
from the Science and Mathematics Teaching Center.

Two departments from the College of Agricultural and Natural
Resources comprise the fourth clique which is the second largest with
fifteen members and 140 links. Interdepartmental relationships
comprise 10 ( %) of the links and 130 (93%) are intradepartmental
relationships. Fourteen members are from Agricultural Engineering
and one is from Food Science. The group has a density of .67 and a
percent of reciprocation of 27.9%. The average strength of all group
links is 1.6 hours per month, with 1.7 the average strength for
within group links, and 1.2 for links between group members and
outside group members.

The final clique is the smallest group with three members and
two links. They are from three departments within the Engineering
College. The density for the group is .67 and the percent of
reciprocation is 0.0%.

Hypotheses

 

H]: The communication behavior of the involved

personnel in energy education will be defined by

94

the network analysis to have regular communication

relationships.

This hypothesis states that regular communication patterns exist
within the energy education personnel. For such patterns to exist,
regular communication concerning energy education must be found
between the participants and each participant must have direct and
indirect linkages with the other participants at least the strength
of one hour per month. Based on the analysis of the data completed
by the NEGOPY computer program, linkages do exist for the energy
education personnel. Following the criteria, a network of
all-function content related to energy education was identified
composed of the 152 members in the study population with 1,692-
linkages. The first hypothesis is supported by the data analysis.

H2: The analysis of the communication content of

the participants based on production, innovation,

and maintenance will identify regular communication

linkages for each content type.

The second hypothesis asserts that linkages based on the
specific energy education communication content of production,
innovation, and maintenance are present within the study population.
The criteria for the existence of these networks are identical to the
all-function content network (i.e. regular communication between the
personnel on these sub-topics and direct and indirect linkages.).
These content specific networks were identified by NEGOPY's analysis
of the data. A network composed of the 152 members in the study
population was determined for each of the function contents of

production, innovation, and maintenance with 1,692, 1,452 and 764

95

linkages for each respective sub-topic. The second hypothesis is
supported as well by the data analysis.

H3: The data analysis will identify cliques

of communication linkages.

This hypothesis states that cliques exist within the network.
Such cliques as defined by Richards (1975) must be composed of three
or more members who have direct linkages to each other and who have
more than 50% of their links within the clique. In addition, no link
nor member may be removed from the clique which would cause the
grouping to not meet the criteria. The NEGOPY's analysis of the data
which was pre-set to this criteria did determine ten cliques within
the networks. The cliques varied in size, number, and density per
content network (see Table 4-5). The data analysis supports the
hypothesis.

H4: Participation in the linkages and the cliques

will not be restricted by departmental membership.

This hypothesis states the existence of multidisciplinary inte-
gration within the energy education personnel. In testing this
statement, the data analysis found all 152 members of the study
population to be in each of the four content networks. In addition,
the study population was from 51 university departments (see Tables 2
and 3). As network participants, all of the population members had
communication access to each other' either: directly' or indirectly
across departmental structures. Indeed, within both the all-function
network and the production network, 1,292 (77%) of the 1,683 linkages

were interdepartmental relationships. The networks for innovation

96

and maintenance had interdepartmental linkages of 1,105 (76%) and 457
(61%) respectively.

In reviewing the cliques for restriction by departmental
membership, the greatest evidence of interdepartmental behavior is
found in the cliques of the all-function, production, and innovation
networks; the networks most directly related to energy education.
The three major cliques of these networks were respectively composed
of 121 (all-function), 121 (production), and 131 (innovation) members
from over 80% of the 51 departments in the study. Because of the
size of the cliques, the interdepartmental and intradepartmental
linkages can not be calculated by the NEGOPY program. Due to the
number of departments represented in the cliques and the strong
percentage of interdepartmental linkages found in the overall
networks, the majority of the clique linkages would be expected to be
interdepartmental. The small cliques in the all-function network and
production network had representation from two departments. The
linkages for these cliques however were primarily intradepartmental
because four of the five members were from the same department.

In addition, only one of the ten cliques identified by the data
analysis within the four networks was composed entirely of members
from one department. This clique was in the maintenance network and
was composed of four members. The remaining nine cliques were
composed of members from two or more departments. The percentage of
interdepartmental linkages for the six cliques which are calculable
were 29%, 29%, 57%, 64%, 7%, and 100%. Friedkin found
multidisciplinary integration to exist. with interdepartmental

linkages of 39%. Using his criteria as a test and in view of the

97

other results noted, the hypothesis is supported by the data
analysis. Membership in the network linkages and cliques was not
restricted by departmental membership.

Chapter V presents conclusions based on an interpretation of the
data analysis and results in relation to Blau's postulation and
Friedkin's findings. The chapter concludes the report of the study

with suggestions for future research.

CHAPTER V
SUMMARY, CONCLUSIONS AND RECOMMENDED RESEARCH

Introduction

 

The purpose of this study was to determine the existence of
multidisciplinary integration within the personnel involved with the
energy education projects at Michigan State University in 1981. The
study extended Friedkin's (1979) finding of multidisciplinary
integration in a research setting against Blau's postulation that
such integration is unlikely in academic settings. The literature
review supported the use of communication network concepts and
analysis procedures in establishing the study hypotheses and the
investigation methodology. The results of the data analysis
indicated the existence of an integrated network of multidisciplinary
energy education personnel. The conclusions of the study and

recommended future research are presented in this final chapter.

Conclusions

 

Several conclusions may be drawn from the results of the study.
First, the determination of the multidisciplinary social network
within the energy education personnel supports Friedkin's suggestion
that multidisciplinary integration can occur, particularly in a
research setting. This is a conclusion which is contradictory with
Blau's postulation that such integration is unlikely in the college

and university setting.

98

99

The identification of multidisciplinary integration in this
study builds a broader foundation that a collegial atmosphere does
exist in the pursuit of academic research. Academic specialization
and departmentalization are not necessarily impediments to
interdepartmental collaboration within research settings. The
concerns cited by Bess (1982), Dressel and Reichard (1970), Blau
(1973), and Mintzberg (1979) about shared communication and
cooperation may not be applicable in the research setting.
Certainly, Hagstrom's (1965) observation is appropriate that
researchers from different disciplines often combine to solve
problems and after completing the task re-form their affiliations.

The study's result is instructive for higher education adminis-
tration. The knowledge that regular interdepartmental communication
can occur and form into a network is in itself an important dimension
of the college setting. The administrator can play a significant
role in encouraging this communication as well as utilizing the
network for furthering academic pursuits. This study and the
literature review indicated how the amount and type of communication
flowing within and across departments can be assessed using
communication network analysis. Depending on the topic or the need,
the analysis can focus on identifying communication barriers, major
communication linkers, or the general communication patterns of the
system studied. The analysis can give insight to the administrator
on how to introduce new ideas and how to improve communication within
the academic setting. Such information would be particularly

beneficial for a campus Research and Development Center.

100

A second conclusion is related to the methodology of the study.
The divergent results found between Blau's postulation and this study
and Friedkin's study may be explained by the differences in the
methodologies of the studies. Blau's postulation was based on a
general study of a sample of American colleges and universities. He
did not focus on research institutions, nor did he stratify his data
analysis to look at communication patterns by type of institution,
academic discipline, or faculty member. Maybe most important of all,
Blau did not use a communication network analysis methodology similar
to Friedkin's study and this study. The two latter studies surveyed
their population members about with whom they communicated on a given
topic and, in the case of the present study, used a computer program
to generate the patterns of communication behavior. Blau's
methodology was not sophisticated and focused enough to identify
actual communication patterns and properties.

The final conclusion concerns the content of the communication
shared by the participants. The finding of networks, particularly a
personal network, is not supported in the literature by Blau (l973),
Dressel and Reichard (1970), and Mintzberg (1979). They describe the
academic setting more as a series of isolated academic behaviors.
Mintzberg noted that professionals in this setting are unwilling to
extend themselves to interact with other non-disciplinary members.
The members are described as staying within their primary work
setting.

The finding of greatest integration within the production
network and the all-function network is, however, supported by

Hagstrom's (1965) comment on scientists' typically attacking a

101

problem in an integrated way. The noteworthy finding, as well, is
the existence of a personal or maintenance network. It is reasonable
to conclude that the researchers in this study share a collegial
atmosphere of friendship in addition to cooperation in their

professional pursuits.

Recommended Research

 

The conclusions of this study have implications for future
research in multidisciplinary integration for studies both at the
present setting, as well as, at other campuses. Because the present
study's conclusion supports Friedkin's suggestion that
multidisciplinary integration may be found and encouraged at research
settings, the same study should be replicated at other research
settings in order to enhance the validity of these findings. Further
research results are needed from other campuses in order to arrive at
a generalization about multidisciplinary integration in research
settings.

The appropriateness of the network analysis methodology as a
dependable investigative strategy adds to the reliability of this
study. The data measurement and analysis methodology used in this
study was found in the literature review to be an effective means for
identifying communication patterns. Of most importance, researcher
error is reduced by the use of algorithms in a data analysis computer
program such as NEGOPY.

In addition, the setting of the present study is comparable to
other campus settings. The type of faculty person, the academic

structure, and the involvement in a research problem, such as energy

102

education, are very similar with many college campuses. As noted in
the first chapter, the limitations of the study are not in the
results or methodology.

Two ancillary studies are also suggested for fUture research.
The first suggestion is to investigate why multidisciplinary
integration can be found in research settings, particularly in the
sciences, and have not been identified in other selected academic
settings. The communication patterns of faculty in different
settings could be surveyed, stratified, and analyzed as to how they
compare with each other. The second suggestion is to do fellow-up
studies comparing personal-friendship networks with production
networks. The personal-friendship networks may be more wide spread
on college campuses than noted in the research. These networks may
also play a significant role in maintaining existing
multidisciplinary networks.

Concerning the research of multidisciplinary integration at the
present setting, the suggestion is to assess the stability of the
integration and the communication patterns over time. The study
population could be investigated to determine if the same research
content is being shared, if new research content is being
communicated, or if any on-going interaction at all is occurring.
Such an investigation could detect new, if any, networks, cliques,
and communication roles.

The faculty who played communication bridging roles could also
be studied for stability as well, to ascertain if they regularly play
such roles in research communication on campus. These persons may be

critical communicators who need to be considered in the dissemination

103

of information. The type of persons in these roles could also be
studied. Variables could be determined for them (i.e. status, length
of tenure, type of research, type of research support, etc.) and
analysis conducted on the relationship of the variables with the
persons' communication roles.

One of the benefits in determining communication networks is the
identification of member information needs and roles. Isolates and
information flow blockages can be located.and organizational inter-
vention strategies developed to open up these communication passages.
A practical reason for replicating this study could be to collect
data for improving the research information flow in this setting. A
secondary value of this study could be to stimulate and to enhance
the collegial atmosphere in the present study's setting.

The enhancement of the research information flow is suggested as
a function of the Research and Graduate Studies Office. This office
has the primary responsibility for coordinating campus research and
has records (M1 current faculty involvement ir1 this activity. The
network analysis of research communication could be an ongoing
process focusing on separate research topics and/or personnel. The
analysis would produce the patterns of who is communicating with
whom, as well as, who is doing what. In this way, the opportunity is
available to identify the current campus research, provide better
coordination of it, and enchance the personnel's flow of

communication through intervention activities.

APPENDIX A
GLOSSARY OF TERMS

Bridges --

Centrality --

Clique --

GLOSSARY

Group members who have linkages to one or more groups
(Albrecht, 1978).

A person is in a central position when that person is
strategically located on the shortest communication path
connecting pars of others (Bavelas, 1948). These po—
sitions can therefore facilitate, impede or bias the
transmission of messages (Freeman, 1977).

A subset of an organization consisting of three or more
members each in a symmetric relation to each other member
of the subset, and provided there is no element outside
the subset that is in a symmetric relation to each of the

elements of the subset (Luce and Perry, 1949).

Communication Network -- A series of linkages based on communication

content (Farace, Monge, and Russell, 1977). The linkages
comprise all possible communication pathways between

members within a system's boundary (Richards, 1975).

Communication Roles -- Group members are differentiated according to

Density --

Dyad --

their range of communication linkages (Albrecht, 1978).
The roles exert influence over the content and flow of
information in the system (Likert, 1961).

The degree to which sets of system members are
interlinked or interconnected (Farace, Monge and Russell,
1977).

Two nodes who are linked to each other (Richards, 1975).

104

105

Energy Education Projects -- Those areas related to the discovery and

dissemination of energy concepts and data. Energy
Education is concerned with research, instruction, as
well as the administration of energy projects. The
energy education field does not include retrofitting of
campus buildings for energy conservation, but it would
include educational materials on energy awareness,
research into alternative energy sources, and the

development of energy management programs.

Invisible Colleges -- An elite network of mutually interacting and

Isolates --

Liaisons --

Links --

Nodes --

Reciprocal --

productive scientists within a research area (Price,
1963).

Individuals who have little if any contact with other
members of' the~ organization (France, 1978). Richards
(1975) further defines a type one isolate as a person
with no links and a isolate type two as a person connect-
ed to only one other person, who is not an isolate.
Individuals who link groups but who are not themselves
group members (Albrecht, 1978).

Measured relationships between individual nodes (Farace
and Mabee, 1978).

The entities comprising the network. Usually nodes are
people, but nodes my also represent other relationships
such as among roles and among organizations (Farace and
Mabee, 1978).

An agreement by the partner to the relationship indicated

by the respondent (Rogers and Kincaid, 1981).

Reciprocity ~-

Rules --

Social Network

Tree Node --

106

The percentage of links a person indicates which are also
agreed upon by those who are nentioned (Farace, Monge,
and Russell, 1977).

Guide the structure of interaction and hence the type of
messages communicated. Selection of topics, the length
of discussion, and the number of disturbances are reg-
ulated by "rules" (Albrecht, 1978).

-- A social structure composed of members connected
through communication of common topic, interest, or
discipline (Granovetter, 1976).

An individual who is not a member of a clique but who has
a relationship with an isolate type two and with a clique

member (Richards, 1975).

APPENDIX 8
INSTRUCTIONS AND NETWORK ANALYSIS QUESTIONNAIRE

107

February 19, 1981

T0: Campus Energy Personnel

FROM: Ron Stump
108 Student Services Bldg.
3-3860

I would like your assistance with a study of the communication pattern bet-
ween the faculty/staff involved with campus energy education, research, and
extension.

The aim of the study is to identify and to improve the quality of information
exchange. It is hoped that as a result of this study infbrmation about re-
search results, available funding grants, services, and proposed studies will
be more readily accessible to the campus energy faculty/staff. To accomplish
this goal you and other faculty/staff in the energy field at Michigan State
University are being asked to describe your campus communication contacts re-
lated to energy. With this information, the information flow related to en-
ergy can be characterized fbr Michigan State University.

The goals and the nature of this study are familiar to Jon Bartholic, Agri-
cultural Experiment Station; Bill Cooper, Zoology; Herman Koenig, Center for
Environmental Quality; Tom Edens, Agriculture Economics; and Adger Carroll,
Cooperative Extension. You may wish to contact me at 3-3860, or one of them
if you have any questions about the study.

The description of your communication will be obtained by the enclosed ques-
tionnaire. Based on ore-testing, it is estimated that it will take you
roughly fifteen minutes to complete the questionnaire. Please be as thorough
and honest as possible in your replies.

Please return the completed questionnaire in the enclosed envelope within a
week to Ron Stump, 108 Student Services Bldg. A summary of the results of
the study will be sent to you, if you indicate an interest in receiving a
copy. Thank you for your assistance.

RS/ds

108

PERSONAL DATA SHEET

Na

 

me .
First MiddleInitial Last

Department
College
Campus Address Campus Phone

Are you the Director ( ). Co-director ( 1. or a staff member of an energy project?

Are your energy-related activities at the University primarily involved with:
(check all that apply)

 

 

 

Teaching Research Public Service
Title of Project(s) 1
or Activity(s):
~ 2.
3.
4.

 

Products or major outcomes expected

 

 

 

Primary audiences/users of project outcomes

 

 

Would an updated directory of MSU energy projects and personnel be of value to you?

 

 

Would you like a summary of the study's results? Yes ( ) No ( )

Please complete the attached Energy Contact Questionnaire and return it with this
personal data sheet in the enclosed envelope within a week.

109
ENERGY CONTACT QUESTIONNAIRE INSTRUCTIONS

On the attached questionnaire, you will be asked to describe your en-
ergy related conmunication with other energy personnel on-campus. You will
be presented with a list of faculty/staff at Michigan State University and
asked to-report how much you communicate with them in energy matters, and
generally what function such communication serves.

The list of names are presented with three columns next to them. These
columns are headed by the description of three communication functions. The
functions are Work-Related (getting the work done), New Ideas-Related (new
ideas and ways to accomplish goals), and Personal-Related (social relations).

Please read through the list of names. When you come to a person with
whom you communicate estimate roughly how many hours you communicate with
this person in an average month about work, ideas, and personal topics. Cons
sider the average month to be during a normal academic term.

Record your estimated number of hours per average month in each corres-
ponding box. Round parts of hours up to the nearest hour. Remember that com-
municating with a person occurs whenever you contact someone or when someone
makes contact with you through face-to-face conversation, meetings, written
work, telephone conversations, etc.

If you have no contact with a particular person in an average month of
a normal term, leave the box blank and go on. Likewise, if you have work-
related contact with a person but no person-related contact, mark the work-
related box with your estimate and leave the person-related box blank. When
you come to your name, circle it and leave the boxes blank.

A sample of the questionnaire is presented on the next page.

110
ENERGY CONTACT QUESTIONNAIRE

SAMPLE

(Please circle your name)

NUMBER OF HOURS COMMUNICATING IN AN AVERAGE MONTH

 

 

WORK RELATED IDEAS RELATED PERSONAL
01 John Daly ;F’
02 Ma Cruso -? / 5"

 

A

03 ( Laura Upiétj

04 Marcy Sachs

 

 

 

 

 

 

EXPLANATION OF SAMPLE:

In this example, Laura Uplet has estimated the number of hours she com-
municates with other energy personnel in an average month (during working and
non-working hours). She first circled her name on the list. Next, she esti-
mated that in an average month, she coumunicated with John Daly about five
hours about work-related matters and not at all about ideas and personal mat-
ters. Next, she estimated that in an average month, she conmnicated with
Mary Cruso for two hours concerning her work, not at all about new ideas, and
15 hours concerning personnal matters. She does not comnunicate with Marcy
Sachs in an average month.

Mr that conmunication contacts occur when you contact someone else

03 they contact you, by phone, memo, or in person.

111
ENERGY CONTACT QUESTIONNAIRE

Q§S_CRIPTION or comumcmou FUNCTIONS
WORK RELATED: I

NEW IDEAS RELATED:
PERSONAL RELATED:

(Please circle your

name)

Writing reports and proposals, research, teaching. etc. '
New ways to do things, innovative things to do, etc.

Interpersonal relationships, soCial interaction, coun-
seling people.

NUMBER OF HOURS COMMUNICATING IN AN AVERAGE MONTH
WORK RELATED IDEAS RELATED PERSONAL

 

 

001

Mike Abkin

 

002

Edith Allen-Schult

 

003

Robert Anderson

 

004

Charles Arntzen

 

005

Jes Asmussen

 

006

Fred Bakker-Arkema

 

007

Robert Bandurski

 

008

Jon Bartholic

 

009

George Bird

 

010

Roy Black

 

011

Robert'Boger

 

012

Georg Borgstrom

 

013

Maureen Bowman

 

014

Roger Brook:

 

015

Gal en Brown

 

016

John Cantlon

 

 

 

 

112
NUMBER OF HOURS COMMUNICATING IN AN AVERAGE MONTH

(Please circl; your WORK RELATED‘ IDEAS RELATED PERSONAL
name

 

017 Adger Carrol

 

018 Duncan Case

 

019 Daniel Chappelle

 

020 Larry Connor

 

021 Bill Cooper

 

022 Eric Crawford

 

023 Gerald Crawl ey

 

024 Doug Cron

 

025 Carol Dahl

 

026 Don Dickerson

 

027 Don Dickman

 

028 David Dwyer

 

029 John Eastman

 

030 Tom Edens

 

031 Don Edwards

 

032 Earl Erickson

 

033 Merle Esmay

 

034 Vince Farace

 

035 Anne Field

 

036 Philip Filner

 

037 Fred Fink

 

033 Cynthia Fridgen

 

039 Joe Fridgen

 

040 Stuart Gage

 

 

 

 

041 Mitchell Geasler

 

113
NUMBER OF HOURS COMMUNICATING IN AN AVERAGE MONTH

(Please circl; your WORK RELATED IDEAS RELATED PERSONAL
name

 

042 John Gerrish

 

043 Norman Good

 

044 Eric Grulke

 

045 Gerald Haarer

 

046 Fred Hall

 

047 Roger Hamlin

 

048 James Hanover

 

049 Craig Harris

 

050 Bud Hart

 

051 Irene Hathaway

 

052 Martin Hawley

 

053 Dean Haynes

 

054 Sherwood Haynes

 

055 Denny Heldman

 

056 Zane Helsel "

 

057 Martin Hetherington

 

058 Donald Holecek

 

059 Allan Hollingsworth

 

060 Hank Huber

 

061 John Huber

 

062 Dennis Hudson

 

063 Peter Kakela

 

064 Dennis Keefe

 

055 Joanne Keith

 

 

 

 

055 Ron Kerber

 

114
NUMBER OF HOURS COMMUNICATING IN AN AVERAGE MONTH

(Please circle your WORK RELATED‘ IDEAS RELATED PERSONAL .
name)

 

067 Kyle Kittleson

 

068 Herman Koenig

 

069 Kathryn Kolasa

 

070 Otto Krauss

 

071 Rex Lamore

 

072 Nancy Landes

 

073 Larry Libby

 

074 Theodore Loudon

 

075 Les Mack

 

075 Dick McLeod

 

077 Tom McPeak

 

073 Martha Molder

 

079 Lewis Moncrief

 

030 Don Montgomery L

 

031 William Mooney:

 

032 Bonnie Morrison

 

083 Denton Morrison

 

084 Robert Muth

 

085 Bob Neumann

 

086 Dick Niehoff

 

087 Steve Orlick

 

088 Gerald Park

 

089 Lucas Parsch

 

O90 Wally Piper

 

 

 

 

091 Carl Ramm

115
NUMBER OF HOURS COWNICATING IN AN AVERAGE mNTli

(Please circlg your WORK RELATED IDEAS RELATED PERSONAL
name . .

 

092 C. A. Reddy

 

093 Bill Reusch

 

094 Alvin Rippen

 

095 Michael Rogers

 

095 Ron Rosenberg

 

097 Lowell Rothert

 

098 Alan Rotz

 

099 Lester Schick

 

100 Ger Schultink

 

101 Gerald Schwab

 

102 James Shaffer

 

103 Robert Snow

 

104 Larry Spinners

 

105 Aji t Sri vastava

 

106 Jim Steffe

 

107 Mil ton Stei nmuel ler

 

108 Bill Stout

 

109 Barbara Stowe

 

110 Otto Suchsland

 

111 William Thomas

 

112 Ed Talbert

 

113 Larry Tombaugh

 

114 R. L. Tummala

 

115 Gary VanEe

 

 

116 Eileen VanRavenswaay

 

 

116
NUMBER OF HOURS COMMUNICATING IN AN AVERAGE MONTH

(Please circl; your WORK RELATED IDEAS RELATED PERSONAL
name .

 

ll7_§ancy_yeenstra

 

118 Maurice Vitosh

 

119 John Waller

 

120 Bruce Wilkinson

 

121 Sylvan Wittwer

 

122_‘[ruman Woodmf f

 

123 David Wright

 

124 Melvin Yokoyama

 

125 Roland Zapp

 

126 James Zuiches

 

 

 

 

 

 

List additional names below if not all of your campus energy contacts are pro-
vided. Energy personnel are those campus full-time faculty/staff directly in-
volved in energy education, research, and extension. 00 NOT include person-
nel doing energy conservation projects nor University facTTTties or those ad-
ministrative officers involved with payroll, personnel, purchasing, etc.

 

 

 

 

 

 

 

 

REFERENCES

REFERENCES

Aiken, M. and Hage, J., "Organizational Interdependence and Intra-
Organizational Structure.“ American Sociological Review, £I3, #6:
912-928, December, 1968.

 

Alba, R.D. and C. Kadushin, "The Introduction of Social Circles: A New
Measure of Social Proximity in Networks." Sociological Methods and
Research, 5: 77-102, 1976.

 

Albrecht, T.L., Communication and Perceptions of Organizational Climate:
An EmpiricaT‘Analysis. Unpublished doctoral dissertation, Depart-
ment of Communication, Michigan State University, 1978.

 

Barnes, J.A., Social Networks. Reading, Massachusetts, Addison-Wesley,
Modular Publications in Anthropology. 26: 1-29, 1972.

 

Barnes, J.A., "Network Analysis: Orienting Notion Rigorous Technique,
or Substantive Field of Study." in P.W. Holland and S. Leinhardt
(Eds.), Perspectives on Social Network Research. New York, New
York: Academic Press, 1979.

 

Bavelas, A., "A Mathematical Model for Group Structure." Applied
Anthropology 7: 16-30, 1948.

 

Berlo, D.K. "Human Communication: The Basic Proposition." Unpublished
manuscript, Department of Communication, Michigan State University,
East Lansing, 1970.

Bernard, H.R. and Killworth P.D., "0n the Social Structure of an
Ocean-Going Research Vessel and Other Important Things." Social
Science Research, 2: 145-184, 1973.

 

Bernard, H.R. and Killworth P.D., "0n the Structure and Effective Socio-
metric Relations in (1 Closed Group Over Time." Connections, 1:
44, 1978.

 

Bernard, H.R., Killworth, P.D., and Sailer, L., "Summary of Research on
Informant Accuracy in Network Data, and on The Reverse Small World
Problem." Connections, 4, #2: 11-25, 1981.

 

Bernstein, P., "Oil a-plenty but OPEC calls shots." The Detroit Free
Press. 88, September 1, 1980.

 

Bess, J.L., University Organization. New York, New York: Human
Sciences Press, Inc. 1982.

117

 

118

Beyer, J.M. and Lodahl, T.M., "A Comparitive Study of Patterns of
Influence in United States and English Universities.“
Administrative Science Quarterly, 21: 104-129, 1976.

 

Blau, P.M., The Organization of Academic Work. New York: John Wiley
and Sons, Inc., 1973.

 

Blau, P.M., "Parameters (H: Social Structure." American Sociological
Review, 39, #5: 615-635, October, 1974.

 

Boissevain, J., Friends of Friends; Networks, Manipulators and
Coalitions. Oxford: Basil Blackwell, 1974.

 

 

Bolman, F.deW., "Needed Research in Administration of Higher
Education." Educational Record. XLVI: 166-176, Spring, 1965.

 

Bott, E., "Urban Families: Conjugal Roles and Social Networks." Human
Relations. 8: 345-384, 1955.

Breiger, R.L. "Career Attributes and Network Structure: A Block Model
Study of a Biomedical Research Specialty." American Sociological
Review, 41: 117-135, 1976.

 

Burgess, R.L., "Communication Networks and Behavioral Consequences."
Human Relations. 22: 137-159, 1968.

 

Burt, R.S., "Models of' Network Structure." Annual Review of
Sociology, 6: 79-141, 1980.

 

Burt, R.S. and Bittner, W.M., "A Note on Inferences Regarding Network
Subgroups." Social Networks, 3, #1: 71-88, 1981.

 

Bushnell, D. and Rappaport, 0. (Eds.), Planned Change in Education.
New York: Harcourt Brace Jovanovitch, Inc., 1971.

 

Champagne, A. and Klopfer, L., Criteria for Effective Energy Education.
Pittsburg, Pennsylvania: University of Pittsburgh, Learning
Research and Development Center, 1978.

 

Chapin, F.S., "Trends in Sociometrics and Critique." Sociometry, 3:
245-262, 1940.

 

Crane, D., "Social Structure in a Group of Scientists: A Test of the
'Invisible College' Hypothesis." American Sociological Review.
34: 335-352, 1969.

 

Crane, 0., Invisible Colleges: Diffusion of Knowledgey in Scientific
Communities. Chicago: University of Chicago Press,71972.

 

 

Crawford, Susan, "Informal Communication among Scientists in Sleep and
Dream Research." An unpublished doctoral dissertation, University
of Chicago, 1970. Cited by N.E. Friedkin, "University Social
Structure and Social Networks among Scientists.“ American Journal
of Sociology, 83: 1444-65, 1978.

 

 

119

Dressel, P. and Reichard, D.J., "The University Department Retrospect
and Prospect." The Journal of Higher Education, XLI, #5: 387-402,
1970.

 

Erickson, B.H., Nosanchuk, T.A., and Lee, E., "Network Sampling
in Practice: Some Second Steps.“ Social Networks, 3, #2: 127-136,
1981.

 

Farace, R.V. "Organizational Communication". in (L Book (Ed.), Human
Communication: Principles, Contexts, and Skills. New York: St.
Martin Press, Inc., 1978.

 

 

Farace, R.V. and Connelly, R.A., "Organizational Communication
Correlates of Herzberg's Theory of Work Satisfaction." Readings in
Organizational Communication. Unpublished manuscript, Michigan
State University, 1970.

 

 

Farace, R.V. and Johnson, J.D., "Comparative Analysis of Human
Communication Networks in Selected Formal Organizations." Paper
presented at the International Communication Association, New
Orleans, 1974.

Farace, R.V. and Mabee, T., "Communication Network Analysis Methods."
Draft of chapter in P. Monge and J. Cappella (Eds.), Multivariate
Analysis Techniques in Communication Research. New York: Academic
Press, 1978

 

Farace, R.V., Monge, P.R., and Russell, H.M., Communicating and
Organizing. Reading, Mass.: Addison-Wesley, 1977.

 

 

Fenemma, M. and Schift, H., "Analyzing Interlocking Directorates:
Theory and Methods." Social Networks, 1: 297-332, 1978/79.

 

Freeman, L.C., "A Bibliography of Social Networks." Monticello,
Illinois, Council of Planning Librarians Exchange Bibliography,
1170-1171, 1976.

Freeman, L.C., "Centrality in Social Networks Conceptual Clarification."
Social Networks, 1: 215-239, 1978/79.

 

Freeman, L.C. "A Set of Measures of Centrality Based on
Betweenness." Sociometry, 40, #1: 1977.

 

Friedkin, N.E. "University Social Structure and Social Networks Among
Scientists." American Journal of Sociology. 83: 1444-65, 1978.

 

Friedkin, N.E. "The Development of Structure in Random Networks: An
Analysis of the Effects of Increasing Network Density on Five
Measures of Structure." Social Networks, 3, #1: 41-52, 1981.

 

Granovetter, M.S. "The Strength of Weak Ties." American Journal of
Sociology, 78: 1360-1380, 1973.

 

Granovetter, M.S., "Network Sampling: Some First Steps." American

120

Journal of Sociology: 81: 1287-1303, 1976.

 

Guetzhow, H., "Communication in Organizations." in March, J.G. (Ed.),
Handbook of Organizations. Chicago: Rand-McNally, 1965.

 

Hagstrom, W. 0., The Scientific Community. New York: Basic Books,
1965.

 

Hammer, M., "Predictability of Social Connections Over Time." Social
Networks, 2:165-180, 1979/1980.

Hicks, H. The Management of Organizations. New York: McGraw-Hill,
1967.

 

Huse, E. and Bowditch, J., Behavior in Organizations: A Systems

 

 

Approach to Management. Reading, Mass.: Addison-Wesley, 1973.

 

Jacobson, E. and Seashore, 5., "Communication Patterns in Complex
Organizations." Journal of Social Issues, 7: 28-40, 1951.

 

Killworth, P.D. and Bernard H.R., "Informant Accuracy in Social Network
Data." Human Organization, 35, #3: 269-296, 1976.

 

Knoke, D. and Kuklinski, J.H., "Network Analysis." in J.L. Sullivain and
R.G. Niemi (Eds.) anntitive Applications in the Social Sciences.
Beverly Hills: Sage Publications, 1982.

Larkin, T.J., "Network Analysis As An Investigative Tool For
Organizational Communication." Unpublished masters thesis, Depart-
ment of Communication, Michigan State University, 1978.

Leavitt, H.J., "Some Effects of Certain Communication Patterns on Group
Performance." Journal of' Abnormal and Social Peychology, 46:
38-50, 1951.

 

Likert, R., New Patterns of Management. New York: McGraw-Hill, 1961.

 

Lin, N., The Study of Human Communication. Indianopolis:
Bobbs-Merrill, 1973.

 

Luce, D. and Perry, A., "A Method of Matrix Analysis of Group
Structure." Psychometrika, 14: 1949.

 

MacDonald, 0., "Communication Roles and Communication Content In A
Bureaucratic Setting.“ Unpublished doctoral dissertation,
Department of Communication, Michigan State University, 1970.

McConnell, T.R., "Needed Research in College and University Organization
and Administration." in T. Lunsford (Ed.), The Study of Academic
Administration. Boulder, Colorado, Western Interstate Commission
for Higher Education, 1963.

 

Millett, J.D., The Academic Community. New York: McGraw-Hill, 1962.

 

Mintzberg, H., The Structuring of Organizations. Englewood Cliffs, N.
J.: Prentice-Hall, 1979.

121

Mitchell, J.C., "The Concept and Use of Social Networks." in J.C.
Mitchell (Ed.), Social Networks in Urban Situations. Manchester,
England, Manchester University Press, 1969.

 

Mitchell, J.C., "Social Networks." Annual Review of Anthrgpology, 3:
279-299, 1974.

 

Monge, P.R., Edwards, J.A., and Kirste, K.K., "The Determinents of
Communication and Communication Structure in Large Organizations.
A Review of Research." A paper presented at the annual meeting of
the International Communication Association, Chicago, 1978.

Moran, w.F., "The Study of University Organizations." Journal of Higher
Education, XXXIX: 144-155, 1968.

 

Moreno, J.L., Nho Shall Survive? Beacon, New York: Beacon House, Inc,
1953.

 

Mouton, J.S., Blake, R.R., and Fruchter, B., "The Reliability of
Sociometric Measures." Sociometry, 18, 1: 7-48, 1955.

 

Mulkay, M.J., Gilbert, G.N. and Woolgar, 5., "Problem Areas and
Research Networks in Science." Sociology, 9: 187-203 1975.

Mullins, N.C., "The Distribution of Social and Cultural Properties in
Informal Communication Networks Among Social Scientists." American
Sociological Review, 33, #1: 786-797, 1968.

 

Mullins, N.C., "The Development of a Scientific Specialty: The Phage
Group and the Origins of Molecular Biology." Minerva, 10: 51-82,
1972.

Nehnevajsa, J., "Sociometry: Decades of Growth." Sociometry, 18, 2:
304-351, 1955.

 

Perkins, J.A., The University As An Organization. McGraw-Hill, New
York, 1973.

Porter, L.w., Lawler, E.E., and Hackman, J.R., Behavior in
Organizations. McGraw-Hill, New York, 1975.

 

 

Price, D.J.deS., "Networks of Scientific Papers." Science, 149: 510-515,
1965.

Richards, w. A Manual for Network analysis: Using the NEGOPY Network
Analysis Prggram. Institute for Communication Research, Stanford
University, 1975.

Rogers, E.M., Diffusion of Innovations. 3rd Edition, New York: The
Free Press, 1983.

Rogers, E.M. and Agarwala-Rogers, R., Communication in Organizations.
New York: The Free Press, 1976.

 

122

Rogers, E.M. and Bhowmik, D.K., "Homophily and Heterophily: Relational
Concepts for Communication Research." The Public Opinion Quarter-
_Ly, 34: 523-538, 1971.

 

Rogers, E. and Kincaid, D.L., Communication Networks. New York: The
Free Press, 1981.

 

Schriver, J., (Ed.), "Energy and The University: A Guide to Energy
Programs and Resources at Michigan State University." Lifelong
Energy Education Project, East Lansing, Michigan: College of
Education, Michigan State University, 1979.

Schwartz, D. and Jacobson, E., "Organizational Communication Network
Analysis: The Liaison Communication Role." Organizational Behav-
ior and Human Performance, 18: 158-174, 1977.

 

 

Shaw, M.E., "Communication Networks." in L. Berkowitz (Ed.), Advances in
Experimental Social Psychology, 1: 111-147, 1964.

 

 

Stern, R.N., "The Development of an Interorganizational Control Network:
The Case of Intercollegiate' Athletics." Administrative Science
Quarterly, 24: 242-266, 1979.

 

Stokman, F.N., Third World Group Formation in the United Nations: A

 

Methodological Analysis, Amsterdam, University of Amsterdam Press,
1977.

 

Taylor, J., "An Empirical Examination of the Referential Function of
Informal Communication Groups." Unpublished masters thesis,
Department of Communication, Michigan State University, 1976.

Thayer, L., Communication and Communication Systems. Homewood, IL:
Richard Irwin, 1968.

 

Tucker, A. Chairing the Academic Department. American Council on
Education, Washington D.C., 1981.

 

Veysey, L.R., The Emergence of the American University. Chicago:
University of Chicago Press, 1965.

 

Weiss, R.S., and Jacobson, E., "A Method for the Analysing of the
Structure of Complex Organizations." American Sociological Review,
20: 661-668, 1955.

 

White, H.C., Boorman, S.A., and Breiger, R.L., "Social Structure from
Multiple Networks. I Blockmodels of Roles and Positions."
American Journal of Sociology, 81: 4, 730-780, 1976.

 

Whitten, N.E. and Wolfe, A.W., "Network Analysis." in J.J. Hongimann
(Ed.), Handbook of Social and Cultural Anthropology, Chicago:
Rand McNally, 1973.

Wolfe, A.W., "The Rise of Network Thinking in Anthropology." Social
Networks, 1, #1: August, 1978.

123

Woodburne, L.S., Principles of College and University Administration.
Stanford: Standford University Press, 1958.

 

"I11111111111111.1111“