COORIENTATIONAI. SIMILARITY

TOWARD PROCEDURAL ASPECTS
OF COMMUNICATION: ‘

A STUDY OF COMMUNICATION

BETWEEN EXTENSION. AGENTS
AND THEIR SUPERVISORS

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
HAMISH MILTON, RUSSELL
1972

 

- ..-._..——.7-.H‘}!

25§NIILH¥

  

IIIIIIIIIIZIIIIIIIIIIIIIIIIIIIIIIIIIIISIII

 

ABSTRACT

COORIENTATIONAL SIMILARITY TOWARD PROCEDURAL
ASPECTS OF COMMUNICATION:
A STUDY OF COMMUNICATION BETWEEN EXTENSION
AGENTS AND THEIR SUPERVISORS

By
Hamish Milton Russell

This research examined the association between
interpersonal attraction between members of extension
agent/supervisor dyads and their level of coorientational
similarity toward the procedural aspects of their communica-
tion.

On the basis of extensions of Newcomb's coorienta-
tion model, predictions were made about the effect of two
levels of similarity--agreement and accuracy.

Data were obtained separately from both members of
one hundred supervisory dyads in the Michigan Cooperative
Extension Service, through questionnaires completed in
group situations. These data related to the members'
background, their opinions of what would be ideal super-
visory communication between field agents and their
supervisors, how they saw their actual dyadic communication,

their predictions of how the other member of their dyad

Hamish Milton Russell

would perceive their communication, and their satisfaction
with role achievement and their relationship.

Using a factor analysis of sixty items relating to
ideal communication, four independent dimensions of pro-
cedural communication were identified. These dimensions
focus on different aspects of sharing or interdependence
in the dyad--new ideas, family and personal activities,
the responsibility for seeing tasks through to completion,
and the responsibility for seeing that the tasks are in
fact undertaken.

In describing their communication, the supervisors
saw the interaction as more interdependent than did their
field staff. The supervisors also tended to be more skilled
at recognizing divergent perceptions.

Both supervisors and field staff wanted more com-
munication about new ideas; while supervisors alone wanted
an overall increase in communication about all aspects of
their interaction.

The most outstanding feature of the study was the
high overall level of both agreement and accuracy with
respect to communication procedures. This high level of
coorientational similarity, combined with high levels of
field agent role satisfaction and high supervisor ratings
of field agents' performance, gives overall support to the
notion of an association between similarity of orientation

and interpersonal attraction.

Hamish Milton Russell

Although a number of significant associations were
found between measures of similarity of orientation toward
communication procedures and measures of attraction, the
degree of similarity did not account for a major part of
the variance in the indicators of attraction.

General support is claimed for the usefulness of
Newcomb's model, but the influence of many other variables
as well as the procedural aspects of communication is
recognized.

Based upon several trends in the data, explicit
discussion between supervisors and field agents about how
they view both their current and their ideal expectations
for communication procedures is recommended as a way to
increase the understanding between the dyad members.

Finally, taking a systems viewpoint, a number of
suggestions are made for further research into the applica-
tion of Newcomb's model to increase our understanding of
communication within organizations. The focus on both the
shared interaction and on the procedural rules that direct
communication is seen as both useful and powerful for

future research.

COORIENTATIONAL SIMILARITY TOWARD PROCEDURAL
ASPECTS OF COMMUNICATION:
A STUDY OF COMMUNICATION BETWEEN EXTENSION

AGENTS AND THEIR SUPERVISORS

By

.9".
‘5’

Hamish Miltod Russell

A THESIS

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

DOCTOR OF PHILOSOPHY
Department of Communication

1972

Accepted by the faculty of the Department of
Communication, College of Communication Arts, Michigan
State University, in partial fulfillment of the require-

ments for the Doctor of Philosophy degree.

Warm/MW

Director of Thesis

Guidance Committee. MM? ”@waiman
% m {0%,‘fi.
«Kim/(i

meb
PM

 

 

To

Rosemary, Caroline and Jennifer

who have shared in this experience with me

iii

ACKNOWLEDGMENTS

I wish to acknowledge the friendship, advice and
interest that I received from each of my guidance committee
members-~Dr. Vincent Farace, Dr. Mason Miller, Dr. Everett
Rogers, Dr. Lawrence Sarbaugh and Dr. Verling Troldahl. I
was privileged to share teaching assignments with three of
these members, and from each I learned much and received
much support.

In particular, my guidance committee chairman and
thesis director, Dr. Vincent Farace, has contributed a
tremendous amount to my learning experience and to my over-
all enjoyment of my time at Michigan State. Dr. David Berlo
also deserves my special thanks as one of the most exciting
and stimulating people that I have been privileged to work
with.

My thanks go to Dr. Denio Caul and to the extension
agents and other members of the Michigan Cooperative Exten-
sion Service who gave so much of their time and interest in
providing the data for this study.

I also wish to acknowledge my debt to the Public
Service Board of Victoria for granting me three years leave
on salary to study at Michigan State University; and to the

Australian Wool Board for awarding me an Overseas

iv

Postgraduate Scholarship. The lack of financial worry
certainly made my task much easier. The support that I
received from Mr. Peter Hyland, Dr. Rodger Watson and
Dr. David Wishart of the Victorian Department of Agriculture,
and the late Dr. George Moule of the Australian Wool Board
was a major factor in my coming to Michigan State and I owe
them a great personal debt.

Finally, I wish to express my gratitude to my family
and friends who have shared what has been a most wonderful
experience. To Rosemary, Caroline and Jennifer--I dedicate

this thesis; and to all of my friends I offer my thanks.

TABLE OF CONTENTS

CHAPTER Page
I. INTRODUCTION . . . . . . . . . . . . . . . . 1
II. THEORETICAL CONTEXT AND HYPOTHESES . . . . . 15
Newcomb's Coorientation Model . . . . . . . lS
Procedural Aspects of Communication . . . . 28
Dimensions of Procedural Communication . . 34
Expression of Interdependence in Procedural
Dimensions . . . . . . . . . . . . . . . 37
Hypotheses . . . . . . . . . . . . . . . . 39
III. RESEARCH METHODOLOGY . . . . . . . . . . . . 42
Sample . . . . . . . 42
Research Setting and Method of. Data
Collection . . . . . 43
The Population and Sampling Method . . . . 46
Pretest . . . . . . . . . . . . . . . . . 50
Timing . . . . . . . . . . . . . . . . . . 50
Variables . . . . . . . . . 51
Operationalization. of Variables . . . . . . 55
Coding . . . . . . . . . . . . . 58
Statistical Analysis . . . . . . . . . . . 59
Index development . . . . . . . . . . . . 59
Inferential analysis . . . . . . . . . . 72
Hypothesis testing . . . . . . . . . . . 75
Descriptive analysis . . . . . . . . . . 78
IV. RESULTS . . . . . . . . . . . . . . . . . . . 80
Characteristics of the Sample . . . . 80
Supervisor and Field Agent Perceptions of
their Communication . . 88
Actual Levels of Coorientational Similarity
Between Supervisors and Field Agents . . 94
Relationship Between Similarity of Orienta-
tion and Attraction . . . . . . . . . . . 97

vi

CHAPTER

V. DISCUSSION OF RESULTS . . .

Factor Analysis .

Communication Between Supervisors and Field
Agents . . . . . . . . .

Relationship Between Coorientational
Similarity and Attraction

Overall Implications

Future Research . .

Summary of Results

REFERENCES

APPENDIX .

vii

Page
109
109
111
118
125
127
128
131

137

TABLE

9a.

10.

11.
11a.
12.
13.

14.

LIST OF TABLES

Page
Dimensions of procedural communication . . . . , 36

Interdependence of selected communication
dimensions . . . . . . . . . . . . . . . . . . 38

Theoretical hypotheses derived from General
Hypothesis 1 . . . . . . . . . . . . . . . . . 54

Relationship of items to conceptual dimensions 56

Factor loadings for selected items relating

to communication rules . . . . . . . . . . . . 62
Communication questions selected for indices . . 63
Inter-item correlations for selected

communication items . . . . . . . . . . . . . 65
Inter—index correlations . . . . . . . . . . . . 66

Factor loadings for selected items relating

to expressed communication . . . . . . . . . . 68
Factor loadings for items relating to output . . 69
Questions relating to output measures . . . . . 69

Inter-item correlations for items relating

to output . . . . . . . . . . . . . . . . . . 70
Factor loadings for background measures . . . . 71
Questions relating to background measures . . . 71
Inter-item correlations for background measures 72

Empirically tested hypotheses that are derived
from General Hypothesis 1 . . . . . . . . . . 77

Empirically tested hypotheses that are derived
from General Hypothesis 2 . . . . . . . . . . 79

viii

TABLE
15.
16.

l7.

18.

19.
20.

21.

22.
23.

24.

25.

26.

27.

28.

29.
30.
31.

32.

Distance of agents from their supervisors

Similarity of extension background in field
agent/supervisor dyads . . . .

Distribution of scores on field agent role
satisfaction . . . . . . . . .

Distribution of scores on perceptions of the
dyadic relationships . .

Percentage salary increases for field agents

Distribution of scores on supervisor
evaluations of field agents . . . . .

Inter-item correlations for background
variables . . . . . . .

Inter-item correlations for output measures

Inter-index correlations between background
and output measures . . . . . . . .

Supervisor/field agent differences in per-
ceptions of communication . . . . . .

Supervisor/field agent differences in index
scores on procedural aspects of communication

Summary of dyadic agreement and accuracy
scores 0 O O O O O O O O O O O O O 0

Combination of agreement and accuracy about
the actual communication . . . . . . .

Effect of deviation from mean group per-
ceptions . . . . . . . . . .

Relationship of coorientation to output measures
Variance explained by measures of coorientation

Multiple correlations with all variables
retained

Multiple correlations of selected variables
with output measures

ix

Page

81

81

82

83

83

84

85
86

87

89

92

95

96

98
99

103

104

107

FIGURE

1.
2.

LIST OF FIGURES

Newcomb's coorientation model . . . . . . .
Types of understanding in a dyad .
Symmetrical, or balanced, ABX systems
Relationship of coorientational similarity

toward procedural aspects of communication to
overall similarity between members of a dyad

Page
18
21
23

31

CHAPTER I
INTRODUCTION

In the last fifteen years or so, social scientists
have been devoting increasing attention to the study of
social systems. This approach has led to a focus on the
interrelationship and interaction between components, and
on factors that enhance or detract from the effectiveness
of the bonds between them. An enduring problem for any
social system is the socialization of new members and the
continued integration of existing members. To a large
degree this becomes a problem of developing and maintaining
effective communication linkages. Therefore, communication
is increasingly being seen by system theorists as having a
crucial place in the understanding of the operation of a
social system. Barnard (1938) predicted this development
when he stated:

. . in an exhaustive theory of organization, communi-
cation would occupy a central place, because the
structure, extensiveness, and scope of an organization
are almost entirely determined by communication
techniques.

If human communication were simply a process like

carrying a brick from one bucket to another, then there

would be little need to do more than set up adequate links

or channels through which to transfer ideas from one person
to another. The process is not as simple as this, however,
and so we must consider the interaction in greater detail.
Berlo (1970) proposes as a basic premise that human communi-
cation involves the transmission of symbolic information or
pattern from one participant to another. However, the sig-
nificance or meaning for that patterning is NOT transferred.
Each one of us interprets symbolic information in our own
unique way which is a result of our total previous experi-
ence. This basic communication proposition underpins the
whole interest in the communication interaction within a
system. Because the significance, or meaning, for a
message rests with each individual participant and not in
the message itself, so it follows that meanings are unique
to each individual and inevitably will differ between the
participants. Therefore, we can see that the probability
of perfect communication (the transfer of an idea in exactly
the same form from one person to another) is zero. If we
assume that perfect consensus (similarity of orientation)
requires two people to have exactly the same meaning for a
message, then perfect consensus is impossible. However, we
can seek maximum consensus (or synonymously, minimum

dissensus)1 if the benefits are sufficiently high.

 

1If we define consensus (or similarity of orienta-
tion) as a continuous variable, then we can denote the two
extremes as complete consensus (100% consensus) and complete
dissensus (0% consensus), thus increasing dissensus is
synonymous with decreasing consensus.

There is a close analogy between this search for a
greater understanding of the processes involved in achieving
acceptable levels of consensus and the engineer's search for
ways to reduce frictional losses of energy in the flow of
a liquid through a pipe. The laws of physics indicate that
some friction is inevitable, but that good design and the
choice of the right materials can minimize this loss. In
considering any communication transaction, or series of
transactions, we can liken the message to the liquid and
dissensus to the frictional loss. In communication science
we are at the stage in development where we need to estab-
lish the costs to a communication system caused by different
levels of consensus (this is analogous to the establishment
of coefficients of friction by the physicist). If empirical
research can demonstrate sufficient benefits through in-
creasing consensus in communication, then significant
system-wide improvements can be sought by determining the
factors in a communication interaction that enhance or
detract from consensus. When the study of communication
phenomena is placed in this type of system model, it becomes
clear that the traditional approach of relating intraper-
sonal attributes of one of the participants in isolation
to the system output will not help us to an understanding
of the interactive process.

Despite the generally accepted viewpoint that Gordon

Allport expressed in Lindzey and Aronson (1968--Vol. 1,}L 3):

. . . with few exceptions, social psychologists regard
their discipline as an attempt to understand and explain
how the thought, feeling, and behavior of individuals
are influenced by the actual, imagined, or implied
presence of others.
Rogers and Bhowmik (1969) noted that in a review of over
1,000 empirical studies, the vast majority of publications
on diffusion dealt with individual characteristics of one
or other category of communication participants and N91
with the communication interrelationship involved in
achieving the communication effects. This focus, that
relates individual characteristics of one class of communi-
cation participant (source or receiver) to communication
effects, implies perfect communication (zero dissensus
between participants) and therefore effectively ignores
any interaction effects. This failure to consider inter-
action clearly violates Berlo's basic proposition and
Allport's viewpoint.

In the hope of developing a richer theory, this
project is designed to yield some basic information about
the levels of consensus in a relatively enduring dyadic
relationship, and to measure the relationship between the
levels of consensus among communication participants and
various system output measures. To return to our earlier
analogy, the proposed focus will be on the measurement of

the effects of various degrees of dissensus, which corre-

sponds to the establishment of coefficients of friction.

When we turn from the study of intrapersonal char-
acteristics to the study of relationships as the basic unit
of analysis, the simplest system that we can consider is a
dyad (a two-person group). Mark (1970), Rogers (1971) and
others have stressed the importance of the dyad as the
basic unit in studying interpersonal communication. Each
of us is a member of many relatively stable and enduring
dyads. For example: husband/wife

parent/child
coworker/coworker
subordinate/superior
superior/subordinate
student/instructor
consultant/client.

We come to each of these dyads with different sets
of expectations and with different norms (prescriptions
for behavior) for communication behavior. Most of these
norms, or rules, are not written down as formal procedures,
but rather are a composite of folkways, mores and laws that
the individual has acquired through a process of indoctrina-
tion and experience in his culture. As Davis notes, in
Biddle and Thomas (1966, p. 110), regardless of whether a
person obeys the norms completely, they are a major influ-
ence on his behavior and thinking.

There are two major implications of this situation.
First, because most norms are not codified as formal pro-
cedures, they will be gradually acquired and will be

constantly subject to minor or major modification as a

result of interaction. Second, there will also be the

opportunity for interacting members of a system to have
different sets of norms with which to develop and evaluate
an interaction.

Of all the enduring human systems that can be dis-
tinguished, the formal organization has been the subject
of by far the greatest amount of research. In any attempt
to explain the accomplishment of an agency's objectives,
Simon (1957) claims that:

. . . the operative employee must be at the focus of
attention for the success of the structure will be
judged by his performance within it. Insight into the
structure and function of an organization can best be
gained by analyzing the manner in which the decisions
and behavior of such employees are influenced within
and by the organization.
Most organizational theorists and researchers would seem
to concur with Simon's viewpoint because the subordinate/
supervisor relationship has long been viewed as a crucial
one by organizational scholars, as the vast leadership
literature will confirm. Some writers, as for example
Likert (1961 and 1967), have given some explicit attention
to communication aspects of this relationship; but in
general, other psychological and sociological concepts
have been developed and stressed in the attempts to explain
and predict performance and satisfaction. Recently, in a
study of a large business concern, Berlo et al., (1970 and
1971a) have demonstrated that communication practices

between a line worker and his supervisor are associated

with that worker's satisfaction and also with his progress

in the organization-~as indicated by turnover, absenteeism,
tardiness and promotion.

The line worker or operative in a business concern
generally has his role so defined that he has minimal com-
munication with others as part of his task. He is seen as
the end of a communication network and generally has little
or no formal contact with personnel in other sections of
the organization or with the clients or suppliers to the
organization. Also, he usually has a readily quantifiable
output--number of units produced, number of errors per 100,
amount of type set, etc.

In contrast, the operative in a service organization
is frequently a professionally trained staff member, with
a prime function of handling communication, or ideas,
rather than material things. As Kornhauser (1962), Torpey
(1970) and others note, professional operatives in organiza-
tions have unique problems and contributions that would
appear to offer a rich area for further elucidation.
Kornhauser (op. cit.) sums up the critical issue as follows:

. . . the work establishment faces the dilemma of
seeking too much integration of its professionals into
the organization and thereby losing their professional
worth, versus granting them too much autonomy and
thereby weakening their contribution to the organiza-
tion.

Examples of professional operatives include the

school or college teacher, the social worker and the

extension agent. In each of these positions, the operative's

prime function is interpreting and communicating an agency's
program to a client system.

The change agent in an extension agency presents a
clear example of the problems alluded to by Kornhauser and
others. Traditionally the change agent role has involved
interpreting and diffusing new technical information from
the agency to the clients in an effort to improve some
aspect of their life, developing and managing informal
educational programs, or assisting local personnel in
reaching and involving others in educational programs.
Clearly, the emphasis is on communication to the client
system. However, as noted above, there are links with
other systems as well. Three distinct systems can be
identified, and to be successful the agent must maintain
adequate links with each. The principal systems and the
communication foci are:

l. The organization--the contact that the agent has

with his supervisor and others
in coordinative and control
positions.

2. The client system--the communication that the agent
has with the members of an identi-
fied client group.

3. The professional organization--involvement with
other professionals in the shar-
ing and development of relevant
technical information.

Given these three systems, Kornhauser (op. cit.) proposes

four basic orientations that a professional agent can have.

These are:

. . professional orientation--functional bureaucrat
organizational orientation-~job bureaucrat
professional and organizational orientation-~specialist
client orientation-~service bureaucrat.

Depending on an individual's orientation, so he will tend
to rate one or other orientation as the crucial one in his
work.

In addition to these conflicting group memberships,
there are a number of other differences in the role of a
change agent as contrasted to that of a 'blue collar'
worker. The one that is of prime concern is his output.
The change agent's output is rarely quantifiable in the
short term--because at least in part, it will usually
involve communication strategies that may not be reflected
in behavioral changes for a considerable period of time.
Also, the agent is rarely the only source of this assist-
ance or information.

These differences between the change agent and the
'blue collar' worker, suggest that the change agent will
be extremely dependent on communication, and also that he
will have to maintain membership in markedly different
dyads in the execution of his work. To concentrate on one
role to the exclusion of the others would be to court
disaster and to destroy the basic function of this liaison
or linking position.

Ascroft (1969), Havelock (1969), Jain (1970), Amend

(1971) and Rogers and Shoemaker (1971) and others have

10

discussed this boundary role between systems in some detail.
In general, these authors have focussed upon the communica-
tion between the agent and his client system, although Jain
does consider the professional ties of extension specialists.
These studies have tended to stress the crucial nature of
the liaison role, homophily/heterophily between agent and
clients, communication strategies and the impact of the
client social system. The emphasis has clearly been on
developing more effective strategies for the communication
of technical information to the client system. With the
limited exception of Preiss (1954), little attention has
been directed to the integration of the change agent into
his organization.

As Havelock (1969), Rogers and Shoemaker (1971),
and others note, several authors2 have observed role con-
flicts due to the change agent's boundary or interface
position. There is a tendency for each system to define
the agent's role in terms of that system's needs alone.
This can lead to conflicting demands for his time and
energy, and to his being viewed as marginal by both systems
because of his inability to completely espouse one to the

exclusion of the other.

 

2Wilkening (1957), Bible and Brown (1963), Bible,
Nolan znni Brown (1961), and others.

11

The agent's professional task is to develop educa-
tional programs and to diffuse technical information to the
client system. He is also frequently called upon to analyze
and make suggestions about modification of existing prac-
tices. A strong commitment to this task may result in the
agent becoming fully immersed in the client system (client
orientation) and resenting or ignoring the constraints and
requirements of the agency. The tasks of planning and
reporting can tend to be viewed as an intrusion or an
unnecessary distraction from the job of serving the clients.

It is this issue of the balance between forces that
draw the agent away from the organization and those that
work to integrate him into the organization, that Kaufman
(1960) discusses. He made a very detailed case study of
the role of the Forest Ranger in the United States Forest
Service. The Forest Ranger position was filled by profes-
sionally trained personnel with the prime responsibility
of converting policy into action, dealing with the general
public on behalf of the Service and in general supervising
the basic functions of national forest administration.

As the lowest ranking professional officers in the
Service, the Forest Ranger was seen as the operative pro-
fessional at the base of an administrative pyramid. This
position has many similarities with that of an extension
agent. Therefore, Kaufman's insights and conclusions offer

a useful clarification of some of the complexities of this

12

type of organizational position. Kaufman noted that the
rangers, in their interface position between the administra-
tion and the non-professional staff, were subject to
opposing influences pulling and pushing them in different
directions.

He identifies five forces that work toward the
disintegration of the system:

1. The need for individual rangers to interpret
general instructions in specific situations.

2. The social and geographic distance between the
rangers and their supervisors.

3. The different behavioral norms that develop in
different face-to-face groups.

4. The ”capture" of the rangers by local populations--
primary identification with the local community as
compared to primary identification with the Forest
Service.

5. Personal differences in preferred ways of doing
things.

All these factors tend to draw the field agent away
from his agency and therefore increase the likelihood that
he will develop different norms and expectations to those
of his supervisor. At the same time, Kaufman identifies a
corresponding set of factors that encourage the agent's
integration into the agency. These integrative factors
include:

1. Procedural rules and devices for preforming
decisions.

2. Sanctions, feedback and correction, movement of
personnel to different centers, and other methods
of discouraging deviation.

13

3. Inservice training, selection and building identi-
fication with the Service to develop a will to
conform.

The minimal integration of the liaison role into
the system, and the conflicting demands and pressures
would all seem to indicate that the potential for dissensus
will be high in this dyad. For these reasons, this rela-
tionship between an extension field agent and his supervisor
would seem to offer an uniquely desirable setting for an
empirical study of the effects of differing degrees of
similarity in orientations.

Given this context, this study is directed toward
increasing our understanding of the expectations and per-
ceived communication between field and supervisory personnel
in an extension agency.

Measures of similarity of orientation toward dif-
ferent aspects of communication are proposed and opera-
tionalized and related to measures of interpersonal
attraction as it is reflected in various measures of the
success of the dyad. The specification of the association
between these differing aspects of a system should allow
some estimate of what are acceptable levels of similarity
of orientation in such a dyad.

Further, it may be possible to predict the benefits
that might be expected to accrue through an increase in
similarity. Because all systems have limited resources

and therefore must make decisions about priorities, it is

l4

hoped that this research will help administrators and others
to assess the importance of members sharing a common under-
standing about communication. If a cost/benefit3 analysis
can be built up with this type of data, then we can return
to our basic analogy and start to provide the social
engineer with the social equivalent of coefficients of

friction.

 

3A cost/benefit analysis is an economist's tool for
comparing the value of increased output with the costs that
are incurred in achieving it.

CHAPTER II
THEORETICAL CONTEXT AND HYPOTHESES
Newcomb's Coorientation Model

Newcomb (1968), in Abelson et al., in introducing
a comprehensive set of readings on the theories of cognitive

consistency noted that:

. . . Often in the history of science, when the time

is ripe, a large number of similar theories are put
forward contemporaneously by researchers who have

little if any direct contact with one another. So it
was a decade or so ago when at least half a dozen of
what we shall call "cognitive consistency" theories
appeared more or less independently in the psycholog-
ical literature. They were proposed under various
names, such as balance, congruity, symmetry, dissonance,
but all had in common the notion that the person behaves
in a way that maximizes the internal consistency of his
cognitive system; and by extension, that groups behave
in ways that maximize the internal consistency of their
interpersonal relations.

It is the latter focus on the internal consistency of dyads
(as a special case of the class of groups) that would seem
particularly suitable as a model for considering communi-
cation between supervisors and their subordinates. In
tracing the development of the range of consistency
theories, Newcomb goes on to note that the field owes a
particular debt to the contributions of Heider (1958),

with his introduction of the concept of balance and his

15

16

basic assumption that an imbalanced set of cognitions is
associated with 'tension' and the arousal of forces that
tend to restore, or to attain, balance. These basic notions
underlie the principal propositions of each of the theories
and are central to the whole interest in cognitive con-
sistency.

Within this set of theoretical approaches, Newcomb's
(1953) theory of coorientation offers a unique perspective
for viewing the consequences of dyadic interaction. As
noted by Cartwright and Harary (1956), Heider's intention
was to describe the balance of cognitive units when the
entities and relations are experienced by a single indi-
ovidual. This focus on cognitive consistency for the
individual is also followed in the dissonance and congruity
theories. Newcomb's particular contribution has been to
take Heider's notions of balance and tension and apply
them to social systems. Through this theoretical contri-
bution, it becomes possible to focus on the achievement and
consequences of similarity of orientations of members of a
dyadic or larger system.

Newcomb (1953) proposes a concept of coorientation,

 

which he defines as the simultaneous orientation of two
persons toward each other and toward some common object.
It is this focus that will be developed as the central

framework for this study. In elaborating this focus on

coorientation, Newcomb (1961) defines two sets of

17

orientations that can constitute a system. These are

firstly, individual systems that focus on the balance of

 

intrapersonal cognitions, and which are essentially indi-

vidual measures of consistency; and secondly, collective

 

systems that focus on the symmetry (or balance) of inter-
personal cognitions.

While perhaps the majority of the applications of
Newcomb's model have focussed on the balance of intra-
personal cognitions; Chaffee and McLeod (1968), Wackman
and Beatty (1971) and others have stressed the importance
of focussing on the coorientation of both persons, which
is Newcomb's collective level. The distinction of which
level is being considered is made by the selection of the
particular cognitions to be contrasted in an analysis.

The importance of this distinction lies in the fundamental
issue of deciding to either focus on the collectivity and
the actual coorientations of the members of the system; or
instead, focussing on the perceptions of only one member
of the system at a time. It would seem that Newcomb's
theory only offers a unique perspective as contrasted to
the other cognitive consistency theories, when the col-
lective level is considered.

While there are clearly applications of both the
individual and the collective perspectives, the social
systems framework would seem to favor a stress on coorienta-

tion as it is defined at the collective level. In line

18

with the social psychological tradition, as it is traced

and advocated by Scheff (1967), the interjpersonal,

 

collective model will be focussed upon as more precisely
fitting the interactive perspective that is suggested in
Chapter I.

If we consider Newcomb's basic dyadic model, it
comprises three primary elements--two interacting persons
(A,B), and their orientations toward each other and an
object X.4 This ABX model, as it has been described, is

presented in Figure l.

interpersonal attractiveness

 

A€—— e>B
A's attitude B's attitude towards
towards object X object X
X 4

Figure 1. Newcomb's coorientation model.

If we direct our attention to the similarity of A's
and B's orientations toward X, we can distinguish three
basic sets of possible relations between the two intra-

Personal orientation systems:

4The object X in this system can be any other
Pelrson, object, norm or rule.

19

1. Agreement, which can be defined as the similarity

 

between A's and B's individual orientations toward X.

2. Congruency, or perceived agreement, which can be

 

defined as the similarity between A's orientation toward
X and A's perception of B's orientation toward X; and vice
versa for B.

3. Accuracy, which can be defined as the similarity
between A's prediction of B's orientation toward X and B's
own orientation toward X; and vice versa for B.

As noted by Chaffee and McLeod (1968), only agree-
ment and accuracy are true coorientational concepts
involving the similarity of inter-personal cognitions. For
the reasons outlined earlier, only these dimensions of
similarity will be considered. Monge and Farace (1972)
present a more detailed discussion of the structure of a
coorientation system.

Scheff (1967), Laing, Phillipson and Lee (1966),
_and others have developed a concept of levels of coorienta-
tion that offers a useful way to relate the concepts of
agreement and accuracy together. They propose that it is
possible to consider an infinite series of reciprocated
levels of similarity or understanding in a dyad. Agreement
is viewed as zero-order coorientation, while accuracy
becomes first-order coorientation, and so on.

This formulation offers the opportunity to view

similarity of orientation (consensus) as a double spiral,

20

with each turn representing another level of mutuality or
understanding. Given this perspective, complete similarity
of orientation would be impossible to achieve; however, it
does allow recognition of the effect of each other's per-
ceptions as a cumulative or hierarchical coorientation.
For example, it is possible to appreciate the added
strength in the statement of coorientation "he knows that
I know that he agrees with rule X” as contrasted to "I
know that he agrees with rule X” as contrasted with "he
agrees with rule X." In considering agreement and accuracy
as two measures of coorientational similarity, this per-
spective of them, as two levels of understanding, will be
used. Taking a similar perspective, Monge and Farace
(1972) add a useful discussion of the logic of the coorienta-
tion system in terms of possible states given changes in
the basic intrapersonal orientations.

Once this perspective of agreement and accuracy,
being two levels of coorientation, is accepted, then
Scheff (1967) offers a particularly useful way of com-
bining the two levels to produce four distinct types of
understanding. Figure 2 represents a distinction between
the four types of understanding, as applied to similarity

of coorientation in a dyad.

21

 

 

Accuracy
High Low
High Monolithic Pluralistic
consensus ignorance
Agreement
Low Dissensus False
consensus

 

 

 

 

Figure 2. Types of understanding in a dyad.

The interpretations that can be made for a dyad
being placed in a particular cell in this 2 x 2 matrix,
are as follows:

a. Monolithic consensus--both members of the dyad
share a similar orientation to the object X,

and they correctly predict that the other
member shares their orientation.

 

b. Pluralistic ignorance-~the two members share a
Similar orientation towards object X, but
they predict that the other member has a
different orientation to their own.

 

c. Dissensus--the two members have differing orienta-
tions toward object X, and they understand
this difference.

 

d. False consensus—-the two members have differing
orientations toward object X, but they incor-
rectly attribute an orientation similar to
their own to the other member of the dyad.5
In this case the members disagree but think
that they agree.

 

Using these notions of coorientational similarity,

it is possible to generate a number of propositions about

 

5This interpretation only follows if orientations
are viewed dichotomously, as done by Scheff and Laing
et al.

 

22

the relationship between the degree of similarity of
orientation between two members of a dyad and their inter-
personal relationship.

Newcomb (1953) in his original statement of the
coorientation model, restricted the application of the
model to intentional, face-to-face communication, between
two members of a group, that is characterized by continued
association. He also stated most of his propositions in
terms of a strain or movement towards symmetry. However,
it would seem useful to extend the application of the
model to a consideration of equilibrium states as well as

movement towards equilibrium.

 

Based on Heider's (1958) model of balance, the
symmetry or balance of an ABX coorientational system can
be assessed by considering the positive or negative sign
of the relationship between the elements. While accepting
this approach for ease of understanding at this point, it
is worth mentioning that one of the extensions of Newcomb's
model that will be advocated is the conceptualization of
orientation as a continuous rather than a dichotomous
variable. If we accept the dichotomy of positive versus
negative orientation, then Heider and Newcomb assess a
system as being balanced, or symmetrical, if the algebraic
product of the three signs, of the relations between
elements in Figure l, is positive, and imbalanced if the

product is negative.

23

Returning to the basic premise of all of the con-
sistency theories, that imbalance creates tension and
therefore movement towards increasing symmetry, it is
possible to define a limited set of possible equilibrium

states for any such dyadic system. These symmetrical

l - \
‘ 7
X

states are presented in Figure 3.
+
Ave 4}B
X
(C) A‘%* >3
X

Figure 3. Symmetrical, or balanced, ABX systems.

B

 

 

(a) (b) A

 

Newcomb (1968) argues that any ABX system where
the A/B relationship is negative is non-balanced. He
develops this argument primarily on the notion that B is
not only an object of A's attitude, but is also a source
of attitude toward A. On this basis, he views the A/B
relationship as not being strictly comparable to the A/X

and B/X relationships. While concurring with Newcomb in

24

defining the A/B relationship as being essentially dif-
ferent to the A/X and B/X relationships; it is proposed
that a fruitful way of viewing this difference is to imply
causality and view the A/B relationship as a consequential
relationship based upon the signs of the A/X and B/X rela-
tionships.

If we accept this approach to the issue of symmetry,
then the system states represented by (b) and (c) above
become balanced or symmetrical consequences of an initial
imbalance between A/X and B/X. The theory that is proposed
is that interpersonal attraction will be a consequence of
coorientation rather than a component of coorientation.

Proposition 1: The greater the similarity of

 

interpersonal orientations toward an object X, the greater

the attraction between members of the dyadic system.

 

This proposition can be expressed mathematically as:
ATTRACTION = + f (COORIENTATIONAL SIMILARITY)
To the extent that an A/B negative relationship engenders
its own tension, that is independent of the notion of
symmetry (or balance), so the symmetrical states repre-
sented by (b) and (c), in Figure 3 above, would be
expected to have a lower probability of continuance than
the system state depicted in (a).
The notion of symmetry is thus determined solely

on the perceptions of the interactants. To the extent

25

that interactants view their interactive system as con-
sistent, then for that system a symmetrical state does in
fact occur.

Some simple examples may help clarify this issue.
In the English language there are several proverbs that
suggest that maintaining a negative A/B relationship does
incur a psychological cost but that members do in fact
weigh this cost against the potential costs incurred in
severing the relationship. These include the sayings that
"It is better the devil that you know than the devil you
don't know" and "Out of the frying pan into the fire" and
"Grin and bear it." Each of these proverbs would seem to
advocate the acceptance of an equilibrium state character-
ized by a negative A/B relationship. Similarly, there is
also considerable evidence that employee turnover is much
greater in times of high employment when the probability
of finding another job is high, than in times of low
employment when the cost (in terms of risk of not finding
another better position) of severance is much higher.

In considering enduring dyads, as contrasted to
more transitory relationships, it seems reasonable to
propose that symmetrical equilibrium states will be found
with negative A/B relationships as well as positive A/B
relationships; and more-over, that it should be possible

to predict the direction and strength of interpersonal

26

relationships on the basis of the coorientational similarity
of the interactants in their attitudes towards an object X.

So far, in considering the application of Newcomb's
ABX model, we have not specified the object X in any detail.
Newcomb (1961), Levinger and Breedlove (1966), and others
have emphasized that object X should have common relevance
as defined by a joint dependence of A and B on the object
X. Levinger and Breedlove incorporate this constraint on
the application of the model when they propose:

. . there is a positive correlation between actual

agreement and attraction, but only to the extent that
such agreement promotes the achievement of the group's

goals.
This stress on high instrumentality6 of object X to the
system, allows considerable latitude for differing A/X and
B/X orientations without any necessary reflection in
reduced attraction providing the objects are regarded as
peripheral or of low instrumentality to the system. In
taking this approach both Newcomb and Levinger and Breed-
love seem to imply a critical threshold of instrumentality,
below which coorientation will not be associated with the
.A/B relationship. An alternative approach, that is con-
sistent with the proposition to treat orientations as
czontinuous variables, is to view instrumentality as a modi-

fFier of the effect of level of coorientational similarity.

6Instrumentality is viewed as a measure of the
CEEntrality of the object to the survival of the system.

27

Proposition 2: The greater the sum of the_product

 

of coorientational similarity and perceived instrumentality

 

over the total range of objects to which coorientation is

 

perceived, the greater the attraction between the members

 

of the dyadic system.

 

This proposition can be expressed mathematically as:

ATTRACTION = f X (COORIENTATIONAL SIMILARITY X
INSTRUMENTALITY)

This proposition extends Newcomb's basic proposition in two
major ways. First, it views the A/B relationship as being
determined by the arithmetic sum of all the perceived
objects that are common to the dyad; and second, it views
instrumentality as a continuous variable acting as a modi—
fier on the effect of each coorientational measure of
similarity. By rejecting the notion of only considering
objects with a system instrumentality above a critical
threshold level, this proposition equates the effect on
attraction of dissimilarity on a large number of issues

of low instrumentality, with dissimilarity on a few issues
of high instrumentality.

If we accept this proposition then the possibility
of completely predicting attraction will be impossible.
However, from a systems viewpoint, the theory will be
supported if attraction and coorientational similarity are
positively associated in systems where a number of highly

instrumental issues are considered.

28

One such set of issues (the object or X in the
model) would appear to be the procedural aspects of com-

munication.

Procedural Aspects of Communication

 

All organizations inevitably will develop a set of
norms and rules about who will have responsibility for
various activities, who will report to whom, who will
coordinate and control other's performance, and so on.

The establishment of rules is thus fundamental to the very
concept of organization.

Ascroft (1969) sums up a consideration of the
opposing forces of organization and entropy by noting:

. . . that man's basic purpose is to maximize the
chances of perpetuating his survival by inducing and
sustaining a locally limited tendency for organiza-
tion in his environment to increase and, thereby,
reduce the characteristic tendency for entropy in
his environment to increase.
This tendency to 'induce and sustain' organization is
largely expressed in human behavior through rules or norms
that delimit acceptable behavior. This restriction on the
range of behaviors that are acceptable in a given situa-
tion clearly increases the probability that certain

behaviors will occur and equally reduces the probability

of other behaviors occurring. Rules through their effect

29

of establishing different probabilities of behavior7 thus
oppose the tendency for entropy.

One set of organizational rules, that would appear
to be of particular concern to communication scientists,
are the rules that govern the communication between members.

Cushman and Whiting (1971) argue for increased
attention to communication rules, as a primary concern for
communication theory, on the basis of the following propo-
sitions:

. . (1) that the transfer of information requires the
interaction of score, message and receivers, guided and
governed by communication rules; and (2) that communica-
tion rules form general and specific patterns which
provide the ground for the scientific explanation, and
prediction of communication behavior.

They distinguish between two classes of communication
rules--procedural and content. Procedural rules are de-
fined as the rules that guide and govern the organization
of symbols and participants within the boundaries of a
communication system.8 Such procedural rules have tradi-
tionally been referred to as social norms.

It is proposed that the procedural aspects of

communication offer a particularly sensitive and appropriate

 

7Entropy assumes equal probability for all possible
outcomes, therefore any differences in probability repre-
sents some degree of organization. The greater the differ-
ences in probability the greater the organization
(predictability).

8The following is an example of a communication
procedural rule: "An extension agent should always confirm
phone conversations with his supervisor with a written
summary.”

30

vehicle for investigating the complementary behavior of the
incumbents of related positions. Figure 4 shows the rela-
tionship between this focus and the overall similarity
between the members of such a dyad.

Before any communication can occur, some minimum
degree of consensus is required about the procedural rules
that will be followed. For this reason, it is only after
the establishment of a set of procedural rules (ground
rules) that content rules (that relate to how concepts will
be used) can be developed and the process successfully
linked together. The place of procedural rules in any
communication transaction can be likened to the Theory of
Limiting Factors. This theory states that the output of a
system will be controlled by the component that is most
limiting; once this component has been brought up to an
acceptable level then another limiting factor will appear.
However, until the first, or most limiting, factor is
attended to, attention to other factors will not increase
system output.

The procedural aspects of communication can be
separated into two primary areas-~(l) the procedural rules,
and (2) their variable expression in actual communication

behavior. While the actual9 communication practices in

 

9The communication behavior that oqcurs within a
dyad can be reported by the participants as they perceive
it, or by an observer. To distinguish between these, in
this paper, participant reports will be referred to as
perceived communication.

Figure 4.

31

 

SIMILARITY

CONSENSUS

COORIENTATION

[PROCEDURAL ‘
ASPECTS OF ‘
COMMUNICATION

 

Relationship of coorientational similarity
toward procedural aspects of communication
to overall similarity between members of a
dyad.

32

themselves may offer considerable predictive power in
understanding the output of a superior/subordinate dyad,
the rules, or limits, of acceptable behavior may be even
more important. As French and Raven (1960) note, a person's
social power is measured by his maximum possible influence
and not the proportion of this that he chooses to exert at
any particular point of time. Similarly, with communica-
tion, the important question may be whether the behavior
falls within certain limits rather than what it specifi-
cally is. A superior may approve of a wide range of
behaviors providing that they are within his range of
acceptability.10

It is proposed that in this study both of these
aspects should be considered. While much of the theoretical
discussion will be restricted to the rules for simplicity,
both aspects should be kept in mind.

In most organizations, many procedural rules
develop to control the communication between members. Some
of these rules are codified into formal written procedures
that specify the precise communication behaviors that are
required or permitted; but, most are informal social norms
that are learned through experience in the organization.

This learning process is frequently described as 'getting

 

10For example, the rule may express Optional be-
havior and then either the presence or absence of the
behavior would be equally acceptable.

33

to know the ropesfi While it may constitute a minor problem
for someone totally integrated into the system and with
limited communication responsibilities, it may be a serious
problem for a change agent whose role keeps him marginal

to the system and who has a wide range of communication
behaviors required by his position.

To the extent that there are different expectations
(different sets of rules) for a given situation so there
exists the opportunity for serious communication breakdown.
If I expect that you will request my ideas if you want me
to contribute to a task, and you expect me to show the
initiative in contributing ideas without having to be asked,
then we run the risk of applying vastly different measures
to judge a situation. If the behavior falls outside the
superior's range of acceptability then loss of confidence
and dissatisfaction may result--not from lack of competence
but from a lack of similar orientations toward the rules
of the interaction.

As Berlo demonstrates very clearly in his film,

”11 it is very easy for

"Avoiding Communication Breakdown,
the participants in an unsuccessful communication dyad to
blame the other member for intellectual failure rather

than consider the possibility of a lack of agreement on

 

1
Maryland.

1This film is distributed by BNA Films, Rockville,

34

what was appropriate communication behavior. The change
agent has to concentrate on and develop his communication
with members of his client system to fulfill his prime
professional task. In addition, he may be stationed in a
center geographically separated from his supervisor, and
therefore, have only restricted opportunities to communicate
with him. This situation would appear particularly sus-
ceptible to communication breakdown, and where output is
measured largely in terms of communication this would seem
to leave the way open for negative performance reviews and
dissatisfaction.

If we accept the importance of procedural aspects
of communication, then the next task becomes one of identi-
fying dimensions, or foci, that the rules relate to. It
seems reasonable to expect that different dimensions will

have differing instrumentality to the system.

Dimensions of Procedural Communication

 

One approach is to look to organizational theory
for rules or norms of communication. Mouzelis (1967) aptly
sums up the state of the literature on organizational
theory when he notes that:

. . . the volume of the literature on this subject and
its rate of increase is as impressive as the magnitude
of the phenomena which it tries to analyze and explain.
Moreover it is not only the sheer volume of studies
which is enormous and bewildering but also the multi-
plicity of points of view from which organizational
phenomena have been examined.

35

During the twentieth century, many theories have been pre—
sented as models for the effective management of organiza-
tions. As I have noted elsewhere, Russell (1971), each
model has tended to stress one particular organizational
output--usually either efficiency of production or employee
satisfaction--and has focussed only on one particular
system level. The importance of these theories for this
study is expressed by Schein (1965) when he states:

. . every manager makes assumptions about people.
Whether he is aware of these assumptions or not, they
operate as a theory in terms of which he decides how
to deal with his superiors, peers, and subordinates.
His effectiveness as a manager will depend upon the
degree to which his assumptions fit empirical reality.

Our interest in organizational theory is directed
to these assumptions which can be viewed as norms or rules
for how people should relate to each other.

Russell (op. cit.), in an earlier analysis,12
considered the communication implications of representative
organizational theories. In making this analysis, a set of
communication dimensions were used that were largely based
on those proposed by Farace and Connelly (1970), and

MacDonald and Farace (1970). Similar conceptualizations

were used in the research reported by MacDonald (1970) and

 

12The specific theories that were considered were:
R. Likert, E. Mayo and Roethlisberger G Dickson,
D. McGregor, H. A. Simon, F. W. Taylor, M. Weber and
social systems theory as explicated by K. Berrien, J. G.
Miller and E. H. Schein.

36

in two current research projects in the Department of
Communication.13 In particular, process and functional
concepts were emphasized. Based primarily on the rationale
presented by Russell (1971), a set of specific dimensions
of communication, that would seem particularly appropriate
to the study of dyadic communication, are presented in

Table 1.14

Table 1. Dimensions of procedural communication.

 

 

Dimension Conceptualization

 

Amount Expectation of volume of interaction between
incumbents of particular positions

Function Expectation on functional purposes that
messages can serve in the system

Initiation Expectations on which member of a hierarchial
dyad should initiate communication

Mode of Implicit or explicit expectation in choice
contact of linking channel
Sequence Receiver pre-involvement in messages

directed to him (distinguishing between
sought and imposed messages)

 

 

13Studies of communication patterns in a large bank-
ing organization and a federal agency. In particular, the
questionnaire dealing with communication relationships and
consensus between superior and subordinate was directly
relevant.‘

14The set of dimensions are based on a set of gen—
eral descriptors that could be applied to any process: what
are the components, what is the direction and sequence of
their linkage, how do the elements move through the system
and what is the volume of elements that are processed? In
this case, the process is the transfer of messages from
extension agents to their supervisors.

37

These dimensions currently represent a 'common
sense' classification of procedural aspects of communica-
tion. Their particular virtue lies in their system-neutral
conceptualization, that allows them to be applied in a wide
range of organizations or systems. The functional dimen-
sion is based on Berlo's (1970) conceptualization, in
which he distinguishes between three major uses that peOple
make of communication in an interdependent system. These
three functions are:

a. Production--getting the job done
b. Innovation--exploring new behavioral alternatives

c. Maintenance--keeping the system and its components
functioning

No claim is made for the independence of these dimensions--
in fact, empirical research will almost certainly suggest
a more parsimonious set of dimensions with which to

describe the procedural aspects of interaction.

Expression of Interdependence
in Procedural Dimensions

 

Having established a set of dimensions that
describe some of the major features of a dyadic communica-
tion process, it then becomes important to develop a basis
for rating positions on these dimensions.

One such basis is an extension of the earlier
analysis, Russell (1971), of the communication implications

of the major organizational theories. As noted earlier,

38

most organizational theories can be viewed as providing a
set of assumptions about how people should relate to each
other--they thus provide sets of contrasting rules or norms
for interactive behavior.

If we take the set of procedural dimensions of
communication, that are outlined above, and apply these to
the implications of the major theories of organization,
then two broad sets of behavior emerge. These correspond
closely to Barrett's (1970) distinction between the

exchange model that stresses a basic independence between

 

the individual and the organization; and his socialization

and accommodation models that stress interdependence.

 

Table 2 represents the major expression of these two basic

orientations.

Table 2. Interdependence of selected communication
dimensions.

 

 

 

Communication dimension Expression

Independent Interdependent
Mode Written Oral
Amount Low High
Initiation Supervisor Subordinate or both
Sequence Imposed Sought
Function Production Production,

Maintenance,
Innovation

 

39

This basis for distinguishing or classifying
various dimensions of communication bears considerable
conceptual overlap with the commonly used notion of com-
munication openness, as used for example by Berlo et al.
(1971a). In general, more interdependent and more open
communication would be synonymous. The notion of inter-
dependence, or sharing, is advocated in this study because
it more nearly describes the essential difference between

the two types of behavior as they are expressed in a dyad.

Hypotheses

 

Procedural aspects that are manifest in the amount,
function, initiation, mode and sequencing of the communica-
tion between supervisors and their field staff in an
extension agency, would seem to fill a highly instrumental
and controlled influence on their interpersonal attraction.

Also, to the extent that their interaction will be
highly restricted due to their lack of proximity and their
particular role demands, it can be expected that there
could be quite marked differences in the orientations of
supervisors and field agents towards these aspects of
communication. As noted in Chapter I, the role of the
field agent requires that he interact in a number of
groups, each of which may possess conflicting role expecta-
tions and demands. Given this potential for conflicting

expectations and the lack of clearly quantifiable measures

40

of task performance; it is proposed that at least part of
the field agent's role satisfaction, evaluation and satis-
faction with his relationship with his supervisor will be
a consequence of the influence of their coorientational
similarity with respect to the procedural aspects of com-
munication.

The theoretical model that would seem particularly
appropriate for predicting and explaining such an influence
is the collective level of Newcomb's (1953) coorientational
theory, as expanded in propositions 1 and 2 (see earlier
in this Chapter). Taking the dyad as a system, the model
serves to relate coorientational similarity of the dyad
members, toward an object (or set of objects), to their
orientation toward each other. Both positive and negative
interpersonal orientations are proposed in the extensions
of Newcomb's formulation, and it is proposed that inter-
personal orientations will serve a balancing function for
the system. Having considered, in Chapter I, the particular
dyadic relationship that might be expected to exist between
field and supervisory staff in an extension agency, and
introduced the procedural aspects of their interaction as
a potentially highly salient or instrumental object of
coorientation, it is now possible to specify two hypotheses

based on the earlier propositions.

41

GENERAL HYPOTHESIS l: Theygreater the degree of
coorientational similarity toward (a) procedural rules of
communication, or (p) theyprocedural expressions of their
interaction, theygpeater the attraction between supervisors
and field egents in an extension service.

GENERAL HYPOTHESIS 2: The greater the summative
similarity of orientations between supervisors and field
egents in an extension service, toward the procedural
aspects of communication, the gpeater the attraction

between them.

 

Specific hypotheses are detailed in Chapter III.

CHAPTER III

RESEARCH METHODOLOGY

As in virtually all current social research that is
conducted in on-going organizations, this study has two
basic aims. First, to provide the cooperating organization,
the Michigan Cooperative Extension Service (MCES), with
feedback on the operation of part of its system and pos-
sible recommendations on changes that might improve its
operation or morale; and second, to obtain empirical data
to test the theoretical hypotheses outlined in Chapter II.

While the focus of this report is almost exclusively
centered on the empirical testing of the hypotheses derived
from theory, there is still an overall cognizance of the

obligation to the MCES.

The Sample

 

Data were collected from 100 field agents and 10
supervisors in the MCES during the period from November 15th
1971 to January 24th 1972. Each field agent completed a

209 item questionnaire relating to his15 communication

 

15Throughout this report only the masculine gender
will be used for simplicity, although two of the super-
visors and twenty-nine of the field agents were women.

42

43

with a nominated supervisor, while each supervisor com-
pleted ten such questionnaires dealing with his communica-
tion with each of ten of the selected field agents. This

gave data on a total of 100 dyads.

Research Setting and Method of Data Collection

 

There are a wide range of possible research settings
that have been proposed and used in social science research.
These range from virtually non-reactive observational
methods as advocated by Webb et a1. (1966), through case
studies and extensive field surveys and more controlled
field experiments to the ‘tightly controlled laboratory
research of such people as Leavitt (1951). There is no
one 'right' way of collecting empirical data. The choice
must ultimately depend on the weighing of relative advan-
tages and disadvantages. The basic issue is one of relative
control as contrasted to reactivity to the experimenter.

The general precedent seems to be to start with
simple observation and increasingly introduce more and
more control until true experimental conditions are
achieved. In communication research most of the reported
research has been of the observational or survey type,
where control is achieved through randomization rather
than elimination of extraneous factors.

Organizational communication research, such as that

conducted in this study, faces two related restraints.

44

These are: first, a need to study intact relationships if
we are to measure enduring behavioral implications of
interaction; and second, the state of knowledge as mentioned
earlier, is only at the exploratory stage. It was thus
decided to make a field survey. To the extent that the
results of this survey provide a clearer understanding of
the implication of coorientational similarity in the dyad,
then it will be possible to initiate the type of field
experimentation proposed by Berlo et al. (1971a).

Given the survey setting, we were still left with
a wide choice of actual methods of data collection. The
main options would appear to be observation of interaction
by a participant or non-participant observers,l6 inter-
viewing, or mail or telephone questionnaires. Given the
disadvantages of mail and telephone questionnaires as

17 and the need to

outlined by Moser (1958) and others,
obtain dyadic members' perceptions rather than their
observed behavior, it was decided that the personal inter-
view was the most appropriate method for this particular

study. The actual technique that was used was the group

 

16Observation can include such techniques as diary
and log keeping as described by Axinn and Axinn (1969) and
Farace et a1. (1969).

17These disadvantages include low response rates
and related response bias, no check on order or interpre-
tation of questions, and no opportunity for supplementary
observation.

45

technique, as described by MacDonald (1970), with groups

of four to sixteen respondents completing the study instru-
ments under the direction of a trained interviewer.18 This
method of data collection has been successfully used with
groups in several organizations and it would appear to
minimize experimenter bias or error and at the same time
provide the advantages of control that accrue from having
the researcher present. Finally, in terms of the major
alternatives in selecting an appropriate method, we were
left with the choice between a cross-sectional study at

one point of time, and a longitudinal study over time.
Given a population about which little was known in terms

of their communication patterns, but who were expected to
have had differential dyadic interaction due to differing
length of administrative relationship, there would seem to
have been real advantages in a longitudinal study where
changes in similarity and their effects could have been
observed and measured over time. However, against the
increased precision that a longitudinal study would offer,
there were the disadvantages of greater expense in reinter-
viewing, the influence of extraneous factors between

repeated measures and reaction to the actual research

 

18Eight field personnel were unable to attend group
sessions and in these cases they were provided with a
written explanation covering the points made in introducing
the instruments in the group sessions.

46

procedure. It was thus concluded that provided the dyads

had a history of over one year in length,19

that they could
be considered as at least approaching a dynamic equilibrium,
with a sample presenting the range of equilibria. Given
the exploratory nature of the study, the cross-sectional

approach at one point of time was chosen as best fitting

the circumstances.

The Population and Sampling Method

 

The study was conducted with personnel in the
Michigan Cooperative Extension Service (MCES). This
organization has been described in some detail by Jain
(1970) and Amend (1971), and in the pamphlet entitled 'A
Career with the Cooperative Extension Service of Michigan
State University'.

The following extensive quote from this pamphlet
outlines the principal aspects of direct concern to this
proposed study:

. . . program organization. The Director of Extension
is in charge of all programs, personnel, finances, and
operations of the Cooperative Extension Service. The
staff consists of more than 400 professionals.

Five program directors are responsible for guiding
programs in Agriculture, Natural Resources, Marketing,
4-H Youth, and Family Living Education. Some program
areas have program leaders to assist the program
directors.

Three assistant field operations directors are
also responsible to the Director of Extension. They

 

19This suggestion is based on Newcomb (1963), which
would indicate some stability by that point of time.

47

counsel with county and district personnel, coordinate
joint program efforts, meet with committees of county
boards of supervisors, handle special problems, etc.

Supervisory teams are under the chairmanship of
the extension program coordinator. Other team members
consist of representatives of the five program areas.
These teams are responsible for supervising all Exten-
sion personnel in their assigned geographical area.

Extension programs are carried to the people in
Michigan's 83 counties through county Extension
offices. A County Extension Director is in charge of
each office. Agents are assigned on a county, multi-
county, area or district basis.

Other functions at the state level include subject
matter specialists who are grouped in departments on
the MSU campus. Specialists are charged to provide
current technical information for field agents. They
also help plan and conduct county Extension programs.

The study focussed on the communication between a
sample of the approximately 250 field personnel with
members of the supervisory team. The field personnel
represented all of the five program areas. The research
population was defined as the total number of supervisory
dyads that the ten designated members of the supervisory
team had with field personnel. The population was thus
defined by direct supervisory interaction. The population
of supervisors was the five program directors, the three
assistant field operations directors and two program
leaders. While it is recognized that there were different
demands in these different dyads, there was no clear
hierarchy of supervision but rather there tended to be
overlapping chains of responsibility. The aspect of multi-
ple responsibility tends to be a feature of many extension

organizations and is in sharp contrast to the strict

48

hierarchical structures in many business concerns. Given
that our concern was with the development of similarity of
orientation in supervisory dyads, the prime concern was
only that the field personnel had some direct subordinate
relationship with the designated supervisors. Therefore
the fact that the supervisory population had different
titles need not be of concern.20
Populations of supervisory-subordinate dyads were
established, excluding any field personnel who had not had
a direct subordinate relationship with a supervisor for at
least one year, and any personnel working from the same
center as their designated supervisor.21 Where field
personnel appear in more than one dyad, they were assigned
at random to one or other dyad (no person should have had
direct supervisory relationships with more than two of the
supervisors). Once the ten populations were established,
ten dyads were selected at random from each supervisory
population, to yield a total of 100 dyads involving ten
supervisors. The sample thus represented the total of

field--one step supervisory dyads (after the exclusions

referred to earlier).

 

20The criterion was that a direct supervisory
relationship be recognized by the MCES.

2 . . .
1As mentloned 1n Chapter I, we are concerned w1th
communication under restricted interaction.

49

This design was neither a true random sample nor a
precise stratified random sample. The procedures outlined
above were selected as a way of sampling a range of rela-
tionships while at the same time ensuring that each
supervisor only had to respond to ten dyadic interactions.
To the extent that some supervisors were involved in many
more dyads than are others, so the proposed sampling method
under-represented these dyads. Given that the prime con-
cern was with establishing as wide a range of dyadic
relationships as possible, while at the same time providing
some information about the organization as a whole, this
would seem to have been an acceptable compromise.

In the case of sickness, transfer or resignation,
or refusal causing a field agent to drop out of the sample,
designated replacement personnel were identified for each
supervisory population. In the actual data collection,
two agents were in fact replaced--one because of a transfer
to a different program area and the other because of a
recent change in supervisor. Given that response to ten
dyads would seem a maximum request to make to any one super-
visor, the loss of a supervisor would have meant the

reduction of the total sample of dyads by ten.22

 

22Given the administrative set up of the MCES, it
did not appear possible to replace a supervisor with another
member of the supervisory staff. Informal discussions with
the Extension Program Coordinator and others suggested that
the ten designated supervisors in fact represented the cur-
rent population of one-step supervisors.

50

The Pretest

 

The field version of the research instrument was
pretested with a small sample of agents. These agents
were the complete professional staff in two counties
adjacent to the state headquarters of the Michigan Coopera-
tive Extension Service. Seven personnel were included in
this pretest sample. The two prime purposes of this pre-
test were to check on the time needed to respond to the
instrument, and to check on the clarity and meaningfullness
of the instruction and items. Thus, in addition to com-
pleting the instruments, group debriefing sessions were
conducted immediately afterwards. On the basis of this

pretest the wording and order of several items were changed.

Timing
The pretest interviews were conducted during the
first half of November 1971, and the main field study was
conducted between November 15th 1971 and January 3rd 1972.
The supervisory staff were individually briefed about the
task of responding to ten dyads and they completed the set
of instruments at their individual convenience, between

December 15th 1971 and January 24th 1972.

51

Variables

 

The hypotheses, outlined at the end of Chapter II,
relate supervisor-field agent coorientational similarity
toward procedural aspects of communication to the attrac-
tiveness of their relationship. Three major categories
of variables can be distinguished that need to be measured
in any attempt to elucidate this hypothesized association--
background, procedural communication, and system output

measures .

Background variables

 

On the basis of Rogers' (1971) discussion of five
dimensions along which dyads can vary, and Newcomb (1963),
Kaufman (1960) and Nix and Bates (1962), it is possible to
suggest a number of variables that would appear particularly
relevant and applicable to this study.

These variables all relate to the shared background
or barriers which might either enhance or detract from
communication and system output. The following four back-
ground variables are proposed as being particularly likely
to be associated both with communication and output
measures:

1. Similarity of background--(Kaufman's social distance)

 

this variable is conceptualized as comprising
the most important dimensions of similarity
with respect to a professional work relation-
ship. Age and previous extension experience
are proposed as being most important.

20

3.

4.

52

Length of interaction—-(Rogers' history)

 

the length of time that the dyad members have
had in a work relationship to each other.

Spatial separation--(Kaufman's geographic distance)

 

the propinquity of the field agents' work
centers to those of their designated super-
visors.

Educational similarity--the similarity in level of

 

formal education between the members of each
dyad.

Procedural communication

 

variables

 

The basic communication variables will be measures

of similarity of orientation toward the procedural aspects

of communication. These measures should incorporate both

the agreement and accuracy levels of coorientation as they

relate to both the procedural rules and their expression

in the actual dyadic communication.

Although it would be possible to generate a more

comprehensive set of variables encompassing more facets of

this coorientation, the following variables are proposed:

1.

2.

Supervisor/Field agent agreement about the pro-
cedural rules.

 

Supervisor/Field agent agreement about the expressed
communication.

 

Field agent accurac in predicting how his super-
visor will v1ew the expressed communicatlon.

Supervisor accuracy in predicting how his field
agent will v1ew t e expressed communication.

53

For each of these measures of coorientational
similarity, it is possible to consider the independent
action of each of the three functions and four other pro-
cedural dimensions of communication. This generates a set
of 28 distinct measures of coorientational similarity as

is shown in Table 3.

System output variables

 

It is not proposed to measure interpersonal attrac-
tion directly, but instead to measure a set of system
output measures that would seem to have some conceptual
overlap with attraction. On the basis of the research
described by Newcomb (1963), the model proposed by Nix and
Bates (1962) and the study by Berlo et al. (1971a), the
following four measures of system output are proposed:

1. Field agent satisfaction with his role achievement.

 

2. The assessment by each member of the dyad of the
formality of the relationship.

3. The evaluation by the supervisor of the field
agent's performance.

4. Social control by the supervisor as it is indicated
by the field agent's relative salary advancement.

 

If we take the two assessments of the relationship
independently, this then gives us five output measures.
When these are combined with the 28 measures of coorienta-
tional similarity, a set of 140 different specific

hypotheses can be generated from the General Hypothesis 1

54

Table 3. Theoretical hypotheses derived from General
Hypothesis 1.

 

 

Coorientation
variables
(increased
similarity)

System output variables

 

Field

Agent

role
satis'n

Field
Agent
relat'
ship

9

0
salary
increase

Super-
visor
evalu-
ation

Super-
visor
relat'
ship

 

Rule agreement

 

amount

mode
initiation
sequence
production
maintenance
innovation

Agreement re

expressed
amount
mode
initiation
sequence
production
maintenance
innovation

 

 

Field agent

accuracy
amount
mode
initiation
sequence
production
maintenance
innovation

 

Supervisor

accuracy
amount
mode
initiation
sequence
production
maintenance
innovation

 

+ ++-+-++ + -++ +-++ +-+ + ++-+-++ +

-++ ++-+-++

+-++ +-++ + -++ +-h+ +-+ + ++-+-++ +

++-+ ++-+ +

+-++ +~++ + -++ +-++ +-+ + +4-+-++ +

+4-+ +i-+ +

+-++ +-++ + -++ +-++ +-+ + ++-+-++ +

++-+ ++-+ +

+-++ +-++ + -++ +-++ +-+ + ++-+-++ +

+4-+ ++-+ +

 

55

that was presented in the previous Chapter. These are
shown in Table 3. The General Hypothesis 1 would suggest
that each of these pairs of variables should be positively

associated.

Operationalization of the Variables

 

The instrument was prepared in two forms, Form A.
for field staff and Form B. for Administrative staff, to
allow the development of reciprocal measures for measuring
the different types of similarity or orientation. Basi-
cally, each instrument was divided into six sections, the
first dealing with control factors; the next four dealing
with description and satisfaction with the communication
relationship, the expectations for field agents' communica-
tion, perceptions of actual communication, and the predic-
tions of other's rating of the field agent's communication;
and the last section dealing with measures of relationship
and evaluation of performance. The actual field agent
instrument is included in the Appendix. Both versions of
the instrument consisted entirely of multiple choiCe items
based on those used by Bible (1959) and MacDonald (1970).

While these two sources were used extensively in
determining the approach to Operationalization, virtually
no items or scales were directly appropriate to this pro-
posed study, without extensive modification. Therefore,

the operationalizations must be considered as untested.

56

In general, Likert type items were used, with four or five
choice points in most cases. Where possible, the phrases

of degree recommended by Dodd and Gerbrick (1960) were
selected as the alternative choice points. The items were
designed to operationalize the five dimensions of procedural
communication that were presented in Table l and the control
and output variables outlined above. Several items were
included to measure each variable, with the intention of
developing various indices. Only face validity was claimed
for most items prior to the analysis. Table 4 below shows
the various conceptual dimensions that the items were

designed to relate to.

Table 4. Relationship of items to conceptual dimensions.

Conceptual dimension Items

 

A. Background variables

 

Similarity of background I.2,3,4

 

Educational background I.5

Length of interaction I.l

Spatial separation (determined independently)l
1

Spatial separation was determined from the in-state
mileage chart in the Michigan State University Travel
Regulations brochure.

Table 4.--Continued

S7

 

 

Conceptual dimension

Items

 

Procedural communication

 

variables

 

Mode of contact

Amount

Initiation

Sequence

Production function

Maintenance function

Innovation function

General interdependence

System output variables

Role achievement

Relationship

Social control

Evaluation

11.2, III.2,10,15,19,25,34,
38,46,47,49,51,52,54,57

II.3,10,13,16, III.S,7,12,16,
17,20,23,32,33,35,39,43,56

11.1,9,12,15, III.4,8,ll,13,
18,27,30,36,37,40,44,55,59

II.5,6, 111.1,3,9,14,21,22,
24,26,28,3l,41,43,45,48,53,6O

11.9,10, III.5,8,10,15,17,18,
24,28,32,35,36,44,46,48,49,53

11.12,13, 111.2,7,9,11,14,19,
20,23,27,30,37,39,45,47,51,60

II.15,16 III.4,6,12,13,25,26,
31,33,38,40,42,43,52,56,57,59

II.1,2,3,4,5,6,7,8, 111.29,
50,58

II.4,11,14,17, VI.l,2,3 8 5
(for field staff)

VI.4,6,7 (for field staff)
VI.4,5,6 (for supervisors)

(Separate statement of rela-
tive salary increments)2

VI.l,2,3 (for supervisors)

 

2

Social control was determined on the basis of the

agent's percentage salary increase from July 1968 to July
1971. Those with less than three years (11 agents) had
their actual percentage increase adjusted by a factor based
on the mean percentage increases granted to personnel in
their program area over the other years.

58

It can be seen from Table 4 that each item related

23 and that most items

to at least one conceptual dimension
in Section II, III, IV and V related to both a function

and another dimension. Given the lack of evidence about
the relative importance and the relative independence of
these conceptual dimensions of procedural communication,

a systematic factorial approach was used in the generation
of items. Given that it would appear probable that some

of the dimensions would have strong pervasive effects, this
design gives maximum replication of each dimension. How-
ever, by combining two dimensions in most items this
actually works against the identification of clear, inde-
pendent factors. The approach should thus yield a minimal
core of orthogonal factors based only on the dimensions

that are sufficiently instrumental to mask out the other

dimension incorporated in each item.

Coding
'A priori' each communication item in Sections II,
III, IV and V of the instrument was rated along the inter-
dependence dimension. Bach item was designed to randomly
favor one or other end of this dimension, so that any

response set need not necessarily bias the results.

 

23The items in Section IV and V correspond to
those in Section III.

59

The items were coded from the completed instruments
onto codesheets, at which stage they were transformed to
represent a consistent direction on the interdependence
dimension. Random checks on coding yielded an error rate
of less than two errors per thousand questions coded.24

From the codesheets the responses were transferred
to punch cards and the punching was independently verified.

All of the subsequent index construction, transformation

and analysis was made from the punch cards.

Statistical Analysis

 

Index develqpment

 

As discussed earlier, one of the principal aims of
the initial analysis was to determine how well the indi-
vidual items represent some underlying dimensions. There
are two related aims in this activity: first, to identify
and remove items that empirically do not correlate with
the underlying dimensions, regardless of their face
validity; and second, to evaluate how independent (orthog-
onal) the proposed dimensions of communication are.
Inter-item correlations and a factor analysis were per-
formed on the responses to all items in Section III of the
instrument. A factor analysis was also performed on the

items in Section VI relating to satisfaction and

 

24Three errors in 2170 items.

60

interpersonal relationship, and the control items in
Section I.’ Although it is recognized that the assumptions
of factor analysis call for interval data and no evidence
is presented to Claim intervality for the items in these
instruments, there is precedent in the writings of such
persons as Vroom (1960) for treating Likert type scales

as if they were in fact interval scales.

To the extent that we can reasonably assume random
variation from linearity in perceptions, so we may choose
to treat scales with equal-appearing intervals as if they
were in fact interval measures. Caution, however, must be
taken against too strict an interpretation of the precise
magnitude of the correlations. In this case, we were not
interested in determining the significance of the inter-
item correlations; but rather, we were interested in
identifying the underlying dimensions (or factors) that
distinguish different groups of items, as discussed by
Nunnally (1967, p. 364-365). Given this restriction, four
independent indices were developed on the basis of inter-
item correlations and factor analysis.

These orthogonal indices are assumed to represent
basic underlying conceptual dimensions of procedural com-
munication, and they are used in the subsequent descriptive
and inferential analysis of the data. Items in the schedule
that did not have a prime loading on one of these factors

were not included in the subsequent analysis.

61

An initial factor analysis of the responses by the
field agents to the sixty items relating to the procedural
rules of supervisory communication (Section III of the
instrument), yielded four factors that accounted for 24
percent of the variance. An inspection of the items with
prime loadings25 on each of these factors (see Tables 5
and 5a) showed that they closely corresponded to the con-
ceptual dimensions of innovation function, maintenance
function, sequencing and initiation. Further iterations
of a varimax rotation failed to isolate any additional
groupings of items that represented any of the remaining
conceptual dimensions. However, while not isolating
additional dimensions, the further iterations made it
possible to check on the stability of prime loadings for
the various items. No movement from one factor to another
was found for any of the items which had a prime loading
on one of the original four factors.

Having identified the four conceptual dimensions
that proved to be essentially orthogonal to each other,
inter-item correlational matrices were developed for all
of the items that were identified in Table 4 as conceptually
relating to one of these four dimensions. Within each of

these matrices, the sum of inter-item correlations of more

 

25A prime loading was defined as a correlation of
> 0.40 with the factor.

62

Table 5. Factor loadings for selected items relating to
communication rules.

 

 

Factor loadings

 

 

Item
number 1 2 3 4
Innovation Maintenance Sequencing Initiation
4 9:45 -0.09 -0.14 0.00
6 OLES -0.07 -0.20 0.14
7 0.25 figggg -0.19 0.25
8 0.38 0.09 -0.06 9111
9 0.25 ‘2111 -0.03 -0.10
11 0.20 -Q;§3 -0.01 0.15
13 0.12 -0.16 0.08 “0:59
23 -0.06 '2;§£ 0.14 -0.01
26 911; 0.00 0.09 -0.16
27 -0.22 -0;4§ 0.08 -0.19
31 911; 0.05 0.31 -0.01
33 QLQI 0.02 0.09 0.00
36 0.27 0.09 0.09 9139
37 -0.12 0.07 0.19 QLQE
38 OLSZ -0.09 0.03 -0.17
39 0.13 TQLQE 0.05 -0.04
40 ‘QLSE -0.09 0.28 0.02
41 0.25 —0.09 0440 -0.47
45 -0.07 -0Lll 219E -0.17
47 0.03 10:49 -0.07 -0.07
48 0.23 0.03 .2119 0.03
51 0.02 -0;§Q 0.08 0.15
52 -0;40 -0.04 -0.13 -0.22
53 0.14 0.15 g;§4_ -0.06
60 0.18 -0.04 0 46 0.18

 

63

 

 

mmho>om .ucmflco>:oo:fi xaamcomhom m“ peg» ucmscmfimmm cm :ofipmmsc ow oumufimez .oo
emuo>om .cmemEoo can» pmcpmu .ufi Mama paw ppoop we: pfium.
soap .ou wocmflmmm soon mm: o: pm:u xuoz one :pflz OOAwmfiuwm you ma e: mH .mm
omhm>om .uomfi>homsm we: ow mcwpwpz cw mmopfi ho meowpfimomopm 3o: wasnsm .Nm
.maonm one ho>o mumpume Hmcomuom mmsomfip on ovum Hmom .Hm
.Efi: cue: pomw>nmm5m we: Eopm mp:mnoEoE msosmfinEm kmfipmau .we
.hOmH>HOm:m we: on mwcwgoEoE a“ mam: Hmcompom opsfiocH .ne
unoEoucm>wm
m.u:owm may now meow“ m.u0mw>uom:m ma: mo Emflowuwho pommo ow wmquOMQ om .me
.mOSmmfi po>Hom0hcs mo Homfi>pomzm mac wcfiw:HEoy mmmx .Hv
.AOmH>hem:m we: :pfiz omfimp ou mmmpfi 3o: pso Meow ou o>fiumflpwcfi may exmb .oe
.hOmH>pomsm was new: mcofipmmho>coo a“ mofiufl>fiuom m.%HflEmm max on homom .mm
.Hmmoaopa m mcwuuwansm ouommn hemw>hom3m nu“: mmopw 3o: no>o Mame .wm
ompu>om .u:oEoo:m>pm
Hmcomuom you momnsou oHnHmmom umomwzm op homfi>hom3m was on pa o>moq .nm
empe>om .mucmz on page muhomop Sam umoscop ow pomfi>hom3m pom paw: .om
.. .Eonu mmopom moEoo on ho>ocesz mmopfl 3o: usonm Homfi>homsm ma; Homecou .mm
.mEmumohm 36: wow mowufipofipm wawuuom a“ acmEm>Ho>cfl Meow .Hm
.mcoflmmsomfiw oucfi mhflwmmm Hmcomhom wchSpoupcfi mo ucfiom m oxmz .nm
.mcofiuwwcoo HmooH oucfl paw ow mamhwohm 3o: wcfixmfiwoe uopfimcou .om
mmhm>om .uOmfi>uOQSm nu“: mcofipmmno>cou can“ whammmw Hmaomuom mcflmcfiun pfio>< .mm
mmno>mm .mampmoum 36: pmomMSm op homfi>nomsm was ou we o>moq .mH
.ponuemou cog: SHwEmm m.homfl>hom:m we: «zoom opwsccm .HH
.mhflmmmm chompom m.pomfl>hom:m we: psonm mcofipmemmsm homwo .m
.poumoscmh “on can: co>m mpoohoum pamphso mo mpHOQOH ca pcom .w
.zfiwewm m.homfi>homsm ma: mo mofiufi>wuom may spa: wHomEfi: chooaou .n
.homfi>hem:m cue: :oquMHoucw pcosc0Hm cmsopnp mEmHmOHQ 36: moHo>OQ .o
.homfi>uomsm new: mqum 3m: o>wumuaou Chasm .e
Mawwou wcwwhoz gonads
EopH

 

 

.moowpcw pom pouomamm mcofiummsc :ofiumUficSEEou .mm OHan

64

than 0.20 was determined for each item, and only those

items with more than three26

such correlations were
retained.

The inter-item correlations for these selected
items are shown in Table 6. On the basis of these inter-
item correlations it became evident that item 52 should be
reversed before inclusion in an index of innovation, and
that item 45 should be dropped from the set of items
relating to the maintenance dimension.

Having selected these items (which are shown in
Table 5a), their factor loadings were checked. As shown
in Table 5, each item that was retained, had a prime load-
ing on at least the factor that related to its conceptual
dimension. Items 8, 31 and 41 show loadings of more than
0.30 on another factor as well, which indicates that they
are not clearly independent. However, because of their
high number of inter-item correlations, these items were
retained, and four indices were constructed based on the
simple sum of scores on each of the four sets of items.
These four dimensions of procedural communication were the
only communication dimensions that were included in the
subsequent analysis. As a final check on the independence,

or orthogonality, of these indices, interindex correlations

 

26For time initiation dimension, the criterion was
relaxed to > 3.

65

 

OH.
NH.
ON.
OO.
«N.
mH.
ON.
ON.

Hm

vo.
OV.
em.
NN.
QH.
ON.
OH.

NO

Om. .
mm. HO.
OH. mm.
Nm Om

XOOGH may
Ommmohw en ow me

NH. .

OH. mm. .

om. Om. HN. .

«O. NO. NO. HN. .

NO.-NH. NO. NH. ON.

OH.-Om. OH. OH. HO.

Ow Om NN mN HH

Nm

Om

. mH
ON. O

OH O EouH

 

eefleeeeHeH HeO

seem
EOOH

Hm
Ne
me
Om
NN
mN
HH
. O
ON. . N

e N EeeH

 

mocmcmuchz HOV

ON.- .
HN.-mm.
mH.-Nv.
ON.-Nm.
OH.-VH.
ON.-mH.
OO.-HH.
Nm ow

OO

mN. . mm
mm. mH. . Ow
NN. Om. ON. . me
HN. HN. ON. ON. . He
OO mm Ow me Hv EouH

 

wdHOGODOom HOV

KOOcH Oz“ :H :onsHocH

OHoHOO Ommnm>eh on on NO EOHH

ON.
HN.
mN.
OM.
Nm.

Om

OH.
ON.
Nm.

Nm
Ow
Om
. mm
ON. . HO
OH. . ON
ON. HN. . O
OH. ON. Om. . v
mm Hm ON O v EmuH

 

:oHum>o::H Hmv

 

 

 

.mEOHH.” HHOHHNUHGHHEEOU UmpUOHmm HOW WCOHHNHOHHOU EOHHIHQHZHH

.O OHan

66

were determined. These are shown in Table 7 below, none
are statistically significantly different from zero at the

p < .05 level.

Table 7. Inter-index correlations.

 

 

 

Index Innovation Maintenance Sequencing Initiation
Innovation . 0.11 0.18 0.16
Maintenance . 0.06 0.05
Sequencing . 0.00
Initiation

 

NS. two tailed test

As a further check on the use of these indices at
different coorientational levels, a similar factor analysis
was made of the field agent responses about how the pro-
cedural aspects of their supervisory communication were
expressed in their actual communication. This analysis
relates to Section IV of the instrument (see Appendix) and
includes the same items as Section III, but with a different
set of foils.

The same four factors were identified using a vari-
max rotation, with the same order as was found for the

items relating to the procedural rules. The four factors

67

(representing the conceptual dimensions of innovation,
maintenance, sequencing and initiation) accounted for 29
percent of the variance.

Table 8 shows the factor loadings for the selected
set of 25 items that were chosen on the basis of the anal-
ysis at the rule level. The factor loading on the assigned
index is underlined in each case. Twenty out of the 25
selected items prime load on their assigned index; while
2 prime load on another factor and 3 do not clearly prime
load on any of the factors.

On the basis of the strong overlap, it was decided
to use the indices that were chosen on the basis of the
original analysis at the rule level.

A similar procedure was used to analyse the
responses by the field agents to the seven items relating
to satisfaction with role achievement and their relation-
ship with their supervisor (Section VI of the instrument).
Two factors accounting for 57 percent of the variance were
identified. These factors corresponded to the conceptual
dimensions of achievement and relationship.

Table 9 shows the factor loadings for the seven
items (which appear in Table 9a) and Table 10 their inter-
item correlations. On the basis of these data, items 2, 3
and 5 were summed to form an index of role achievement and

items 4 and 7 were summed to form an index of perception of

68

Table 8. Factor loadings for selected items relating to
expressed communication.

 

 

Factor loadings

 

 

Item
number 1 2 3 4
Innovation Maintenance Sequencing Initiation
4 2191 0.17 0.11 -0.07
6 gypg 0.17 -0.23 0.30
7 0.04 'QLQQ -0.07 0.13
8 0.64 0.06 -0.07 ‘0L03
9 0.04 QLii 0.03 0.08
11 0.00 ‘QLSQ -0.07 -0.08
13 0 16 0.11 0.20 QLiQ
23 -0 29 ‘QLSZ 0.25 0.02
26 -9;02 -0.23 0.44 -0.17
27 -0.18 0:58 0.01 -0.31
31 9491 -0.01 0.22 -0.27
33 Egg; -0.04 0.00 0.32
36 0.17 0.34 -0.10 .9112
37 0.26 -0.10 0.28 QOiQ
38 915; 0.06 -0.07 -0.26
39 -0.03 0:55 0.20 -0.18
40 Qggg 0.14 0.30 -0.08
41 0.09 0.18 0452_ -0.10
45 0.00 -0.30 QLQi 0 20
47 -0.02 .QLiZ 0 16 0.07
48 0.35 -0.09 Qggp 0.22
51 0.03 Qggp 0.11 -0.25
52 -0;40 0.09 -0.14 -0.06
53 0.12 0.18 0 51 0.34

O
H
N
O
H
I'—‘

60 0.01 0.22

 

69

Table 9. Factor loadings for items relating to output.

 

 

Factor loadings

 

 

Item number Role achievement Relationship
1 0.63 0.42
2 gypg -0.06
3 .Qggg -0.05
4 -0.20 HOBO
S QLél -0.15
6 0.58 0.44
7 0.08 0.85

 

Table 9a. Questions relating to output measures.

 

 

Item
number

Wording

 

1.

How satisfied are you that you have sufficient com-
munication with your supervisor to do your job well?

How satisfied are you with your present job when you
compare it to similar county extension positions in
the state?

How satisfied are you with the progress that you are
making towards the goals that you have set for your-
self in your present position?

How would you describe your overall relationship with
your supervisor?

How satisfied are you with your present job when you
consider the expectations you had when you took the
job?

How satisfied are you with your job relationships

with your supervisor in so far as he is helping to
make your work a rewarding and successful experience?

How would you describe your overall interaction with
your supervisor?

 

1

70

Table 10. Inter-item correlations for items relating to

 

 

 

output.

Item 1 2 3 4 5 6 7
.26
.39 0.43

.02 -0.13 -O.lO .

.25 0.50 0.72 -0.10

.67 0.25 0.27 0.09 0.25 .
0.27 0.12 0.08 0.60 0.00 0.21

VOMJ>CNNH
OOOOO

 

relationship. Items 1 and 6 were dropped from the sub-
sequent analysis.

Essentially the same procedure was used for the
five items (shown in Table lla) relating to background of
the field agents (Section I of the instrument). Two
factors, accounting for 71 percent of the variance were
identified. A consideration of the factor loadings shown
in Table 11 and the inter-item correlations shown in
Table 12, indicates that items 2, 3 and 4 do in fact cor-
respond to the conceptual dimension of similarity of
background; Mfifilf! items 1 and 5 are relatively independent.
On this basis all three items were retained to form an
index of similarity of background, and two single item
measures of length of interaction and educational back-

ground.

71

Table 11. Factor loadings for background measures.

 

 

Factor loadings

 

 

Item number Background ~-
1 0.42 0.60
2 0L1; 0.41
3 9O§Z 0.23
4 QLSE -0.01
5 0.02 0.88

 

Table 11a. Questions relating to background measures.

 

 

Item
number Wording

 

1. How many years have you worked under this person's
authority?

2. How many years have you been working in your
present position?

3. How many years have you been in extension work?
4. How old are you?

5. If you have received graduate training, please
specify below.

 

72

Table 12. Inter-item correlations for background

 

 

 

measures.
Item 1 2 3 4 5
l
2 0.47
3 0.49 0.62
4 0.21 0.48 0.64
5 0.26 0.31 0.22 0.19

 

Inferential analysis

 

In considering the major statistical analysis of
the study, certain general issues will be considered first,
then the detailed procedures for testing the two theoretical
hypotheses outlined in Chapter II. As discussed in con-
sidering the development of indices, individual items were
generally not used directly in hypothesis testing, but
rather as components of summation indices developed from
factor analyses.

Lacking any 'a priori' regression information,
these indices were applied as simple unweighted summations
of the items showing high inter-item correlations and/or
prime loading on orthogonal factors.

Each of the communication indices represents a

particularly crucial aspect of the study and therefore a

73

number of items were included in the instrument and were
retained in the indices (see Tables 4 and 5). The range
for these communication indices varied from a minimum of
4 to 20 for initiation, to a maximum of 8 to 40 for both
innovation and maintenance.27 Given these ranges, the
indices were treated as interval data and the inferential
hypothesis testing was based upon the Pearson Product
Moment Correlation coefficient,28 which is the interval
measure most appropriate to survey data.

In using the product moment correlation coefficient,

the linear rule was taken as a highly desirable approxima-

 

tion in our quest to predict and explain human behavior.
The advantages of the linear rule are outlined by Hays
(1963, p. 565), and any small deviations from linearity
are treated as error effects rather than true effects.
Where the assumption of linearity is significantly vio-
lated, then neither zero order nor partial correlations
can be validly calculated using r. For this reason etas
are reported and also any statistically significant

curvilinear relationships.29

 

27For Sections IV and V, the range for each item is
only 1 to 4 and therefore the index ranges are reduced
accordingly.

28The assumptions and computation of r are out-
lined in detail in McNemar (1962).
29 2 2
F = (n 'r )/(G‘2) - (McNemar (1962, pp. 275-281)).

 

(1-n21/(N-G)

74

One other issue of concern is the question of
inference when we are dealing with a finiteypopulation.
McNemar (1962, p. 93) and Edwards (1968, pp. 40-43) both
discuss this issue. In this study, we included 100 out of
a total population of about 500 dyads, which gives a sample
size of about 20% of the population. This is at the level
that Edwards suggests is reasonable to neglect the need for
a finite correction factor.

An a level of 0.05 was established for the rejection
of the null hypotheses in this study (the null hypotheses
being stated in terms of r not differing significantly from
zero). The t test for testing the significance of the dif-
ference of an achieved r from zero is outlined in Hays
(1963, p. 520) and McNemar (1962, pp. 137-138).30

Another issue of general application is the Eyeps-
formation of index scores. The focus of the hypotheses is
on the coorientational similarity between index values.

The general way of achieving a measure of similarity is to
determine a difference score between pairs, as noted by
Wackman (1969). However, this yields a measure of dis-

similarity rather than similarity.

 

30For N greater than 50, McNemar suggests the use
of the 2 distribution:

9H”

75

For ease of interpretation, each index difference
score was thus subtracted from the maximum score for that
index plus one, so that larger index values reflect greater
similarity.

= + - - I

c (Imax 1) (II1 IZI)

where:

similarity score

maximum possible score
on this index

I1 Individual l's score

HO
II II

I2= Individual 2's score

Using this formula, similarity scores were developed
for each of the four levels of similarity that were des-
cribed earlier in this chapter:

1. Agreement on the rules of procedural communication.
2. Agreement on the actual dyadic communication.

3. Field agent accuracy in predicting supervisor
perceptions.

4. Supervisor accuracy in predicting field agent
perceptions.

Hypothesis testipg

 

General Hypothesis 1 (see previous chapter) can be

rephrased in the Ho form as: That there will be a zero

 

correlation between coorientational similarity (as demon-

 

strated by agreement and accuracy) with respect to

 

procedural aspects (rules and expression) of supervisopy

 

76

communicationLyand inteppersonal relationship and evalua-

 

tion between field agents and their supervisors in the

 

extension service.

 

This hypothesis can be applied to each of the four
dimensions of procedural communication; and thus it gener-
ates sixteen research hypotheses for each dependent
variable, or sixty-four possible relationships with this
set of data. These are depicted in Table 13.

These are one-tailed hypotheses (because of the
directional prediction in the theoretical hypothesis)
requiring the correlation of similarity indices with the
dependent variables. The significance of the difference
of the achieved r's from zero was determined using the 2
test. Because of the assumption of linearity inherent in
the use of the product moment correlation coefficient-~r,
eta values are also computed for all r's that are signifi-
cantly different from zero and the significance of any
curvilinearity tested. In addition, partial correlation
coefficients are reported to allow greater interpretation
of any significant relationships.

General Hypothesis 2 (see previous chapter) can
be rephrased in the Ho form as: That there will be zero
correlation between the summative similarity of orientations
(es demonstrated by agreement and accuracy) with respect

to procedural aspects of supervisory communication, and

77

Table 13. Empirically tested hypotheses that are derived
from General Hypothesis 1.

 

 

System output variables

 

Coorientational Field Field Super- Super-
variables Agent Agent % visor visor
(increased role relat' salary evalu- relat'
similarity) satis'n ship increase ation ship

 

Agreement re

 

 

 

 

 

 

rules
innovation + + + + +
maintenance + + + + +
sequence + + + + +
initiation + + + + +

Agreement re

expressed
innovation + + + + +
maintenance + + + + +
sequence + + + + +
initiation + + + + +

Field agent

accuraey
innovation + + + + +
maintenance + + + + +
sequence + + + + +
initiation + + + + +

Supervisor

accurac
innovation + + + + +
maintenance + + + + +
sequence + + + + +
initiation + + + + +

 

78

inteppersonal relationship and evaluation between field

 

agents and their supervisors in the extension service.

 

This hypothesis can be applied to each of the five
output measures as shown in Table 14. It is a two tailed
hypothesis requiring testing the significance of the
multiple regression equation. In this case a Least Squares
Delete regression program is used to determine the collec-
tive effect of the sixteen measures of coorientational
similarity, and to determine the most parsimonious set of

measures that should be taken into account.

Descriptive analysis

 

Section II of the instrument is primarily used to
provide descriptive data on the communication interaction
of field agents and their supervisors. Summary statistics
in the form of means and standard deviations are reported
where-ever appropriate and the t test is used to determine
the significance of differences between field agents and
supervisors. As well as summary statistics on Section II,
descriptive statistics are also reported for the various

levels of coorientational similarity between dyads.

79

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 14. Empirically tested hypotheses that are derived
from General Hypothesis 2.
System output variables
a a. a.
<u- uua -H
Gcn ax: Ht: HO:
OH4 men oc: ocn
mu O: NO mH m:
Coorientation < 3‘; 4.2 :3 St; '93
and '9 9:: 2:2: 23.3 3::
background '32 EH mu 9.: §.H
varlables ,2 8 LL, 2,’ 1.5 {3.1, m2
Background
Similarity
Interaction
Distance apart
Education level
+ (1) Agreement re rules
InnovatiOn
Maintenance
Sequencing
Initiation + + + + +
Background + (2) Agreement re expressed
Innovation
Maintenance
Sequencing
Initiation + + + + +
Background + (3) Field agent accuracy
innovation
Maintenance
Sequencing
Initiation + + + + +
Background + (4) Supervisor accuracy
Innovfition
Maintenance
Sequencing
Initiation + + + + +
(1) Agreement on rules
+ (2) Agreement re expressed
+ (3) Field agent accuraey
+ (4) supervisor accuraey + + + + +

 

 

CHAPTER IV

RESULTS

The results of this study are presented in four
main sections. First, the characteristics of the sample;
second, the major differences in perceptions of communica-
tion between supervisors and subordinates; third, the
actual levels of coorientational similarity that were
found; and fourth, the results for each hypothesis will

be shown.

Characteristics of the Sample

 

The 100 field agents in the study were stationed
in 63 of the 82 counties in Michigan, and therefore the
sample included agents dealing with both rural and urban
problems and with a considerable range in their proximity
to their supervisors. Table 15 shows the distribution of
distances between agents' work centers and their nominated

31 The median distance apart was 97 miles.

supervisors.
All of the supervisors had a Masters or Doctoral

degree, while approximately two-thirds of the field agents

 

31Nine of the supervisors were stationed in East
Lansing and one in Marquette, Michigan.

80

81

had Masters degrees and the remaining one-third had less

than a Masters degree.

Table 15. Distance of agents from their supervisors.

 

 

 

Distance Percentage of sample
0 - 49 miles 11
50 - 99 miles 41
100 - 149 miles 22
150 - 199 miles 11
200 - 549 miles 15

 

In general, the field agents and their supervisors
had highly similar backgrounds in extension. The similarity
of background index (based on age, time in current position
and years in extension work) had a possible range from 4 to
15, with a score of 15 representing complete similarity.
Table 16 shows the level of similarity that was found in

this study.

Table 16. Similarity of extension background in field
agent/supervisor dyads.

 

 

 

Similarity score Percentage of sample
14-15 33
12-13 25
10-11 28

less than 10 14

 

82

Exactly half of the field agents had worked for
less than five years under their supervisors' direct
authority and half for over five years. Fourteen percent
reported less than two years working under the particular
supervisor.

The field agents were generally highly satisfied
with their role achievement. The index of role achieve-
ment had a maximum range of 3 to 15. The mean score was
12.53. Table 17 shows the distribution of index scores

on role satisfaction.

Table 17. Distribution of scores on field agent role

 

 

 

satisfaction.
Role satisfaction score Percentage of sample
14-15 35
12-13 51
10-11 5
less than 10 9

 

Both field agents and supervisors viewed their
relationships as somewhat more formal and business-like
than personal and social. The index had a maximum range
from 2 to 10, with a score of 2 representing a completely
formal and business-like relationship and a score of 10

representing a completely personal and social relationship.

83

The mean score for field agents was 4.83, while the mean
for supervisors was 4.96. Table 18 shows the distribution

of relationship scores.

Table 18. Distribution of scores on perceptions of the
dyadic relationships.

 

 

Percentage of sample

 

 

Relationship score Field Agents Supervisors
2-3 13 10
4-5 55 54
6-7 32 30
8-10 0 6

 

Percentage salary increases, of July 1971 salaries
over July 1968 salaries, ranged from 9 to 51 percent, with
a mean increase of 24 percent. Table 19 shows the distri-

bution of salary increases.

Table 19. Percentage salary increases for field agents.

 

 

 

Salary increase Percentage of sample
0 - 19.9 % 14
20.0 - 24.9 32
25.0 - 29.9 22
30.0 - 34.9 18

35.0 and greater l4

 

84

Just as the field agents in general were highly
satisfied with their role achievement, so the supervisors
tended to be highly satisfied with the performance of the
field agents. The supervisor evaluation index had a pos-
sible range of 2 to 10, with a score of 10 representing
complete satisfaction. The mean index score was 8.16.
Table 20 shows the distribution of supervisor evaluations

of field agents.

Table 20. Distribution of scores on supervisor evaluations
of field agents.

 

 

 

Evaluation score Percentage of sample
9-10 50
7-8 32
5-6 10
3-4 8

 

Finally, in considering the characteristics of the
sample, the association between the various background and
output variables is presented. Table 21 shows the inter-
item correlations between the four background measures.

The field agents' level of education and the simi-
larity of their background in extension to that of their
supervisors are both significantly associated with their
length of interaction with their supervisors. This may be

a reflection of a common dependence on time. Those who

85

Table 21. Inter-item correlations for background

 

 

 

variables.
Inter- Educa- Dis- Simi-
Index action tion tance larity
Length of interaction .
Level of education 0.26** .
Distance between centers 0.01 -0.16 .
Similarity of background 0.35** 0.11 -0.23* .

 

*p < .05 two-tail.
**p < .01 two-tail.

have been in the service for more than six years inevitably
will have had a long interaction with the particular super-
visor and will have had a longer time within which to have
completed a Masters degree. The other significant associa-
tion is between similarity of background and distance
between centers. This would seem to suggest that personnel
are promoted to counties closer to the supervisory centers.
Table 22 shows the inter—item correlations between the
various measures of the attractiveness of the supervisor/
field agent relationship.

Supervisor evaluation of the field agents' role
jperformance is significantly associated with the field
agents' role satisfaction, with the percentage salary
increase granted to the field agents, and with the super-

‘visors' perceptions of their relationship with their field

agents.

86

Table 22. Inter-item correlations for output measures.

 

 

Role Field Sup.
satis- relat' Evalu- relat'
Index faction ship Salary ation ship

 

Role satisfaction .

Field agent relationship -0.02 .

Salary increase 0.00 0.13 .

Supervisor evaluation 0.30** 0.15 0.29** .
Supervisor relationship 0.04 0.09 -0.09 0.37**

 

**p < .01 two-tail,

Table 23 shows the inter-item correlations between
the background variables and the output measures. Both
zero-order and partial correlations are presented to help
clarify the independent association between the variables.
The partial correlations control on the other background
variables and the communication variables that relate to

the actual agreement level.

Level of education has a 'true' significant associa-
tion with field agent role satisfaction; and supervisor
perceptions of their relationship is associated with the
degree to which they have a similar extension background

to that of the particular field agents.

87

.meu OOHHmu-03p

as. v gee

 

 

 

 

 

 

 

 

 

.pmmu OOHHmu-03u mo. v Ox
«HN.O «NN.O N0.0- N0.0 OH.O- OH.O- H0.0- N0.0 O0.0- m0.0- NuHhmHHEHm
Ooclommxe
:oncopxm
m0.0- O0.0- N0.0- m0.0- H0.0- O0.0 OH.O- NH.O- OH.O- HH.O- whoucmo
aeozumn
mocmumHO
N0.0- O0.0 O0.0 OH.O V0.0 H0.0 H0.0- OH.O «NN.O «eON.O :oHpmoscm
mo Ho>og
NH.O «NN.O OH.O eHN.O O0.0- NH.O- OH.O «OH.O N0.0 OH.O :oHuomhouaH
mo nuwcmq
Hmehmm h HmHuhmm u HmHuamm h HmHuhmm u HmHuhmm h OHanhm>
unsoumxomm
mHamcoHumHou :oHumsHm&m thHmm mHnmaoHpmHOu :oHpomHmHumm
homH>Homsw ucom< OHOHO oHom
.mOHDmmoe “amuse wam Ocsopwxomn comzuon mcoHumHOAuooxbwcH-uou:H .mN OHOmH

88

Supervisor and Field Agent Perceptions
of theirTCommunication

 

 

The major differences in perceptions of communica-
tion are shown in Table 24. Supervisors perceived less
frequent overall communication than did the field agents;
although, field agents perceived significantly less com-
munication about personal matters and new ideas than their
supervisors did. Overall the supervisors were less satis-
fied with the amount of communication than were their
field agents and in particular significantly more super-
visors wanted more communication about the existing program.
Field agents perceived significantly more of the communica-
tion than they received as being imposed by supervisors,
rather than being sought by them. Similarly, the field
agents also perceived significantly more of the communica-
tion, coming from their supervisors, as relating to getting
the job done (the existing program) rather than relating to
personal matters and new ideas.

Table 25 summarizes the index scores for Supervisors
and field agents on the four dimensions of procedural com-
munication. On all four indices, supervisors expressed
significantly greater interdependence (higher index scores)
on the procedural rules of communication between field
agents and their supervisors. There were no significant

differences between the two groups in their perceptions of

89

Table 24. Supervisor/field agent differences in
perceptions of communication.

 

 

Super- Field

 

 

 

 

Communication visors Agents 2 Signifi-
dimension % % x cance
Initiation
supervisor initiates 20 26
field agents initiate ll 15 2.18
both about equally 69 S9 df=2 N.S.
Mode
more written l6 17
1/2 and 1/2 49 35 4.40
more oral 35 48 df=2 N.S.
Frequency
once a week or more 6 15
once a month or more 47 54 7.62
less often 47 31 df=2 p < .05
Satisfaction with amount
satisfied1 25 47 9.57*
like more 75 53 df=l p < .01

 

Sequencing of communication
to field agents
more directed by

supervisors 17 39
1/2 and 1/2 59 48 13.04
more sought by field
agents 24 13 df=2 p < .001

 

Sequencing of communication
to supervisors
more directed by field

staff 31 31
1/2 and 1/2 54 50 0.62
more sought by
supervisors l5 l9 df=2 N.S.

 

*Yates correction applied-

1Includes one field agent who would like less
communication.

90

Table 24.--Continued.

 

 

Super~ Field

 

 

 

 

 

Communication visors Agents Signifi-
dimension % % x cance
Percentage of supervisor
initiated communication
about the existing
program
0-40% 51 23
41-60% 29 37 18.23
61-100% 20 40 df=2 p < .001
Percentage of field
agent initiated communi-
cation about the
existing program
0-40% 51 40
41-60% 29 31 3.05
61-100% 20 29 df=2 N.S
Initiation about the
existing program
supervisor initiates 23 20
field agent initiates 12 15
both about equally 64 62 1.57
don't communicate
about existing 1 3 df=3 N.S
Frequency of communica-
tion about the existing
program
once a month or more 46 39 0.74*
less often2 54 61 df=1 N.S.
Satisfaction3 with
communication about the
existing program
would like more 68 50 5.68*
satisfied 31 48 df=1 p < .05

 

*Yates correction applied.

2Includes those who don't communicate.

3Excludes those who don't communicate.

91

Table 24.--Continued.

 

 

Super- Field
Communication visors Agents 2 Signifi-
dimension % % x cance

 

Initiation about personal

matters
supervisor initiates 16 11
field agent initiates 16 24
both about equally 56 57 3.33
don't communicate
about personal 12 8 df=3 N.S.

 

Frequency of communica-
tion about personal

 

 

 

 

matters
once a monEh or more 20 9 4.03*
less often 80 91 df=1 p < .05
Satisfaction3 with communi-
cation about personal
matters
would like more 20 16 0.55*
satisfied1 68 77 df=1 N.S.
Initiation about new ideas
supervisor initiates 19 6
field agent initiates 19 36 14.24
both about equally 61 54 df=3 p < .01
don't communicate
about new ideas 1 4
Frequency of communication
about new ideas
once a month or more 38 22 5.36*
less often2 62 78 df=1 p < .05
Satisfaction3 with communi-
cation about new ideas
would like more 75 62 2.40*
satisfied 24 34 df=1 N.S.

 

*Yates correction applied.

1Includes one supervisor who would like less communi-
cation.

zIncludes those who don't communicate.

3Excludes those who don't communicate.

92

 

 

 

 

 

 

 

 

 

 

OH.N Hm.OH meomfl>eeasm
m2 mN.H HN.OH OH-e museum uHeeO :eeeeeefieH
NN.N om.mH meeefi>eemsm
mz NN.N ee.mH ON-m eeeeme efieem meeeeeseem
OH.O Oe.OH whomH>ummsm muse a cam
mz efi.m mm.mH Nm-m meeeme efieem e e . z
Hm.m NO.HN whomH>Hmmzw :oHpmoHQDEEou
mz mO.N No.NN mm-m meeeme efieem eeeee>eeeH Heseb<
oo.~ oe.NH beemfl>eeesm
eee Nm.N Om.mH om-e meeeme efleem eefleeeeeeH
ON.N ma.o~ meeme>eeesm
eee mH.m NO.NH mN-m meeeme efiewm mefieeeseem
Nm.m mm.m~ meome>eemsm
eee mH.e NN.eH oe-w meeeme caeflm eeeeeeeeeez
me.~ em.Nm meeme>eeazm
. em.m ew.Hm oe-w meeeme uHeem eeeee>eeeH mefiem
mamas coozuon .MHHO .O.m awe: mmcmh HO>OH xOOGH m3oom I
Ho mocmonchHm ESEmeE
.coHumoHcSEEoo mo muoommm
Hmpswouopm co mOhoum XOOGH :H mmo:opemeO ucomm OHOHm\pomH>Homsm .mN OHOmH

93

Hoe. v a

 

 

 

 

 

«kt
Ho. v a «e
me. v a e
omfiume peep e
eN.H Nm.OH meeme>eemsm see me He
mz oo.N NH.OH efi-e meeeme efiefim .e .e. H
eo.~ NH.mH meomfi>eemsm
mz mN.N me.m~ om-m meeewe efieflm meebeeseem
OH.O NH.OH whomH>homsm menace cam :oHumoH::EEoo
mz mN.m mm.eH mm-w meeewe efiefim e . 2 Heeebe we
:oHpmouhmm
NO.N Nm.mN mHOmH>pmmsm m.honuo Ho
ee Nm.N NO.HN Nm-w meeeme efiewm eofiee>oeeH eefieeflweem
mcmoe ceozuon .HHHO .Q.m awe: omcmh Ho>oq xOO:H msoom
mo oocmuHchmHm EDEmeZ

 

 

veneeeeou--.mN banee

94

the actual communication; but the supervisors attributed
significantly more interdependence to the field agents'
innovative communication than did the field staff in

predicting to their superviSors.

Actual Levels of Coorientational Similarity
Between SuperViSors andFieldTAgents

 

 

The levels of achieved similarity are presented in
Table 26. The outstanding overall feature is the high
level of similarity on all communication dimensions at both
the agreement and accuracy levels. Another major feature
is the stability of both the means and standard deviations
for a particular communication dimension over the three
measures--actual perceptions, field agent predictions of
supervisor perceptions, and supervisor predictions of field
agent perceptions.

As would be expected from Table 25, the least
degree of similarity of orientation, for all dimensions,
occurred at the level of rules of communication.

Using Scheff's (1967) model of types of under-
standing (discussed in Chapter II) it is possible to
consider the combined influence of agreement and accuracy.
Table 27 presents the 2 x 2 matrices combining agreement
and accuracy, with the distributions dichotomized at the
category next to that containing the mean. There is a

very significant tendency on all dimensions for the

95

 

 

 

 

 

 

Table 26. Summary of dyadic agreement and accuracy
scores.
Maximum Adjusted
Focus Index range Mean S.D. mean
Agreement on Innovation 8-412 37.09 2.70 88.1
rules Maintenance 8-41 35.31 4.31 82.7
Sequencing 5-26 22.22 2.66 82.0
Initiation 4-21 18.10 2.17 82.9
Agreement on Innovation 8-33 30.14 2.64 88.5
actual .
communication Malntenance 8-33 29.41 2.85 85.6
Sequencing 5-21 18.66 1.93 85.3
Initiation 4-17 15.20 1.44 86.1
Field agent Innovation 8-33 30.39 2.26 89.5
accuracy in . _
predicting Malntenance 8 33 29.54 2.70 86.1
supervisor Sequencing 5-21 18.91 1.78 86.9
reSP°nses Initiation 4-17 15.23 1.33 86.3
Supervisor Innovation 8-33 30.06 2.39 88.2
accuracy in . _
predicting Ma1ntenance 8 33 29.57 2.64 86.2
field agent Sequencing 5-21 18.87 1.97 86.6
respOnses Initiation 4-17 15.28 1.36 86.7

 

1Means transformed to their position on a range

from 0-100.
2
orientation.

Higher scores represent greater similarity of

.He>eH Hoo.o v a pa eeeeflmfiemwmeee

.OOHHOOO :oHuompNoo mmum>

 

 

 

 

 

96

 

 

 

 

H
OH OH 30g
«ea ee.eH me a OOH: meemfi>eegam
OH OH 304
nee OH.ON OO N OOH: maceme OHOHO :oHumHuH:H
O OH 304
eee NN.eH Ne NH OOH: meemfi>eeenm
O OH 304
«ea O¢.OH OO ON OOH: mucmmm OHOHO OaHocmscmm
N NN 304
eee Om.OO OO O :OH: muomH>Nomsm
O ON 30H
«he mm.Nm mm N OOH: muaeme uHeNm beeeeeeefiez
O HN 304
eee mo.wH em OH OOH: meemfl>eemam
OH NH BOO
«tt ON.OH HO O :OH: madame OHon coHum>occH
oocmonHcmHm oxx OOHm 304 acmsmopw< Ho>OH xOOGH
N Nomusoo<

 

.coHmeH::EEoo Hampom one psonm Nomhsoom Ocm unoEomuOm mo :oHumcHOEoo .NN OHOmN

97

respondents to view the system as consistent with their
particular orientation. This trend is indicated by the
high proportion of responses in the High-High and Low-Low
cells. These cells represent Scheff's categories of
monolithic consensus and false consensus.

Finally, in describing the achieved levels of
similarity of orientation, Table 28 relates centrality of
perceptions of actual communication to the levels of
coorientation. Perceptions that deviated more than one
standard deviation from the mean score for that group,
were treated as deviant. Deviations from mean scores on
perceptions of actual communication by field agents were
not significantly associated with level of dyadic agree-
ment except with respect to the sequencing dimension.
However, for all dimensions of communication, supervisor
deviation from mean perceptions was significantly associ-

ated with low dyadic agreement.

Relationship Between Similarity
of Orientation and Attraction

 

 

Table 29 shows the zero order and partial correla-
tions between the sixteen measures of coorientation toward
the procedural aspects of communication and the five
measures of system performance. Agreement on procedural
rules of initiation of communication was the only measure

at the rules level that was significantly associated with

.Ho>OH HOO. v Q 0:“ pw pcwuHmHamHm«¢«
.He>eH Ho. v a one He peeuHOHeOHmee

.OOHHOOO :oHuomHHoo mmpm>

 

 

 

 

 

98

 

 

 

 

H
NH N ucmH>eO
««« HO.NH OO OH Hmpucou mH0mH>Homsm
OH O uamH>OO
.O.z om.o OO ON Heeuaeu meeeme cHeHO aoHHeHHHeH
HH OH ucmH>mO
«« He.O OO HH Happcoo whomH>HOmsm
OH NH pcmH>OO
«« O0.0H «O O Hwhuceu mpcomm OHOHO OcHocmsdom
O OH ucmH>OO
*«e NO.NH NO OH Heeeeeu OuomH>eemsm
OH O pcmH>mO
.m.z O0.0 OO ON Hmupcmo mucemm OHOHO mocmcounHmz
O OH pcmH>OO
«ex OO.HN «O OH Hmuucmu OHOOH>HOOSO
HH O ucmH>om
.O.z O0.0 OO ON Heeeeeu Ogeeme OHeHO :eHee>oeeH
mocmonHcOHO oNx OOH: Bog NuHHmHucmo Ozone xowcH
H N zomhsoo<

 

 

.mcoHummopom Ozonm :mmE thm cOHumH>OO mo pumwmm .ON OHOmH

99

.mep OOHHmu-o:o HO. v OOO

.umou OOHHmH-oco OO. v m«

 

 

 

 

 

 

 

 

 

 

OH.O mH.O OH.O mH.O O0.0- H0.0. <0.0- «0.0n O0.0 O0.0 :oHumHuHcH
N0.0 mH.O O0.0 mH.O O0.0 N0.0 O0.0- O0.0 OH.O- O0.0- OcHocosOom
OH.O- NH.O- OH.O- O0.0- O0.0 H0.0 «H.O- OH.O- H0.0- O0.0- mocmcopaHmz
HN.O «N.O O0.0 O0.0 O0.0 N0.0 OH.O OH.O OH.O NH.O :oHpm>o::H
an {as {as «us
cOHmeHGSEEoo
Hm3pom
:o psoEOOHO<
N0.0 N0.0 mm.o mm.O O0.0- O0.0- OH.O- NH.O- MN.O mm.o :oHumHuHcH
OH.O- NH.O- OH.O- O0.0- O0.0- O0.0- O0.0- HH.O- O0.0 O0.0 OcHocmscom
O0.0 H0.0 O0.0 N0.0 O0.0 O0.0 OH.O- OH.O- O0.0- O0.0- ooamcouaHmz
OH.O OH.O O0.0- O0.0- O0.0- O0.0- N0.0 OH.O N0.0- O0.0- :oHum>occH
OOHSH
no pzosomhm<
Hmepmm H HmHuHmm N HmHuNmO H HmHunm h #mHuth H xOOGH HO>OH
mHzmaoHuwHOH :oHpmsHm>m NumHmm mHzmcoHpmHOH :oHuomHmHumm
mo mcoHOQOOHOO Ho mcoHumOOHoO OHom
HomH>Homam pcmm< OHOHO
mmpsmmoz usmuso mousmmez
:OHpmoHasesou
.mopzmmoe usmuso ou :OHHmpaloooo mo OHgmcoHumHom .ON OHOmh

100

.meu OOHku-o:o HO. v O«*

.HOOH OeHHep-eco OO. v me

 

 

 

 

 

 

 

OH.O mH.O OH.O OH.O H0.0 H0.0- O0.0- O0.0- O0.0- O0.0- :OHpmHuHcH
OH.O mm.O OH.O mH.O O0.0- OH.O- OH.O OH.O OH.O- OH.O- OcHocmscom
OH.O- OH.O- OH.O- NH.O- O0.0- H0.0- OH.O- ON.O- OH.O- NH.O- mocmnouchz
OH.O OH.O O0.0- N0.0 O0.0- N0.0- O0.0 O0.0 O0.0 O0.0 :oHum>occH
HumHsoom
HomH>Hom3m
mN.O OH.O HH.O NH.O N0.0- O0.0- O0.0 O0.0 O0.0 N0.0 :oHpmeHcH
N0.0 NH.O N0.0 OH.O O0.0 O0.0 N0.0- O0.0 HH.O- NH.O- OnHocesdom
H0.0- O0.0- O0.0 H0.0 N0.0- H0.0- HN.O- ON.O- H0.0 O0.0- oucmcouchz
OH.O «N.O ON.O O0.0 N0.0- H0.0- OH.O NH.O H0.0- N0.0 :oHum>o::H
O *« «« «x a
HomHsoom
mmmum OHon
HmHuHmm H HmHuHmm H HwHuHmm H HmHunO H HmeHmm H xOO:H Ho>OH
mHnmaoHumHOH :oHumsHm>m HHmHmw OHHOGOHHOHOH :oHuommOHumm
mo mcoHummuHmm mo mcoHHQOOHO oHom
HomH>HOQ3m usew< OHon
OOHSOOOZ usmpno OOHsmmmz
:oHumquseaou

 

 

 

weseHHeeu--.ON eHOeH

101

any of the criterion measures. At the level of agreement
on the actual dyadic communication, innovation, sequencing
and initiation were significantly correlated with two of
the output measures--supervisor evaluation of the field
agents and supervisor perceptions of the relationship.
Field staff accuracy in predicting supervisor
perceptions of the actual communication with respect to
innovation was significantly associated with field agent
and supervisor perceptions of the relationship; and with
supervisor evaluations. Field staff accuracy in predicting
supervisor perceptions of initiation was also significantly
associated with supervisor perceptions of the relationship.
Finally, supervisor accuracy in predicting field agent
perceptions of sequencing and initiation was significantly
associated with supervisor evaluations and supervisor
perceptions of the relationship, and with field agent
perceptions of relationship in the case of sequencing.
Coorientation toward maintenance aspects of com-
munication was consistently negatively associated with the
criterion measures at all levels except that of rules.
Also, none of the sixteen measures of coorientation were
significantly associated with percentage salary increase.
In general, Table 29 indicates little difference
between the zero order and partial correlations. However,
there is a trend for the partial correlations to be lower

than the zero order correlations. This suggests that in

102

these cases, the zero order correlations are boosted through
the communication and control measures working through each
other.

Table 30 presents the eta2 for each significant
association, and also a measure of the degree of curvi-
linearity. In only two cases, agreement on rules of
initiation with field agent role satisfaction, and field
staff accuracy in predicting supervisor perceptions of
innovation, were the relationships significantly curvi-
linear. Nevertheless, Table 30 indicates that the eta
values allow considerably greater prediction than the zero
order correlations. On the basis of the array means for
the communication measures, that showed significant correla-
tions in Table 29, it is possible to predict between five
and twenty-seven percent of the variance in the output
measures, as shown in the eta2 scores in Table 30.

Table 31 presents the evidence relating to
Hypothesis 2. Only seven of the twenty-five regression
equations were statistically significant at the p < .05
level. None of the regression equations relating to role
satisfaction, field agent relationship with supervisor or
salary increase were significant. However, seven of the
ten regression equations relating to supervisor evaluations
of field staff and supervisor relationships with field

staff were significant.

103

Table 30. Variance explained by measures of coorientation.

 

 

 

 

 

 

Curvi-
Output 2 linear-
variable Communication measure n ity F
Field staff Agreement on rule on initiation 0.24 2.25*
role satis-
faction
Field staff Field staff accuracy in pre-
relationship dicting innovation 0.10 0.84
with Supervisor accuracy in pre-
supervisor dicting sequencing 0.07 0.52
Supervisor Agreement on rule on initiation 0.21 1.38
evaluation Agreement on actual communication
of field on innovation 0.27 1.93
staff Agreement on actual communication
on sequencing 0.17 1.35
Agreement on actual communication
on initiation 0.08 0.78
Field staff accuracy in predict-
ing innovation 0.26 2.30*
Supervisor accuracy in predict-
ing sequencing 0.14 1.64
Supervisor accuracy in predict-
ing initiation 0.10 1.38
Supervisor Agreement on actual communication
relationship on innovation 0.12 0.65
with field Agreement on actual communication
staff on sequencing 0.09 0.63
Agreement on actual communication
on initiation 0.05 0.39
Field staff accuracy in predict-
ing innovation 0.10 0.48
Field staff accuracy in predict-
ing initiation 0.09 1.75
Supervisor accuracy in predicting
sequencing 0.13 1.44
Supervisor accuracy in predicting
initiation 0.06 0.50

 

*Curvilinearity significant at p < .05.

104

 

 

 

.m.z OO.O OH.O NO.O NomHzoom Ocm HHOEOOHOO mo mHm>OH HH<
.m.z O0.0 O0.0 NH.O HomHsuom HomH>HOO5m + OHDOHOHOOO
.O.z HO.O O0.0 ON.O HomHsoom Hcmmm OHOHO + OHSOHOHomm
.O.z NO.O O0.0 ON.O HO>OH unoEOOHOm HmSHo< + OHSOHOHomm OmmeHocH
.m.z OO.O O0.0 ON.O HO>OH HHOEOOHOO OHsm + OHSOHOHomm OOOHMMMHMM
.m.z HH.H OH.O NO.O OumHsoom Ocm psoEOOHOm Ho mHo>OH HH<
.m.z OO.H NH.O OO.O HomHsoom HomH>Hmmzm + OcsoHOHomm
.O.z ON.H OH.O OO.O HomHsoum ucemm OHon + OQDOHOHomm
.m.z OO.H HH.O OO.O Ho>oH H:OEOOHOO Hm3uu< + OHDOHOHomm HomH>He93m :qu
mHnmcoHumHmH
.O.z NO.H NH.O OO.O He>eH useseeeme eHsm + weseHOHbem Omeem OHeHO
.m.z ON.H ON.O OO.O NomHsoom Ocm acoEOOHOm mo mHo>OH HH<
.m.z OO.H NH.O OO.O OomHzoom H0mH>Homsm + OGSOHOHomm
.m.z NN.H OH.O HO.O HomHsoom ucowm OHOHO + OHDOHOHomm
.O.z NO.H HH.O OO.O Ho>mH unoEOOHOm Hmzpo< + OHSOHOHOOO
:oHuommOHumm
.O.z ON.H OH.O O0.0 HO>OH HCOEOOHOO OHDO + OHSOHOHomm oHOH Hmmum OHOHO
mocmuHchOHm O Nu m mOHOmHHm> OHDwmmE Hzmuso

:oHumscm :onmmHOOm

 

 

.OmchHOH meHOmHHm> HHm :HHB mcoHpmHOHHoo OHOHHHDE

.HO OHan

105

 

 

 

:oHpmscm :onmOHOOO

.He>eH Ho. v 9 ea peeeHOHemHOeO
.He>eH Oo. v a em weeuHOHemHOO
.O.z ON.H ON.O OO.O NomHsoom Ocm ucosemHOm mo mHO>mH HH<
« OO.N NH.O H0.0 Hostoom HomH>HOmsm + Ocsonxomm
« OH.N OH.O O0.0 NomHsoom pcomm OHon + OHDOHOHOOO
ucomm
«« OO.N ON.O O0.0 Ho>OH HOOEOOHOO Hmzuo< + OOSOHOHOOO OHOHM :HHS
mHgmcoHumHOH
.m.z OO.H HH.O O0.0 Ho>mH unoEOOHOm oHsm + OHSOHOxomm HomH>Hmm3m
«« OO.N NO.O NO.O Nomquom Ocm H:OEOOHOO Ho mHO>mH HH<
.m.z OO.H NH.O OO.O NomHsoum HOOH>HOQSO + OHSOHOHOOO
O OO.N NH.O HO.O HomHsoom ucomm OHOHO + OCSOHOHOOO
«O NN.N ON.O Oe.o He>eH geeseeeme Heseb< + wesegmxeem Omeem uHeHO
mo :oHumsHm>o
« NH.N OH.O O0.0 Ho>OH HHOEOOHOO OHDO + OHDOHOHOOO HomH>Hmasm
moamonHcmHm O NO x moHOmHHm> OHSOOOE usmuso

 

 

.eeneHeeeo--.HO eHOeH

106

A least squares delete program was used to reduce
these equations to a minimal set of variables. Table 32
presents the significant sets of variables. None of the
regression equations reached the p < .05 level of signifi-
cance in relation to salary increase. With each of the
other criterion measures, some subset of two or three back-
ground and communication measures was able to explain

approximately ten percent of the variance.

107

.He>eH Hoe. v O OOH He HeeeHOHemHOOOO
.He>eH Ho. v 9 exp ee eeeeHHHeOnge
.He>eH Oo. v H eHe we peeeHOHemHOO

 

 

 

 

 

 

 

 

 

 

 

 

eee Oe.eH OH.O :eHee>eeeH Hampee a6 eeeEeeHm< OHeeO OHeHO
we :oHuwsHm>o
«ue OO.HH HH.O .Hmw coHHmHuHcH OHSO HomH>Hmmsm
ocoz OmmmHocH meHmm
OOmH:OOHOO
« N0.0 O0.0 mommaeuchE HomHsoom
mmmum OHOHO ”coHum>o::H Hanuom :o HGOEOOHO<
« O0.0 OH.O OucmcmchmE xumHsoom
.mzm muHmmm mocmumHO maoHuomHouaH mo OHOHOH
« N0.0 O0.0 mocmamuchE
NomHzoom mmmpm OHOHO ”coHuomHmch mo OHO:OH
« O0.0 O0.0 :oHum>o::H Hmsuom co “desoeHO<
muHmOm oozmumHO “coHuomHmuaH mo OHOHOH HomH>Homsm
OHHZ
e H0.0 O0.0 ..Hmm :oHueHOHeH eHsm OHOOcoHpeHeH
muHmmm OucmumHO ”:oHuomHeuaH mo Ouwcmq Hmmum OHOHO
«« NH.O HH.O OOHmGOHchE HomHsoom A§$OM.HOO :oHHmHuHcH OHzm :oHuommeumm
OHOH
«« ON.O HH.O .HOm :oHumeHcH OHSO maoHpmosvm Hmmum OHOHO
OuaonmHHOHm m m OOHHOHOH mOHOmHHm> OH3mmoE
coHumscm :oHOOOHOOO N coHumoHGDEEoo Ocm OHSOHOHOOO usmuzo

 

 

. mmhflmwwE

usmuso OHHB moHOmHHm> OeuooHom mo mcoHHmHOHHoo OHQHHHDZ

.NO OHLMH

108

.HO>OH HOO. O Opp Hm ucmuHchmmexg
.Hm>mH HO. a may um HdOUHmHamHm««

.Ho>OH OO. a may um unmonchHm«

 

 

 

 

 

 

 

 

 

«« OO.O NH.O OcHocoscem HomHsoum .Osm ”momm:OpaHmE
NomHsoom Hmmum OHOHO ”coHpm>occH HomHsoom
mmmum OHOHO ”Oucmc0Hche HOSHOO :o ucoEOOHO<
«e O0.0 OH.O :oHumHuHcH HomHsoom .Osm
”OcHocmscmm HomHsuum .Ozm HOHHHOHHEHO coHOOOme
«« ON.O OH.O :oHumHuHcH Nostuom mmmum OHOHO ”coHum>o::H
HomHsoom mmmum OHOHO ”HHHHOHHEHO :on:0pxm
«% mO.© mH.o GOHHN>OCGH
Hanuom co psoEOOHO< ”NHHHmHHEHm :oncouxm “comm
OHOHH OHHS
« OO.N O0.0 .HOm aoHum>occH OHsm mHamcoHumHmH
”NHHHOHHEHO :oncouxo ”coHuomHmucH mo AuwcmH H0mH>Homsm
*«« HO.NH HN.O coHpm>occH
HomHsoom mmwpm OHOHO m.HOw :oHuwHHHcH mHsm
« NO.O NO.O :oHpmHHHcH NomHsoow .mzm ”coHpomHOHcH mo AHOHOH
«ee OO.OH OH.O coHum>occH NomHzoom Hmmpm OHOHO
moconchme m m OochpOH OOHOOHHO> OHSOOOE
coHHmscm :onmOHOOO N :oHHOOHHSEEoo Ocm OHSOHOHOOO usmuso

 

 

.OmaeHeeeu--.NO eHOeH

CHAPTER V
DISCUSSION OF RESULTS

The major findings of this study can be grouped

into three sets:

1. A factor analysis of conceptual dimensions of
communication.

2. The actual communication between supervisors
and field agents, differences in their percep-
tions and their degree of coorientational
similarity.

and 3. The relationship between coorientational

similarity and interpersonal attraction.

Factor Analysis

 

One of the intermediate tasks in this study was to
develOp a set of indices of procedural communication.
While several authors have suggested dimensions of organiza-
tional communication, there were no reported attempts to
assess the independence or relevance of them to the inter-
action in a system like the Michigan Cooperative Extension
Service. Four independent dimensions were identified

through a factor analysis of sixty items relating to

109

110

procedural aspects of communication. These four factors
accounted for one-quarter of the variance; and they were
clearly distinct from subsequent factors, which showed

virtually no isomorphism with the conceptual dimensions.

The dimensions represented by the four factors
were the innovation function, the maintenance function,
the sequencing of interaction and the initiation of inter-
action.

The questions that prime loaded on the innovation
factor focussed on involvement or sharing in new ideas.
The stress here seems to be on the distribution of new
ideas, with a contrast between field agents being actively
involved on the one hand and merely implementing without
discussion on the other.

The questions relating to the maintenance function
all focus on the extent to which personal matters should
be brought into supervisory communication. Both the field
agent's personal activities and those of his supervisor
are included. The distinction appears to be between
restricting the communication strictly to work topics and
involving the members in each other's family life as well.

The third factor isolated items relating to the
sequencing dimension. The focus in these questions was on
the field agent sharing the responsibility for seeing that
tasks get completed. Thus they stress questioning, remind-

ing, clarifying and debating various issues.

111

Finally, the fourth factor was made up of items
that conceptually related to the initiation dimension.
These questions again stress the sharing of the responsi-
bility for undertaking particular tasks.

Overall, there is a clear stress on the notion of
interdependence between the dyad members. In each case the
questions focus on a different aspect of sharing--new ideas,
family and personal activities, the responsibility for
seeing tasks through to completion, and the responsibility
for seeing that tasks are in fact undertaken.

While no claim is made for the general application
of these procedural dimensions in other systems; their
independence and their relative power in this study sug-
gests that they should be considered as possibly powerful
dimensions in future studies of organizational communica-

tion.

Communication Between Supervisors
and Field Agents

 

In contrasting the field agents and their super-
visors certain major differences stand out. The field
agents as a group tended to view the communication with
their supervisors as more frequent, more directed by them,
and more oriented toward getting the job done than did the
supervisors. They were also more satisfied with the

present levels of communication.

112

These perceptions suggest that the field agents
view their supervisory communication in a narrower and
less central way than do their supervisors. Given their
different roles within the organization this difference
would seem understandable. The prime role of the field
agents is to communicate with their clients and various
county officials, and supervisory communication essentially
fills a support or facilitating function for them. On the
other hand, for the supervisors their communication with
the field agents represents a much more central aspect of
their work. These differences suggest that the field
agents generally do not feel cut off from their organiza-
tion; and that any effort on the part of supervisors to
increase communication may be seen by many field agents as
undesirable overload. The one major exception is communi-
cation about new ideas. Two-thirds or more of both
supervisors and field staff would like to have more com-
munication about new ideas. Given that the prime purpose
of the organization is to introduce new ideas to the people
of Michigan, the generation, discussion and sharing of new
ideas would seem to be one area where increased supervisory
communication and supportive rules would increase satis-
faction and enhance performance.

Approximately two-thirds of both the field agents
and their supervisors reported discussing new ideas

together less than once a month. This would seem to be

113

far below the optimum for an organization where change is
meant to be the prime focus. Such a lack of searching for
and discussing new ideas suggests an unfortunate level of
complacency or misdirected emphasis in professional inter-
action. Steps to increase the searching and sharing of
new ideas would seem very desirable.

One way to achieve such an increase in the sharing
of new ideas would be to encourage the supervisors to act
as facilitators--taking ideas from center to center.

When the four dimensions of procedural communication
are taken together the supervisors as a group are seen to
advocate more interdependent communication than do the
field agents. This ties in with their view of the communi-
cation as being less imposed and less production oriented
than it is seen by the field agents. Supervisors also
attributed more interdependence to the field agents' per-
ceptions of innovative communication. Taken together,
these tendencies to advocate and see more highly inter-
dependent communication procedures appear to represent an
important difference between supervisors and their field
agents. Two possible interpretations can be offered to
explain this trend. First, it may be a reflection of their
greater involvement due to the centrality of this inter-
action to their roles. On the other hand, it may represent
a basic difference in philosophy of supervision. If in

fact, supervisors do advocate more interdependent

114

communication because of a more 'human resource' or 'human
relations' philosophy, then the seeds may be developing
for a coorientation gap based on different rules of what
is appropriate procedural communication.

Although there are these signs of minor differences
in orientation between the supervisors as a group and the

field staff, the most outstanding feature of the whole

 

study is undoubtably the high degree of both egreement

and accuracy that was found between the dyad members.

 

Although the indices for individuals were normally
distributed, with means approximately at the mid-points
of the distributions (see Table 21), the agreement and
accuracy scores (for dyads) were all highly skewed with
almost all dyads demonstrating very high similarity of
orientation.

Also, despite the limited amount of communication
between the dyad members, and some differences in experience
and educational background, this study shows that almost
all of the dyads in the sample were able to achieve a high
level of agreement. Perhaps the most important reason for
such a result is the great similarity in basic type and
level of education and prior experience. The results seem
to indicate that common types of experience and background
were much more important in this case than the pressures

toward disintegration that are proposed by Kaufman (1960).

A major implication that follows is that geographic

115

separation need not lead to serious problems of dissimilar
orientations provided that the organization is relatively
stable and the members share highly similar background
experiences.

By and large, both field agents and their super-
visors seem to have come from the same mold--to have
attended similar mid-West colleges, to have similar views
on the philoSOphy and place of extension, and to have
similar interests and views. Deviant views in general
were severely criticized and those who 'rocked the boat'
reported having considerable opposition. Many respondents
expressed concern about participation in the study because
of a fear of reprisal and several reported unfortunate
previous incidents where 'confidential' information was
released. All of this would seem to point to considerable
rigidity in approach and to an unfortunately high level of
'back-biting.'

Conformity does appear to be rewarded and probably
almost all who stay in the organization have had to adopt
the 'correct' way of doing things.

Again, in an organization of the size of the MCES
and one associated with the principles of a University,
some freeing of the system to meet new and unknown needs
would seem to be very desirable.

While the overall feature is one of high similarity

and understanding, the combination of agreement and accuracy

116

scores for each dyad, reveals some cause for concern.

Using Scheff's (1967) combination of high and low levels

of agreement and accuracy to yield four types of under-
standing, it was found that both supervisors and field
agents tended to predict that the other member of their
dyad would share their particular view of the communication
interaction.

This trend is similar to that reported by Berlo
et al. (1971b and c). Given high levels of agreement,
this feature means a high level of understanding--the
members of the dyads agree and know it. However, to the
extent that responses in the low agreement cells indicate
real differences in orientation, then a significant number
of dyads actually have low agreement while they think that
they have high agreement--this is the type of understanding
that Scheff calls false consensus.

This state of misunderstanding would seem to be
particularly undesirable for most systems. Whereas, with
dissensus the members disagree and are aware of their dis-
agreement, with false consensus the members think that they
understand the other's orientation but do not. False con-
sensus may thus mask differences and encourage the develop-
ment of even lower agreement.

While over half of the responses indicated high
agreement and high accuracy--the pairs agreed and accu-

rately predicted that the other member would share their

117

view-—between one-third and one-half of the responses show
disagreement or misunderstanding.

It is this group that gives cause for concern.
Certainly the organization should not claim nor assume that
everyone shares the same idea about how agents and super-
visors communicate. In at least one-third of the pairs
that were studied one or other of the members suffered
from some type of misunderstanding about the communication.

This need not be always undesirable provided that
it is known and taken into account; but if not it will
almost certainly contribute to communication breakdowns
and strained relationships. With the rotation of super-
visors from one group of field agents to another, different
expectations do occur and field agents then go through a
period of 'trying to psych out' what the new supervisor
will want and how he would like things done. Such uncer-
tainty must be dysfunctional to the organization, and
informally it seemed to contribute to considerable lack of
morale in individual cases.

As both supervisors and field agents were equally
likely to make this type of wrong judgment, some form of

explicit feedback on the procedural aspects of communica-

 

tion would seem highly desirable. This feedback would
hopefully serve as a check on whether the agreement that

the members perceive is based on actual agreement or

118

instead on a false extrapolation of each individual's par-
ticular point of view.

Another reason for suggesting such an explicit
discussion of procedural aspects of communication, is the
tendency for dyads in which the supervisor holds different
views, to those of other supervisors, to have low agreement.
This could be a result of field staff viewing the inter-

action as they expect it should occur rather than being

 

aware of its unique features; or equally, it could indi-
cate a lack of awareness by the supervisors of the actual
interaction. In either case, explicit feedback on views
of the interaction would clarify the situation and that
should lead to greater understanding.

Given this difference in ability between super-
visors and field staff in predicting unusual views, it
would seem that the supervisory staff have less difficulty
than their field staff in treating each dyad as a unique
system. This may well be another reflection of a basic
difference in centrality of this communication to their

perceived roles.

Relationship Between Coorientational
Similarity and Attraction

 

The major purpose of the study was to test two
extensions of Newcomb's (1953) coorientation model. These

extensions were: first, that increased similarity of

119

orientation would lead to greater interpersonal attraction;
and second, that attraction would be determined by the net
effect of coorientations within the dyad. Both hypotheses
suggest that increased coorientational similarity should

be associated with increased interpersonal attraction. The
first seeks to establish one to one relationships; while
the second suggests that the level of attraction will be a
reflection of the combined effects of many different
orientations within the dyad. For objects of orientation
of high instrumentality to the relationship both hypotheses
should allow a more precise estimate of attraction than
could be obtained by considering individual orientations
one by one.

At a gross level, the high levels of coorientational
similarity (agreement and accuracy) for all dimensions of
procedural communication, combined with the high levels of
field agent role satisfaction and high ratings of agent
performance by their supervisors would seem to support
both hypotheses. To some extent this becomes the most
significant support for the hypotheses, because the strong
skewing of the similarity distributions sharply curtails

32

the variance on the communication variables. With this

type of distribution it could be argued that almost the

 

32This restricted range on one variable leads to a

lower correlation between the variables, see McNemar
(1962, p. 144).

120

entire sample exhibits satisfactory similarity, and that
differences are mainly a reflection of response patterns
or other confounding factors.

With this important reservation in mind, support
is offered for both hypotheses. Given the skewed distribu-
tions on similarity of orientations, those relationships
that are still significantly different from zero would
seem to represent the smallest and strongest core of
relationships that do in fact occur. The essentially nor-
mal distributions of the individual indices, and the
evidence of consistent patterns across levels supports
this point of view, as contrasted to the notion of the
significant relationships were a simple reflection of
various error components.

With respect to hypothesis 1, weak relationships
were found between a few dimensions of similarity of
orientation and field agent role satisfaction and field
agent perceptions of their relationships with their super-
visors. Stronger relationships, both in number and size,
were found between the measures of coorientational simi-
larity and both supervisor ratings of field agent perform-
ance and supervisor perceptions of the formality of their
relationship with the particular agents.

Similarity of orientations toward the procedural

aspects of communication do in fact seem to contribute to

the supervisors' evaluations of the system. More similar

121

orientations are associated with better evaluations and
more social and informal views of the relationship. The
partial correlation coefficients give added support to

this conclusion, as they show that the associations occur
over and above any influence of such factors as similarity
of background or geographic separation, and so on. However,
despite statistical significance, the individual correla-
tions only explain a very small proportion of the variance
in the interpersonal attraction.

These results suggest that while coorientation
toward several aspects of procedural communication is
associated with dyadic satisfaction and interpersonal
perceptions, that many other factors are also contributing
to these states.

In particular, no significant correlations were
found between any of the dimensions of similarity and per-
centage salary increase. This occurred in spite of an
expectation that at least some part of salary increment
would be a reflection of the relationship between the
supervisor and the field agent. There could be at least
two reasons for this lack of association. First, several
people are involved in making the recommendation and
decision about salary increases for a particular agent, and
the particular supervisor nominated for this study may only
have a minor input. Second, other factors such as actual

performance as a change agent, promotion or correction of

122

prior inequities will almost certainly have important
influences on salary increases that are granted, and these
may have masked any communication effects.

If a systems viewpoint is taken, the lack of asso-
ciation between similarity of orientations toward communica-
tion and salary increase would seem to indicate a healthy
state. If there was a strong association, that would
indicate that 'knowing the ropes' was more important than
actual job performance!

Similarity of orientation in terms of accuracy and

agreement about the actual dyadic interaction was more

 

strongly associated with the satisfaction measures than
was similarity towards the ideal rules of supervisory
communication. This does not seem to be an artefact of
greater variance or differing means. Instead it indicates
that, at least for this organization, the actual communica-
tion is more important and more instrumental than the

ideal rules.

One finding that is difficult to interpret, is the
consistent negative association between similarity of
orientation toward maintenance aspects of communication
and the various measures of attraction and satisfaction.

It is difficult to see why increased agreement and/or
accuracy about how much to share personal and family matters
with the other member of the dyad should lead to less satis-

faction and more formal impressions of the relationship.

123

One possible explanation may be that both field and
supervisory staff feel that they pughp to talk about non-
work matters, such as each other's family life, but that
they in fact do not feel comfortable disclosing these
aspects to a person that they define in a work relationship.
If people are disclosing more of their personal or intimate
selves than they feel comfortable with, then it could be
argued that this would lead to dissonance and ultimately
to a more strained relationship.

Again, if this is the case, if non-threatening
situations could be set up to discuss communication pro-
cedures, it may well be discovered that both field and
supervisory staff may be happier to make a clear distinction
between their professional and personal lives. Both groups
may be involving themselves in each other's personal lives
simply because they think they should, rather than because
they want this sort of involvement.

Hypothesis 2 suggests that combining several dimen-
sions and several levels of similarity should lead to
increased accuracy in predicting interpersonal attraction.
Although the multiple correlation coefficients (R) were
consistently larger than the zero order or partial correla-
tion coefficients for individual items (r), these differ-
ences were generally not statistically significant. The
hypothesis was not supported in relation to the field

agent aspects of the relationship nor for salary increase,

124

but the net effect of combining all the dimensions did
allow more accuracy in predicting how the supervisors would
rate the field agent and how formal they would see the
relationship. However, only one-third or less of the vari-
ance is explained by the combination of similarity scores.
Therefore, considerable influence on field agent satis-
faction and salary increase in particular, and all the
system measures to a large degree, must be coming from
other factors in the system.

Reducing the number of variables through a least
-squares delete program identifies a more limited set of
measures that contribute the major part of the explanation.
However, there is no clear subset of instrumental dimen-
sions that are important for all of the measures of inter-
personal attraction. This suggests that the five measures
of attraction--field agent role satisfaction, field agent
perceptions of the formality of the relationship, per-
centage salary increase, field agent evaluation by super-
visors and supervisor perceptions of the formality of the
relationship--have little isomorphism.

All of these measures were included because it
seemed reasonable to expect that they would be affected by
the degree of liking or attraction between the members of
each dyad. The results suggest that each is also affected

by many other factors within the system.

125

The associations that were found, and the overall
gross association, between level of coorientational simi-
larity and attraction, suggest that coorientation of
members of supervisory dyads toward communication procedures
in their interaction will be an important contributor to
their satisfaction but NOT the sole or even the major
determinant in systems like the MCES.

The system relationships may well be considerably
more complex than those suggested in the hypotheses.
Different aspects of procedural communication seem to be
linked with different aspects of dyadic attraction, and
many other aspects of the system are almost certainly
operating at the same time to help to determine such states

as role satisfaction or salary increase.

Overall Implications

 

This study was undertaken to determine the levels
of coorientational similarity between members of an exten-
sion service; and to assess the effect of differing levels
of similarity on the integration and success of the agents
in the field.

In Chapter I we suggested that some field agents
may be penalized for not 'knowing the ropes.‘ From the
sample that we studied, we can conclude that this does not
constitute a major problem for the Michigan Cooperative

Extension Service at this time. There was what appears to

126

be a high degree of agreement between the supervisor/field
agent pairs, and they held accurate perceptions of each
other's views.

Nevertheless, support was found for the hypothesized
association between level of similarity toward communication
procedures and the relationship between the dyad members.
While this association appears to be mediated by other
aspects of the system, it does indicate one area where
issues not directly related to job performance were influ-
encing the evaluation and the satisfaction of the field
agents.

If we accept this association, then it would seem
desirable to commit some time and energy to increasing
mutual understanding of communication procedures. We have
suggested that such a commitment might take the form of
explicit discussion of expectations and individual orienta-
tions.

At this stage, we have little evidence of how
important these differences might be in other organizations.
However, if we accept the predictions of such people as
Toffler (1970), that organizations and their components
will become increasingly short lived with a changing
structure built on highly transitory human relationships,
then the issue may become critical. If relationships

change more rapidly, then differences may become the order

127

of the day and more and more people may get hurt through
misunderstanding the communication rules. Such an outcome
would seem untenable.

Communication scientists have a challenge to clarify
the association between similarity of orientations toward
communication procedures and system output measures; and

to develop ways to minimize these differences if they arise.

Future Research

 

The results presented in this study provide suf-
ficient evidence of a link between dyadic similarity in
orientation toward the procedural aspects of communication
and such system measures as role satisfaction, formality
of interpersonal relationships and interpersonal evaluations
to justify continued research based on the proposed exten-
sions of Newcomb's model.

Three extensions would appear particularly valuable.
First, to repeat the study in a system where a greater
range in levels of similarity might be expected, thereby
overcoming the problems caused by restricted variance.

Such a system may be found in commercial organizations
between line and staff, or between labor and management
negotiators, or between different ethnic or class groups.

Second, to conduct a longitudinal study. Newcomb's
model is essentially a dynamic one, and a major development

to test the notions of causality would be to conduct a

128

field experiment. This should involve measuring orienta-
tions and relationships at two points of time, with an
intermediate, controlled manipulation of similarity of
orientation through such strategies as group discussion.

Third, to repeat the study with a wider range of
system variables. As Blalock (1969) and others suggest,
the solution to a set of simultaneous equations, relating
a large set of system variables to each of a set of crite-
rion measures, offers one way to explain the influence of
a number of variables. Such an approach would seem to
offer valuable insights on the importance and influence
of coorientation toward different aspects of interaction,
and on the importance of other direct influences on the
criterion measures.

These three extensions are not mutually exclusive.
In fact, the ideal extension of this research (if resources
were unlimited) would be to incorporate each of these modi-
fications into a major field experiment.

Both the notion of coorientation and communication
procedures would seem to offer considerable promise to the

field of communication.

Summary of Results

 

This research examined the association between
interpersonal attraction between members of extension

agent/supervisor dyads and their level of coorientational

129

similarity toward the procedural aspects of their communica-
tion.

On the basis of extensions of Newcomb's coorienta-
tion model, predictions were made about the effect of two
levels of similarity--agreement and accuracy.

Using a factor analysis of sixty items relating to
the ideal communication, four independent dimensions of
procedural communication were identified. These dimensions
focus on different aspects of sharing or interdependence
in the dyad--new ideas, family and personal activities,
the responsibility for seeing tasks through to completion,
and the responsibility for seeing that the tasks are in
fact undertaken.

In describing their communication, the supervisors
saw the interaction as more interdependent than did their
field staff. The supervisors also tended to be more skilled
at recognizing divergent perceptions.

Both supervisors and field staff wanted more com-
munication about new ideas; while supervisors alone wanted
an overall increase in communication about all aspects of
their interaction.

The most outstanding feature of the study was the
high overall level of both agreement and accuracy with
respect to communication procedures. This high level of
coorientational similarity, combined with high levels of

field agent role satisfaction and high supervisor ratings

130

of field agents' performance, gives overall support to the
notion of an association between similarity of orientation
and interpersonal attraction.

Although a number of significant associations were
found between measures of similarity of orientation toward
communication procedures and measures of attraction, the
degree of similarity did not account for a major part of
the variance in the indicators of attraction.

General support is claimed for the usefulness of
Newcomb's model, but the influence of many other variables
as well as the procedural aspects of communication is
recognized.

Based upon several trends in the data, explicit
discussion between supervisors and field agents about how
they view both their current and their ideal expectations
for communication procedures is recommended as a way to
increase the understanding between the dyad members.

Finally, taking a systems viewpoint, a number of
suggestions are made for further research into the applica-
tion of Newcomb's model to increase our understanding of
communication within organizations. The focus on both the
shared interaction and on the procedural rules that direct
communication is seen as both useful and powerful for

future research.

REFERENCES

REFERENCES

Allport, G. W., The historical background of modern
sociology, Chapter in Lindzey, G. andiAronson, E.,
The Handbook of Social Psychology, 2nd. edition,
vOlume 1, Reading, Massachuseffs: Addison-Wesley,
1968.

 

 

 

Amend, E. H., Liaison Communication Roles of Professionals
in a Researcthissemination Organization, unpub-
lished dOctoral dissertation, MiCHiganiState
University, 1971.

 

 

Ascroft, J. R., Modernization and Communication: Con-
trollinngnvironmental Change, unpublished doctoral
dissertation, Michigan State University, 1969.

 

 

Axinn, G. H. and Axinn, N. W., Communication among the
Nsukka Igbo: A Folk-Village Society, Journalism
Quarterly 46 (1969): 320-24.

 

 

Barnard, C. I., The Functions of the Executive, Cambridge,
Mass.: Harvard University’Press, 1938.

 

Barrett, J. H., Individual Goals and nganizational
Objectives, Ann Arbor, Michigan, Institute for
Social Research, University of Michigan, 1970.

 

 

Berlo, D. K., Essays on Communication, Department of
Communicatibn, Michigan State University mimeo,
1970.

 

, Farace, R. V., MacDonald, D. S., Buley, J., and
Connelly, R., nganizational Communication: A
first-line managerial communicatiOn system,
Department ofTCommunication, MiChigan State Univer-
sity mimeo, 1970.

 

 

, Farace, R. V., Connelly, R. A., and Russell, H. M.,
Relationships between Supervisor-Subordinate Com-
munication Practices and Employee Turnover, Attend-
anceLyand Perfbrmance Evaiuations, Department off
COmmunication,‘MiChigan State University mimeo,
1971a.

 

 

131

132

, Farace, R. V., Connelly, R. A., and Monge, P. R.,
Supervisor-subordinate communication relations:
Shmmapy of basic data, Department of COmmunication,
Michigan State University, mimeo 1971b.

 

, Farace, R. V. and Monge, P. R., Supervisor-
subordinate communication relations in the Office
of’Civil‘DEfense: ’Summary of basic data, Department
6f7Communication, MiEHigan State University, mimeo
1971c.

 

Bible, B. L., Role Expectations of County Executive Com-
mittee‘Members’in the COOperative Agricultural and
Home Economics Extension Service of Pennsylvania,
unpublished doctoral diSsertation, Pennsylvania
State University, 1959.

, and Brown, E. J., Role consensus and satisfaction
of extension advisory committee members, Rural
Sociology 28 (1963): 81-90.

 

, Nolan, F. A., and Brown, E. J., Consensus on role
definition of the county extension executive com-
mittee member, Rural Sociology 26 (1961): 146-156.

 

Blalock, H. M., Theory Construction, Englewood Cliffs,
N. J.: Prentice-Hall, 1969.

 

Cartwright, D. and Harary, F., Structural balance: A
generalization of Heider's theory, Psychological
Review 63 (1956): 277-293.

 

 

Chaffee, S. H. and McLeod, J. M., Sensitization in Panel
Design: a Coorientational Experiment, Journalism
Quarterly 45 (1968): 661-669.

 

 

Cushman, D. and Whiting, 6., An Approach to Communication
Theory: Toward Consensus on Rules, Draft ongaper
fOr Speech Communication Association Convention,
1971.

Davis, K., Social Norms, Chapter in Biddle, B. J. and
Thomas, E.‘JI Editors, Role Theory: Concepts and
Research, New York: Johfi Wiley 8 Sons, 1966.

 

Dodd, S. and Gerbrick, T. R., Word Scales for Degrees of
Opinion, Language and Speech 31 (1960): 18-31.

 

Edwards, A. L., Experimental Design in Psychological
Research, 3rd7Edition, New York: Holt, Rinehart
Winston, 1968.

133

Farace, R. V. and Connelly, R. A., Organizational Communica-
tion Correlates of Herzberg's Theory of Work
Satisfaction, Department ofTCommunication, Michigan
State‘University mimeo, 1970.

 

 

 

, et al., The Communication System of Justin Morrill
Colle e, Department of Communication,7Michigan State
Unlversity mimeo, 1970.

 

 

French, J. R. P., Jr. and Raven, B., The Bases of Social
Power, Chapter in Cartwright, D. and Zander, A.
(editors), Group Dynamics: Research and Theory,
London: Tavistock—PublicatiOns, 1960.

 

 

Havelock, R. G., Planning_for_ipnovation through Dissemina-
tion and UtilizationcHFKnowledge, Ann Arbor,
MiChigan: *Institute for Socia Research, University
of Michigan, 1969.

 

 

Hays, W. L., Statistics, New York: Holt, Rinehart and
Winston, 1963.

 

Heider, F., The Psychology of Interpersonal Relations,
New York: John Wiley and Sons, 1958.

 

Jain, N. C., Communication Patterns and Effectiveness of
Professionals PerfOrming Linking Roles in a Research
Dissemination Orgepization, unpublished doctoral
dissertation, Michigan State University, 1970.

 

 

 

Kaufman, H., The Forest Ranger: A Study in Administrative
Behavior, Baltimore, Md.: The John Hopkins Press,

 

Kornhauser, W., Scientists in Industry, Berkeley,
California: University of California Press, 1962.

 

Laing, R. D., Phillipson, H. and Lee, A. R., Interpersonal
Perception, New York: Springer PubliShing Company,
I966.

 

 

Leavitt, H. J., Some effects of certain communication
patterns on group performance, J. Abnormal and
Social Psychology 46 (1951):. 38-50.

 

Levinger, G. and Breedlove, J., Interpersonal attraction
and agreement: A study of marriage partners, J.
Personnel 8 Social Psychology 3 (1966): 367-372.

 

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

 

134

, The Human Organization: Its Management and Value,
New Y6rk: MCGraw Hill, 1967.

 

 

MacDonald, D., Communication Roles and Communication Content
in a Bureaucratic Setting, unpublished doctoral
dissertation, MiChigan State University, 1970.

 

 

, and Farace, R. V., Approaches to the study of
communication in organizations: An example of
applied theory in a large financial’institutiOn,
Paper presented to the Speech Communication
Association Convention, 1970.

 

 

 

McNemar, Q., Psychological Statistics, 3rd edition, New
York: JOhn Wiley 8 Sons,71962.

 

Mark, R. A., Parameters of Normal Family Communication in
the D ad, unpublished dOctoral dissertation,
Mlchlgan State University, 1970.

 

Michigan Cooperative Extension Service, A Career with the
Coeperative Extension Service of Michigan State
University, pamphlet lP3R-4-69-3M.

 

 

 

Monge, P. R. and Farace, R. V., Transformations and messege
linkages for a theory of communiCation coorienta-
tion, paper submitted to the International Communi-
catlon Association convention, Atlanta, 6a., 1972.

Moser, C. A., Survey Methods in Social Investigation,
London: Heinemann, 1958.

Mouzelis, N. P., Organization and Bureaucracy: An analysis
of modern theories, Chicago, Illinois: Aldine
PfibliShing Co., 1967.

 

Newcomb, T. M., An approach to the study of communicative
acts, Psych Rev 60 (1953): 393-404.

 

, The Aeguaintance Process, New York: Holt,
RinEhart G Winston, 1961.

 

, Stabilities underlying changes in interpersonal
attraction, J Abnormal and Social Psych 66 (1963):
376-386.

, Interpersonal Balance, chapter in Abelson, R. P.
et al., Theories of Cognitive Consistency: A
Sourcebook, Chicago: Rand MENally, 1968.

 

 

135

Nix, H. L. and Bates, F. L., Occupational role stresses:
A structural approach, Rural Sociology 27 (1962):
7-17.

 

Nunnally, J. C., Psychometric Theory, New York: McGraw
Hill, 1967.

 

Preiss, J. J., The functions of relevantepower and authority
groups in the evaluation of county agent performance,
unpubli§hed doctoraI’dissertation,iMiChigan State
College, 1954.

 

 

Rogers, L. Edna, Dimensions of Dyadic Systems and Inter-
personal Communication, unpublished paper,
Department ofFCOmmunication, Michigan State Univer-
sity, 1971.

 

 

Rogers, E. M. and Bhowmik, D. K., Homophily-Heterophiiye
Relational concepts for communication research,
Paper presented to Associaiion fOr Education in
Journalism, Berkeley, California, 1969.

 

 

, and Shoemaker, F. F., Diffusion of Innovations:
A Cross-Cultural and CommuniEation Approach,
New YOfk: Free Press of Glencoe, 1971.

 

Russell, H. M., Communication Implications of Selected
Organizational Theories, Department of Communica-
tion, Michigan State University mimeo, 1971.

 

 

Scheff, T. J., Toward a sociological model of consensus,
Am Sociol Review 32 (1967): 32-46.

 

Schein, E. H., Organizational Theory, Englewood Cliffs,
New Jersey: Prentice-Hall, 1965.

 

Simon, H. A., Administrative Behavior, 2nd edition,
New York: *Macmillan 8 Co., 1957.

 

Toffler, A., Future Shock, New York: Random House, 1970.

 

Torpey, W. G., Optimum Utilization of Scientific and
Engineering Manpower, Richmond, Virginia: Whittet
and Shepperson, 1970.

 

Vroom, V. H., Some Personality Determinants of the Effects
of Participation, Englewood Cliffs, New Jersey:
Prentice-Hall Inc., 1960.

 

136

Wackman, D. B., A for a new measure of coorienta-
tional accuracy or empathy, paper presentedtot The
Association ibr Educatlon in Journalism, Berkley,
Calif., 1969.

 

, and Beatty, D. J. F., A Comparison of Balance and

Consensus Theories for ExplainingChanges in ABX
Systems, Paper presented to the International
ommunlcation Association, Phoenix, Arizona, 1971.

 

Webb, B. J., Campbell, D. T., Schwartz, R. D., and
Sechrest, Unobtrusive Measures: Nonreactive

research in the social sciences, Chicago: Rand
MENally 8 Co., 1966.

 

Wilkening, E. A., Consensus in role definition of county
extension agents between the agents and local
sponsoring committee members, Rural Sociology 23
(1958): 184-197.

APPENDIX

FIELD AGENT QUESTIONNAIRE

Department of Communication

Michigan State University

Study of Communication between Field
and Administrative Staff in the Michigan
Cooperative Extension Service

Respondent:

 

This questionnaire deals with your communication

with

 

To help us ensure your anonimity, we have already
coded the questionnaire to identify the pairs. No one
within the extension service will have access to the list
of code numbers.

Please tear off and destroy this sheet when you
have finished.

137

Jflflfifli
a

 

138

Department of Communication

Michigan State University

Form A. Field Staff

 

STUDY OF COMMUNICATION BETWEEN FIELD AND ADMINISTRATIVE STAFF
IN THE MICHIGAN COOPERATIVE EXTENSION SERVICE

1971

STRICTLY CONFIDENTIAL

 

 

 

 

------ office use ------

Supervisor: Schedule number:

 

 

Date of interview:
Field Agent:

 

Checked:

 

Coded:

 

 

 

 

139

INSTRUCTIONS

 

This questionnaire deals with your communication
with one of your immediate supervisors in the Michigan
Cooperative Extension Service. We are interested in your
present communication with this person, the sort of communi-
cation that you would like and what you believe is appro- i
priate communication. Mi

While code numbers are required so that we can

match your responses with those of your supervisor, we can

 

assure you that no one within the extension service will

 

have access to any of the completed questionnaires.

 

How you answer the questions will neither help nor
hinder you in your future in the Service. However, we hope
that the results of the survey will provide valuable leads
in making the Service as rewarding as possible to work in.

Please answer each question by filling in the blank
or checking the alternative that is closest to the way you
feel. There are no 'right' answers--we are interested in
how YOU feel and in how YOU see the situations.

If you have any questions, please feel free to
discuss them with the member of the Department of Communica-

tion who is present.

PLEASE ANSWER ALL QUESTIONS

140

I. GENERAL INFORMATION

 

First we would like some general information about
your background and experience in the Michigan Cooperative
Extension Service.

1. How many years have you worked under this person's
authority?
less than 1 year

1 to 2 years

3 to 4 years
5 to 6 years

 

more than 6 years

 

2. How many years have you been working in your present
position?

 

less than 1 year
1 or 2 years

3 or 4 years
5 or 6 years

 

more than 6 years

 

3. ,How many years have you been in extension work?
5 years or less

 

6 to 10 years

11 to 15 years
16 to 20 years

 

more than 20 years

 

4. How old are you?
30 years old or less
31 to 40 years old
41 to 50 years old
over 50 years old

 

 

 

141

5. If you have received graduate training, please specify
below.

none received

 

less than 1 year

 

Masters or Doctoral degree

 

special training (please
specify)

 

 

 

 

 

142

II. COMMUNICATION BETWEEN YOU AND YOUR IMMEDIATE SUPERVISOR

 

Now lets turn to the main questions. We need to
learn about the communication between you and your super-
visor. We want you to describe several aspects of this
communication, and to tell us ways in which you are satis-
fied or dissatisfied.

In describing this communication, we want you to
take into account face-to-face conversations, formal or
informal meetings, telephone conversations, memos, letters,
and any informal interaction out of office hours.

1. When two people begin to communicate, one of them has
to ask for it to happen. On the average, when you and
your supervisor communicate, who usually initiates it?

I usually initiate it

 

He usually initiates it

 

We initiate it about equally

 

2. On the average, would most of your communication with
your supervisor be written (through letters or memo-
randa) or oral (either phone or face-to-face conversa-
tion)?

All written

More written than oral

About 1/2 written and 1/2 oral
More oral than written

All oral

 

 

 

 

 

3. On the average, how often do you and your immediate
supervisor communicate?

Several times a day

 

Once a day

 

Once a week or more

 

Once a month or more

 

Less often

 

4. How satisfied are you with the overall frequency with
which you and your supervisor communicate?

I'd like us to have less communication
I'd like us to have more communication
I'm satisfied the way that it is.

143

5. Of the communication that yeu reeeive from your super-
visor, how much is sought by you as contrasted to
merely being directed to you?

All directed to me
More directed than sought

About 1/2 directed and 1/2
sought

 

More sought than directed
All sought
6. Of the communication that you initiate with your super-

visor, how much is sought by him as contrasted to being
merer directed to him?

All sought by supervisor

"infifl’
.I

 

More sought than directed I

About 1/2 sought and 1/2
directed

 

More directed than sought
All directed on my choice

We would like you to distinguish between three
major functions that communication could serve in the
interaction between you and your supervisor:

a. communication that relates to the implementation
of existingprograms of work -- coordination,
directions, reporting on
progress, administrative
procedures, general implementa-
tion.

 

 

b. communication that relates to personal matters --
relations with other staff]
personal advancement, family
activities, career opportunities,
general health and interests.

 

c. communication that relates to the seeking and
developing of new ideas -- new things to do,
new ways of doing things, new
findings, new programs, new
areas of application.

 

144

Given these three major functions, we would now

like you to evaluate the communication between you and
your supervisor.

7.

10.

On the average, what percentage of the communication
initiated byeyour supervisor with you, would relate to

 

each Oi fhese functions?

 

Implementation of existing program % THESE
Personal matters % SHOULD
________ ———— TOTAL
Seeking and developing new ideas % 100%

 

On the average, what percentage of the communication
that you initiate with your supervisor, would relate
to each Of these functions?

 

 

Implementation of existing program % THESE
SHOULD

Personal matters % TOTAL

Seeking and deveIOping new ideas % 100%

 

When you and your immediate supervisor communicate about
the EXISTING PROGRAM who usually brings it up?

He does
I do

It's split about evenly between
us

We don't communicate about the
existing program

 

If you don't communicate about the existing program,
please turn to question 12, otherwise please continue
with question 10.

 

If you do communicate about the existing program, on the
average how often do you discuss these topics?

Several times a day

 

Once a day

 

Once or twice a week

 

Once or twice a month

 

Less often

11.

12.

13.

14.

145

How satisfied are you about the amount of communication
that you have about the existing program?

I'd like to communicate more
about the existing program

I'd like us to have less com-
munication about the existing
program

I'm satisfied the way it is now

 

When you and your immediate supervisor communicate
about PERSONAL matters, who usually brings them up?

He does

I do

It's split about evenly between
us

We don't communicate about
personal matters

 

If you don't communicate about personal matters, please
turn to question 15, otherwise please continue with
question 13.

 

If you do communicate about ersonal matters, on the
average how often do you discuss these topics?

Several times a day

 

Once a day

 

Once or twice a week

 

Once or twice a month

 

Less often

 

How satisfied are you with the amount of communication
that you and your supervisor have about personal matters?

I'd like to communicate more
about personal matters.

I'd like us to have less com-
munication about personal
matters.

I'm satisfied the way it is now

 

15.

16.

17.

146

When you and your immediate supervisor communicate
about NEW IDEAS who usually brings them up?

He does

I do

It's split about evenly between
us

 

We don't communicate about new
1deas

If you don't communicate about new ideas, please turn
to the nexf section, otherwise please continue with
question 16.

If you do communicate about new ideas, on the average
how often do you discuss these topics?

 

Several times a day

 

Once a day

Once or twice a week

 

Once or twice a month

 

Less often

 

How satisfied are you with the amount of communication
that you have with your immediate supervisor about
new ideas?

 

I'd like to communicate more
about new ideas

I'd like us to have less com-
munication about new ideas

I'm satisfied the way it is now

147

III. EXPECTATIONS FOR COMMUNICATION WITH SUPERVISOR

 

We would now like you to indicate how obligated an
extension egent should be to do the following things in his
communicaiion with hissupervisor.

 

 

For each statement, please write the number that
indicates your opinion about the statement.

1 = Definitely should 4 = Probably should not
2 = Probably ShOUId 5 = Definitely should not
3 = May or may not

 

1. Avoid interrupting his supervisor if a member of
the public wants assistance.

2. Share his personal goals with supervisor when in
a meeting.

3. Plan ahead to save his supervisor time.
4. Share tentative new plans with supervisor.

5. Apart from routine reports, report on current
projects only when there is a change in them.

6. Develop new programs through frequent inter-
action with supervisor.

7. Concern himself with the activities of his
supervisor's family.

8. Send in reports of current projects even when
not requested.

9. Offer suggestions about his supervisor's
personal affairs.

10. Phone in summaries of successful field days
or projects.

11. Enquire about his supervisor's family when
together.

12. Bring up some new ideas with supervisor at
least once a year.

13. Leave it to his supervisor to suggest new
programs.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.
26.

27.

28.

148

 

Definitely should
Probably should

May or may not
Probably should not
Definitely should not

MhMND-J
II II II II II

 

 

 

Talk with supervisor about how he gets along
with other district staff.

Encourage supervisor to personally inspect
local projects.

Avoid contacting supervisor unless important
issues arise.

Avoid contacting his supervisor about the
current program more than once a week.

Take responsibility for preparing budgets and
reports before they are due.

Discuss family activities or problems only when
face-to-face.

Seek to add some note of personal feeling into
discussion.

Forget unacceptable demands rather than debate
them with supervisor.

Interrupt supervisor if unclear about a point
that he has made.

Avoid bringing personal affairs into conversations
with supervisor.

Discuss projects with supervisor only if they are
going well.

Discuss new programs with supervisor orally first.

Consider modifying new programs to fit into local
conditions.

Make a point of introducing personal affairs into
discussion.

Contact his supervisor only if a project is
going badly wrong.

29.

30.

31.

32.

33.

34.
35.

36.

37.

38.

39.

40.

41.

42.

43.

149

 

Definitely should
Probably should

*May or may not
Probably should not
Definitely should not

OMNH

 

 

 

Feel free to contact supervisor out of office
hours.

Discuss personal affairs only after his super-
visor has introduced this subject.

Seek involvement in setting priorities for new
programs.

Send in reports to supervisor about current
projects as things happen.

Contact his supervisor about new ideas whenever
he comes across them.

Confirm phone conversations with written summaries.

Keep supervisor regularly informed on current
projects.

Wait for supervisor to request any reports that
he wants.

Leave it to his supervisor to suggest possible
courses for personal advancement.

Talk over new ideas with supervisor before sub-
mitting a proposal.

Refer to his family's activities in conversations
with his supervisor.

Take the initiative to seek out new ideas to
raise with his supervisor.

Keep reminding his supervisor of unresolved
issues.

Seek to implement all new ideas suggested by
supervisor.

Avoid bringing up new issues at every meeting
with superv1sor.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.
56.

57.

58.

150

 

Definitely shOuld
Probably should

May or may not
Probably should not
Definitely should not

m'hCNNH
II II II II II

 

 

 

Avoid sending in routine reports and forms until
they are requested.

Be prepared to offer criticism of his supervisor's

ideas for the agent's advancement.

Write reports on the existing program rather than
phoning them in.

Include personal news in memoranda to his super-
visor.

Clarify ambiguous memoranda from his supervisor
with him.

Discuss progress of current projects face-to-face
rather than sending in a report.

Wait for his supervisor to terminate conversa-
tions.

Feel free to discuss personal matters over the
phone.

Submit new propositions or ideas in writing to
his supervisor.

If he is not satisfied with the work that he has
been assigned to, then 'grit his teeth and bear
it' rather than complain.

Have secretary phone in reports to supervisor if
busy.

Invite his supervisor to come to local meetings.

Try to present all new projects as a single
package.

Raise new ideas only at formal meetings with
supervisor.

Raise only one issue in any one memorandum to his
supervisor.

59.

60.

151

 

Definitely should,
Probably should

May or may not
Probably should not
Definitely should not

U'I-hQNNI-J
II II II II II

 

 

 

Make suggestions about how others could improve
their work.

Hesitate to question an assignment that is
personally inconvenient.

IV.

152

ROLE BEHAVIOR IN COMMUNICATION WITH SUPERVISOR

 

For the same questions we would now like you to rate

the extent to which you do the following things.

 

For each statement, please write the number that

indicates how you communicate with this supervisor.

10.

11.

12.

13.
14.

15.

= Never 1 = Seldom 2 Usually 3 = Always

 

Avoid interrupting my supervisor if a member of
the public wants assistance.

Share my personal goals with supervisor when in a
meeting.

Plan ahead to save my supervisor time.
Share tentative new plans with supervisor.

Apart from routine reports, report on current
projects only when there is a change in them.

Develop new programs through frequent interaction
with supervisor.

Concern myself with the activities of my super-
visor's family.

Send in reports of current projects even when not
requested.

Offer suggestions about my supervisor's personal
affairs.

Phone in summaries of successful field days or
projects.

Enquire about my supervisor's family when together.

Bring up some new ideas with supervisor at least
once a year.

Leave it to my supervisor to suggest new programs.

Talk with supervisor about how I get on with other
district staff.

Encourage supervisor to personally inspect local
projects.

153

 

Lo

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.
26.

27.

28.

29.

30.

31.

32.

= Never 1 = Seldom 2 Usually 3 = Always

 

Avoid contacting supervisor unless important
issues arise.

Avoid contacting my supervisor about the current
program more than once a week.

Take responsibility for preparing budgets and
reports before they are due.

Discuss family activities or problems only when
face-to-face.

Seek to add some note of personal feeling into
discussions.

Forget unacceptable demands rather than debate
them with supervisor.

Interrupt supervisor if unclear about a point
that he has made.

Avoid bringing personal affairs into conversations
with supervisor.

Discuss projects with supervisor only if they are
going well.

Discuss new programs with supervisor orally first.

Consider modifying new programs to fit into local
conditions.

Make a point of introducing personal affairs into
discussions

Contact my supervisor only if a project is going
badly wrong.

Feel free to contact supervisor out of office
hours.

Discuss personal affairs only after my supervisor
has introduced this subject.

Seek involvement in setting priorities for new
programs.

Send in reports to supervisor about current pro-
jects as things happen.

154

 

 

O

= Never 1 = Seldom 2 = Usually 3 = Always

 

 

33.

34.
35.

36.

37.

38.

39.

40.

41.
42.

43.

44.

45.

46.

47.

48.

49.

Contact my supervisor about new ideas whenever I

come

across them.

Confirm phone conversations with written summaries.

Keep supervisor regularly informed on current
projects.

Wait for supervisor to request any reports that
he wants.

Leave it to my supervisor to suggest possible

courses for personal advancement.

Talk over new ideas with supervisor before sub-
mitting a proposal.

Refer to my family's activities

with my supervisor.

Take the initiative to seek out
with my supervisor.

Keep reminding my supervisor of
Seek to implement all new ideas
supervisor.

Avoid bringing up new issues at
supervisor.

in conversations

new ideas to raise

unresolved issues.

suggested by

every meeting with

Avoid sending in routine reports and forms until

they

are requested.

Be prepared to offer criticism of my supervisor's
ideas for my advancement.

Write reports on the existing program rather than
phoning them in.

Include personal news in memoranda to my super-
visor.

Clarify ambiguous memoranda from my supervisor

with

him.

Discuss progress of current projects face-to-face

rather than sending in a report.

155

 

 

0

= Never 1 = Seldom 2 = Usually 3 = Always

 

50.

51.

52.

53.

54.

55.
56.

57.

58.

59.

60.

Wait for my supervisor to terminate conversations.

Feel free to discuss personal matters over the
phone.

Submit new propositions or ideas in writing to my
supervisor.

If I am not satisfied with the work that I have

been assigned to, then 'grit my teeth and bear it'

rather than complain.

Have secretary phone in reports to supervisor if
busy.

Invite my supervisor to come to local meetings.

Try to present all new projects as a single
package.

Raise new ideas only at formal meetings with
supervisor.

Raise only one issue in any one memorandum to my
supervisor.

Make suggestions about how others could improve
their work.

Hesitate to question whether to accept an
assignment that is personally inconvenient.

 

156

V. PREDICTING SUPERVISOR'S RATINGS

 

You have described and evaluated your communication
relationship with your immediate supervisor. We will also
ask him the same questions we've asked you.

We would now like you to predict how he will rate
your behavior.

 

Please write the number that you think your super-
visor will use in ratingeyourbehavior as it relates to
each of the fOIIOWing statements.

 

0 = You predict that he will rate your behavior as never

1 = You predict that he will rate your behavior as seldom
2 = You predict that he will rate your behavior as usually

3 = You predict that he will rate your behavior as always

 

1. Avoid interrupting him if a member of the public
wants assistance.

2. Share your personal goals with him when in a
meeting.

3. Plan ahead to save him time.
4. Share tentative new plans with him.

5. Apart from routine reports, report on current
projects only when there is a change in them.

6. Develop new programs through frequent interaction
with him. . __

7. Concern yourself with the activities of his family.

8. Send in reports of current projects even when not
requested.

9. Offer suggestions about his personal affairs.

10. Phone in summaries of successful field days or
projects.

11. Enquire about his family when together.

12. Bring up some new ideas with him at least once a
year.

157

 

 

 

29.

0 = You predict that he will rate your behaviOr as never

1 = You predict that he will rate your behavior as seIdom

2 = You predict that he will rate your behavior as usuaIIy

_3 = You_predict that he will rate your behavior as always

13. Leave it to him to suggest new programs.

14. Talk with him about how you get on with other
district staff.

15. Encourage him to personally inspect local projects.

16. Avoid contacting him unless important issues arise.

17. Avoid contacting him about the current program
more than once a week.

18. Take responsibility for preparing budgets and
reports before they are due.

19. Discuss family activities or problems only when
face-to-face.

20. Seek to add some note of personal feeling into
discussions.

21. Forget unacceptable demands rather than debate
them with him.

22. Interrupt him if unclear about a point that he
has made. ____

23. Avoid bringing personal affairs into conversations
with him.

24. Discuss projects with him only if they are going
well.

25. Discuss new programs with him orally first.

26. Consider modifying new programs to fit into local
conditions.

27. Make a point of introducing personal affairs into
discussions.

28. Contact him only if a project is going badly wrong.

Feel free to contact him out of office hours.

158

 

 

 

0'= You predict that he will rate your behavior as never

1 = You predict that he will rate your behavior as seIdom

2 = You predict that he will rate your behavior as usually

'3 = You predict that he will rateeyour behavior as a ways

30. Discuss personal affairs only after he has intro-
duced this subject.

31. Seek involvement in setting priorities for new
programs.

32. Send in reports to him about current projects
as things happen.

33. Contact him about new ideas whenever you come
across them.

34. Confirm phone conversations with written summaries.

35. Keep him regularly informed on current projects.

36. Wait for him to request any reports that he wants.

37. Leave it to him to suggest possible courses for
personal advancement.

38. Talk over new ideas with him before submitting a
proposal.

39. Refer to your family's activities in conversations
with him.

40. Take the initiative to seek out new ideas to raise
with him.

41. Keep reminding him of unresolved issues.

42. Seek to implement all new ideas suggested by him.

43. Avoid bringing up new issues at every meeting with
him.

44. Avoid sending in routine reports and forms until
they are requested.

45. Being prepared to offer criticism of his ideas
about your advancement.

46. Write reports on the existing program rather than

phoning them in.

159

 

 

 

 

0'= You predict that he will rate your behavior as never
1 = You predict that he will rate your behavior as seIdom
2 = You predict that he will rate your behavior as usuaIIy
3 = You predict that he will rate your behavior as alweys
47. Include personal news in memoranda to him.
48. Clarify ambiguous memoranda from him with him.
49. Discuss progress of current projects face-to-face
rather than sending in a report.
50. Wait for him to terminate conversations.
51. Feel free to discuss personal matters over the
phone. ____
52. Submit new propositions or ideas in writing to him.
53. If you are not satisfied with the work that you
have been assigned to, then 'grit your teeth and
bear it' rather than complain.
54. Have secretary phone in reports to him if busy.
55. Invite him to come to local meetings.
56. Try to present all new projects as a single
package.
57. Raise new ideas only at formal mqetings with him.
58. Raise only one issue in any one memorandum to him.
59. Make suggestions about how others could improve
their work.
60. Hesitate to question whether to accept an assign-

ment that is personally inconvenient.

VI.

160

JOB SATISFACTION

 

Finally, we would like you to indicate how you feel

about various aspects of your present job and your relation-
ship with your supervisor.

Please check the alternative that best indicates

your degree of satisfaction.

1.

How satisfied are you that you have sufficient communi-
cation with your supervisor to do your job well?

Very satisfied
Satisfied
Indifferent
Dissatisfied

 

 

 

Very dissatisfied

 

How satisfied are you with your present job when you
compare it to similar county extension positions in
the state?

Very satisfied
Satisfied
Indifferent
Dissatisfied

 

 

Very dissatisfied

 

How satisfied are you with the progress that you are
making towards the goals that you have set for your-
self in your present position?

Very satisfied
Satisfied
Indifferent
Dissatisfied

 

 

 

 

Very dissatisfied

 

161

How would you describe your overall relationship with
your supervisor?

Completely business-like

More business-like than social

 

Intermediate between business-
like and social

More social than business-like
Completely social
How satisfied are you with your present job when you
consider the expectations you had when you took the job?
Very satisfied
Satisfied
Indifferent
Dissatisfied

 

 

 

Very dissatisfied

 

How satisfied are you with your job relationships with
your supervisor in so far as he is helping to make your
work a rewarding and successful experience?

Very satisfied
Satisfied
Indifferent
Dissatisfied

 

 

Very dissatisfied

 

How would you describe your overall interaction with
your supervisor?

Completely formal
More formal than personal

Intermediate between formal
and personal

More personal than formal
Completely personal and informal

HICHIGQN STATE UNIV. LIBRRRIES
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
31293102672957