Citation
An exploration of the perceptions of power and symptoms of job related stress among department chairpersons in four-year institutions of higher education in Colorado

Material Information

Title:
An exploration of the perceptions of power and symptoms of job related stress among department chairpersons in four-year institutions of higher education in Colorado
Creator:
Spaulding, Warren A
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
viii, 163 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Job stress -- Colorado ( lcsh )
College department heads -- Colorado ( lcsh )
College department heads ( fast )
Job stress ( fast )
Colorado ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 154-163).
Thesis:
Submitted in partial fulfillment of the requirements for the degree, Doctor of Philosophy, Public Administration
General Note:
School of Public Affairs
Statement of Responsibility:
by Warren A. Spaulding.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
32077769 ( OCLC )
ocm32077769
Classification:
LD1190.P86 1994d .S63 ( lcc )

Downloads

This item has the following downloads:


Full Text
AN EXPLORATION OF THE PERCEPTIONS OF POWER
AND SYMPTOMS OF JOB-RELATED STRESS
AMONG DEPARTMENT CHAIRPERSONS IN FOUR-YEAR INSTITUTIONS
OF HIGHER EDUCATION IN COLORADO
by
Warren A. Spaulding
B.S.E, United States Military Academy, 1952
M.S.E., University of Michigan, 1956
M.S., Public Administration, University of Colorado, 1976
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Administration
1994


This thesis for the Doctor of Philosophy
degree by
Warren A. Spaulding
has been approved for the
Graduate School of
Public Affairs
by


//i
Robertf W/ Gage
2+LfH



Spaulding, Warren A. (Fh.D., Public Administration)
An Exploration of the Perceptions of Power and Symptoms of
Job-Related Stress Among Department Chairpersons in
Four-Year Institutions of Higher Education in Colorado.
Thesis directed by Professor Robert W. Gage
ABSTRACT
This study analyzed the relationship between perceptions of
power, job-related stress and organizational/personal
characteristics of academic department chairpersons in four-year
institutions of higher education in Colorado. The primary
hypotheses were that a statistically significant percentage of
these individuals would exhibit job-related Stress and that
job-related stress would be more prevalent among those
chairpersons who perceived they had inadequate power to carry out
the responsibilities of their positions and/or were identified
with certain organizational or personal characteristics.
The absence of any theoretical basis for the existence of a
power-stress dependency relationship for individuals in this
particular managerial role gave rise to the study. The data were
gathered by losing three self-administered questionnaires. Two
were previously validated questionnaires on power and the results
of job-related stress (burnout), and one focused on
organizational and personal characteristics of the position and
iii


the department chairperson. The total target population was all
department chairpersons in four-year institutions of higher
education in Colorado (n = 441); the usable number of responses
was 166 (35.3%).
Analysis of the data, using chi-square tests and analyses
of variance, resulted in the following major conclusions:
Within any group of department chairpersons in
institutions of higher education, a significant number will be
experiencing burnout;
Expert and referent power form the primary perceived
power bases for the influence exerted by department chairpersons
over their faculty, peers and deans;
Those chairpersons either uncertain as to their power
bases, or perceiving an absence of power relative to their
faculty, are more likely to experience burnout than those who do
perceive such power.
This abstract accurately represents the content of the
candidate's thesis. I recommend its publication.
Signed ____________________________
Robert W. Gage
iv


ACKNOWLEDGMENTS
My most sincere appreciation is extended to Janet
Spaulding my wife, for the unflagging encouragement and
support she provided me through the 6 years it took me to
obtain this degree (as well as during the 36 years preceding
those); to Robert W. Gage my committee chairperson, for
taking on that arduous task and providing me his wise counsel
and assistance throughout; to T.inda deLeon, R. Wayne Boss, and
John C. Buechner my QJ Denver committee members, who each
took the time to read and criticize constructively both my
research and my writing; to R.E.D. Woolsey my committee
member from CSM and mentor, without whose encouragement, and
prodding, I would never have finished this thesis; to Ira
Russianoff who helped me organize and manipulate my data so
that it made sense; and to William Astle who led me through
the confusing trails of statistical analysis to reasonable
conclusions.
To all these individuals and to the many department
chairpersons who responded to my survey, I give my sincerest
thanks for their help along the way in my long journey toward
the accomplishment of a very long time desire.
v


CONTENTS
CHAPTER
1. INTRODUCTION........................................1
Overview of the Study...............................1
Definitions.........................................5
Objectives of the Research..........................6
Hypotheses..........................................8
Limitations of the Study......................... 11
2. LITERA3URE REVIEW..................................16
Introduction.......................................16
Power.............................................16
Definitions.....................................16
General......................................18
Types and Sources of Power...................20
Research on Power in General.................23
Research on Power in Higher Education ... .26
Job-related Stress ............................... 30
Definition.................................... 30
Relationship between Stress and Burnout .... 31
Research on Burnout........................... 33
Department Chairpersons............................40
General.........................................40
Workplace Characteristics.......................43
Research........................................46
Burnout.........................................47
vi


Power...........................................50
Summary.........................................51
3. METHODOLOGY........................................52
General............................................52
Data Collection....................................54
Data Analysis..................................... 57
Survey Instruments.................................60
Statistical Inquiry................................61
Summary............................................64
4. FINDINGS...........................................66
Univariate Analyses................................67
Burnout Component Scores........................67
Power Perceptions...............................72
Power Perceptions by Power Base. ...............74
Department Chairperson Characteristics..........79
Bivariate Analyses.................................90
Power Perceptions and Stress....................90
Faculty and Power............................92
Peers and Power..............................93
Deans and Power..............................94
Organizational/Personal Characteristics .... 95
Explanatory Variables...........................96
Hypotheses Testing................................101
Hypotheses on Stress...........................101
Hypotheses on Stress versus Power Perceptions. 103
vii


Hypotheses on Power Perceptions ..............104
Hypotheses on Organizational/Personal
Characteristics and Stress..................106
Summary ..........................................106
5. CONCLUSION........................................109
Introduction......................................109
Conclusions.......................................110
General........................................110
Typical Chairpersons...........................Ill
Stress Levels .................................114
Power Perceptions..............................115
Stress Relationships...........................117
Strengths and Weaknesses..........................119
Contribution .....................................122
Recommendations...................................123
Actions by Vice-Presidents and Deans...........124
Further Research...............................127
APPENDIX
A. Department Chairperson Questionnaire..............129
B. Chi-Square Test Results: Power Perceptions and
Burnout Components.............................141
C. Chi Square test results: Organizational/Personal
Characteristics and Burnout Components .... 146
D. Analyses of Variance on Dependency Pairs..........149
BIBLIOGRAPHY..............................................154
viii


CHAPTER 1
INTRODUCTION
Overview of the Study
The department chairman or head is the
foreman in higher educationthe person who
sees that the work gets done. It is a
difficult and ambiguous role, and so
ill-defined that at many colleges no
description of his duties appears on paper.
And he is the man or woman caught in the middle
Brann, 1972
When Brann likens department chairpersons to foremen, he
is only partly accurate. Certainly these academic managers do
have many similarities to foremen they are, typically,
responsible for work (teaching) assignments; they evaluate
their departmental faculty; they act as links between their
faculty and the institution's administration, representing at
different times, one group to the other; and they frequently
assume their positions after serving as one of that
department's regular teaching faculty. But they are
significantly different in an important way from foremen and
the great majority of non-academic managers that difference
1


is in their relationship to those they supervise. Even though
in name they are the "heads" of their departments, their faculty
still consider them peers. The relationship has been illustrated
by Tucker in his book, Chairing the Academic Department, wherein
he characterizes the academic chairperson as both a manager and a
faculty colleague, an advisor and advisee, a soldier and a
captain, and a drudge and a boss. The chairperson is a leader,
yet is seldom given "the scepter of undisputed power." The
chairperson is also first among equals, yet finds that any strong
coalition of those equals can severely restrict his ability to
lead (1981, 4).
The collegial aspects of the position are extremely
important to the exercise of power. Typically chairpersons exert
leadership through the power of their ideas, not through the
authority that goes with the position. They are colleagues as
well as and perhaps more so than supervisors.
And this is not an easy life to live. As Robert Merton has
pointed out, "(A) conception basic to sociology holds that
individuals have multiple social roles and tend to organize their
behavior in terms of the structurally defined expectations
assigned to each role. Every sociological textbook abounds with
illustrations of incompatible demands made of the multiselved
person" (Merton 1957, 116). He further states that frequent
conflict between roles is dysfunctional for the society as well
as the individual" (116).
2


The particular conflict in the chairperson role makes the
latter extremely interesting from a theoretical standpoint.
While theory abounds on leadership and management, there exists
little, if any, on the exercise of influence when a formal leader
is strongly recognized as and is particularly oriented toward
being a peer of those led. Likert (1961), in identifying the
"linking pin" function of managers, does not address it.
Neither, for example, do Katz and Kahn (1966), Sayles (1979) or
Bums (1978).
The paradoxical role of the department chairperson in
particular has received only limited study and analysis. In
fact, as Norton points out (1978, 1), not only is there "a dearth
of literature in the field relating to the work of the academic
department (chairperson), but there also appears to be no rising
interest in the position despite certain evidence of erosion in
this administrative role" (Norton 1978, 1). Frank Dilley, Vice
Provost for Instruction, University of Delaware, has also noted
that "there is so very little good literature on the chairmanship
available" (Dilley 1972, 29).
What literature there is primarily attempts to apprise
the incumbent or aspiring department chairperson as to the
responsibilities, roles, procedures, powers and problems
particular to this position. Examples include Bennett's Managing
the Academic Department (1983) and Tucker's Chairing the Academic
Department (1981).
3


In the majority of cases the procedures, powers, and
problems discussed in these writings are identical to those
encountered by any manager. Norton's study (1978) of department
chairpersons in colleges of education, however, did add to what
is known about these positions by analyzing the time individuals
typically allot to specific jobs and the degree of reward and
satisfaction they obtain from various facets of those jobs.
None of the literature, however, has given major
consideration to the department chairpersons' perceptions of
their personal power and the stresses that may result from the
paradoxical nature of their jobs, particularly in relationship to
those power perceptions.
And the importance of this position cannot be overemphasized.
"A brilliant university or college administration with inept
chairpersons cannot survive; an inept administration, with the help
of brilliant chairpersons, usually can" (Tucker 1981, 4). One dean
of faculty interviewed in the course of this research described the
chairperson role as harder and more important than his. He
characterized the job as similar to herding cats they (the
departmental faculty) always want to go in different directions and
they're quite likely to bite the hand that feeds them.
Department chairpersons in public institutions are, to a
degree, public administrators. Also, the environments within which
4


these chairpersons must function, and which may foster their
burnout (Maslach 1982a, 118), are much affected by public
administration's policies and practices. In particular, they are
highly related to the resource allocation aspects of public
administration. Inadequate resources to hire sufficient faculty to
teach necessary courses at appropriate class sizes and to provide
the teaching materials and support needed by faculty, legislated
faculty workloads, enrollment limits, denials of degree program
requests; all of these can create strong tensions and significant
frustrations contributing to the onset of burnout.
Definitions
Definitional difficulties exist with the key constructs
involved in this research. Those difficulties will be addressed
later in this dissertation. In brief, however, power, in the
context of this investigation, is "the ability to produce intended
change in others, to influence them so that they will be more
likely to act in accordance with one's own preferences" (Bimbaum
1988, 17). Stress, for this research effort, refers to a
nonspecific response of the body to any demand, particularly a
demand to which the body is not adapted. In this research,
job-related stress will be operationalized as "burnout," further
defined as one specific psychological and behavioral outcome of
stresses in the workplace. As Farber (1983) states, burnout is
5


not a stressor nor the perception of stress but more the final
step in a progression of unsuccessful attempts to cope with a
variety of stress factors. Primarily for research purposes, the
burnout syndrome has been further divided into three components:
emotional exhaustion, the depersonalized treatment of others, and
a feeling of reduced personal accomplishment.
Simply stated, the process is as follows (a more detailed
model is presented in Chapter 2). Conditions in the person and
conditions in the job interrelate to create job stress. This
stress then creates physiological strain to which the body reacts
by developing coping mechanisms to deal with the stress. These
mechanisms are exhibited in the depersonalized treatment of
others and a lowered feeling of personal accomplishment. When
the strain surpasses an individual's normal coping limits, the
result is the energy deficit encompassed by the notion of
emotional exhaustion (Maslach 1982a, 24). The burnout syndrome
and its components will be discussed in more detail later in this
dissertation.
Objectives of the Research
It is the purpose of this research to investigate the power
perceptions and symptoms of job-related stress (burnout)
evidenced by department chairpersons in four-year institutions of
higher education in Colorado and to determine if any dependence
6


exists between those variables and their organizational/personal
characteristics. This research is intended to add to the general
theoretical bases of both burnout and the department chairperson
role by providing data on the occurrence of burnout in these
individuals. The department chairpersons' perceptions of their
power bases relative to their faculty, peers and administrative
superiors will also be examined so as to identify what those
perceptions are and how, if at all, they may explain variance in
their burnout scores. Accordingly, the objectives of this
research are:
1. To determine the relative degree to which academic
department chairpersons in four year institutions of higher
education in Colorado exhibit symptoms of job-related stress
(burnout);
2. To determine the perceptions these individuals have
regarding the power they possess relative to their faculty,
their peers, and their deans;
3. To identify the relationships between symptoms of
job-related stress exhibited by these individuals, their power
perceptions and selected organizational/personal characteristics.
7


Hypotheses
Several general hypotheses guided the investigation. The
first was designed to discover whether or not the department
chairpersons surveyed exhibited the adverse result of
job-related stress characterized as burnout. The conditions
surrounding the role appeared to provide justification to
expect that such is the case. This led to the first hypothesis
and its two sub-hypotheses:
HI. That burnout is exhibited by a statistically
significant percentage of the individuals performing
the role of department chairperson in four-year
institutions of higher education in Colorado;
Hla. The average burnout ccnponent scores of these
individuals exceed those displayed by post-secondary
educators in the results of research by Maslach
(1981, 3);
Hlb: More of these individuals will score High
rather than Low cn the various ccnponents of burnout.
8


The second hypothesis is based on the idea that there is a
relationship between the perceptions of power by these
individuals and the level of their burnout. Research on other
groups has shown that burnout can be a consequence of "not
perceiving control over important outcomes in one's job..."
(Maslach 1982a, 40). The second hypothesis and its three
sub-hypotheses, then, are:
H2. That burnout is more prevalent among those
chairpersons who perceive they have inadequate power
to properly carry out the responsibilities of their
positions;
H2a. Subjects perceiving lower levels of power
relative to their departmental faculty have higher
burnout ccnpcnent scares than those who do not;
H2b. Subjects perceiving lower levels of power
relative to their peers (other department
chairpersons) have higher burnout component scores
than those who do not;
H2c. Subjects perceiving lower levels of power
relative to their administrative superiors have
9


higher burnout ccnponent scores than those who do
not.
The literature typically characterizes power as having
different bases. In the educational area, particularly, the more
coercive bases are considered undesirable. Accordingly, the
objective of the third hypothesis was to investigate how
department chairpersons perceived the relative strength of their
power bases. This hypothesis and its two sub-hypotheses, then,
are:
H3: That subjects perceive they have stronger power
bases of seme types than others;
H3a. Subjects perceive that they have more power due
to their expertise than to the authority associated
with their position;
H3b. Subjects perceive that they have more power due
to their ability to reward than to their ability to
punish.
Other research (Maslach 1982a, 58-63; Golembiewski and
MUnzenrider 1988, 64) has shown robust associations between
10


features of the worksite and personal characteristics, albeit
without "causal arrows." It would be expected that similar
associations, as well as others unique to the department
chairperson situation, would be demonstrated in this study.
Accordingly, the final hypothesis states, as a "null" hypothesis:
H4. That the organizational and personal characteristics
of the individuals occupying these positions do not
contribute to the differences in burnout component
scares reported.
limitations of the Study
It is to be expected that questions will be raised as to
the external and internal validity of this research.
Specifically, do the conclusions drawn from questionnaires
administered to these individuals accurately reflect the true
situation existing and can those conclusions be extended to
individuals occupying similar positions in other institutions of
higher education? The author would argue that, despite the
research's correlational design and its reliance on personal
perceptions, affirmative answers can be given to both questions.
The current version of the primary instrument used in this
research, the Maslach Burnout Inventory (MBI), was developed over
a period of approximately eight years. The developers of the
11


Inventory first administered it in a preliminary form, consisting
of 47 items, to 605 people from a variety of occupations
considered to have a high potential for burnout based on earlier
research. A factor analysis was then applied to the data
obtained from the first sample, resulting in a reduction in the
number of items, first to 25 and then to the currently used 22.
Convergent validity was established through correlation of
individuals7 MBI scores with independent behavioral ratings, with
the presence of certain job characteristics expected to
contribute to burnout, and with outcomes hypothesized to be
related to burnout (Maslach and Jackson 1981 1986, 10). The job
characteristics included the number of clients dealt with and key
dimensions on the Job Diagnostic Survey (Hackman and Oldham 1974,
1975). Among the outcomes analyzed were difficulties with family
and friends and the desire to leave the job.
The particular form of the Maslach Burnout Inventory used
in this research was MBI Form Ed, a modification of the basic
MBI by changing the word "recipient" to "student." The writer
further modified the instrument, with the permission of
Consulting Psychologists Press, Inc., by changing the word
"student" (in six items) to "my departmental faculty."
Two studies have established the validity and reliability
of the MBI Form Ed. Cronbach alpha estimates in each of the two
studies were approximately .90 for Emotional Exhaustion and .75
12


for both Depersonalization and Personal Accomplishment.
As in other such research, some self-selection bias may
have existed. Even though all department chairpersons in the
state of Colorado were given the opportunity and asked to
participate, individuals suffering from high levels of burnout
could have been disproportionately represented either among those
choosing not to do so or those choosing to do so. Also, despite
same researchers' findings to the contrary, the validity of
self-reported data as contrasted to that obtained through trained
observation is always suspect.
Finally, a comment in regard to generalizability of the
findings to department chairpersons outside the state of
Colorado. It has been the observation of this researcher, who
has been a teacher and administrator in four different
institutions of higher education across the country over a 21
year period, and has had close contact with department
chairpersons during that time (particularly as an administrator
in the last 13 years), that the problems faced by department
chairpersons and their relationships with their faculty, peers
and supervisors do not differ significantly among institutions.
Research to establish this observed generalizability, if such is
desired, would be a logical follow-on to this research.
The author would argue that, if job-related stress does
occur among department chairpersons in institutions of higher
13


education, the results of such stress are a workplace cost those
institutions cannot afford. This nation educates its best and
brightest, at great expense to society, so they can push the
frontiers of knowledge forward through research and pass
knowledge on to others through teaching. Same of these
well-educated individuals who are motivated toward, and capable
of, leadership choose to move into the plethora of roles required
of academic department chairpersons. Included in these roles
are: "fashioning the direction of the department, distributing
burdens and resources, coping with paperwork and dealing with
conflict," to name only a few (Bennett 1983, 2). Such roles
cannot be effectively accomplished, however, if department
chairpersons exhibit the classic symptoms of job-related stress;
i.e., became emotionally exhausted, treat others inappropriately,
and/or develop feelings of personal inadequacy. And, if
department chairpersons do not accomplish their roles
effectively, a strong argument can be made that the entire
institution may be adversely affected.
It is believed that, as a minimum, this study provides
information useful to those interested in improving the
department chairperson environment. The detailed look this
study takes at the power perceptions of department chairpersons
and the results of their job-related stress fills a void
14


currently existing in the literature concerning a very important
role in this nation's educational system.
15


CHAPTER 2
UTERAIURE REVIEW
Introduction
Two important constructs are material to this
investigation: (1) Power and (2) Job-related stress.
While the literature is relatively rich on the subjects of
stress and power, definitional difficulties exist with both
concepts and complicate the problem of measuring and relating
them. Also, the role of department chairperson and its
paradoxical nature needs to be fully appreciated in the context
of these constructs. Accordingly, this review of the related
literature will not only explicate the constructs that will be
studied in this investigation but will also discuss the roles
required of department chairpersons and the results of studies on
this group of managers.
Power
Definitions
Unfortunately, there is no one uniformly accepted
definition of power in the social science literature. One
dictionary of the social sciences lists 6 definitions of the
term.
16


- The ability to control the actions of others.
- Unauthorized ability to control the actions of
others.
- Illegitimate influence.
- The production of intended effects.
- The ability to produce effects.
- An authority of specified scope attached to an
office (Reading, 1977).
The Oxford Dictionary of the English Tananaae devotes 15
columns to different definitions of power, under headings which
include:
- Quality or property,*
- Person, body or thing;
- Technical term.
Dorwin Cartwright, a major writer in the field of power,
notes that "anyone reading the literature on power is bound to be
troubled by the absence of a generally accepted definition of
power. Most authors felt compelled to invent one of their own"
(1959, 185). In James McGregor Bums' epic book on
transformational leadership, he titles his first chapter "The
Power of leadership" and discusses its "multifarious, ubiquitous,
and subtle forms" but he never specifically defines it. John
Kbtter reports (1985, 1-2) that less than three percent of the
material in 10 popular management/organizational behavior texts
17


dealt explicitly with power, influence and authority and same did
not even mention them. The U.S.Army, in its series of field
manuals on leadership, skillfully avoids all reference to the
term. Accordingly, the literature gives one a broad choice of
power definitions.
General. The most general power definitions are generic
references either to "the ability to command, to apply force"
(Rue and Byars 1989, 254), or to "the basic energy needed to
initiate and sustain action" (Bennis and Nanus 1985, 9). These
definitions, however, could apply equally as well to the analysis
of an engine or of a pipeline problem as to a problem in
organizational behavior. Management theorists therefore
typically define power in a social context; their interest is in
the behavior of individuals. In this context the most general
definitions include "the ability to control the behavior of
others" (Sites 1973, 1), "an influence relationship" (Stogdill
1974, 275) and "the influence on a person produced by another
person, role norm or group" (French and Raven 1959, 608).
More specific definitions include an intent; e.g.,
"intentional influence over the beliefs, emotions and behaviors
of people" (Siu 1979, 31) and the "ability of one person to
produce (consciously or unconsciously) intended effects on the
behavior or emotion of another person" (Winter 1973, 5).
18


Same definitions include resistance by the recipient of the
power action. For example, Baron (1986, 436) states that power
can be defined as "the ability to control the actions of others
to promote one's own goals even without their consent, against
their will, and without their knowledge or understanding." Weber
(1947, 152) defines it as "the probability that one actor within
a social relationship will be in a position to carry out his own
will despite resistance [from another]." In contrast, however,
Stogdill (1974, 275) argues that many behavioral theorists assert
that power can be exerted without intention. And Hotter includes
in his definition an individual's ability to resist influence,
stating that power is a "measure of a person's potential to get
others to do what he or she wants them to do, as well as to avoid
being forced by others to do what he or she does not want to do
(1985, Preface).
A further problem in clearly defining power is presented by
the negative attitude many have toward the concept. Individuals
exercising power or acting so as to acquire power are often
viewed negatively. "People are suspicious of a man who wants
power" (McClelland 1975, 167).
In How Colleges Work. Bimbaum (1988, 17) defines power as
"the ability to produce intended change in others, to influence
them so that they will be more likely to act in accordance with
one's own preferences." Further, he stated that "power is
19


essential to coordinate and control the activities of people in
groups and universities as it is in other organizations." This
definition also incorporates the "horizontal power" concept
defined by Salancik and Pfeffer (1974) in their research on the
bases (or types) and use of power in decisions concerning
resource allocations in a large American university. Since
Birribaum's definition includes the emphasis on intent expressed
in Siu's and Winter's definitions, while avoiding the coercive
connotations of Baron's and Weber's, and has been specifically
used in analyzing governance of higher education, it will also be
used in this research.
Types and sources of power. Nineteen different types of
power are listed by Richman and Farmer (1974), referencing a
study by Cohen and March (1974) on the American college
president. Still, there is more consensus on the different types
of power and its sources than there is on its definition. Most,
if not all, writers on the subject of power recognize the five
pcwer types/sources delineated by French and Raven in 1959 as "a
basis for controlled research" (Stogdill 1974, 275). Those five
types will be used for this investigation of a power-stress
relationship and are defined (per French and Raven 1959, 613-20)
as follows:
Reward power: The power whose basis is the ability to
"administer positive valences and to remove or decrease negative
20


valences," the ability to reward. The strength of this type of
pcwer depends not only on the magnitude of the reward and one's
desire for it but on the strength of the perception by one person
that another can and will provide the desired reward if the
influence is accepted.
Coercive rower: The power which stems from the
expectation of punishment by one person if he fails to respond
correctly to an influence attempt by another. Its strength is
similarly dependent on one's perceptions and desires and the
magnitude of the punishment, punishment which can range from
physical harm to social rejection.
Legitimate power; The power which is categorized as
the power of position, as consisting of the right to influence;
i.e., authority. It is power which stems from internalized
values in one person which dictate that another has a legitimate
right to influence him and that he has the obligation to accept
that influence. Among the bases for this type of power are
cultural values, acceptance of a political or social structure,
and designation by a legitimizing agent; i.e., delegation of
authority to one person by a legitimizing agent accepted by
another.
Referent rower: The power which has its basis in the
desire by one person for identification (a feeling of oneness)
with another. Referent power does not include pressures toward
21


conformity; those are better associated with the other kinds of
power. Membership in a prestigious group, however, if valued by
an individual, would be an example of referent power.
Expert rower: The power which exists when one person
evaluates another's expertness in relation to her own as well as
against an absolute standard. Its primary influence is on one's
cognitive structure and hence an associated behavioral change is
normally categorized as having resulted from secondary social
influence. The range of expert power is more limited than that
of referent. The power of an expert is typically limited to
those areas in which she either has or is perceived to have
superior knowledge or ability. However, occasionally some "halo
effect" may occur whereby individuals are assumed to have
expertise in one area simply because they are highly respected
for their expertise in another.
In organizations such as colleges and universities, it
should be expected that both faculty and administrators will
primarily rely on referent and expert power, as those types of
power are less likely to cause alienation and more likely to
produce committed participants. Even though rewards,
particularly in the form of money, position and recognition, are
important to faculty, those individuals should tend to be
influenced more by internalized principles and collegial
relationships than by threat of administrative sanctions
(Birribaum 1988).
22


There are overlaps in the above distinctions; e.g., one
with expert power can influence another by promising to share
with the other that expert information he possesses or can
obtain. Also, Lasswell and Kaplan (1950, 97) state that "forms
of power and influence are agglutinative; those with some forms
tend to acquire other forms also." In spite of this, the
definitions developed by French and Raven provide a basis for
analyzing the sources of an individual's power that serves as the
common thread which ties together most of the empirical research
on the construct.
Research on Power in General. The concept of power has
been characterized as "perhaps the most fundamental in the whole
of political science; the political process is the shaping,
distribution and exercise of power" (lasswell and Kaplan 1950,
75). As mentioned previously, James McGregor Bums (1978, 40)
refers to it as "ubiquitous" and Siu calls it "the universal
solvent of human relations" (1979, 1).
Needless to say, then, the literature on power is
extensive. A search for the word in one college library's
catalog system found 5,345 references. But, while writings on
the subject are voluminous, many of the authors are concerned
primarily with expounding on hypotheses, assertions and personal
observations. James MacGregor Bums (1978) is one of these,
23


relying on historical demonstrations of power for the most part
to support his theses and rarely referring to any empirical
studies.
This is not to say that the research is not extensive also;
it is. That research is, however, quite recent in origin. In
Cartwright's 1959 review of the research literature he listed
only 33 references, not all experimental; in 1965 he listed 180,
mostly experimental (Stogdill 1974, 277-78). And in 1992,
Pfeffer listed 230 references in his in-depth look at the role of
power and influence in organizations (1992).
Same of the earliest instances of power research are
recounted in Studies and Social Power, edited by Dorwin
Cartwright in 1959. Among these is the report by Zander, Cohen
and Stotland (Cartwright 1959, 15) of interviews conducted to
analyze "the beliefs which members of three organizations have
about one another and the way in which those beliefs are
determined by an individual member's role and power." The
compilation also includes French and Snyder's 1959 studies on
leadership and interpersonal power in small groups in the U.S.
Air Force in which referent power was found to have the greatest
range (had the broadest behavioral influence).
At the time of its publication, Stogdill's classic survey
of the theory and research on leadership, Handbook of Leadership:
A Survey of Theory and Research (1974), was recognized as the
24


best compilation of the research on power, covering community
power structures as well as power in formal organizations and
experimental groups. More recently, however, Pfeffer (1992) has
provided a comprehensive evaluation of the power concept which
includes consideration of the different sources of power, the
specific strategies through which power and influence are used
and issues of power dynamics: how power is lost, its role in
organizational change and its positive and negative effects in
organizations. Pfeffer7s personal typology places all power
sources into two one of two categories: personal attributes,
including social adeptness, competency, ambition and sensitivity;
and structural sources, including control over resources and
position in the communication system of the organization. These
categories essentially subsume French and Raven's categories of
sources but contain more specificity. His support for his
delineation of sources includes Winter's study of the power
motive, need for achievement, and the affiliation-intimacy motive
(1987), Halberstam's analysis of the rise of finance personnel to
top leadership in large organizations (1986) and Krackhandt's
examination of the importance of one's position centrality in the
friendship network in a firm (1990). The traditional importance
of hierarchial position is not neglected in Pfeffer's treatment;
in this regard he refers to the classic works of Weber (1947)
Mechanic (1962) and Kanter (1977), among others. He concludes
25


his evaluation with a most appropriate reminder that, while it is
one thing to understand power, it is "quite another thing to use
that knowledge in the world at large" (1992, 337).
The reader is referred to Cartwright, Stogdill and Pfeffer
for detailed references to research on any particular aspect of
the general power construct.
Research on Power in Higher Education. In contrast to
research on power in general, the literature and research on
power within the area of higher education is not extensive. The
literature tends either to address itself to a broad look at
govemance/administration, typified by the work of Corson (1960),
Baldridge (1971) and Blau (1973), or to be prescriptive
concerning the specific duties Of department chairpersons,
typified by the work of Brann (1972), Dilley (1972), and Monson
(1972).
Corson's work is an "observation" of the governance of
colleges and universities, primarily discussing the university as
an "administrative exercise." He acknowledges that the
initiative for a great deal of educational policy, for personnel
appointment and evaluation, and for budgeting of equipment and
educational facilities has shifted to departments and that "the
basis of the power of departments lies in this initiative" (1960,
87). He further notes how much the roles of department
chairpersons vary at different institutions, from Gaucher College
26


where chairpersons "refrain from 'running7 their departments" to
George Washington University where they are usually younger and
are not accorded the "formal responsibilities normally associated
with department heads." He also refers to a study by Doyle
(1953) of 33 liberal arts colleges in which 69 percent of the
departmental chairpersons participated directly in budgeting
formulation and faculty selection, promotion and retention
(Corson, 1960). Baldridge's case study of New York University
(1971) contains virtually no mention of the power of department
chairpersons. He describes that university's power as having
been shared only among trustees, presidents, faculty and
students, with large shifts occurring in its primary location
over the period analyzed in the study. Blau (1973) recognizes
the existence of departments as part of the institution's
organizational system but primarily addresses the effects of
their creation, abolition, size and number within that system,
without comment on the ability of department chairpersons to
influence those actions and characteristics.
On the prescriptive side, Brann, while including a list of
29 functions of a department chairperson from a Pennsylvania
State University faculty, handbook (1972, 7-10), makes no mention
of the power needed to carry out these functions. Dilley does
include "fighter" in his list of eight functions for a department
chairperson and mentions the "need to build up power which can be
27


used to counter-balance power exerted against you" (1972, 35) but
gives no details on how to do it. Monson's training program for
department chairpersons (1972) includes nothing on power.
Two areas of research on department chairperson power have
been conducted that do relate directly to this study. Those
include an analysis of the perception of power of department
chairpersons by professors Winston Hill and Wendell French (1967)
and analyses of power in resource allocation decisions in
colleges and universities made by Salancik and Pfeffer (1974) and
by Hackman (1985). Hill and French, in their 1967 study of five
state-supported four year colleges, found that the greater the
power of the chairpersons, the greater the professors' level of
satisfaction. They also found that "professors consider
departmental chairmen as having less influence than any other
groups in the colleges, even less than the professors" (1967,
558). They attributed the low power position of departmental
chairpersons to a lack of influence over higher administrative
groups and concluded that the chairpersons were subject to more
control by professors than were the professors by the
chairpersons (1967, 558-59).
Salancik and Pfeffer's 1970 examination of a large
Midwestern state university included interviews with department
chairpersons in which they were asked to rate the power of their
own subunits, ranging from "a great deal" to "very little." They
28


found that power, used most in the allocation of graduate
fellowships, was most highly correlated with the department's
ability to obtain outside grants and contracts. Further,
"graduate education and research were empirically found to be the
best predictors of subunit power" (1974, 460).
In an extension of that work in 1980, they asked department
chairpersons at the University of Illinois and the University of
California how much power they thought various departments
possessed. There was great consistency in their responses: no
department that was judged to be in the top third overall was
rated by any chairperson as being lower than the top third and,
similarly, no department that was judged to be in the bottom
third overall was rated by any department chairperson as being
higher than the bottom third. A particularly interesting
observation was that only one department chairperson asked for a
definition of the power to be evaluated (Pfeffer and Moore 1980)
A research-based theory on internal resource allocation by
colleges and universities was developed by Hackman (1985) from
interviews at six institutions. Her findings supported and added
to the work done previously by Salancik and Pfeffer, emphasizing
the importance of centrality (core positions versus peripheral
positions) in resource allocation decisions.
29


Job-related Stress
Definition
Stress has been called "one of the most imprecise terms in
the dictionary" (Ivancevich and Matteson 1980, 8). Beehr and
Newman (1978, 667) have stated that "there is no universally
agreed-upon meaning [of stress] among behavioral scientists, let
alone scientists in general."
Writers on the subject of stress consistently reference
Selye (1983) who, in 1936, developed a conceptualization that
still provides the keystone for understanding the stress
construct. He entitled the conceptualization "The General
Adaptation Syndrome" (GAS). According to Selye, the GAS is
composed of three stages: an alarm reaction in the organism in
response to diverse stimuli to which it is not adapted, a
resistance stage in which the organism's physiology tries to
Oounteract the effects of the alarm reaction, and a stage of
exhaustion in which resistance fails (Selye 1983). Stress, as he
defined it, is a nonspecific response of the body to any demand,
essentially stage one of the GAS. Similarly, Quick and Quick
(1984, 3) have defined stress as "the generalized, patterned,
unconscious mobilization of the body's natural energy resources
when confronted with a stressor," where a stressor is "any
demand, either of a physical or psychological nature, encountered
in the course of living."
30


Rather than attempting to deal with all the various ways in
which stress and stressors can occur and the associated behaviors
which can be exhibited by individuals experiencing stress, a
number of researchers have concentrated on the particular
behaviors of persons whose work requires frequent interpersonal
contact of an intense psychological nature. The name they have
applied to these behaviors is "burnout," an outcome of
work-related stress.
Relationship between Stress and Burnout
While the stress and burnout concepts are sometimes
ccanmingled, clear distinctions can, and should, be drawn between
them. These distinctions are demonstrated by a combination of
two models: a general model of work stress (Farber 1983) and a
model of the burnout process (Gaines and Jermiar 1983). It is
shown in Figure 2.1.
As indicated in the figure, burnout is one specific
psychological and behavioral outcome of stresses in the
workplace. It is not a stressor nor the perception of stress but
more the final step in a progression of unsuccessful attempts to
cope with a variety of negative stress conditions (Farber 1983).
Burnout, as Farber, Gaines, Jerman and others have agreed
to define it, is a syndrome involving emotional exhaustion,
depersonalized treatment of others, and feelings of reduced
personal accomplishment (Maslach 1982). The unique aspect of
31


Figure 2.1. Burnout (One specific psychological and behavioral
outcome of stress in the workplace)
turnout is that, by definition, the stress leading to it arises
exclusively from social interaction in the individual's work
environment and not from relationships occurring outside that
environment, such as with family members or social acquaintances.
32


The burnout syndrome was initially investigated primarily
among those in the helping professions, particularly health care
providers, teachers, social and human services workers, and
prison and law enforcement personnel. Typically, individuals in
these professions who experienced burnout had negative
perceptions of their workplace processes (Tennis 1986);
tended toward cver-commitment, authoritarian personalities and
ideal .ism (Freuderiberger 1977); perceived responsibility without
authority or resources to accomplish the task (Emener 1982);
and/or encountered problems of role conflict (Eunham 1978) and
role overload (Chemis et al. 1985).
The most widely accepted definition of burnout results from
research by Maslach and Jackson (1984) on the common
characteristics displayed by "burned out" individuals. In their
definition, burnout consists of any or all of three components.
Emotional exhaustion usually a loss of feeling,
trust, interest, and/or spirit.
Depersonalization a negative attitude toward
clients and coworkers.
Reduced feeling of personal accomplishment -
feelings of depression, failure and incompetence;
a devaluation of self.
Research on Burnout
Research conducted by Golembiewsk and Munzenrider beginning
33


in the early 1980's, which they admittedly began with a mind-set
biased toward "debunking" the burnout phenomenon, has shown that
"burnout is where you find it; that is, everywhere, whether at
high levels of organization or low, in various demographic
aggregates, and so on" (1988, 12). These researchers made a
major contribution toward a better understanding of the
phenomenon by their development of an eight-phase model of
burnout, varying from the demonstration of low levels of all
three components by an individual to the demonstration of high
levels of all three. Maslach and Jackson (1981, 1986), however,
while noting that this model proposes an "interesting way" of
combining MBI scores, believe that "the validity of this or any
other phase model will depend on far more extensive theorizing
and research" (14).
Little was written about burnout before the early 1970s
when, for all intents and purposes, Freudenberger and, later,
Maslach "discovered" it. They also discovered it was a major
concern for many people. Same of this concern may have been
merely faddish. For example, that was the time for many
self-analysis best sellers, e.g., When I Sav No. I Feel Guilty
(Dyer 1981a) and Your Erroneous Zones (Dyer 1981b). But as
research was conducted, primarily among those who worked in the
helping professions nurses, doctors, counselors, and prison
administrators the existence of the burnout syndrome was found
34


to be pervasive. Maslach found that, while personality plays a
part in burnout, the syndrome is best understood in terms of
job-related interpersonal stress (Maslach 1982). Chemis et al.
(1985) proposed that the major causes of burnout in public sector
professionals are beliefs about their jobs that lead to
unrealistic expectations and disillusionment. Warnath and
Shelton's research (1976) also concluded that situational factors
contribute greatly to burnout, particularly such factors as a gap
between ideals and job realities and a lack of reinforcement from
colleagues. Where such interpersonal (and intrapersonal)
stressors exist, evidence can frequently be seen of the emotional
exhaustion, depersonalization and feelings of reduced personal
accomplishment that are symptomatic of burnout.
The research on burnout by Christina Maslach, together with
Susan Jackson and A. Pines, is extensive (see Bibliography). To
summarize her findings, particularly in respect to burnout in
educators:
- Men typically show more of one aspect of burnout and
women more of another. Specifically, women score higher on
Emotional Exhaustion while men are more apt to score higher on
Depersonalization;
- Compared to whites, blacks do not bum out as much.
Burnout among Asian-Americans is very similar to that among
whites. Research on other ethnic minorities has been
insufficient to provide a basis of comparison;
35


- Burnout varies inversely with age, being more
prevalent with youth;
- The greatest amount of burnout is found in providers
who have completed college but have no postgraduate training.
Those with postgraduate training demonstrate higher levels of
Emotional Exhaustion but score lowest on the other components
Maslach 1982, 58-61).
Although role characteristics have been investigated for
possible contribution to burnout, Clarke, in her 1991 research
into burnout, found the results so far to be inconclusive. As
she reported, Kahn (1974) developed evidence that job
satisfaction decreases in the presence of role conflict and role
ambiguity. Chemiss (1985) found among human service
professionals that role ambiguity and role overload contributed
to burnout. Both Chiarmonte's (1983) research with priests and
Jackson's (1986) with teachers supported role conflict as an
antecedent of burnout. In a study of professional women by Pines
and Kafry (1981), however, the number of roles performed did not
promote burnout and, in fact, the variety of roles added interest
and stimulation and was negatively related to burnout. Also,
role ambiguity was found in a study by Leiter and Meedhan (1986)
to raise rather than lower feelings of personal accomplishment.
Further, a study of burnout by Jayaratne and Chess (1984) found
no contribution from role ambiguity. Clarke concluded her review
36


of burnout sources by stating "the role of personal and
environmental variables in the development of burnout is far from
clear" (1991, 54).
A highly comprehensive integration of research on job
burnout is provided in an October 1993 review by Cordes and
Dougherty (1993). Even at this late date they found that "little
has been done thus far to establish burnout's generalizability in
industry" (1993 621). Their review thoroughly examines the
construct of burnout, describes its evolution and provides an
extensive overview of the literature on the antecedents and
consequences of burnout, many of which are also contained in this
paper's bibliography.
Cordes and Dougherty's review of the burnout research
caused them to make a number of propositions specifically
intended to clarify the dynamics of the antecedents of burnout so
as to assist further research. These included:
1. High levels of work and personal demands are
critical determinants of the components of burnout;
2. The key determinants of emotional exhaustion
reflect both organizational and personal demands placed on
employees;
3. If demands from one's supervisor conflict with the
demands from clients, role conflict may occur and lead to
burnout;
37


4. Boundary spanners, who function as information
processors or filters between the organization and the client and
act as agents in influencing the decisionmaking of the client,
are caught in a difficult position when they perceive that client
demands cannot or will not be met by the organization;
5. Employees enter the workplace with certain
expectations regarding, among other things, what they will be
able to accomplish professionally. The contrast between those
expectations and daily realities will influence their perceptions
of personal accomplishment (1993, 631-47).
The findings of Cordes and Daughterty will, wherever
applicable, be contrasted with the conclusions of this research
effort.
As mentioned previously, the process of burnout, usually
expressed as a sequencing of various burnout components has been
conceptualized primarily by Maslach with a variation developed by
Golembiewski and Munzenrider (1981, 1984). Studies by Leiter and
Maslach (1988) and others support Maslach's model but a number of
researchers in the late 1980s have used Golembiewski and
Munzenrider's phase model successfully in their studies.
Much research has substantiated the validity of the Maslach
Burnout Inventory (e.g., Maslach and Jackson in 1981 and 1986,
and Wolpin, Burke and Greenglass in 1991), but it should be noted
that several authors (e.g., Kbeske and Koeske 1989 and lee and
38


Ashforth 1990) have reported correlations between burnout
components.
While the primary research on the antecedents of burnout
has been conducted by Maslach and Jackson (1981, 1984),
significant research on antecedents such as role conflict and
role ambiguity has been accomplished by Kahn (1978) and Schwab
and Iwanicki (1982). Also, Russell, Altmeier and Van Velzen
(1987) found an association between burnout scores and higher
staff-child ratios in school classrooms (i.e., quantitative
overload).
Recently, the effect of job context as well as gender on
stress has been investigated by Pretty, McCarthy and Catano
(1992), providing findings that women experienced more emotional
exhaustion and depersonalization if they were nonmanagers while
men did so if they were managers. The effects of social support
on stress has been most recently studied by Kirmeyer and
Dougherty (1988) who found that "social support buffers the
relationship of stressors with use of adaptive coping" (Cordes
and Dougherty 1993, 633).
Finally, in 1988 Kahili grouped the consequences of burnout
into five categories: physical, emotional, interpersonal,
attitudinal and behavioral. Research involving these
consequences includes work by Maslach and Jackson (1981, 1982,
1984), Burke and Deszca (1986), and Firth and Britton (1989) as
39


well as others but there is little empirical evidence to date
supporting these relationships.
Department Chairpersons
General
One group of individuals who have not been investigated as
to burnout propensity, but who appear to have work environments
likely to give rise to occurrence of the syndrome, are academic
department (or division) chairpersons in four-year institutions
of higher education.
On the one hand, these individuals might logically be
expected to be self-assured, enthusiastic about their work, and
supportive of those with wham they are in contact. They have
academic credentials that are envied and admired by American
society and they occupy important positions in a respected
American institution: higher education. They would not logically
be expected to fit the profile of what Maslach terms a
"burnout-prone individual": weak, nonassertive in dealing with
people, submissive, anxious, lacking in self-confidence, lacking
in ambition, or fearful of involvement (Maslach 1982, 62).
Individuals of that description would normally not be expected to
seek out or accept the position of department chairperson.
Department chairpersons are, however, "the men [people] in
the middle." They are the linking pins between the administration,
40


those responsible for operating the business of education and
research, and the faculty, those responsible for accomplishing
that education and research. As linking pins, each has a focal
role, "sitting in the middle of a group of people, with all of
wham he interacts in some way in that situation" (Handy 1983,
54). This situation may lead, however, as described by Calvin
Lee, Chancellor of the University of Maryland, to a department
chairperson who is ".. .quite often in conflict as to whether his
role is one of spokesman for his colleagues in the department or
whether it is one of an administrator who must make the decisions
not only for the welfare of his department but for the welfare of
the college and university as a whole" (in Brann and Emmet 1972,
54-55). Interestingly, Dr. Lee did not acknowledge that
chairpersons could be spokespersons for their faculty colleagues
and make decisions in the best interests of the department as
well as the college/university as a whole. One example of the
difficult situation faced by persons in an academic hierarchy is
provided by the dean at one Colorado institution included in this
study. The dean was removed from his position a few years ago
because, in the eyes of the department chairpersons and faculty,
he failed in the performance of the faculty spokesman role. His
decisions were seen, at least in the eyes of the faculty, as
being neither in the best interests of the faculty nor,
accordingly, in the best interests of the institution as a whole.
41


The "balancing act" required of educational administrators
is a tenuous one. Admittedly, as well described by Likert (1961)
and Handy (1983), the "linking pin" role of managers can be
expected to present ambiguity. The "first among equals" status
of department chairpersons, however, provides the potential for
aggravating that ambiguity. Movement up in the managerial
hierarchy in organizations (other than on the academic side of
post-secondary educational institutions) brings with it new
expectations and perceptions. Those in positions now subordinate
to the new leader are ejected to recognize the change in that
person's role. They are expected to perceive and behave toward
that individual differently, acknowledging her as no longer their
equal, even if only as a result of a mandate to that effect by
higher management. Not so in post-secondary education; that
difference is not so perceived by the institution's faculty nor
is its acknowledgement required by higher level administrators.
The new chairperson is typically viewed by both as still a
faculty member, only with additional duties.
The method of selection of chairpersons may also affect
their attitudes toward the position. Sometimes they are
unilaterally appointed by a dean after same discussion with
department members. Sometimes they are selected by a committee
composed of department members, subject to veto of the dean.
Another practice is to nominate by ballot, with the dean
42


having authority to accept or reject the nominee. Many
departments practice rotation among senior faculty, with terms
varying from as little as one year to as many as five (Demerath
1967, 185).
Workplace Characteristics of Department Chairpersons
Department chairpersons -typically see themselves as first
among equals, as faculty members and academicians first and only
secondarily as "linking pins" between the administration and
faculty members in their departments. Their primary
responsibilities, as many perceive them, consist of enhancing
their departments' reputations, insuring their departments get
fair shares of always scarce resources, and protecting their
departments from incursions by administrators which interfere
with the teaching, research, and public service their faculty
want to accomplish (Woolsey et al. 1990).
The position of academic department chairperson typically
encompasses many responsibilities. At the Colorado School of
Mines, the Faculty Handbook states that:
Department heads represent their faculty and are
responsible for their vigor and productivity but at
the same time are members of the administrative
team. It is important for department heads and
faculty members to clearly recognize this two-way
role involving responsibilities to both the
department and to the school as a whole. Department
heads are also responsible for administering and
operating their respective academic units.
Consequently, department heads need to possess
administrative and personnel skills as well as to be
leaders among peers in their areas of academic
training. (CSM Faculty Handbook, 62)
43


Among the tasks involved in "administering and operating"
their respective academic units are:
- Planning department curricula, policies and
procedures;
- Determining activities required to achieve
departmental goals;
- Recruiting, selecting, recommending, promoting
and retaining faculty;
- Assuring intradepartment and
department-institution communication;
- Participating in establishment of new
departmental programs;
- Motivating faculty to achieve departmental
goals;
- Representing the department to the academic and
administrative units of the institution,
professional associations, accrediting
agencies, government bodies, foundations, and
the business community;
- Obtaining funds from institutional and other
sources and allocating them within the
department;
- Engaging in public relations functions;
- Participating in academic counseling of
enrolled students and making decisions in
instances of departure from departmental
norms;
- Recommending salaries;
- Coordinating cooperative programs with other
departments, educational institutions,
government agencies, and/or private industry
(Fletcher, 1969).
Other authors list similar and additional
responsibilities and identify no particular differences in
the department chairperson job and culture at four year
institutions across the country (Doyle 1953, Norton 1978,
Tucker 1981).
44


In view of the responsibilities listed above and the
interdependence of departmental and institutional objectives,
it would certainly not be surprising if senior administrators
in same institutions of higher education regarded the
chairpersons of their academic departments as subordinate
administrators. These would be those individuals who viewed
their department chairpersons as first line supervisors
responsible for carrying out the objectives of the
institution through proper management of the faculty in their
various departments. Same faculty, on the other hand, could
be expected to regard their department chairpersons more as
black sheep, as former members of the flock who have not
exactly deserted and joined "the other side" in the classic
administration-faculty conflict but whose loyalty to them,
the faculty, is still somewhat suspect. In view of the
nature of academe, however, it would not be expected that
many faculty would view their department chairpersons as
supervisors, as individuals responsible for carrying out
specific responsibilities by managing/leading them, their
faculty. Accordingly, a situation can easily obtain wherein
the role conflict and role ambiguity previously discussed,
can occur and possibly result in burnout.
45


Research on Department Chairpersons
Kathryn Martin's study of department chairpersons and
deans in colleges of fine arts (1966) concluded that
department chairpersons were often in the situation of
fulfilling a multiplicity of roles, with higher stress levels
and more probable role ambiguity and conflicts for department
chairpersons than for their faculty.
In 1978, Norton investigated the role of department
chairpersons in colleges of education nationally. The
foreword to his report states, "This is indeed an area
wherein little research has been done." His findings
covered factors underlying resignations, difficulties in job
responsibilities, satisfaction and related job factors, the
state of the chairperson's position and the changing nature
of the role, to name just a few.
Among the 202 participants in his study, Norton found
that:
55% held the rank of full professor at appointment;
75% were department members prior to their
selection;
The mean age at time of appointment was
approximately 42 years of age;
42% were equally distributed between first year in
the position and five to ten years;
46


More than 50% chaired a department with 6-10
full-time equivalent faculty;
50% saw themselves as academic faculty members and
46% as equally a faculty member and an
administrator;
Over 50% gave as a primary reason for accepting the
position; "the opportunity it presented for
providing leadership to the academic program"
(Norton 1978, 2-5).
Hie Norton study also found that "the adequacy of
department resources" provided the highest degree of
frustration for chairpersons. General job frustrations and
negative faculty relationships were found to be the primary
factors leading to resignation. On the other hand, support
from department faculty, with regard to decisionmaking and
curriculum development, received the highest ratings of
significance as related to job satisfaction (Norton 1978,
21-27).
Department Chairpersons and Burnout
Various investigators (Cresswell 1986; Singleton 1987;
Saton-Spicer and Spicer 1987) have identified the position of
department chairperson as key in the administration of
today's institutions of higher education. The work situation
of the academic department chairperson,
47


however, provides opportunities for many of the stresses
leading to burnout: unrealistic expectations,
disillusionment, gaps between ideals and job realities, and
laclc of collegial reinforcement. Also adding to the general
malaise seemingly associated with the position, Wilbert
MdKechnie, a professor of philosophy at the University of
Michigan, has described department chairpersons as
"...generally ill-prepared, inadequately supported, and more
to be pitied than censured" (Brann and Emmet 1972, 43). It
would not be unlikely, then, if a department chairperson
suffered from role ambiguity, role incompatibility, role
conflict, and/or role overload or underload each of which
has been argued to be a cause of burnout.
The most recent survey concerning department
chairperson stress is that which was conducted by Gnelch and
Bums (1993). These researchers, director and associate
director, respectively, of the Center for the Study of the
Department Chair at Washington State University, surveyed
over eight hundred chairs from 101 doctorate granting and
research universities using the Department Chair Stress
Index. Similar to previous studies and to the premise of
this research, they found chairpersons to be "in a
paradoxical situation, feeling double pressure to be an
effective leader and productive faculty member" (1993, 259).
48


They stated, however, that "the occupational stress
associated with the dual administrator-faculty role of the
academic department chair has not been investigated" (1993,
260).
In Gnelch and Bums' study, chairpersons identified the
most serious stressors as having too heavy a workload,
obtaining program approval and financial support, keeping
current in their disciplines and complying with rules and
regulations. No significant differences were found, however,
in the severity of stress experienced by age, gender and
discipline variables.
Faculty and administrative job strain and its impact on
"quality of life" indicators have also been examined by
Blackburn and others (1986), resulting in the identification
of a significant association between the satisfaction of
administrators with their supervisor and their job strain.
Given the foregoing, it is quite possible that a
department chairperson's "struggle to cope with the
traditional ambiguity of the position" (Bennett 1983, 2) is
an antecedent to the individual's exhibiting the burnout
syndrome. If unrelieved by some action such as frequent
rotation of faculty members through the position, the
stresses which may result from performing the department
chairperson's role can lead to the occurrence of the burnout
syndrome in these individuals.
49


Power and Department Chairpersons
Burnout is a greater risk when people feel powerless, when
they have the "sense that they are at the mercy of the situation
and that there is nothing to do about it" (Maslach 1982, 146).
This can be a chairperson's lot. Chairpersons are designated to
be leaders, yet are seldom given "the scepter of undisputed
power" and are first among equals, yet find that "any strong
coalition of those equals can severely restrict [their] ability
to lead" (Tucker 1981, 4).
The collegial aspects of the position are extremely
important to the exercise of power. Typically the chairperson
"exerts leadership through the power of his ideas" (Heimler, in
Brann and Emmet 1972, 198), not through the authority that goes
with the position. They are colleagues as well as and perhaps
more so than supervisors.
As Bennett (1983) states: "Ample opportunities exist for
department chairpersons to perceive a shortage of the power they
need to carry out their responsibilities" (1983, 146). And
without power, without the ability "to produce intended change in
others [their faculty, their peers, and their deans], to
influence thorn so that they will be more likely to act in
accordance with [the department chairperson's] own preferences,"
it is most unlikely that they will be able to accomplish their
departmental objectives. As previously noted, "power is
essential to coordinate and control the activities of people and
50


groups in universities, as it is in other organizations"
(Bimbaum 1988, 12-13).
Additionally, the work environment of the academic
chairperson also contains, as part of its potential for causing
stress, the "horizontal power" problems identified by Salancik and
Pfeffer in their research. They defined this type of power as
"the use of influence among coacting peers to obtain benefits for
themselves," a necessary but difficult part of the department
chairperson's role if "themselves" is interpreted to include their
departments (Salancik and Pfeffer 1974, 453). The power to which
Salancik and Pfeffer referred is the social power discussed by
French and Raven and is the subject of this study, a concept the
latter authors acknowledged as possessing "more intuitive appeal
than empirical precision" (1959, 456).
Summary
As discussed, the department chairpersons' constant "person
in the middle" situation can easily lead to role ambiguity, role
conflict and a feeling of powerlessness. The protected (tenured)
status of the faculty they must "manage" and the frustrations they
encounter in obtaining the resources they believe are necessary
for proper functioning of their departments can also lead to
feelings of powerlessness. In all cases the result can be stress;
in many, the stress may result in burnout burnout which may be
highly related to a chairperson's perceptions of his or her power
as well as other factors in the work environment.

51


CHAPTER 3
MEJIHODOIDGY
General
Burnout has been analyzed chiefly in the helping
professions, among physicians, nurses, mental health workers,
and police officers in particular (Ginsburg 1974 and Chemis, et
al. 1985). In the educational area, burnout has been analyzed
primarily among teachers and counselors, with only general
analyses of its occurrence in educational staff members
(Freuderiberger 1977 and Carroll 1980). Studies on power in a
university, such as Salancik and Pfeffer's (1974) Hackman's
(1985), and Baldridge's (1971), have involved investigations only
on the relationship of a department chairperson's power to affect
resource allocation or on the more general issue of institutional
governance. This investigation looks at both burnout and power,
with, as its unit of analysis, academic department chairpersons
in four year institutions of higher education; those individuals
who are at once teachers, counselors, staff and executives.
The individuals studied occupy (or had recently occupied)
the position of department (or division) chairperson in one of
fourteen four-year institutions of higher education in Colorado.
An analysis of the college structure in Colorado identified 441
such positions at the time the study was conducted. Since these
52


structures are always changing, subsequent studies could find
either more or less such positions.
The work environments of these individuals vary widely. At
the institutional level these situations can vary depending on
whether the institution is privately or publicly funded; whether
it is focused on teaching, research or a combination of the two;
whether it serves a broad or a narrow community; and whether its
students are residential or commuting. At the departmental level
the situations can differ depending on the discipline involved
(and, therefore, variations in funding levels and emphasis on job
preparation or educational breadth), whether the department is
primarily undergraduate or graduate student oriented, whether it
grants degrees or serves only in a support role, and the
department's age and size.
These differences make generalizations of the findings to
specific departments suspect. But, while the explanatory effects
of selected variables in those environments on the occurrence of
burnout were investigated, the intent of this research from the
outset was to analyze those environments from primarily a power
perspective, without strongly differentiating between the types
of institutions or departments contributing to the study.
Accordingly, generalizability should be based primarily on
similar power perceptions.
53


The investigation was quasi-experimental. That is, it
involved an experiment that had a treatment (a possible cause),
outcome measures (possible effects of the treatment) and
experimental units, but it did not use random assignments to
create the comparisons from which the treatment which causes
change is inferred (Cook and Campbell 1979, 295-98). Its results
were exploratory and investigative only, involving the absence of
independence, not the inference of causation. Data were
collected from the chairpersons mentioned by use of previously
tested and validated written questionnaires, including personal
and organizational questions as well as questions about power
perceptions and burnout. Among the personal and organizational
characteristics analyzed were age and gender; reasons for taking
the position; whether the chairperson's power includes assigning
courses to be taught; and whether the chairperson recommends
tenure, promotion and salary increases, to name a few. The
intent of the questions on power perceptions and burnout was to
obtain data on subjects' perceptions of their stress levels and
of their influence on their faculty, their peers and their deans.
Data Collection
The folder in Appendix A was mailed to all identifiable
department chairpersons in four-year institutions of higher
54


education in Colorado. The folder consisted of an explanatory-
letter requesting their participation in a research study,
thirty-one organizational/personal data questions, the Maslach
Burnout Inventory Educators Survey (twenty-two questions), and
three fifteen-item questionnaires addressing their power
perceptions.
This choice of a research sample was made for a number of
reasons. Among these was that the author had direct contact with
each institution through his state professional association
(Rocky Mountain Association of College Registrars and Admissions
Officers), that they provided a sufficiently large group for
statistical significance, and that they included rural and urban
as well as large and small institutions. While the data were not
gathered prior to the impact on state budgeting resulting from
the passage of Amendment 1 to the state constitution in fall
1992, that passage probably only exacerbated the resource problem
ever present in any educational institution.
Letters were also sent to deans/directors of the department
chairpersons, asking them to encourage their chairpersons to
complete and return the surveys. Four hundred and forty-one
folders were sent out and one hundred and sixty-six usable
replies were received a 35.3% response rate.
The first questionnaire asked for biographical and
organizational data.
- Reasons for becoming a department chairperson.
55


- The individual's goals for the department.
- The process by which he/she became a department
chairperson.,
- The type, age, size, stability and role of the
department.
- The levels of the degree programs offered.
- The availability of resources to the department.
- The degree of influence on various institutional
actions affecting faculty.
- The individuals' age, sex, ethnicity, highest degree
held, years in position, years on the faculty,
academic rank and salary (both in dollars and in
relation to salaries of faculty in the department).
The second questionnaire was entitled "Educators Survey"
and consisted of the Maslach Burnout Inventory adapted to
education. It contained 22 statements of job-related feelings
and asked how often, if at all, the individual had had this
feeling. Forced choices for answers ranged from "never" or "a
few times a year or less" to "every day."
The third questionnaire contained 15 statements as to the
department chairperson's perceptions concerning the attitude of
departmental faculty regarding the chairperson's ability to
influence their behavior. Each statement was related to one of
the five bases of power defined in the French and Raven
56


typology. For each statement the department chairpersons were
asiced to indicate the degree to which they concurred with the
statement. Five choices of degree were offered, ranging from
"strongly agree" to "strongly disagree." Examples of the
statements included "Respect me and want to act in ways to merit
my respect and admiration" and "Cooperate with me because they
wish to be identified with me."
The fourth and fifth questionnaires were identical to the
third, except the statements in the fourth related to the
attitudes of other department (division) chairpersons in their
college and in the fifth to the attitudes of the dean/vice
president or other individual to whom the chairperson was
responsible.
All responses were coded with numerical values: 0 to 6 for
the Educators Survey, 1 to 5 for the power perception statements
and from 1 to a variety of numbers for the organizational and
biographical data.
Data Analysis
The survey data were entered into a Minitab program on a
VAX 4500. Individual scores on the various items in the
Educators Survey were then combined appropriately to provide
measurements of Emotional Exhaustion, Depersonalization and (a
reduced feeling of) Personal Accomplishment. Responses to
57


statements in the power questionnaires were added together in
groups of three, determined by which of the French and Raven
power bases they concerned. The resulting values, ranging from 3
to 15, represented the degree to which each power base was
perceived to exist. Preliminary chi-square tests were then run
on contingency tables made up of the data in the Minitab table.
These revealed a need to combine some of the data so as to avoid
discontinuities in the distribution of the expected frequencies
obtained, thereby negating the need for applying Yates'
correction for continuity and improving the value of the
chi-square tests.
For a portion of the statistical inquiry, the power data
were aggregated into a three-stage distribution: perceived to
exist, neutral, and perceived not to exist. This distribution
was made by assigning the first state to the case where the power
type was distinguished with a value of 3 through 6, the second 7
throuc£i 11, and the last 12 through 15.
In the case of burnout data, Golembiewski, Munzenrider and
Carter initially used median scores to distinguish High from Low
burnout scores. In subsequent comparisons of two different
sites, they used percentile rankings, categorizing scores above
the median of each component as High and below as low (1973,
28). Maslach, however, prefers dividing the normative
distribution within a particular sample into thirds: High,
58


Average and low. To categorize the department chairpersons
surveyed in this study, the burnout scores obtained by Maslach
for 635 post-secondary educators were used as a baseline:
Table 3.1. Maslach's Burnout Scores for 635 Educators
High Averaae Low
Emotional Exhaustion >23 14 23 <14
Depersonalization > 8 3 8 < 3
Personal Accomplishment <36 36 42 >42
(Maslach and Jackson 1981 1986, MBI Manual, 3)
Depending on the score of each chair, then, for each component of
burnout, it was translated into a 3 for High, a 2 for Average,
and a 1 for Low.
Aggregation of organizational/personal data varied between
questions, depending on the breadth of the answers. Written-in
personal reasons (primary and secondary) for taking the position
were collapsed into six codes and combined with the six choices
provided for check-off, with the value of 12 representing
"Other." The department chairpersons' perceived levels of
influence, compared to the levels exerted by most of the
department's faculty, were combined into three: much more,
somewhat more, and approximately the same or less. Leadership
style ratings of 1-3, representing the autocratic end of the
spectrum, were combined into the value 3 while ratings of 7-9,
59


the laissez-faire end, were combined into a 7. Like compressions
were necessary for the availability of institutional and
noninstitutional resources questions. Ethnic minorities had to
be combined into "non-Caucasian" due to the few occurrences of
individuals in ethnic groups other than Caucasian in the study.
Because nearly all the chairpersons had more than 11 years as
full-time faculty members, responses to the question on this area
in the 1-4 and 5-7 year ranges were combined with the 8-11 year
responses to give a "less than 12 years" category. Other such
aggregations were made as necessary to create chi-square cells in
which the expected frequencies were less than 5 in no more than
20% of the cases (per MINITAB Handbook 1985, 274).
Survey Instruments
The two instruments used to analyze power perceptions and
burnout had been previously used in other studies and thoroughly
validated by their developers. One, the Maslach Burnout Inventory
(MBI), was a twenty-two item measure of the three aspects of the
burnout syndrome: emotional exhaustion, depersonalization and
reduced feeling of personal accomplishment.
The reliability coefficients that have been demonstrated for
the burnout scales of emotional exhaustion, depersonalization and
personal accomplishment are .90, .79 and .71, respectively.
Convergent validity has been demonstrated by two methods: first,
60


through behavioral ratings made by persons who knew the tested
individuals well and second, through correlations made between MBI
scores and the personal assessment of certain characteristics
expected to contribute to burnout, as well as between the scores
and measures of various outcomes expected to be related to
burnout; e.g., dissatisfaction with opportunities for personal
growth and development.
The second instrument was a fifteen item questionnaire
adapted from one developed by Spekman to operationalize French and
Raven's bases of social power typology. The questionnaire has a
high degree of face validity and Spekman reported a reliability
coefficient of .78 for the scale items used in his questionnaire
(1979, 110). It was also used by Fiorelli in a 1988 study of
power in work groups. He modified Spekman's items so that the
nouns and pronouns were applicable to his research setting. This
research used a similar modification and obtained reliability
coefficients ranging from .81 (power perceptions relative to
faculty) to .86 (power perceptions relative to peers and
deans/vice presidents).
Statistical Inquiry
The statistical inquiry was based on a correlational
analysis of categorical and ordinal (frequency) data, not
measurement (interval) data. That is, the data represent either
61





names (such as sex, reasons for taking the position, current
salary level, etc.) or comparative amounts of feeling (agree,
strongly agree, disagree, etc.) or time (never, once a week,
etc.). They do not consist of measurements such as length and
weight where differences between numbers are accurate
representations of the values represented by the numbers.
Accordingly, it was decided to use the chi-square statistic and
one way analysis of variance to test hypotheses about the
population studied. (Although multiple regression analyses were
run on the various variables in the study, none explained an
amount of variance considered to be significant.)
The conditions for use of the chi-square statistic were all
met: frequency, not measurement, data; individual answers by
subjects in the population studied were independent of each other;
expected (theoretical) frequencies were less than 5 in no more
than 20% of the cells in each table analyzed; the data were
categorized on logical bases; and the sum of expected and observed
frequencies were the same.
Chi-square has its greatest usefulness in testing for
significance of differences between groups. The statistic is
derived by obtaining the deviations of each observed frequency
from its corresponding expected frequency, squaring the
deviations, dividing each squared deviation by its expected
frequency and summing the quotients. The results are then
62


compared with a table of chi-square values for different degrees
of freedom to determine the level of significance of the
difference between the observed and expected values and hence the
probability of the difference having been obtained only from
chance alone. If the number of degrees of freedom involved were 1
and the expected frequency in one or more of the cells was less
than 10, Yate's correction for continuity would be required. This
was not the case in this study so no correction was necessary.
The intent of an analysis of variance is to decide whether
observed differences among means can be attributed to chance or
whether they are indicative of actual differences among the means
of the corresponding populations. For the analyses of variance in
this study then, the dependent variables were the three components
of burnout; Emotional Exhaustion, Depersonalization, and (a
reduced feeling of) Personal Accomplishment. The independent
variables were the various kinds of perceived power and the
organizational and personal data. The objectives in each analysis
were to determine whether differences in the means of each
department chairperson group reporting a particular level of a
burnout component were due to chance or would suggest associations
between the burnout component and the type of power or the item of
organizational or personal data involved. These results would
then support or provide evidence in opposition to the various
hypotheses posed.
63


Summary
The data gathered were all self-reported. The results
should therefore be interpreted with this limitation in mind.
Research has indicated, however, that such data can provide valid
information. In a study by Howard et al. (1991), it was found
that "self-report measures can be as good as behavioral measures
and, in some instances, may be superior." That study, at least,
encourages the assignment of a reasonable degree of credibility to
these results.
Also, in regard to representativeness, it is realized that
some self-selection bias may have occurred. It is possible that
the majority of those department chairpersons who did not respond
failed to do so due to high levels of burnout. And the majority
of those that did respond may have done so because they felt less
pressure on their time reflecting lower levels of burnout. Just
as possibly, however, many of those who responded may have done so
to express their frustrations and stress. Similarly, many of
those who did not may have felt their stress was too low to bother
filling out a survey regarding it. Accordingly, the study results
will only be indicative of the stress levels and power perceptions
of those who participated, leaving the extent of generalizability
open for argument.
As will be discussed in the final chapter, further research
would be helpful in determining the degree of representativeness
64


as well as clarifying other issues that arose in the study. Other
statistical tools could also be profitably used in analyzing the
data obtained. For the specific purposes of this study, however,
the surveys returned, the instruments used and the analyses made
of the data obtained through their use are considered adequate.

65


CHAPTER 4
FINDINGS
This chapter presents the findings of research on the
stress faced by department chairpersons in four-year institutions
of higher education. The conclusions drawn from these findings
will be discussed in Chapter 5.
The specific objectives of this study were:
1. To determine the relative degree to which academic
department chairpersons in four year institutions of higher
education in Colorado exhibit symptoms of job-related stress;
2. To determine the perceptions these individuals have
regarding the power they possess relative to their faculty, their
peers, and their deans;
3. To identify the relationships between symptoms of
job-related stress exhibited by these individuals, their power
perceptions and selected organizational/personal characteristics.
The findings related to these objectives are presented in
sequence. The first section of this chapter, "Univariate
Analyses," contains a summary of findings relating to the
66


characteristics of department chairpersons in institutions of
higher education in Colorado relative to their stress levels
(measured by their degrees of burnout); their perceptions of
power relative to their faculty, peers and deans; and their
differing situations (institutional/departmental/personal). The
following section, "Bivariate Analyses," reviews the results of
several chi-square tests and one-way analyses to burnout. The
first series of chi-square tests identifies any associations
which might exist between the various power perceptions of these
individuals and their reported stress levels. A second series
identifies the associations between organizational and personal
characteristics and the individuals' stress levels. The final
part of the section provides the results of analyses of variance
run to examine the variables identified by the chi-square tests
as evidencing a relationship to burnout. The section entitled
"Hypotheses Testing" provides findings related to the hypotheses
presented in Chapter 1. The final section, "Summary,"
summarizes the variables which primarily associate with the
symptoms of job-related stress reported by the typical department
chairperson in a four-year institution of higher education in
Colorado.
Univariate Analyses
Burnout Component Scores
The first objective, a major thrust of the study, was
67


addressed by determining scores on the three different components
of burnout for each of the 166 department chairpersons responding
to the survey. Accordingly, Emotional Exhaustion scores were
compiled as shown in Figure 4.1 from answers to questions
1,2,3,6,8,13,14,16, and 20 in the Maslach Educators Survey;
Depersonalization from questions 5,10,11,15 and 22; and Personal
Accomplishment from questions 4,7,9,12,17,18,19, and 21.
HOW OFTEN: 0 1 2 3 4 5 6
Never A few times a year or less Once a month or less A few times a month Ones a week A few times a week Every cay
HOW OFTEN
0 6 Statements:
1 EE
2. EE
3. PP
4. PA
5. DP
6. EE
7. PA
a. EE
9. PA
10. DP
11. DP
12. PA
13. EE
-* EE
. 15. DP
16. EE
5 PA
18. PA
19. PA
20. EE
21. PA
22. DP
I feel emotionally drained from my work.
I feel used up at the end of the workday.
I feel fatigued when I get up in the morning and have to face another
day on the gab.
I can easily understand hat my departmental faculty feel about things.
I feel I treat sene departmental faculty as if they were impersonal
objects.
Working with people all day is really a strain for me.
I deal very effectively with the problans of my departmental faculty.
I feel burned out frtm my work.
X feel I'm positively influencing other people's lives through my werk.
I've beoane more rallrnra toward people since I took this jcb.
X worry that this job is hardening me emotionally.
X feel very energetic.
X feel frustrated by my job.
X feel X'm working too hard an my jab.
X don't really care what happens to sane departmental faculty.
Working with people directly puts too midi stress on me.
I^can_easily create a relaxed atmosphere with my departmental
X feel exhilarated after working closely with my departmental faculty.
X have aoceuplished many wrthwhile things in this job.
X feel like X'm at the end of my rope.
In my work, X deal with emotional problems very calmly.
X feel departmental faculty blame me for sane of their problems.
Figure 4.1. Educators Survey
68


Then, following the practice used by Maslach in her study
of 635 post-secondary educators, the scores for these 166
department chairpersons were segregated into the categories of
low, average, and high as indicated in Figure 4.2. It should be
noted that burnout component scores are considered separately and
are not combined into a single, total score (Maslach and Jackson
1981, 1986; 2.)
Percent
Emotional Dsporionallzatlon Pori.
Ezhaaitlon Accomplishmont
Components
High Mil Average I .J Low
Figure 4.2. Burnout Component Scores
The reported levels of Emotional Exhaustion were spread
fairly evenly over the chairpersons surveyed, with the largest
percentage categorized as low and the least percentage as
69


average. In the case of both Depersonalization and Personal
Accomplishment, however, those with average levels dominated. Of
the three components, Personal Accomplishment made the largest
contribution to burnout, with 38% reporting that they experienced
high levels of feelings of low personal accomplishment and only
19% reporting low levels.
Hie ranges and means of the burnout component scores
reported by the 166 chairpersons responding to the survey and the
maximum possible reported scores are displayed in Figure 4.3.
(Note that a low score for personal accomplishment represents a
high contribution to burnout for that component whereas low
scores for the other components represent low contributions.)
60
50
40
30
20
10
0
Emotional Depersonalization Pers.
Exhaustion Accomplishment
Scores
s '
; l
, . M

H
. M : h
- -
.
L I. 1_£
Components
Figure 4.3. Burnout Component Ranges
70


For Emotional Exhaustion and Depersonalization, the maximum
scores reported (H) closely approached or equaled the maximum
scores possible (P). Die least Personal Accomplishment score
reported, however, shown also as H and representing a high
burnout contribution, was well above zero. That is to say, some
individuals indicated by their survey answers that they
experienced feelings nearly every day that they had lost their
spirit and interest and/or had negative attitudes toward others.
Those reporting the greatest sense of reduced personal
accomplishment, however, tended to indicate that those feelings
occurred less often than daily.
On the low stress side, some of the respondents reported
"Never" to all questions involving any loss of spirit or
interest, some to all questions involving negative attitudes
toward others, and some to nearly all those involving senses of
reduced accomplishment. This is reflected in the figure above in
the form of minimum scores (L) of zero for Emotional Exhaustion
and Depersonalization and the maximum reported score (L) for
Personal Accomplishment approaching 55, the maximum possible
score (representing a low contribution to burnout). Mean scores
fell in the average ranges (14 23 for Emotional Exhaustion, 3 -
8 for Depersonalization and 36 42 for Personal Accomplishment)
with one standard deviation including scores in both the high and
low ranges for each component. Evaluations of these data will be
found in Chapter 5.
71


Power Perceptions
The second and third objectives of the study involved the
power perceptions of these individuals; specifically, how they
perceived their influence in relation to each of the three groups
within the institutional structure with which department
chairpersons typically need to exert influence: their faculty,
their peers and their deans. The chairpersons were requested to
indicate on a questionnaire (Figure 4.4) the degree to which they
concurred with fifteen statements concerning power relationships
REF 1
EXP 2
EWD 3
COE 4
LEG 5
EXP 6
COE 7
LEG 8
EWD 9
EXP 10
REF 11
COE 12
LEG 13
REF 14
RWD 15
Faculty in my department:
Respect me and want to act in ways that merit my respect
and admiration.
Respect my competence about things in which I have more
experience.
Believe I can give special help to those who cooperate
with me.
Believe I can apply pressure on those who do not
cooperate with me.
Perceive me as having a legitimate right to expect that
my suggestions will be carried out.
Defer to my judgment in areas with which I am more familiar.
Believe I am able to make things difficult for them if
they fail to follow my advice.
Feel obligated to follow my suggestions.
Believe they can personally benefit by cooperating with me.
Believe that following my advice will result in better
decisions.
Cooperate with me because they have high regard for me
as an individual.
Believe I can penalize those who do not follow my
suggestions.
Feel they have to cooperate with me.
Cooperate with me because they wish to be identified
with me.
Perceive that cooperating with me can positively impact
on their performance.
Figure 4.4. Questionnaire on Perceived Power Relative to Faculty
72 .


that involved the actions, beliefs and perceptions of those
individuals. Hie five types of power defined by French and Raven
(reward, coercive, legitimate, referent and expert) were each
represented by three of the fifteen statements as shown, coded
RWD, 00E, IDG, REF, and EXP, respectively. Degrees of
concurrence could be: strongly agree, agree, neutral, disagree,
and strongly disagree. Analyses of these responses were made by
assigning a weight of 2 to strongly agree, 1 to agree, 0 to
eutral, -1 to disagree and -2 to strongly disagree and combining
each chairperson's answers to the three questions. Hie combined
results ranged from +485 (of a possible +996) to -426 (of a
possible -996), with 8 out of the 15 combined totals for all
chairpersons falling between +185 and -156. They were then
categorized for general findings as follows:
Greater than 300 An indication of a strongly held
perception that the power did exist (Strongly agree);
From 300 to 200 A definite perception of the existence
of the power (Same agreement);
From 200 to -200 An uncertainty as to the existence of
the power (Neutral);
From -200 to -300 A definite perception, of the
nonexistence of the power (Disagree);
Less than -300 Indication of a strongly held perception
that the power did not exist (Strongly disagree).
73


Power Perceptions bv Power Base
Figures 4.5 through 4.19 on the pages following show the
department chairpersons' perceptions in regard to the existence
of each type of power relative to a particular group.
In regard to their faculty, chairpersons perceived the
existence of both referent and expert power (strongly so in the
latter case) but were uncertain as to the existence of* reward,
coercive and legitimate power. That is, the chairpersons
believed their faculty were influenced by the respect and regard
they had for their chairpersons as well as by their belief in the
chairpersons' management competence.
Figure 4.5. Faculty
- Referent Power
Figure 4.6. Faculty
- Expert Power
74


They were, however, not sure as to whether or not their
faculty were influenced by the chairpersons' ability to reward or
punish them for their actions or by any sort of internalized
values creating an obligation to accept the chairpersons'
influence attempts because of the letters' positions.
Itaaatir tea* JfMtMl DtMpw *atr ioeair im n*rjt*a Mnwir taoiaiy kM aumm ioatr
9gtm NMl <1TW N ffwim dlKOTM 4PM ^Tllit 41K0T**
AQiMBml that pawn xliti that pow*r sziib that power axlM
Figure 4.7. Faculty Figure 4.8. Faculty
- Reward Power - Coercive Power
Figure 4.9. Faculty
- Legitimate Power
(Note in Figure 4.9 the uncertainty regarding the existence
of a legitimate ability to influence their faculty. This may
result from a widely held perception that the position is one of
first among equals where the exercise of legitimate power is not
desired by either the faculty or the chairperson. In general,
the concept of academic freedom, so key to the culture of higher
education, embodies an inherent resistance to authority.)
In relation to their peers, these chairpersons also
perceived the existence of both referent and expert power, with
the latter being the stronger of the two.
75


Aenamant Ibal povar axliti
Figure 4.10. Peers
- Referent Power



11
4
H MM wk
Itreaglr (oat* Katral DUagr** ttiooalr
ogiM agi**n&t dUagi**
Agreement that power exists
Figure 4.11. Peers
- Expert Power
They were uncertain as to the existence of reward power
relative to their peers, and they believed that they had no
Figure 4.12. Peers
- Reward Power
76


coercive or legitimate power over their peers (strongly so in the
former case).
A0MOMI tt Figure 4.13. Peers
- Coercive Power
ibnwtr ottmtm acyir
AgrHQNt Ibd powit tslftti
Figure 4.14. Peers
- legitimate Power
In regard to their deans, the chairpersons' general
perceptions of referent and expert power were similar to those
they held in regard to their peers and their faculty.
Figure 4.15. Deans
- Referent Power
Figure 4.16. Deans
- Expert Power
77


They again perceived that coercive power did not exist (but
not as strongly).
Figure 4.17. Deans
- Coercive Power
Interestingly, however, while they perceived the existence
of reward power relative to their deans, they were uncertain as
to the existence of ary legitimate power. While a logical
argument can be made for the reward perception (a quid pro quo,
perhaps), the uncertainty regarding legitimate power (as opposed
to perceiving nonexistence) is surprising. In regard to their
deans, it would be expected that the authority direction should
be clearly seen as down, not up. Also, there was a much smaller
percentage perceiving the existence of legitimate power in
relation to their peers than there was in relation to their deans
(13% vs 25%).
78


Figure 4.18. Deans
- Reward Power
Figure 4.19. Deans
- legitimate Power
Table 4.1 summarizes these power perceptions.
Table 4.1. Power Perceptions by Power Type and Subject
Referent Expert Power Tvoe Reward Leoitimate Coercive
Facultv Exists Exists Uncertain Uncertain Uncertain
Peers Exists Exists Uncertain Nonexistent Nonexistent
Deans Exists Exists Exists Uncertain Nonexistent
Department Chairperson Characteristics
To achieve the third objective, after having identified the
burnout levels and power perceptions of these individuals, it was
necessary to determine the organizational/personal
characteristics of the academic department chairpersons involved
in this research. The figures and tables in this section provide
79


details as to the characteristics of the department chairpersons
surveyed and their institutions.
The "typical" department chairpersons in four-year
institutions of higher education in Colorado responding to the
survey were 46 50 year old Caucasians. Only 3 of the 166
chairpersons responding were Asian-American, 5 were Native
Americans, 1 was Hispanic and 3 specified no ethnicity. Only 1%
were under 35 years of age.
100%
80*
60*
40*
SO*
0*
Figure 4.21. Ethnicity
These chairpersons reported to deans and had reported
leadership styles characterized as democratic (84% and 85% of the
cases, respectively). They were male and tenured. (Only 22% of
the chairpersons were female.)
Paraent
Ethnic group
Age groups
Figure 4.20. Age
80


P*ro*nt
P*rwnt
Figure 4.22. Sex
Figure 4.23. Tenure
They held terminal degrees and had been full-time faculty
members for over 19 years. (Only 9 of the chairpersons reported
less than 9 years in that category.)
Ymt
Figure 4.24. Highest Degree Figure 4.25. Full-time as
Faculty Member
81


They were full-time as department chairpersons and had
occupied the position for from one to four years.
Percent
Position Status
Percent
Figure 4.26. Current Status Figure 4.27. Time in Position
They were full professors and had been at the time of their
appointment.
Figure 4.28. Faculty Rank Figure 4.29. Faculty Rank
- Currently - When Appointed
82


Their salaries fell in the $50,000 $55,000 range with
less than 10% of their faculty receiving higher salaries.
SlOOO's
Figure 4.30, Current Salary Figure 4.31. Percent of
Faculty Paid More
Their departments were in public institutions and had been
in existence for at least 25 years.
5 25
Ysars
Figure 4.32. Institution Type Figure 4.33. Department Age
83


Their departments currently consisted of 7 12 members and
had remained within 20% of that number over the past three years.
Mrmi
Compared to Current Number
Figure 4.34. Department Size Figure 4.35. Department Size
- Currently - During last 3 Years
Their departments played what were seen as central roles in
a university and offered degrees at the BS, MS, and FhD levels.
Figure 4.36. Role Figure 4.37. Inst
In Institution Classification
Figure 4.38. Degrees
Offered by Department
84


These typical department heads viewed the availability of
institutional resources as being barely adequate and the
difficulty of obtaining noninstitutional resources as being
slightly more than moderate.
Hiont
Figure 4.39. Resources
- Institutional

Difficulty of Obtalnlna
Figure. 4.40. Resources
- Noninstitutional
They perceived that their influence, as compared to that of
their faculty, was greatest in regard to faculty promotions and
the allocation of institutional resources.
P*rc*nt
Chair's Influence compared to faculty's
Chair's Influence compared to faculty's
Figure 4.41. Influence
- Fac. Premotions
Figure 4.42. Influence
- Inst. Resources
85


Influence was also seen as significant in regard to the
determination of faculty salaries and teaching assignments.
mon mor* th kzm
Chatr'i tntlosDcs compared to faculty'!
Prcat
Chalr'i Influence compared to lacnlty'a -
Figure 4.43. Influence Figure 4.44. Influence
- Fac. Salaries - Teaching Asgmnts
Their influence over tenure was seen as somewhat less than
that over the previous areas, and that over curriculum
development and class meeting times was seen as being the least
they possessed.
Fig. 4.45. Influence Fig. 4.46. Influence Fig. 4.47. Influence
- Fac. Tenure - Curriculum - Class Times
86


The two primary reasons they gave for having became
department chairpersons were to represent the faculty to the
administration or to respond to faculty encouragement to take the
position.
Table 4.2.Reasons for Becoming a Department Chairperson
Primarv Secondary
43 Represent faculty to administration 35
42 Respond to faculty encouragement 28
19 My turn 7
18 Prevent less capable person from taking 29
11 Wanted challenge 9
9 Build program 16
8 Prepare for a higher position 11
4 No one else could/would do it 1
3 Maintain stability 1
2 Appointed by president 2
2 A one person department 1
2 Money 0
1 None 8
1 Break from teaching and research 7
1 Recruited from outside 2
0 Represent administration to faculty 9
166 166
87


In the prior table, 4.2, the number of times the phrase
"prevent less capable person from talcing position" appeared as a
primary or secondary reason is of particular interest. It
undoubtedly reflects a strong sense of self-confidence, a
self-confidence that is also indicated by the fact that the
existence of expert and referent power was perceived by these
chairpersons relative to all groups and by the generally low
degree of burnout reported for each burnout component. (Note on
page 61 that on any burnout component, the percentage of
chairpersons reporting in the High category did not exceed 40%.)
Typical chairpersons had been appointed to their positions
with faculty consultation. In fact, in 80% of the cases, the
selection process had involved either faculty consultation or
faculty election.
Table 4.3.Process by Which They Became Chairpersons
Appointed w/faculty consultation
Faculty election
76
57
Rotation
13
Appointed without faculty consultation
12
National search
8
166
88


Their primary goal for their departments, upon becoming
chairpersons, was to improve department status in their
institutions with a secondary goal of improving that status
outside their institution.
The distribution of goals is of particular interest.
Approximately 50% selected the aforementioned primary and
secondary goals of improving department status either in or
outside the institution. Relatively equal percentages
(approximately 13% each) selected "improve the curriculum,"
"improve departmental teaching", or "better distribute
resources/workload". (The "Other" goals shown in Table 4.4 were
write-ins.)
Table 4.4 Individual Chairperson's Goals for the Department
PriTnary Secondary
62 Improve dept' s status in the inst.
23 Improve the curriculum
21 Improve departmental teaching
37
28
16
19 Better distribute resources/workload 22
19 Improve dept's status outside inst.
22 Other (including: Build program,
49
14
improve working environment,
improve reputation/quality,
improve product, and none)
166
166
89


The foregoing has summarized findings related to the
characteristics of department chairpersons in institutions of
higher education in Colorado relative to their job-related stress
levels (measured by their degrees of burnout); their perceptions
of their power relative to their faculty, peers, and deans; and
their institutional/departmental/personal situations. The next
section will review the results of analyses conducted to identify
associations among these three sets of data.
Bivariate- Analyses
Power Perceptions and Stress
Chi-square tests were run on all combinations of power
perceptions and burnout components. Using a significance level
of .05, evidence was found to be sufficient to reject the
hypothesis of independence between burnout components and power
perceptions in ten instances. Six of these related to faculty,
two to peers and two to deans. The data constituting this
evidence are reported in Appendix B.
To illustrate these data, the test data for Depersonalization
and Referent Power relative to Faculty are shown In Table 4.5
(Table B.l in Appendix B).
90


Table 4.5.Perception of Referent Power Relative to Faculty
Level of Power Existence of Power
Deoerson- Perceived Uncertain or None Perceived
alization Obs./Exn. Obs./Exn. Total
Low 40/32.84 7/14.16 47
Average 62/63.59 29/27.41 91
High 14/19.57 14/ 8.43 28
All 116 50 116
Chi-square = 10.567 p < .01 df = 2
In Table 4.5, the first value in each cell represents the number
of the 166 respondents observed in that category while the second
value (following the slash) represents the number expected to be
in that category if there were independence between the two
variables. For example, the number of respondents who reported
High Depersonalization and were (statistically) expected either
to be uncertain as to the existence of referent power relative to
their faculty or to perceive none to exist was 8.43. In
actuality, 14 of the respondents were observed to be in that
category.
Ihe value of the chi-square statistic in this table was
10.567. Using a chi-square distribution table for the two
degrees of freedom represented by these data, the chances of
getting this statistic with independence existing between the
variables studied would be less than 01 (exactly 0051).
Accordingly, the hypothesis of independence in the case of these
91


two variables is rejected at a significance level of .01.
It was also noted that the group uncertain about or
perceiving no referent power to exist and reporting a High level
of Depersonalization made the largest contribution to the
chi-square statistic.
Faculty and Power. The instances where independence between
burnout components and power perceptions was rejected in the case
of faculty included all power bases except coercive. (Tables
B.1-B.7, Appendix B.)
Table 4.6.Independence Rejected Between Power Type and Burnout
Component (Faculty)
Power Tvoe Burnout Component Sianificance Level
Referent EmotionalExhaustion .0518a <.01~ <*s
Referent Depersonalization
Referent Personal Accomplishment
Expert Depersonalization <01?
Expert Personal Accomplishment <02?
Reward Personal Accomplishment <.05b
Legitimate Emotional Exhaustion <.01c
a. Although independence was not rejected in this
instance, the significance level was sufficiently close to .05 to
be of interest.
b. Among those chairpersons who were uncertain as to the
existence of the power or perceived that it did not exist, more
reported high stress than would have been predicted to do so in
the absence of a dependency relationship.
c. Among those chairpersons who perceived that the power
did not exist, more reported high stress than would have been
predicted to do so in the absence of a dependency relationship.
92


Full Text

PAGE 1

AN EXPIDRATION OF T.HE PERCEPI'IONS OF rowER AND SYMPitMS OF JOB-RElATED srnESS AM'JNG DEPARIMEl\'T rnAIRPERSONS m FOUR-YFAR mSTI'IUI'IONS OF ffiGHER EIXJCATION m OJIDRAOO by Warren A. Spaulding B.S.E, United States Military Academy, 1952 M.S.E., University of Michigan, 1956 M.S., Public Administration University of Colorado, 1976 A thesis submitted to the Faculty of the Graduate School of the University of COlorado at Denver in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Administration 1994

PAGE 2

This thesis for the Doctor of Philosophy degree by Warren A. Spaulding has been approved for the Graduate School of Public Affairs by R. Wayne Boss

PAGE 3

Spaulding, Warren A. (Ph. D. Public Administration) An Exploration of the Perceptions of Power and Symptoms of Job-Related Stress Among Department Chairpersons in Four-Year Institutions of Higher Education in Colorado. 'lhesis directed by Professor Robert W. Gage ABSTRACI' '!his study analyzed the relationship between perceptions of power, job-related stress and organizational/personal characteristics of academic department chairpersons in four-year institutions of higher education in Colorado. '!he pri.mru:y hypotheses were that a statistically significant percentage_ of these individuals would exhibit job-related stress and that job-related stress would be more prevalent among those ch.ail:persons who perceived they had inadequate power to cany out the responsibilities of their positions andjor were identified with certain organizational or personal characteristics. '!he absence of any theoretical basis for the existence of a power-stress dependency relationship for individuals in this particular managerial role gave rise to the study. '!he data were gathered by using three self-administered questionnaires. Two were previously validated questionnaires on power and the results of job-related stress (bw:nout), and one focused on organizational and personal characteristics of the position and iii

PAGE 4

the deparbnent chai.l:person. 'Ihe total target population was all department chairpersons in four-year institutions of higher education in Colorado (n = 441); the usable number of responses was 166 (35.3%). Analysis of the data, using chi-square tests and analyses of variance, resulted in the following major conclusions: Within any group of deparbnent chairpersons in institutions of higher education, a significant number will be experiencing burnout; Expert and referent power fonn the primary perceived power bases for the influence exerted by deparbnent chairpersons over their faculty, peers and deans; '!hose chaizpersons either uncertain as to their power bases, or perceiving an absence of power relative to their faculty, are more likely to experience burnout than those who do perceive such power. 'Ibis abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed ________________________ __ Robert W. Gage iv

PAGE 5

My most sincere appreciation is extended to Janet Spaulding -my wife, for the unflagging encouragement and support she provided me through the 6 years it took me to obtain this degree (as well as during the 36 years preceding those) ; to Robert W. Gage -my committee chail:person, for taking on that arduous task and providing me his wise counsel and assistance throughout; to Linda dei.eon, R. Wayne Boss, and John c. Buechner -my aJ Denver committee members, who each took the time to read and criticize constructively both my research and my writing; to R.E."D. Woolsey -my committee member from CSM and mentor, without whose encouragement, and prodding, I would never have finished this thesis; to Ira Russianoff -who helped me organize and manipulate my data so that it made sense; and to William Astle -who led me through the confusing trails of statistical to reasonable conclusions. To all these individuals and to the many department chail:persons who responded to my smvey, I give my sincerest thanks for their help along the way in my long journey toward the accomplishment of a very long time desire. v

PAGE 6

CHAPl'ER 1. 2.
PAGE 7

Power 50 summary 51 3. MEIHOOOI..OOY .52 General .52 Data Collection .54 Data Analysis .57 SUrvey Instnnnents 60 Statistical Inquil:y .61 summary .64 4. FINDmGS 66 Univariate Analyses. 67 Burnout Component Scores .67 Power Perceptions 72 Power Perceptions by Power Base. .74 Department Chairperson Characteristics .79 Bivariate Analyses. 90 Power Perceptions and stress. 90 Faculty and Power .92 Peers and Power .93 Deans and Power .94 Organizational/Personal Characteristics 95 Explanatocy Variables 96 Hypotheses Testing .101 Hypotheses on Stress 101 Hypotheses on Stress versus Power Perceptions. 103 vii

PAGE 8

5. Hypotheses on Power Perceptions Hypotheses on Organizational/Pel:-sonal Cllaracteristics and stress SUmmcuy CX>NCWSION Introduction Conclusions General Typical Cllaiipersons Stress Levels Power Perceptions stress Relationships Strengths and Weaknesses Contribution Recamrnendations Actions by Vice-Presidents and Deans F\lrther Research APPENDIX .104 .106 106 .109 .109 110 .110 111 .114 .115 117 .119 .122 123 124 .127 A. Department Chairperson Questionnaire 129 B. Chi-Square Test Results: Power Perceptions and Burnout Components 141 C. Chi Square test results: Organizational/Personal Cllaracteristics and Burnout Components 146 D. Analyses of Variance on Dependency Pairs .149 BIBLICGRAmY .154 viii

PAGE 9

CHAPI'ER 1 mrROOOCI'ION ovm:view of the study '!he deparbnent chainnan or head is the foreman in higher education-the person who sees that the work gets done. It is a difficult and ambiguous role, and so ill-defined that at many colleges no description of his duties appears on paper. Ani he is the man or woman caught in the middle Brann, 1972 When Brann likens deparbnent chail:persons to foremen, he is only partly acicurate. Certainly these academic managers do have many similarities to foremen -they are, typically, responsible for work (teaching) assignments; they evaluate their deparbnental faculty; they act as links between their faculty and the institution 1 s administration, representing at different tilnes, one group to the other; and they frequently assume their positions after sm:ving as one of that deparbnent 1 s regular teaching faculty. But they are significantly different in an important way from foremen and the great majority of non-academic managers -that difference 1

PAGE 10

is in their relationship to those they supervise. Even though in name they are the "heads" of their deparbnents, their faculty still consider them peers. The relationship has been illustrated by TUcker in his book, Cllairing the Academic Department, wherein he characterizes the academic chairperson as both a manager and a faculty colleague, an advisor and advisee, a soldier and a captain, and a drudge and a boss. The chairperson is a leader, yet is seldom given "the scepter of undisputed power." The chairperson is also first among equals, yet finds that any strong coalition of those equals can severely restrict his ability to lead (1981, 4). The collegial aspects of the position are extremely important to the exercise of power. Typically chairpersons exert leadership through the power of their ideas, not through the authority that goes with the position. They are colleagues as well as and pertlaps more so than supervisors. And this is not an easy life to live. As Robert Merton has pointed out, "(A) conception basic to sociology holds that iniividuals have multiple social roles and tend to oi:gani.ze their behavior in tenns of the structurally defined expectations assigned to each role. Evecy sociological textbook abounds with illustrations of incompatible demands made of the mu1 tisel ved person" (Merton 1957, 116) He further states that ... frequent conflict between roles is dysfunctional for the society as well as the individual" (116). 2

PAGE 11

'!he particular conflict in the chairperson role makes the latter extremely interesting from a theoretical starrlpoint. While theocy abounds on leadership and management, there exists little, if arr:t, on the exercise of influence when a fo:rmal leader is strongly recognized as and is particularly oriented towani being a peer of those led. Likert (1961), in identifying the "linking pin" function of managers, does not address it. Neither, for example, do Katz and Kahn (1966), Sayles (1979) or Bunls (1978) '!he paradoxical role of the deparbnent chairperson in particular has received only limited study and analysis. In fact, as Norton points out (1978, 1), not only is there "a dearth of literature in the field relating to the work of the academic deparbnent (chairperson) but there also appears to be no rising interest in the position despite certain evidence of erosion in this administrative role" (Norton 1978, 1). Frank Dilley, Vice Provost for Instruction, University of Delaware, has also noted that "there is so vecy little good literature on the chainnanship available" (Dilley 1972, 29). What literature there is primarily attempts to apprise the incumbent or aspiring deparbnent chairperson as to the responsibilities, roles, procedures, :powers and problems particular to this position. Examples include Bennett's Managing the Academic Deparbnent (1983) and TUcker's Chairing the Academic Department (1981). 3

PAGE 12

In the majority of cases the procedures, powers, and problems discussed in these writirgs are identical to those encountered by ,gny manager. Norton's study (1978) of department chairpersons in colleges of education, however, did add to what is known about these positions by analyzirg the time individuals typically allot to specific jobs and the degree of reward and satisfaction they obtain from various facets of those jobs. None of the literature, however, has given major consideration to the department Chairpersons' perceptions of their personal power and the stresses that may result from the paradoxical nature of their jobs, particularly in relationship to those power perceptions. And the importance of this position cannot be overemphasized. "A brilliant university or college administration with inept chairpersons cannot survive; an inept administration, with the help of brilliant chairpersons, usually can" ('1\lcker 1981, 4). One dean of faculty intel:viewed in the course of this research described the chairperson role as harder and more important than his. He characterized the job as similar to herding cats -they (the departmental faculty) always want to go in different directions and they're quite likely to bite the hand that feeds them. Department chairpersons in public institutions are, to a degree, public administrators. Also, the environments within which 4

PAGE 13

these chairpersons nn.ist function, an:i which may foster their bumout (Maslach 1982a, 118) are much affected by public administration's policies an:i practices. In particular, they are highly related to the resource allocation aspects of public administration. Inadequate resources to hire sufficient faculty to teach necesscu:y courses at appropriate class sizes an:i to provide the teaching materials and support needed by faculty, legislated faculty workloads, enrollment limits, denials of degree program requests; all of these can create strong tensions an:i significant frustrations contributing to the onset of burnout. Definitions Definitional difficulties exist with the key constructs involved in this research. '!hose difficulties will be addressed later in this dissertation. In brief, however, power, in the context of this investigation, is "the ability to produce intended change in others, to influence them so that they will be more likely to act in accordance with one's own preferences" (Bi.nlbatnn 1988, 17) Stress, for this research effort, refers to a nonspecific response of the body to any demand, particularly a demand to which the body is not adapted. In this research, jab-related stress will be operationalized as "burnout, 11 further defined as one specific psychological an:i behavioral outcome of stresses in the workplace. As Farber (1983) states, burnout is 5

PAGE 14

not a stressor nor the perception of stress but more the final step in a progression of unsuccessful attempts to cope with a variety of stress factors. Primarily for research purposes, the bumout symrome has been further divided into three components: exhaustion, the depersonalized treatment of others, and a feeling of reduced personal accamplishment. S.inply stated, the process is as follows (a more detailed model is presented in Olapter 2) Conditions in the person and conditions in the job interrelate to create job stress. This stress then creates physiological strain to which the body reacts by developing coping mechanisms to deal with the stress. 'lhese mecllanisms are exhibited in the depersonalized treatment of others and a lowered feeling of Personal accamplishment. When the strain smpasses an in:lividual's nonnal coping limits, the result is the enEm3Y deficit encampassed. by the notion of emotional exhaustion (Maslach 1982a, 24). The bmnout syndrome and its components will be diSClJSsed in more detail later in this dissertation. Objectives of the Research It is the pw:pose of this research to investigate the power perceptions and syrrptoms of job-related stress (bmnout) evidenced by deparbnent chab:persons in four-year institutions of higher education in Colorado and to detennine if any dependence 6

PAGE 15

exists between those variables and their organizational/personal characteristics. '!his research is intended to add to the general theoretical bases of both bumout and the deparbnent chairperson role by providing data on the occurrence of bumout in these inlividuals. '!he department chairpersons' perceptions of their power bases relative to their faculty, peers and administrative superiors will also be examined so as to identify what those perceptions are and how, if at all, they may explain variance in their bumout scores. Accorclingly, the objectives of this research are: 1. To detennine the relative degree to which academic department chairpersons in four year institutions of higher education in COlorado exhibit symptoms of job-related stress (bumout); 2. To detennine the Perceptions these individuals have regai'ding the power they possess relative to their faculty, their peers, and their deans; 3. To identify the relationships between symptoms of job-related. stress exhibited by these individuals, their power perceptions and selected organizational/personal characteristics. 7

PAGE 16

Hypotheses Several general hypotheses guided the investigation. The first was designed to discover whether or not the department chail:persons surveyed exhibited the adverse result of job-related stress characterized as burnout. The conditions surrourxling the role appeared to provide justification to such is the case. This led to the first hypothesis arxl its two sub-hypotheses: m. 'Blat hmnit is exhibited by a statistically significant pe.roent.age of the inli.viduals perfoDilin:.J the role of department chahpersan in four-year instibiticns of higher education m Colorado; ma. 'nle average bmx:JUt o "lonent scores of these inli.vidnal s exceed those displayed by post--seccn)acy in the results of :researdl by Masladl (1981, 3); mb: M:xe of these :inli.vidnals will scme High rather than I.c.Jw' on the various
PAGE 17

'!he second hypothesis is based on the idea that there is a relationship between the perceptions of power by these imividuals and the level of their burnout. Research on other groups has shown that burnout can be a consequence of "not perceiving control over important outcomes in one's job 11 {Maslach 1982a, 40). The second hypothesis and its three sub-hypotheses, then, are: H2. 'lhat blmcut :is m:>:m prevalent aiiDlXJ these dla:b:perscns who perceive tlley have inadequate pa.1er to pzope.rly carey cut the :resoonsibilities of their positions: H2a. SUbjects peroeiv:inj la.er levels of pa.er relative to their depaL btental faculty have higher blmcut CCIIpJl1eiit scm:es than' these who do not: H2b. SUbjects peroeiv:inj la.er levels of pa.er relative to their peers (ether department dla:b:perscns) have higher hmxJut cxmp:ment scm:es than those who do not: H2c. SUbjects peroeiv:inj la.er levels of pa.1er relative to their administrative superiors have 9

PAGE 18

higher birnout ca:opanent scores than those who do mt. '!he literature typically characterizes power as having different bases. Iri the educational area, particularly, the more coercive bases are considered undesirable. Accordingly, the objective of the third hypothesis was to investigate hOW' department chail:persons perceived the relative strength of their power bases. '!his hypothesis am its two sub-hypotheses, then, are: ID: '!bat subjects perceive they have sLtouqer pcYNe.r bases of sane types than ot:hel:s: IDa. SUbjects perceive that they have m:>re pcYNe.r due to their expertise than to the authority associated with their position: IDb. SUbjects perceive that they have :mre pcYNe.r due to their ability to :reward than to their ability to pmish. other research (Maslach 1982a, 58-63; Golembiewski and Mun.zenrider 1988, 64) shovm :robust associations between 10

PAGE 19

features of the worksite and personal characteristics, albeit without "causal arrows." It would be expected that similar associations, as well as others unique to the deparbnent C'llallperson situation, would be detrDnstrated in this study. Accordingly, the final hypothesis states, as a "null" hypothesis: H4. '!bat the organizational arxl pe.rsanal dlaracteristics of the .inli.vidnal s occupyjnJ these p:JSitions do mt cant:ribrt:e to the differences in bunx:ut CXIllpOilel1t socn:es zeported. Limitations of the Study It is to be expected that questions will be raised as to the external and internal validitY of this research. Specifically, do the conclusions drawn from questiomaires administered to these inlividuals accurately reflect the true situation existing and can those conclusions be extended to individuals occupying similar p:>Sitions in other institutions of higher education? '!he author would argue that, despite the research's correlational design and its reliance on personal perceptions, affinnative answers can be given to both questions. '!he current version of the prilnary instnnnent used in this research, the Maslach Burnout Inventory (MBI), was developed over a period of approximately eight years. '!he developers of the 11

PAGE 20

Inventory first administered it in a preliminary fo:nn, consisting of 4 7 items, to 605 people from a variety of occupations considered to have a high potential for burnout based on earlier research. A factor analysis was then applied to the data obtained from the first sample, resulting in a reduction in the rnnnber of items, first to 25 and then to the currently uSed 22. Convergent validity was established through correlation of individuals' MBI scores with indeperrlent behavioral ratings, with the presence of certain job charilcteristics expected to contribute to burnout, and with outcomes hypothesized to be related to buznout (Maslach and Jackson 1981 1986, 10). The job characteristics included the rnnnber of clients dealt with and key dimensions on the Job Diagnostic SW::vey (Hackman and Oldham 1974, 1975). Among the outcomes analyzed were difficulties with family and friends and the desire to leave the job. '!he particular fo:nn of the Maslach Burnout Inventory used in this research was MBI Fo:nn Ed, a modification of the basic MBI by changing the word "recipient" to "student." The writer further modified the i.nst.rurnent, with the pe:nnission of Consulting Psychologists Press, Inc. by changing the word "student" (in six items) to "my departmental faculty." 'lWo studies have established the validity and reliability of the MBI Fo:nn Ed. Cronbach alpha est:ilnates in each of the two studies were approximately 90 for Emotional Exhaustion and 75 12

PAGE 21

for both Depersonalization ani Personal Accomplishment. As in other such research, some self-selection bias may have existed. Even though all department chairpersons in the state of COlorado were given the opportunity ani asked to participate, irxiividuals suffering from high levels of bmnout could have been disproportionately represented either among those choosing not to do so or those choosing to do so. Also, despite same researchers' findings to the contrary, the validity of self-reported data as contrasted to that obtained through trained observation is always suspect. Finally, a comment in regard to generalizability of the findings to department chairpersons outside the state of COlorado. It has been the observation of this researcher, '\Nho has been a teacher ani administrator in four: different institutions of higher education across the country over a 21 year period, ani has had close contact with department chairpersons during that tilre (particularly as an administrator in the last 13 years) that the problems faced by department chairpersons ani their relationships with their faculty, peers ani supervisors do not differ significantly among institutions. Research to establish this observed generalizability, if such is desired, would be a logical follow-on to this research. 'llle author would argue that, if job-related stress does occur among department chairpersons in institutions of higher 13

PAGE 22

education, the results of such stress are a workplace cost those institutions cannot afford. '!his nation educates its best and brightest, at great expense to society, so they can push the frontiers of knowledge fo:tWa:rd through research and pass knowledge on to others through teaching. same of these well-educated individuals who are motivated toward, and capable of, leadership to move into the plethora of roles required of academic deparbnent chai:tpersbns. Included in these roles are: "fashioning the direction of the deparbnent, distributing burdens and resources, coping with papen10rk and dealing with conflict, 11 to name only a few (Bennett 1983, 2) SUch roles cannot be effectively accomplishEd, however, if deparbnent chail:persons exhibit the classic symptoms of job-related stress; i.e., become emotionally exhausted, treat others inappropriately, arrljor develop feelings of personal inadequacy. And, if deparbnent chairpersons do not accompli$ their roles effectively, a strong can be made that the entire institution may be adversely affected. It is believed that, as a mi.nimum, this study provides infonnation useful to those interested in improving the deparbnent chail:person environment. The detailed look this study takes at the power perceptions of deparbnent chairpersons and the results of their job-related stress fills a void 14

PAGE 23

currently existing in the literature concerning a very important role in this nation's educational system. 15

PAGE 24

OlAPI'ER 2 LITERA'IURE REVIEW Introduction TWo important cxmstructs are material to this investigation: (1) Power arrl (2) Job-related stress. While the literature is relatively rich on the subjects of stress am power, definitional difficulties exist with both concepts am complicate the problem of measuring am relating them. Also, the role of department chairperson and its paradoxical nature needs to be fully appreciated in the context of these constru.cts. Accordingly, this review of the related literature will not only explicate the constru.cts that will be studied in this investigation but will also discuss the roles required of deparbnent chairpersons am the results of studies on this group of managers. Dafinitions Unfortunately, there is no one unifonnly accepted definition of power in the social science literature. One dictionary of the social sciences lists 6 definitions of the term. 16

PAGE 25

-'!he ability to control the actions of others. -Unauthorized ability to control the actions of others. -Illegitimate influence. '!he production of intended effects. '!he ability to produce effects. -An authority of specified scope attached to an office (Reading, 1977) '!he Oxford Dictionazy of the English Ianguage devotes 15 columns to different definitions of power, under headings which include: Quality or property; Person, bcxiy or thing; -Technical tenn. Darwin Cartwright, a major writer in the field of power, notes that "anyone reading the literature on power is bound to be troubled by the absence of a generally accepted definition of power. Most authors felt carrpelled to invent one of their own" (1959, 185) In James McGregor Burns' epic book on transfonnational leadership, he titles his first chapter "'!he Power of leadership" and discusses its ''multifarious, ubiquitous, and subtle foms" but he never specifically defines it. John Kotter reports (1985, 1-2) that less than three percent of the material in 10 popular managementjorganizational behavior texts 17

PAGE 26

dealt explicitly with pc:Mer, influence and authority and same did not even mention them. '!he U.S.Anny, in its series of field manuals on leadership, skillfully avoids all reference to the tenn. Accordingly, the literature gives one a broad choice of pc:Mer definitions. General. '!he most general pc:Mer definitions are generic references either to "the ability to conunand, to apply force" (:Rue and Byars 1989, 254) or to "the basic energy needed to initiate and sustain action" (Bennis and Nanus 1985, 9). 'lhese definitions, however, could apply equally as well to the analysis of an engine or of a pipeline problem as to a problem in organizational behavior. Management theorists therefore typically define pc:Mer in a social context; their interest is in the behavior of individuals. In this context the most general definitions include "the ability to control the behavior of others" (Sites 1973, 1) "a!?influence relationship" (Stogdill 1974, 275) and "the influence on a person produced by another person, role nann or group" (French and Raven 1959, 608). More specific definitions include an intent; e.g., "intentional influence over the beliefs, emotions and behaviors of people" (Siu 1979, 31) arrl the "ability of one person to produce (consciously or unconsciously) interrled effects on the behavior or emotion of another person" (Winter 1973, 5) 18

PAGE 27

same definitions include resistance by the recipient of the power action. For example, Baron (1986, 436) states that :power can be defined as "the ability to control the actions of others to promote one's own goals even without their consent, against their will, and without their knowledge or understanding." Weber (1947, 152) defines it as "the probability that one actor within a social relationship will be in a position to carry out his own will despite resistance [from another]." In contrast, however, stogdill (1974, 275) argues that many behavioral theorists assert that :power can be exerted without intention. And Kotter includes in his definition an intividua.l's ability to resist influence, stating that :power is a ''measure of a person's potential to get others to do what he or she wants them to do, as well as to avoid being forced by others to do what he or she does not want to do ( 1985, Preface) A further problem in clearly defining power is presented by the negative attitude many have toward the concept. Individuals exercising :power or acting so as to acquire :power are often viewed negatively. "People are suspicious of a man who wants :power" (McClelland 1975, 167) In How Colleges Work, Bimbaum (1988, 17) defines power as "the ability to produce inten::led change in others, to influence them so that they will be likely to act in accordance with one's own preferences." Further, he stated that "power is 19

PAGE 28

essential to coordinate and control the activities of people in groups and universities as it is in other ol:'ga.llizations." '!his definition also incorporates the "horizontal power" concept defined by Salancik and Ffeffer (1974) in their research on the bases (or types) and use of power in decisions concerning resource allocations in a lcm]e American university. Since Birnbaum's definition includes the on intent expressed in Siu's and Winter's definitions, while avoiding the coercive connotations of Baron's and Weber's, and has been specifically used in analyzing governance of higher education, it will also be used in this research. Types and sources of power. Nineteen different types of power are listed by Richman and Fanner (197 4) referencing a study by Cohen and March (1974) on the American college president. still, there is more consensus on the different types of power and its sources than there is on its definition. Most, if not all, writers on the subject of power recognize the five power types/sources delineated by French and Raven in 1959 as "a basis for controlled research" (stogdill 1974, 275). '!hose five types will be used for this investigation of a power-stress relationship and are defined (per French and Raven 1959, 613-20) as follows: Reward power: '!he whose basis is the ability to "administer positive valences and to remove or decrease negative 20

PAGE 29

valences," the ability to reward. The strength of this type of p:JWer depends not only on the magnitude of the reward and one's desire for it but ori the strength of the perception by one person that another can and will provide the desired reward if the influence is accepted. Coercive power: The power which stems from the expectation of punishment by one person if he fails to respond correctly to an influence attempt by another. Its strength is similarly dependent on one's perceptions and desires and the magnitude of the punishment, punishment which can range from physical harm to social rejection. I.egit:i.nate p::mer: The power which is categorized as the power of position, as consisting of the right to influence; i.e. authority. It is power which stems from internalized values in one person which dictate that another has a legit:i.nate right to influence him and that he has the obligation to accept that influence. Among the bases for this type of power are cultural values, acceptance of a political or social structure, and designation by a legitimizing agent; i.e., delegation of authority to one person by a legitimizing agent accepted by another. Referent power: 'lhe power which has its basis in the desire by one person for identification (a feeling of oneness) with another. Referent power does not include pressures toward 21

PAGE 30

c:onfonnity; those are better associated with the other kinds of power. Membership in a prestigious group, however, if valued by an inlividual, would be an exanple of referent power. Expert power: '!he power which exists when one person evaluates another's expertness in relation to her own as well as against an absolute st:armrd. Its pr.imal:y influence is on one's cognitive structure and hence an associated behavioral change is nonnally categorized as having resulted from secondary social influence. '!he range of expert power is more limited than that of referent. '!he power of an expert is typically limited to those areas in which she either has or is perceived to have superior knowledge or ability. However, occasionally some "halo effect" may occur whereby individuals are assumed to have expertise in one area simply because they are highly respected for their expertise in another. In organizations such as colleges and universities, it should be expected that both faCulty and administrators will primarily rely on referent and expert power, as those types of power are less likely to cause alienation and more likely to produce canunitted participants. Even though rewards, particularly in the fonn of money, position and recognition, are important to faculty, those individuals should tend to be influenced more by internalized principles and collegial relationships than by threat of administrative sanctions (Birnbaum 1988). 22

PAGE 31

'!here are overlaps in the above distinctions; e.g., one with expert power can influence another by promising to share with the other that expert infonnation he possesses or can obtain. Also, Iasswell arxl Kaplan (1950, 97) state that "fonns of power arrl influence are agglutinative: those with some fonns tern to acquire other fonns also." In spite of this, the definitions developed by French and Raven provide a basis for analyzing the sources of an individual's power that serves as the common thread which ties together most of the enpirical research on the construct. Research on Power in General. '!he concept of power has been characterized as "perllaps the most fundamental in the whole of political science; the political process is the shaping, distribution and exercise of power" (Iasswell and Kaplan 1950, 75). As mentioned previously, James McGregor Burns (1978, 40) refers to it as "ubiquitous" arxl Siu calls it "the universal solvent of human relations" (1979, 1) .. Needless to say, then, the literature on power is extensive. A search for the word in one college library's catalog system found 5,345 references. But, while writings on the subject are voluminous, many of the authors are concerned primarily with expouming on hypotheses, assertions and personal observations. James MacGregor Burns (1978) is one of these, 23

PAGE 32

relying on historical demonstrations of power for the most part to support his theses and rarely referring to any empirical studies. 'lhi.s is not to say that the research is not extensive also; it is. '!hat research is, however, quite recent in origin. In cartwright's 1959 review of the research literature he listed only 33 references, not all experimental; in 1965 he listed 180, mostly experimental (Stogdill 1974, 277-78). And in 1992, Pfeffer listed 230 references in his look at the role of power and influence in organizations (1992). Same of the earliest instances of power research are reco\m.ted in studies and Social Power, edited by Darwin cartwright in 1959. Among these is the report by Zander, Cohen and stotland (cartwright 1959, 15) of interviews conducted to analyze "the beliefs which members of three organizations have about one another and the way in which those beliefs are detennined by an individual member's role and power." '!he compilation also includes French and Snyder's 1959 studies on leadership and int:expersonal power in small groups in the U.s. Air Force in which referent power was found to have the greatest range (had the broadest behavioral influence). At the time of its publication, Sto:Jd.ill 's classic smvey of the theory and research on leadership, Handbook of leadership: A survey of 'Iheo:cy and Research (197 4) was recognized as the 24

PAGE 33

best compilation of the research on power, covering community power structures as well as power in fonnal. organizations and experimental groups. More recently, however, Pfeffer (1992) has provided a comprehensive evaluation of the power concept which includes consideration of the different sources of power, the specific strategies through which power and :i,nfluence are used and issues of power dynamics: how power is lost, its role in ozganizational change and its positive and negative effects in ozganizations. Pfeffer's personal typology places all power sources into two one of two categories: personal attributes, including social adeptness, competency, ambition and sensitivity; and structural sources, including control over resources and position in the communication syStem of the organization. These categories essentially subsume French and Raven's categories of sources but contain more specificity. His support for his delineation of sources includes Winter's study of the power motive, need for achievement, and the affiliation-intimacy motive (1987) Halberstam's analysis of the rise of finance personnel to top leadership in large ozganizations ( 1986) and Krackhardt' s examination of the importance of one's position centrality in the frienjship network in a finn (1990) The traditional importance of hierarchial position is not neglected in Pfeffer's treatment; in this regard he refers to the classic works of Weber (194 7) Medlanic (1962) and Kanter (1977), among others. He concludes 25

PAGE 34

his evaluation with a IOOSt appropriate reminder that, while it is one thing to l.mderstand power, it is ''quite another thing to use that knowledge in the world at large" (1992, 337). '!he reader is referred to cartwright, Stogdill and Ffeffer for detailed references to research on any particular aspect of the general power constnict. Research on Power in Higher Education. In contrast to research on power in general, the literature am research on power within the area of higher education is not extensive. '!he literature tends either to address itself to a broad look at governancejadministration, typified by the work of Corson (1960), Baldridge (1971) an:i Blau (1973), or to be prescriptive concerning the specific duties of department d'lairpersons, typified by the work of Brann (1972), Dilley (1972), and Monson (1972). Corson's work is an "observation" of the govemance of colleges and mti.versities, primarily discussing the university as an "administrative exercise." He acknowledges that the initiative for a great deal of educational policy, for personnel appointment and evaluation, and for budgeting of equipment and educational facilities has shifted to deparbnents and that "the basis of the power of deparbnents lies in this initiative" (1960, 87). He further notes how IIDJ.ch the roles of deparbnent d'lairpersons vary at different institutions, from Goucher College 26

PAGE 35

where chai.l:persons "refrain from 'running' their deparbnents" to George Washington University where they are usually younger and are not aocorded the "fonnal responsibilities nonnally associated with department heads." He also refers to a study by Doyle (1953) of 33 liberal arts colleges in which 69 percent of the departmental chail:persons participated directly in budgeting fonnul.ation and faculty selection, promotion and retention (COrson, 1960}. Baldridge's case study of New York University (1971} contains virtually no mention of the power of deparbnent chail:pe:rsons. He describes that university's power as having been shared only among trustees, presidents, faculty and students, with large shifts occurring in its primal:y location over the period analyzed in the study. Blau (1973} recognizes the existence of departments as part of the institution's organizational system but prilnarily addresses the effects of their creation, abolition, size and nmnber within that system, without canunent on the ability of department chail:persons to influence those actions and characteristics. On the prescriptive side, Brann, while including a list of 29 ftmctions of a deparbnent chai.l:person from a Pennsylvania state University faculty. handbook (1972, 7-10), makes no mention of the power needed to carry out these functions. Dilley does include "fighter" in his list of eight functions for a department chairperson and mentions the "need to build up power which can be 27

PAGE 36

used to counter-balance power exerted against you" (1972, 35) but gives no details on heM to do it. Monson's training program for department (1972) includes nothing on power. Two areas of research on deparbnent chall:person power have been comucted that do relate directly to this study. Those include an analysis of the perception of power of deparbnent chall:persons by professors Winston Hill and Wendell French (1967) and analyses of power in resource allocation decisions in colleges arrl universities made by Salancik and Ffeffer (1974) and by Hackman (1985) Hill arrl French, in their 1967 study of five four year colleges, found that the greater the power of the chall:persons, the greater the professors' level of satisfaction. '!hey also found that "professors consider depart:me.nt.cil chainnen as having less influence than any other groups in the colleges, even less than the professors" (1967, 558). '!hey attributed the low power position of departmental chall:persons to a lack of influence over higher administrative groups and concluded that the chall:persons were subject to more control by professors than were the professors by the (1967, 558-59). Salancik arrl Ffeffer' s 1970 examination of a large Midwestern state university included inteJ:views with department in which they were asked to rate the power of their own subunits, ranging from "a great deal" to ''very little." They 28

PAGE 37

found that power, used most in the allocation of gradUate fellowships, was most highly correlated with the department's ability to obtain outside grants and contracts. Further, "graduate education and research were empiri.ca.lly found to be the best predictors of subunit power" (1974, 460). In an extension of that work in 1980, they asked department chairpersons at the University of Illinois and the University of California how much power they thought various departments possessed. '!here was great consistency in their responses: no deparbnent that was judged to be in the top third overall was rated by any chairperson as being lower than the top third and, similarly, no department that was judged to be in the bottom third overall. was rated by any department chairperson as being higher than the bottom third. A particularly interesting observation was that only one department chairperson asked for a definition of the power to be evciluated (Pfeffer and Moore 1980) A research-baSed theory on internal resource allocation by colleges and universities was developed by Haclanan (1985) from interviews at six institutions. Her findings supported and added to the work done previously by Salancik and Pfeffer, errphasizing the importance of centrality (core positions versus peripheral positions) in resource allocation decisions. 29

PAGE 38

Job-related Stress Definition stress has been called "one of the most imprecise tenns in the dictionacy" (Ivancevich and Matteson 1980, 8). Beehr and Newman (1978, 667) have stated that "there is no universally agreed-upon meaning [of stress] among behavioral scientists, let alone scientists in general." writers on the subject of stress consistently reference Selye (1983) who, in 1936, developed a conceptualization that still provides the keystone for understanding the stress construct. He entitled the conceptualization "The General Adaptation syndrome" (GAS). According to Selye, the GAS is ccmp:>Sed of three stages: an alann reaction in the organism in response to diverse sti.nn.U.i to which it is not adapted, a resistance stage in which the organism's physiology tries to OOlmteract the effects of the alann reaction, and a stage of exhaustion in which resistance fails (Selye 1983). stress, as he defined it, is a nonspecific response of the body to any demand, essentially stage one of the GAS. Similarly, Quick and Quick (1984, 3) have defined stress as "the generalized, patten"led, l.U'lCOnscious mobilization of the body's natural energy resources when confronted with a stressor," where a stressor is "any demand, either of a physical or psycholcx:;Jical nature, encountered in the course of living." 30

PAGE 39

Rather than attempting to deal with all the various ways in which stress and stressors can occur and the associated behaviors which can be exhibited by irrlividuals experiencing stress, a number of researchers have concentrated on the particular behaviors of persons whose work requires frequent interpersonal c::ontact of an intense psychological nature. '!he name they have applied to these behaviors is "bumout," an outcome of work-related stress. Relationship between stress and Bl.unout While the stress and bumout concepts are sometimes commingled, clear distinctions can, and should, be drawn between them. 'Ihese distinctions are demonstrated by a combination of two models: a gelieral model of Wc>rk stress (FCU'ber 1983) and a IOOdel of the bw:nout process (Gaines and Jenniar 1983). It is shown in Figure 2.1. As irrlicated in the figure, bw:nout is one specific psychological and behavioral outcome of stresses in the workplace. It is not a stressor nor the perception of stress but more the final step in a progression of unsuccessful attenpts to cope with a variety of negative stress conditions (FCU'ber 1983) Bunlout, as Fal:Der, Gaines, Jennan and others have agreed to define it, is a syndrome involving emotional exhaustion, depersonalized treatment of others, and feelings of reduced personal accomplishment (Maslach 1982) '!he unique aspect of 31

PAGE 40

Corrli. tions in Person I )ol Gen:ier Race kJe Tenure Marital status Personali' I o( I COrrlitions in Environment Relationships Tasks str\lctures Roles J ::s.,..,.. I r J Copin:J Mecbanisms -U-R-N Disen;Jagement Detachment Avoidance ,' ' ' ....... '' &.avcu.uatJ.on of Self Figure 2.1. Burnout (One specific psychological and behavioral outcome of stress in the workplace) bumout is that, .Qy definition, the stress leading to it arises exclusively from social interaction in the individual's work environment and not from relationships cx::curring outside that environment, such as with family members or social acquaintances. 32

PAGE 41

'!he burnout syndrome was initially investigated primarily axoong those in the helping professions, particularly health care providers, teachers, social and human workers, and prison and law enforcement personnel. Typically, individuals in these professions who experienced burnout had negative perceptions of their workplace processes (Tennis 1986); tended tcMard over-conuni:bnent, authoritarian personalities and idealism (Freudenberger 1977); perceived responsibility without authority or resources to accomplish the task (Emener 1982); andjor encountered problems of role conflict (Dunham 1978) and role overload (Chernis et al. 1985). '!he most widely accepted definition of burnout results from research by Maslach and Jackson (1984) on the conunon characteristics displayed by "bumed out" individuals. In their definition, burnout consists of any or all of three components. Emotional exhaustion -usually a loss of feeling, trust, interest, andjor spirit. Depersonalization -a negative attitude toward clients and coworkers. Reduced feeling of personal accomplishment -feelings of depression, failure and incompetence; a devaluation of self. Research on Burnout Research conducted by Golembiewsk and Munzenrider beginning 33

PAGE 42

in the early 1980's, which they admitte::Uy began with a mind-set biased toward "debunking" the burnout phenomenon, has shown that ''bumout is where you find it; that is, everywhere, whether at high levels of organization or low, in various dem::graphic aggregates, and so on" (1988, 12). 'lhese researchers made a major contribution toward a better understanding of the phenomenon by their development of an eight-phase model of bunlout, varying from the demonstration of lO'irl levels of all t:hree COll'p)nents by an individual to the demonstration of high levels of all t:hree. Maslach and Jackson (1981, 1986), however, while noting that this model proposes an 11 interesting way" of combining MBI scores, believe that "the validity of this or any other phase model will depend on far more extensive theorizing ani research" ( 14) Little was written about burnout before the early 1970s when, for all intents and pmposes, Freudenberger and, later, Maslach "discovered" it. '!hey also discovered it was a major CO:ncenl for many people. Same of this concern may have been merely faddish. For example, that was the time for many self-analysis best sellers, e.g., When I Say No, I Feel Guilty (Dyer 1981a) and Your Erroneous Zones (Dyer 1981b). But as research was conducted, primarily among those who worked in the helping professions -nurses, doctors, counselors, and prison administrators -the existence of the burnout syndrome was found 34

PAGE 43

to be pe.t.VaSive. Maslach found that, while personality plays a part in burnout, the syndrome is best tmderstood in tenns of job-related intet:personal stress (Maslach 1982). Chemis et al. (1985) proposed that the major causes of burnout in public sector professionals are beliefs about their jobs that lead to tmreal.istic expectations am disillusiomnent. Wcunath and Shelton's research (1976) also concluded that situational factors contril:nrt:e greatly to burnout, particularly such factors as a gap between ideals am jab realities am a lack of reinforcement from colleagues. Where such intet:personal (and intrapersonal) stresso:rs exist, evidence can frequently be seen of the emotional exhaustion, depersonalization am feelings of reduced personal acx:::atrplishment that are symptomatic of burnout. The research on burnout by O'lristina Maslach, together with Susan Jackson and A. Pines, is extensive (see Bibliography). To sununarize her findings, particularly in respect to bumout in educators: Men typically show more of one aspect of bumout and women more of another. Specifically, women score higher on Emotional Exhaustion while men are more apt to score higher on Depersonalization; Compared to whites, blacks do not bum out as much. Burnout among Asian-Americans is very similar to that among whites. Research on other etlmic minorities has been insufficient to provide a basis of comparison; 35

PAGE 44

Burnout varies inversely with age, being more prevalent with youth; '!he greatest all'Ount of burnout is found in providers who have completed college but have no postgraduate training. '!hose with postgraduate training demonstrate higher levels of EnvXional Exhaustion but score lowest on the other corrponents Masladh 1982, 58-61). Although role characteristics have been investigated for possible contribution to burnout, Clarke, in her 1991 research into bumout, fourd the results so far to be inconclusive. As she reported, Kahn (1974) developed evidence that jab satisfaction decreases in the presence of role conflict and role ambiguity. Chemiss (1985) fourd all'Ong human service professionals that role ambiguity and role overload contributed to bumout. Both Chiannonte's (1983) research with priests and Jackson's (1986) with teaChers supported role conflict as an antecedent of bumout. In a study of professional women by Pines and Kafry (1981), however, the number of roles perfonned did not promote bumout and, in fact, the variety of roles added interest and stimulation and was negatively related to burnout. Also, role ambiguity was found in a study by leiter and Meechan (1986) to raise rather than lower. feelings of personal accomplishment. Further, a study of burnout by Jayaratne and Chess ( 1984) found no contribution from role ambiguity. Clarke concluded her review 36

PAGE 45

of burnout sources by stating "the role of personal and envirornnenta.l variables in the development of bmnout is far from clear" (1991, 54) A highly cx:mprehensive integration of research on jab burnout is provided in an October 1993 review by Cordes and Dougherty (1993). Even at this late date they found that "little has been done thus far to establish bumout' s generalizability in irrlusb:y" (1993, 621). Their review thoroughly examines the construct of bumout, describes its evolution and provides an extensive overview of the literature on the antecedents and consequences of bumout, many of which are also contained in this paper's bibliography. Cordes and r::t>ugherty' s review of the bumout research caused them to make a number of propositions specifically interxied to clarify the dynamics of the antecedents of bmnout so as to assist further research. These included: 1. High levels of work and personal demands are critical detenninants of the cxnnponents of burnout; 2. The key detenninants of emotional exhaustion reflect both and personal demands placed on employees; 3. If demands from one's supez:visor conflict with the demands from clients, role conflict may occur and lead to burnout; 37

PAGE 46

4. Boundary spanners, who flmction as infonnation processors or filters between the organization and the client and act as agents in influencing the decisionmaking of the client, are caught in a difficult position when they perceive that client demands cannot or will not be met by the organization: 5. F.rrployees enter the workplace with certain expectations regarding, azoong other things, what they will be able to accomplish professionally. '!he contrast between those expectations and daily realities will influence their perceptions of personal accompliShment (1993, 631-47). '!he findings of COrdes and Daughterty will, wherever applicable, be contrasted with the conclusions of this research effort. As mentioned previously, the p:rc:cess of burnout, usually expressed as a sequencing of various burnout C0111p0nents has been conceptualized primarily by Maslach with a variation developed by Golembiewski and Munzenrider (1981, 1984). Studies by leiter and Maslach (1988) and others support Maslach 1 s model but a m.nnber of researchers in the late 1980s have used Golembiewski and Munzenrider1s phase nalel successfully in their studies. Much research has substantiated the validity of the Maslach Bumout Irwento:ry (e.g., Maslach and Jackson in 1981 and 1986, and Wolpin, Burke and Greenglass in 1991), but it should be noted that several authors (e.g. 1 Koeske and Koeske 1989 and lee and 38

PAGE 47

Ashforth 1990) have reported correlations between bm:nout components. While the primal:y research on the antecedents of bumout has been conducted by Maslach and Jackson (1981, 1984), significant research on antecedents such as role conflict and role ambiguity has been accomplished by Kahn (1978) and Schwab and Iwanicki (1982) Also, .Russell, Altineier ani Van Velzen (1987) found an association between burnout scores ani higher staff-child ratios in school classrooms (i.e., quantitative overload). Recently, the effect of job context as well as gender on stress has been irwestigated by Pretty, McCarthy ani catano (1992), providing. findings that women experienced more emotional exhaustion and depersonalization if they were nomnanagers 'While men did so if they were nanagers. '!he effects of social support on stress. has been IrOSt recently studied by Kinneyer ani Dougherty (1988) who fourxi that "social support buffers the relationship of stressors with use of adaptive coping" (COrdes ani Dougherty 1993, 633). Finally, in 1988 Kahil! grouped the consequences of bm:nout into five categories: physical, emotional, intel:personal, attitudinal and behavioral. Research involving these consequences includes work by Maslach and Jackson (1981, 1982, 1984), Burke and Deszca (1986), and Firth and Britton (1989) as 39

PAGE 48

well as others but there is little empirical evidence to date supporting these relationships. Deparbnent Chairpersons General One group of individuals who have not been irwestigated as to burnout propensity, but who appear to have work envirornnents likely to give rise to occurrence of the syndrome, are academic deparbnent (or division) chail:persons in four-year institutions of higher education. On the one hand, these individuals might logically be expected to be self-assured, enthusiastic about their work, and supportive of those with whom they are in contact. '!hey have academic credentials that are envied and admired by American society and they occupy i.nportant positions in a respected American institution: higher education. '!hey would not logically be expected to fit the profile of what Maslach terms a ''burnout-prone individual": weak, nonassertive in dealing with people, submissive, anxious, lacking in self-confidence, lacking in ambition, or fearful of irwolvement (Maslach 1982, 62). Individuals of that description would nonnally not be expected to \ seek out or accept the position of deparbnent chail:person. Deparbnent chail:persons are, however, "the men [people] in the middle." They are the linking pins between the administration, 40

PAGE 49

those responsible for operating the business of education and research, ani the faculty, those responsible for acc::oiTplishing that education ani research. As linking pins, each has a focal role, "sitting in the middle of a group of :people, with all of wham he interacts in some way in that situation" (Handy 1983, 54). '!his situation may lead, however, as described by calvin Lee, Chancellor of the University of Maryland, to a department chairperson who is 11 quite often in conflict as to whether his role is one of spokesman for his colleagues in the deparbnent or whether it is one of an administrator who 1lUlSt make the decisions not only for the welfare of his department but for the welfare of the college ani university as a whole" (in Brann and Ennnet 1972, 54-55) Interestingly, Dr. Lee did not acknowledge that chahpersons could be spokespetsons for their faculty colleagues ani make decisions in the best interests of the department as well as the college/university as a whole. One exanple of the difficult situation faced by persons in an academic hierarchy is provided by the dean at one Colorado institution included in this study. The dean was removed from his position a few years ago because, in the eyes of the department chairpersons and faculty, he failed in the perfonnance of the faculty spokesman role. His decisions were seen, at least in the eyes of the faculty, as being neither in the best interests of the faculty nor, accordingly, in the best interests of the institution as a whole. 41

PAGE 50

'!he "balancing act" required of educational administrators is a tenuous one. Admittedly, as well described by Likert (1961) and Handy (1983) 1 the "linking pin" role of managers can be expected to present ambiguity. The "first among equals" status of deparbnent chairpersons 1 however, provides the potential for aggravating that ambiguity. Movement up in the managerial hierarchy in organizations (other than on the academic side of educational institutions) brings with it new expectations and perceptions. 'lhose in positions nCM subordinate to the new leader are expected to recognize the change in that person's role. They eire expected to perceive and behave towal:d. that individual differently, acknowledging her as no longer their equal, even if only as a result of a mandate to that effect by higher management. Not so in poSt-secondary education; that difference is not so perceived by the institution's faculty nor is its ackilowledgement required by higher level administrators. 'lhe new chairperson is typically viewed by both as still a faculty member, only with additional duties. 'Ihe method of selection of chairpersons may also affect their attitudes toward the position. Sametilnes they are tmilaterally appointed by a dean after same discussion with deparbnent members. Sometimes they are selected by a canunittee c:amposed of deparbnent members, subject to veto of the dean. Another practice is to nominate by ballot, with the dean 42

PAGE 51

having authority to accept or reject the nominee. Many deparbnents practice rotation among senior faculty, with tenns varying from as little as one year to as many as five (Demerath 1967, 185). Workplace Characteristics of Department Chail:persons Department chail:persons typically see themselves as first among equals, as faculty members and academicians first and only secondarily as "linking pins" between the administration and faculty members in their departments. '!heir pri.macy responsibilities, as many perceive them, consist of enhancing their deparbnents' reputations, insuring their departments get fair shares of always scarce resources, and protecting their departments from incursions by administrators which interfere with the teaching, research, and public service their faculty want to accomplish(Woolsey et al. 1990). 'Ihe position of academic department chairperson typically encampasses many responsibilities. At the Colorado School of Mines, the Faculty H.arx3book states that: Department heads represent their faculty and are responsible for their vigor and productivity but at the same time are members of the administrative team. It is inportant for department heads and faculty members to clearly recognize this two-way role involving responsibilities to both the department and to the school as a whole. Department heads are also responsible for administering and operating their respective academic units. Consequently, department heads need to possess administrative and Personnel skills as well as to be leaders among peers in their areas of academic training. (CSM Faculty H.arx3book, 62) 43

PAGE 52

Among the tasks involved in 11administering and operating11 their respective academic units are: -Planning department curricula, policies and procedures; Detennining activities required to achieve deparbteltal goals; Recruiting 1 selecting 1 recammending 1 promoting and retaining faculty; Assuring intradepartment and department-institution communication; -Participating in establishment of new deparbnental programs; -Motivating faculty to achieve departmental goals; -Representing the department to the academic and administrative units of the institution, professional associations, accrediting agencies, government bodies, fomldations, and the business community; -Obtaining funds from institutional and other sources and allocating them within the department; Engaging in public relations functions; -Participating in academic counseling of enrolled students and making decisions in instances of departure from departmental nonns; Rec:onuneming salaries; -Coordinating cooperative programs with other departments, educational institutions, government agencies, andjor private industry (Fletcher, 1969). other authors list similar and additional responsibilities and identify no particular differences in the department chaizperson job and culture at four year institutions across the country (Doyle 1953, Norton 1978, 'l\lcker 1981). 44

PAGE 53

In view of the responsibilities listed above and the interdeperrlence of departmental and institutional objectives, it would certainly not be smprising if senior administrators in some institutions of higher education regarded the chaiJ:persons of their academic departments as subordinate administrators. 'lhese would be those individuals who viewed their department chairpersons as first line supervisors responsible for can:ying out the objectives of the institution through proper management of the faculty in their various deparbnents. Same faculty, on the other hand, could be expected to regard their deparbnent ch.ahpersons more as black sheep, as fanner members of the flock who have not exactly deserted and joined "the other side" in the classic administration-faculty conflict but whose loyalty to them, the faculty, is still somewhat suspect. In view of the nature of academe, however, it would not be expected that many faculty would view their deparbnent chairpersons as supezvisors, as individuals responsible for can:ying out specific responsibilities by managing/leading them, their faculty. Accordingly, a situation can easily obtain wherein the role conflict and role ambiguity previously discussed, can occur and possibly result in burnout. 45

PAGE 54

Research on Department 01airpersons Kathryn Martin 1 s study of department cllairpersons and deans in colleges of fine arts (1966) concluded that deparbnent chairpersons were often in the situation of fulfilling a multiplicity of roles, with higher stress levels and more probable role ambiguity and conflicts for department chairpersons than for their faculty. In 1978, Norton investigated the role of department cllairpersons in colleges of education nationally. The foreword to his report states, "'!his is indeed an area wherein little researdl has been done. 11 His findings covered factors urderlying resignations, difficulties in job responsibilities, satisfaction and related job factors, the state of the cllairperson 1 s position and the changing nature of the role, to name just a few. that: Among the 202 participants in his study, Norton found 55% held the rank of full professor at appointment; 75% were department members prior to their selection; The mean age at time of appointment was approximately 42 years of age; 42% were equally distributed between first year in the position and five to ten years; 46

PAGE 55

More than 50% cllaired a deparbnent with 6-10 full-time equivalent faculty; 50% saw themselves as academic faculty members and 46% as equally a faculty member and an administrator; over 50% gave as a pri.mal:y reason for accepting the position: "the opportunity it presented for providing leac;iership to the academic program" (Norton 1978, 2-5). '!be Norton study also found that "the adequacy of deparbnent resources" provided the highest degree of frustration for cllaiipersons. General job frustrations and negative faculty relationships were found to be the pri.mal:y factors leading to resignation. on the other hand, support from department faculty, with regard to decisiomnaking and curriculum development, received the highest ratings of significance as related to job satisfaction (Norton 1978, 21-27). Department Olairpersons and Burnout Various investigators {cresswell 1986; Singleton 1987; Saton-Spicer and Spicer 1987) have identified the position of deparbnent challperson as key in the administration of today' s institutions of higher education. The work situation of the academic deparbnent cllaiiperson, 47

PAGE 56

however, provides opportunities for many of the stresses leading to burnout: unrealistic expectations, disillusiomnent, gaps between ideals and job realities, and lack of collegial reinforcement. Also adding to the general malaise seemingly associated with the position, Wilbert McKechnie, a professor of philosophy at the University of Michigan, has described department chairpersons as 11 generally ill-prepared, inadequately supported, and more to be pitied than censured" (Brann and Ermnet 1972, 43). It would not be unlikely, then, if a department chairperson suffered from role ambiguity, role incon'patibility, role conflict, andjor role overload or underload -each of which has been argued to be a cause of bumout. '!he most recent smvey concerning department chairperson stress is that which was conducted by Gmelch and B.Jms (1993) 'lhese researchers, director and associate director, respectively, of the Center for the Study of the Deparbnent Clair at Washington state University, smveyed over eight hundred chairs from 101 doctorate granting and research universities using the Department Chair stress In:iex. Similar to previous studies and to the premise of this research, they found chairpersons to be "in a paradoxical situation, feeling double pressure to be an effective leader and productive faculty member" (1993, 259). 48

PAGE 57

'!hey stated, however, that "the occupational streSs associated with the dual administrator-faculty role of the academic deparbnent chair has not been investigated" (1993, 260). In Gmeldl and Bums' study, cha.il:persons identified the nDSt serious stressors as having too heavy a workload, obtaining program approval and financial support, keeping current in their disciplines and canplying with rules and regulations. No significant differences were found, however, in the severity of stress experienced by age, gender and discipline variables. Faculty and administrative job strain and its inpact on "quality of life" indicators have also been examined by Blackburn and others (1986), resulting in the identification of a significant association between the satisfaction of administrators with their supeJ:Visor and their job strain. Given the foregoing, it is quite possible that a department chairperson's "struggle to cope with the traditional ambiguity of the position" (Bermett 1983, 2) is an antecedent to the individual's exhibiting the burnout syndrome. If unrelieved by some action sudl as frequent rotation of faculty members through the position, the stresses whidl may result from perfonning the department cha.il:person' s role can lead to the occurrence of the burnout syndrome in these individuals. 49

PAGE 58

Power and Department Cllairpersons Bunlout is a greater risk when people feel powerless, when they have the "sense that they are at the mercy of the situation and that there is nothing to do about it" (Maslach 1982, 146) '!his can be a challperson' s lot. Chaizpersons are designated to be leaders, yet are seldom given "the scepter of undisputed power" and are first among equals, yet find that "any strong coalition of those equals can severely restrict [their] ability to lead" ('1\lcker 1981, 4) '!he collegial aspects of the position are extremely :ilnportant to the exercise of power. Typically the chaizperson "exerts leadership through the power of his ideas" (Heimler, in Brann and Emmet 1972, 198), not through the authority that goes with the position. '!hey are colleagues as well as and perhaps more so than supet:Visors. As Bermett (1983) states: "Ample opportunities exist for deparbnent chaizpersons to perceive a shortage of the power they need to can:y out their responsibilities" (1983, 146). And without power, without the ability "to produce intended change in others [their faculty, their peers, and their deans] to influence them so that they will be more likely to act in accordance with [the department chaizperson' s] own preferences, it is most unlikely that they will be able to accomplish their deparbnental objectives. As previously noted, "power is essential to coordinate and control the activities of people and 50

PAGE 59

groups in universities, as it is in other organizations" (Birnbaum 1988, 12-13). Additionally, the work envirornnent of the academic ch.aizperson alSO cOntains 1 as part Of its potential for causing stress, the "horizontal power" problems identified by Salancik ani Pieffer in their researc:h.. They defined this type of power as "the use of influence among coacting peers to obtain benefits for themselves," a necessary b1lt difficult part of the department chairperson's role if "themselves" is interpreted to include their deparbnents (Salancik ani Pieffer 1974, 453). The power to which Salancik ani Pieffer referred is the social power discussed by French and Raven ani is the subject of this study, a concept the latter authors acknowledged as pOssessing ''more intuitive appeal than empirical precision" (1959, 456) SUmmazy As discussed, the department chairpersons' constant "person in the middle" situation can easily lead to role ambiguity, role conflict ani a feeling of powerlessness. The protected (tenured) status of the faculty they nrust ''manage" and the frustrations they encormter in obtaining the resources they believe are necessacy for proper functioning of their departments can also lead to feelings of powerlessness. In all cases the result can be stress; in many, the stress may result in bm:nout bm:nout which may be highly related to a chairperson's perceptions of his or her power as well as other factors in the work envirornnent. 51

PAGE 60

aJAPl'ER 3 MElliOOOI.OGY General Burnout has been analyzed chiefly in the helping professions, among physicians, nurses, mental health workers, am police officers in particular (Ginsburg 1974 and Chernis, et al. 1985). In the educational area, bmnout has been analyzed primarily among teachers and counselors' with only general analyses of its occurrence in educational staff members (Freudenberger 1977 am carroll 1980). studies on power in a university, SUch as 5alancik am Pfeffer'S (1974) 1 Hackman'S (1985), ani Baldridge's (1971), have involved investigations only on the relationship of a deparbnent chairperson's power to affect resource allocation or on the more general issue of institutional governance. '1his investigation lOOks at both bumout am power 1 with, as its unit of analysis, academic department chairpersons in four year institutions of higher education; those individuals who are at once teachers, CO\.ll'lS9lors, staff and executives. 'lhe individuals studied occupy (or had recently occupied) the position of department (or division) chairperson in one of fourteen four-year institutions of higher education in Colorado. An analysis of the college stnlcture in Colorado identified 441 such positions at the time the study was corrlucted. Since these 52

PAGE 61

structures are always cl'lanJinJ, subsequent studies could find either IrOre or less such positions. '!he work environments of these .i.rdividuals vary widely. At the institutional level these situations can vary depending on whether the institution is privately or publicly fun:led; whether it is focused on teachi.n:J, research or a combination of the two; whether it serves a broad or a narrow conmrunity; and whether its students are residential or cammutin]. At the deparbnental level the situations can differ depending on the discipline involved (arxi, therefore, variations in funding levels and enphasis on job preparation or educational breadth) whether the deparbnent is primarily undergraduate or graduate student oriented, whether it grants degrees or serves only in a support role, and the department's age arxi size. 'Ihese differences make generalizations of the findings to specific departments suspect. But, while the explanatory effects of selected variables in those enviroriments on the occurrence of burnout were investigated, the intent of this research from the outset was to analyze those environments from primarily a power perspective, without strongly differentiatin] between the types of institutions or departments contributing to the study. Accordingly, generalizability should be based primarily on silnilar power perceptions. 53

PAGE 62

The :investigation was quasi-experimental. '!hat is, it :involved an experiment that had a treatment (a possible cause), outcame measures (possible effects of the treatment) and experimental units, but it did not use random assigrnnents to create the cx:nnparisons from which the treatment which causes change is inferred (Cook and campbell 1979, 295-98). Its results were exploratory and :investigative only, :involving the absence of irxiependence, not the inference of causation. Data were collected from the chairpersons mentioned by use of previously tested and written questionnaires, including personal and organizational questions as well as questions about power perceptions and burnout. Among the personal and organizational characteristics analyzed were age and gender; reasons for taking the position; whether the chairperson's power includes assigning courses to be taught; and wnether the chairperson recarmnends tenure, promotion and salary increases, to name a few. The intent of the questions on power perceptions and burnout was to obtain data on subjects' perceptions of their stress levels and of their influence on their faculty, their peers and their deans. Data Collection '!he folder in Appendix A was mailed to all identifiable department chairpersons in four-year institutions of higher 54

PAGE 63

education in COlorado. '!he folder consisted of an explanatory letter requesting their participation in a research study, thirty-one organizational/personal data questions, the Maslach Burnout Inventory Educators survey (twenty-two questions)' and three fifteen-item questionnaires addressing their power perceptions. '!his choice of a research sample was made for a number of reasons. A1I1on;J these was that the author had direct contact with eac::h institution through his state professional association (Rocky Motmtain Association of COllege Registrars and Admissions Officers), that they provided a sufficiently large group for statistical significance, and that they included rural and w:ban as well as large and small institutions. While the data were not gathered prior to the impact on state budgeting resulting from the passage of Alnerrlment 1 to the state constitution in fall 1992, that passage probably only exacel:Dated the resource problem ever present in any educational institution. Letters were also sent to deans/directors of the department chairpersons, asking them to encourage their chairpersons to complete and retum the surveys. Four hundred and forty-one folders were sent out and one hundred and sixty-six usable replies were received -a 35.3% response rate. '!he first questionnaire asked for biographical and organizational data. Reasons for becoming a .department chairperson. 55

PAGE 64

'!he irrlividual 1 s goals for the department. '!he process by which hejshe became a department chairperson., '!he type, age, size, stability and role of the deparbnent. '!he levels of the degree programs offered. '!he availability of resources to the department. '!he degree of influence on various institutional actions affecting faculty. '!he irrlividuals1 age, sex, ethnicity, highest degree held, years in position, years on the faculty, academic rank and salary (both in dollars and in relation.to salarieS of faculty in the department). '!he second questionnaire was entitled "Educators SUrvey" and consisted of the Maslach Bw:nout Inventocy adapted to education. It contained 22 statements of job-related feelings and asked haw often, if at all, the irrlividual had had this feeling. Forced choices for answers ranged from "never" or "a few times a year or less" to "evecy day. 11 '!he third questionnaire contained 15 statements as to the department chail:person1 s perceptions concerni.ng the attitude of deparbnental faculty regarding the chail:person 1 s ability to influence their behavior. Each statement was related to one of the five bases of pcMer defined in the French and Raven 56

PAGE 65

typology. For each statement the deparbnent chairpersons were asked to indicate the degree to which they concurred with the statement. Five choices of degree were offered, ranging from "strongly agree" to "strongly disagree." Exanples of the statements included "Respect me and want to act in ways to merit my respect and admiration" and "COo_perate with me because they wish to be identified with me. 11 '!he fourth ani fifth questiormaires were identical to the third, except the statements in the fourth related to the attitudes of other department (division) Chairpersons in their college and in the fifth to the attitudes of the deanjvice president or other individual to whom the chairperson was responsible. All responses were coded with numerical values: 0 to 6 for the Educators SUrvey, 1 to 5 for the power perception statements and from 1 t6 a variety of numbers for the o:rganizational and biographical data. rata Analvsis '!he survey data were entered into a Minitab program on a VAX 4500. Irrlividual scores on the various items in the Educators SUrvey were then combined appropriately to provide measurements of Emotional Exhaustion, Depersonalization and (a reduced feeling of) Personal Accamplishment. Responses to 57

PAGE 66

statements in the power questionnaires were added together in groups of three, detennined by which of the French and Raven power bases they concerned. '!he resulting values, ranging from 3 to 15, represented the degree to which each power base was perceived to exist. Preliminary chi -square tests were then nm on contingency tables made up of the data in the Minitab table. 'lhese revealed a need to combine some of the data so as to avoid discontinuities in the distribution of the expected frequencies obtained, thereby negating the need. for applying Yates' correction for continuity and improving the value of the chi -square tests. For a portion of the statistical inquiry, the power data were aggregated into a t:hree-stclge distribution: perceived to exist, neutral, and perceived not to exist. '!his distribution was made by assigning the first state to the case where the power type was distinguished with a value of 3 through 6, the second 7 through 11, and the last 12 through 15. In the case of bunlout data, Golembiewski, Munzenrider and carter initially used median scores to distinguish High from I..oN bunlout scores. In subsequent c:anprrisons of two different sites, they used percentile rankings, categorizing scores above the median of each component as High and below as IDw (1973, 28). Maslach, however, prefers dividing the nonnative distribution within a particular sample into thirds: High, 58

PAGE 67

Average and Il:M. To categorize the department chairpersons surveyed .in this study, the burnout scores obtained by Maslach for 635 post-secondal:y educators were used as a baseline: Table 3 .1. Maslach 1 s Bl1rnoUt Scores for 635 Educators High Average Il:M Emotional Exhaustion >23 14 -23 <14 Depersonalization > 8 3 -8 < 3 Personal .Acc:an'plishment <36 36 42 >42 (Maslach and Jackson 1981 1986, MBI Manual, 3) Deperx:ling on the score of each chair, then, for each camponent of bumout, it was translated .into a 3 for High, a 2 for Average, and a 1 for Il:M. Aggregation of organizational/personal data varied between questions, depending on the breadth of the answers. Written-in personal reasons (primacy and secondal:y) for taking the position were collapsed into six codes and combined with the six choices provided for check-off, with the value of 12 representing "other. 11 'Ihe department chairpersons' perceived levels of influence, campared to the levels exerted by InOSt of the department 1 s faculty, were combined into three: much more, somewhat more, and approximately the same or less. leadership style ratings of 1-3, representing the autocratic end of the sPectrum, were combined into the value 3 while ratings of 7-9, 59

PAGE 68

the lai ssez-faire end, were combined into a 7. Like compressions were necesscuy for the availability of institutional and noninstitutional resources questions. Ethnic minorities had to be combined into "non-caucasian" due to the few occurrences of in:lividuals in etlmic groups other than caucasian in the study. Because nearly all the chah'persons had more than 11 years as full-time faculty members, responses to the question on this area in the 1-4 and 5-7 year ranges were combined with the 8-11 year responses to give a "less than 12 years" categocy. other such aggregations were made as necessary to create chi -square cells in which the expected frequencies were less than 5 in no more than 20% of the cases (per MINITAB Harrlbook 1985, 27 4) Smvey Instrtnnents The two instruments used to analyze power perceptions and bmnout had been previously used in other studies and thoroughly validated by their developers 0 one, the Maslach Bun'lout Inventocy (MBI) was a twenty-two item measure of the three aspects of the bun1out syndrome: emotional exhaustion, depersonalization and reduced feeling of personal accomplishment. '!he reliability coefficients that have been demonstrated for the burnout scales of emotional exhaustion, depersonalization and personal accomplishment are .90, 79 and 71, respectively. Convergent validity has been demonstrated by two methods: first, 60

PAGE 69

through behavioral ratings made by persons who knew the tested Wividuals well and second, through correlations made between MBI scores and the personal assessment of certain characteristics expected to contribute to bumout, as well as between the scores and measures of various outcomes expected to be related to bunlout; e.g. dissatisfaction with opportunities for personal growth and developnient. The second instnnnent was a fifteen item questionnaire adapted from one developed by Spekman to operationalize French and Raven's bases of social power typology. The questionnaire has a high degree of face validity and Spekman reported a reliability coefficient of 78 for the scale items used in his questionnaire (1979, 110) It was also used by Fiorelli in a 1988 study of power in work groups. He m::xlified Spekman's items so that the nouns and pronouns were applicable to his research setting. This reSearch used a similar modification and obtained reliability coefficients ranging from 81 (power perceptions relative to faculty) to .86 (power perceptions relative to peers and deans/vice presidents) Statistical Inquiry '!he statistical inquiry was based on a correlational analysis of categorical and o:rdinal (frequency) data, not measurement (interval) data. That is, the data represent either 61

PAGE 70

names (such as sex, reasons for taking the position, current salary level, etc.) or c:xmparative amounts of feeling (agree, strongly agree, disagree, etc.) or tilne (never, once a week, etc.) '!hey do not consist of measurements such as length and weight where differences between rnnnbers are accurate representations of the values represented by the rnnnbers. Accordingly, it was decided to use the chi -square statistic and one way analysis of variance to test hypotheses about the population studied. (Although multiple regression analyses were :run on the various variables in the study, none explained an amount of variance considered to be significant.) for use of the chi-square statistic were all met: frequency, not measurement, data; individual answers by subjects in the population stalled were independent of each other; expected (theoretical) frequencies were less than 5 in no more than 20% of the cells in each table analyzed; the data were categorized on logical bases; and the sum of expected and obsel:ved frequencies were the same. Chi-square has its greatest usefulness in testing for significance of differences between groups. The statistic is derived by obtaining the deviations of each obsel:ved frequency from .its corresponding expected frequency, squaring the deviations, dividing each squared deviation by its expected frequency and sunnning the quotients. The results are then 62

PAGE 71

ccanpared with a table of chi-square values for different degrees of freedom to detennine the level of significance of the difference between the observed and expected values and hence the probability of the difference having been obtained only from chance alone. If the number of degrees of freedom involved were 1 and the expected frequency in one or toore of the cells was less than 10, Yate's correction for continuity would be required. '!his was not the case in this study so no correction was necessaey. '!he intent of an analysis of variance is to dec:ide whether observed differences among means can be attributed to chance or whether they are indicative of actual differences among the means of the corresponding populations. For the analyses of variance in this study then, the dependent variables were the three components of bumout; Emotional Exhaustion, Depersonalization, and (a reduced feeling of) Personal Accomplishment. '!he independent variables were the various kinds of perceived power and the organizational and personal data. '!he objectives in each analysis were to detennine whether differences in the means of each deparbnent dlai.J:person group reporting a particular level of a bumout component were due to chance or would suggest associations between the burnout component and the type of power or the item of organizational or personal data involved. '1hese results would then support or provide evidence in opposition to the various hypotheses posed. 63

PAGE 72

SUrrnnary 'lhe data gathered were all self-reported. '!he results should therefore be interpreted with this limitation in mind. Research has indicated, however, that such data can provide valid infonration. In a study by Howard et al. (1991), it was found that "self-report measures can be as good as behavioral measures ani, in some instances, may be superior." That study, at least, encourages the assignment of a reasonable degree of credibility to these results. Also, in regard to representativeness, it is realized that some self-selection bias may have occurred. It is possible that the majority of those deparbnent chail:persons who did not respond failed to do so due to high levels of burnout. And the majority of those that did respond may have done so because they felt less pressure on their time -reflecting lower levels of bumout. Just as however, many of those who responded may have done so to express their frustrations and stress. Similarly, many of those who did not may have felt their stress was too low to bother filling out a sw:vey regarding it. Acco:rdingly, the study results will only be indicative of the stress levels and power perceptions of those who participated, leaving the extent of generalizability open for argument. As will be discussed in the final chapter, further research would be helpful in detennining the degree of representativeness 64

PAGE 73

as well as clarifying other issues that arose in the study. other statistical tools could also be profitably used in analyzing the data obtained. For the specific purposes of this study, however, the surveys returned, the instruments used and the analyses made of the data obtained through their use are considered adequate. 65

PAGE 74

aJAPl'ER 4 FINDINGS '!his chapter presents the findings of research on the stress faced by deparbnent chail:persons in four-year institutions of higher education. '!he conclusions drawn from these findings will be discussed in Qlapter 5. '!he specific objectives of this study were: 1. To determine the relative degree to which academic department chail:persons in four year institutions of higher education in Colorado exhibit of job-related stress; 2. To determine the perceptions these individuals have regarding the power they possess relative to their faculty, their peers, and their deans; 3. To identify the relationships between of job-related stress exhibited by these individuals, their power perceptions and selected organizational/personal characteristics. '!he findings related to these objectives are presented in sequence. '!he first section of this chapter, "Univariate Analyses, 11 contains a summaxy of findings relating to the 66

PAGE 75

characteristics of department chaiipersons in institutions of higher education in Colorado relative to their stress levels (measured by their degrees of burnout); their perceptions of pc::Mer relative to their faculty, peers and deans; and their differing situations ( institutionaljdeparbnental/personal) The follOW'ing section, "Bivariate Analyses," reviews the results of several chi -square tests and one-way analyses to bun1out. The first series of chi-square tests identifies any associations which might exist between the various power perceptions of these in:lividuals and their reported .stress levels. A second series identifies the associations between 01:ganizational and personal characteristics and the in:lividuals' stress levels. The final part of the section provides the results of analyses of variance nm to examine the variables identified by the chi -square tests as evidencing a relationship to burnout. The section entitled ''Hypotheses Testing" provideS findings related to the hypotheses presented in Olapter 1. The final section, "SUmmal:y, 11 smmnarizes the variables which primarily associate with the synptoms of job-related stress reported by the typical deparbnent chairperson in a four-year institution of higher education in Colorado. Univariate Analyses Bw:nout Component Scores The first objective, a major thrust of the study, was 67

PAGE 76

addressed by dete:rmini.ng scores on the three different c::onponents of burnout for each of the 166 department chairpersons responding to the sw:vey. Accordingly, Emotional Exhaustion scores were compiled as shown in Figure 4 .1 from answers to questions 1,2,3,6,8,13,14,16, ani 20 in the Maslach Educators SUrvey; Depersonalization from questions 5, 10, 11,15 and 22: and Personal AcoampliShment from questions 4,7,9,12,17,18,19, and 21. HOW OFTEN: HOW OFTEN 0-6 EE .... __ 3 .....E.E....,__ 4._P_A_ s._o_P_ 6._E_E_ 7,_P_A_ a.__!!_ 10._D_P_ u._o_P_ ,13 ...!!__ : 14 ...,!!_ 15._D_P_ 16 ....!!__ 17._P_A_ 18._P_A_ 19._P_A_ 20 ....!!__ 21._P_A_ 22 ...E!__ 0 Never A lew times a year or less Statements: 2 Once a month or less 3 A lew times a month I feel E!I!Ctionally drained. fran wrlc I feel used up at the end of the wrkday. 4 Once a week 5 A lew times a week 6 E"ery cay I feel when I qet up in the mornin; am have to face ar:ct.her day on the J .ob. I can easily un:ierstancl hew departlpantal faallty feel about t.'l.irqs. I I treat sc:me depart:nental faallty as i! they were obJects. WOrkin; with pelq)le all day is really a strain for 112. I c!eal very effectively with the problems departmental fac:Uty. I feel Curnecl cut fran wrlc. I feel I'm positively influencil'Jl other pecple's rrrt w'Crk. I've l:lec::anB more callous towartl pelq)le sin:a I took this job I worry that this job is hardenin; 112 euctionally. I feel very erm;etic. I feel trustrata::l job. I feel I'm wcrlcin; too hard job I dcn't really care what haR;lens to sane departmental faculty. Worltin; with pecple clirer:::tJ.y puts too llllCh stress an 112. I can easily c:reata a relaxed depart:nental faallty. I feel exhilarata::l after wrltin; closely departmental faculty. I have acc:ai1Plishai many wrthwhile things in this job. I feel like I m at the end of rrrt rcpe. rn work, I c!eal with e=tional problems very c:alml.y. I feel departmental faall ty blam me for sc:me of their probla::s. Figure 4 .1. Educators SUrvey 68

PAGE 77

'!hen, following the practice used by Maslach in her study of 635 post-secondaey educators, the scores for these 166 department chaizpersons were segregated into the categories of lCM, average, an:i high as imicated in Figure 4. 2. It should be noted that bun1out component scores are considered separately an:i are not combined into a single, total score (Maslach and Jackson 1981, 1986; 2.) P 60..so Emotional Depenonallzatlcn Pen. Bxhauatlon Accomplishment Components -High -Average D Low Figure 4. 2. Burnout component SCOres '!he reported levels of Emotional Exhaustion were spread fairly evenly over the chaizpersons sw::veyed, with the largest percentage categorized as low and the least percentage as 69

PAGE 78

average. In the case of both Depersonalization and Personal Accamplishment, however, those with average levels dominated. Of the three components, Personal Accomplishment made the largest contribution to burnout, with 38% reporting that they experienced high levels of feelings of low personal acc::ozrplishment and only 19% reporting low levels. '!he ranges and means of the burnout component scores reported by the 166 chairpersons responding to the survey and the max:i.mum possible reporterl scores are displayed in Figure 4. 3. (Note that a low score for personal accomplishment represents a high contribution to burnout for that component whereas low scores for the other components represent low contributions. ) Scores 50 L 40 M 30 ...... gP .. . H 20 M K 10 M EmotlODCll Oepenonallzallon Pen. EzhCIUUion Accomplishment Components Figure 4. 3. Bumcut Component Ranges 70

PAGE 79

For Emotional Exhaustion and De:personalization, the maximum scores reported (H) closely approached or equaled the maximum scores possible (P) 'lhe least Personal Ac:carrq;:>lishment score reported, however, shown also as H and representing a high bn:nout contribution, was well above zero. '!hat is to say, same in:tividuals in:ticated by their smvey answers that they experienced feelings nearly every day that they had lost their spirit and interest andjor had negative attitudes toward others. '!hose reporting the greatest sense of reduced :personal acc:cmplislnnent, however, tended to indicate that those feelings occurred less often than daily. on the low stress side, same of the respondents reported ''Never11 to all questions irwolving any loss of spirit or interest, same to all questions irwolving negative attitudes toward others, and some to nearly all those irwolving senses of reduced accomplishment. '!his is reflected in the figure above in the fonn of minimum scores (L) of for Emotional Exhaustion and Depersonalization and the maximum reported score (L) for Personal .Accanplishment approaching 55, the maximum possible score (representing a low contribution to bw:nout). Mean scores fell in the average ranges (14 23 for Emotional Exhaustion, 3 8 for De:personalization and 36 42 for Personal Accomplishment) with one standard deviation including scores in both the high and low ranges for each component. Evaluations of these data will be found in Cllapter 5. 71

PAGE 80

Power Perceptions '!he secorrl and third objectives of the study irwol ved the p:JWer perceptions of these individuals; specifically, how they perceived their influence in relation to each of the three groups within the institutional structure with which deparbnent chairpersons typically need to exert influence: their faculty, their peers and their deans. The chahpersons were requested to indicate on a questionnaire (Figure 4. 4) the degree to which they concurred with fifteen statements concerning power relationships REF 1. EXP 2. I&ID 3. CDE 4. ux; s. EXP 6. CDE 7. ux; 8. I&ID 9. EXP 10. REF 11. CDE 12. ux; 13. REF 14. I&ID 15. Faculty in nr:1 deparbnent: Respect me and want to act in ways that merit nr:1 respect and admiration. Respect nr:1 competence about things in which I have ll\Ore experience. Believe I can give speeial help to those 'Who cooperate with me. Believe I can apply pressure on those who do not cooperate with me. Perceive me as having a legitimate right to expect that nr:1 suggestions will be carried out. Defer to nr:1 judgment in areas with which I am ll\Ore familiar. Believe I am able to make things difficult for them if they fail to follCYN nr:1 advice. Feel obligated to follCYN nr:1 suggestions. Believe they can personally benefit by cooperating with me. Believe that follCYNing nr:1 advice will result in better decisions. COOperate with me becauSe they have high regard for me as an individual. Believe I can penalize those 'Who do not follow nr:1 suggestions. Feel they have to cooperate with me. Cooperate with me because they wish to be identified with me. Perceive that cooperating with me can positively impact on their performance. Figure 4.4. Questionnaire on Perceived Power Relative to Faculty 72

PAGE 81

that involved the actions, beliefs and perceptions of those inlividuals. '!he five types of power defined by French and Raven (rewaro, coercive, legitimate, referent and expert) were each represented by three of the fifteen statements -as shown, coded RWD, CDE, Im, REF, and EXP, respectively. Degrees of concurrence could be: strongly agree, agree, neutral, disagree, and strongly disagree. Analyses of these responses were made by assigning a weight of 2 to strongly agree, 1 to agree, 0 to eutral, -1 to disagree and -2 to strongly disagree and combining each chairperson's answers to the three questions. '!he combined results ranged fram +485 (of a possible +996) to -426 (of a possible -996), with 8 out of the 15 combined totals for all chaiJ:persons falling between + 185 and -156. '!hey were then categorized for general findings as follows: Greater than 300 -An indication of a strongly held perception that the power did exist (Strongly agree) ; From 300 to 200 -A definite perception of the existence of the power (Same agreement) ; From 200 to -200 -An uncertainty as to the existence of the power (Neutral) ; From -200 to -300 -A definite perception. of the nonexistence of the power (Disagree); less than -300 -Indication of a strongly held perception that the power did not exist (Strongly disagree). 73

PAGE 82

Power Perceptions by Power Base Figures 4. 5 through 4 .19 on the pages following show the deparbnent chail:persons' perceptions in regard to the existence of each type of power relative to a particular group. In regard to their faculty, cbail:persons perceived the existence of both referent arrl expert power (strongly so in the latter case) but were uncertain as to the existence o'frewam, coercive and legitimate power. '!hat is, the chail:persons believed their faculty were influenced by the respect and regard they had for their cbail:persons as well as by their belief in the chairpersons' management COII'petence. .... 11' ... --..... 11' _, "-p-a1ltl Figure 4. 5. Faculty -Referent Power IIGIY to-Mnl111l D....,_ .fDDGif __ -AG ... IMDI tbat power ftllll Figure 4. 6. Faculty Expert Power 74

PAGE 83

'!hey were, however, not sure as to whether or not their faculty were influenced by the chairpersons' ability to reward or punish them for their actions or by air:f sort of intemalized values creating an obligation to accept the chairpersons' influence attempts :because of the latters' positions. Figure 4. 7. Faculty Reward Power lfiD .. IJ lo-Bnta D"-"-aiD.IT -AglweJMDI thai 'POW.f ftllll Figure 4. 8. Faculty -Coercive Power lfiDa.GlT lo-Wnt111l DiMIIJ'M -Agtwemeot thai J)ow.r mall Figure 4. 9. Faculty Legitimate Power (Note in Figure 4. 9 the uncertainty regarding the existence of a legitimate ability to influence their faculty. '!his may result from a widely held perception that the position is one of first among equals where the exercise of legitimate power is not desired by either the faculty or the chairperson. In general, the concept of academic freedom, so key to the culture of higher education, embc:xlies an inherent resistance to authority. ) In relation to their peers, these chairpersonsalso perceived the existence of both referent and expert power, with the latter being the stronger of the two. 75

PAGE 84

lllo .. IY ... D .,IIGir -AGJwemeDt tbal pow.r abll Figure 4.10. Peers Referent Power 60S,------------------------, Some Keutral DllarJrM ltroDQ"lf _.. a;r .. mnl ClllaCP'" Aqreemenl lbal power ftlala Figure 4.11. Peers Expert Power '!hey were uncertain as to the existence of reward power relative to their peers, and they believed that they had no Figure 4.12. Peers Reward Power 76

PAGE 85

coercive or legitimate power over their peers (strongly so in the fonner case) __ ., -.... If -lllal.,_-Figure 4 .13. Peers -Coercive Power thai pawwr eiu111 Figure 4.14. Peers -Legitimate Power In regard to their deans, the chail:persons' general perceptions of referent arrl expert power were similar to those they held in regard to their peers and their faculty Figure 4.15. Deans -Referent Power .,, ,,, ..... tNJ. aa.,... .,,.,,, ..... ......... 4 .... ... Figure 4.16. Deans Expert Power 77

PAGE 86

'!hey again perceived that coercive power did not exist (but not as strongly). IDo .. tr ... D -4-thai.,_ allll Figure Deans -Coercive Power Interestingly, hOW'ever, while they perceived the existence of reward power relative to their deans, they were uncertain as to the existence of any legitimate power. While a logical argument cal?be made for the reward perception (a quid pro auo, pertlaps), the uncertainty regarding legitimate power (as opposed to perceiving nonexistence) is sm:prising. In regard to their deans, it would be expected that the authority direction should be clearly seen as dCY.m, not up. Also, there was a nn.1ch smaller percentage perceiving the existence of legitimate power in relation to their peers than there was in relation to their deans (13% vs 25%). 78

PAGE 87

ltii .. IT M-..... D&MIIW l? -4-Figure 4.18. Deans Reward l?ower AqrHmeal tllal atall Figure 4.19. Deans legitimate Power Table 4 .1 summarizes these power perceptions. Table 4 .1. l?ower Perceptions by l?ower Type and SUbject Power Tvoe Referent Expert Reward Isait:ilnate Coercive Faculty Exists Exists Uncertain Uncertain Uncertain Peers Exists Exists Uncertain Nonexistent Nonexistent Deans Exists Exists Exists Uncertain Nonexistent Department Chairperson Characteristics To achieve the third objective, after having identified the burnout levels and power perceptions of these individuals, it was necessary to detennine the organizational/personal characteristics of the academic deparbnent chairpersons involved in this research. '!he figures and tables in this section provide 79

PAGE 88

details as to the characteristics of the deparbnent cllairpersons sw:veyed and their institutions. 'Ihe "typical" deparbnent cllairpersons in four-year institutions of higher education in Colorado respond:in;J to the sw:vey were 46 -so year old caucasians. Only 3 of the 166 chaiJ:persons responding were Asian-American, 5 were Native Americans, 1 was Hispanic and 3 specified no ethnicity. Only 1% were under 35 years of age. p"""""-3BIIr-'-....;,_::;-----------, Age groups Figure 4. 20. Age Ethnic group Figure 4. 21. Etlmicity 'lhese cllairpersons reported to deans and had reported leadership styles characterized as derocratic (84% and 85% of the cases, x:espectively) '!hey were male and tenured. (Only 22% of the chaizpersons were femal!! ) 80

PAGE 89

100 ..... ,..;;..;.'";;.;;".:...' ------,------., Peroent 100 .. ..-'-.;_:_-------------, Sex Tenure Figure 4.22. sex Figure 4.23. '!hey held terminal. degrees and had been full-time faculty members for over 19 years. (Only 9 of the chairpersons reported less than 9 years in that categoi:y.} BSIIJA MSIMA Ph.D/Ed.D Degree Lewl Figure 4. 24. Highest Degree ..... Figure 4.25. FUll-time as Faculty Member 81

PAGE 90

I '!hey were full-time as deparbnent chairpersons and had occupied the position for from one to four years. Position Status Figure 4. 26. CUrrent status Peraent 3B!Ir------------r Years Figure 4. 27. Time in Position 'Dley were full professors and had been at the time of their appointlnent. ----Ia -... --F..utrAu* Figure 4.28. Faculty Rank -CUrrently ----I Gil 11 -""'-Figure 4. 29. Faculty Rank -When Appointed 82

PAGE 91

'lheir salaries fell in the $501 ooo -$551 ooo range with less than 10% of their faculty receiving higher salaries. ----------------, 2"i-! ------------i UfWitt .. $1,000'1 a .. .. Figure 4. 30 CUrrent salcu:y "''""' Ncn 1-1011o oo" Percent of faculty pald more than chalr Figure 4.31. Percent of Faculty Paid More 'lheir departments were in public institutions and had been in existence for at least 25 years ,.n:eat _________ __, .. l'CeDI Figure 4. 32. Institution Type Figure 4. 33. Department Age 83

PAGE 92

'!heir departments currently consisted of 7 12 memberS and had remained within 20% of that number over the past three years. ,. ..... 40S 2SS t-I.. ISS IDS J:\1 l\l as le r-11 11-W l .. U 26-al Mte taall Number of Pull-Time Pax:ulty Figure 4.34. Department Size Olrrently IDGSr .. :::..,..:;::._l _________ -, 2ft )2DS )'lOS Compcmld to cunent Number Figure 4.35. Department Size -During Iast 3 Years '!heir departments played what were seen as central roles in a university and offered degrees at the :as, MS, and PhD levels. Figure 4.36. Role In Institution kDOOI Class!!icalion Figure 4.37. Inst. Classification 84 Figure 4.38. Degrees Offered by Department

PAGE 93

'lhese typical deparbnent heads viewed the availability of institutional resources as being barely adequate and the difficulty of obtaining noninstitutional resources as being slightly more than moderate ...... Mequcz:y "'"' Figure 4. 39. Resources Institutional DUIIc:alty 01 Obla1nlnQ' LJtUe Figure. 4.40. Resources -Noninstitutional '!hey perceived that their influence, as compared to that of their faculty, was greatest in regard to faculty promotions and the allocation of institutional resources. Percea.t 1011....--------------, Figure 4. 41. Influence -Fac. Promotions .. tblt tame Chair's influence compantd to lacalty's Figure 4. 42. Influence -Inst. Resources 85

PAGE 94

Influence was also seen as significant in regard to the detennination of faculty salaries and teaching assignments. Pn:eot Percent W:ac: .UOUI LIM mo.. tle Mme Chalr'a lllllaenc:e compared to lac:ully'a Cba1r' a tnllaenc:e compared to lac:ully'a Figure 4. 43. Influence -Fac. Salaries Figure 4.44. Influence -Teaching Asgmnts '!heir influence over tenure was seen as somewhat less than that over the previous areas, and that over curriculum development and class meeting times was seen as being the least they possessed Cba1r'l IDIJUeDc:e compared to lac:ully'l Fig. 4.45. Influence -Fac. Tenure l .. dl a.a .. Det eo,. IDe..a Cha1r'l tnllaence compare4 to ICIICUI!y's Fig. 4.46. Influence curriculum 86 Mea ..... aet ,. m..a Cba1r'l tnllaence compare4 to ICIICUI!y'l Fig. 4. 4 7. Influence -Class Times

PAGE 95

'!he two primal:y reasons they gave for having became department chairpersons were to represent the faculty to the administration or to respond to faculty encouragement to take the position. Table 4. 2. -Reasons for Becoming a Deparbnent Clairperson Primazy Secondazy 43 Represent faculty to administration 35 42 Respond to faculty encouragement 28 19 My turn 7 18 Prevent less capable person from taking 29 11 Wanted challenge 9 9 Build program 16 8 Prepare for a higher position 11 4 No one else could/would do it 1 3 Maintain stability 1 2 Appointed by president 2 2 A one person department 1 2 Money 0 1 None 8 1 Break from teaching and research 7 1 Recruited from outside 2 0 Represent administration to faculty 9 166 166 87

PAGE 96

In the prior table, 4. 2, the rn.nnber of times the phrase "prevent less capable person from taking position" appeared as a primary or secomary reason is of particular interest. It llirloubtedly reflects a strong sense of self-confidence, a self-confidence that is also indicated by the fact that the existence of expert and referent power was perceived by these challpersons relative to all groups and by the generally low degree of bumout reported for each burnout component. (Note on page 61 that on any bumout component, the percentage of challpersons reporting in the High catego:cy did not exceed 40%. ) Typical chairpersons had been appointed to their positions with faculty consultation. In fact, in 80% of the cases, the selection process had involved either faculty consultation or faculty election. Table 4. 3. -Process by Which '!hey Became Olai.tpersons Appointed wjfaculty consultation Faculty election Rotation Appointed without faculty consultation National search 88 76 57 13 12 8 166

PAGE 97

'!heir prilnary goal for their departments, upon becoming chairpersons, was to improve deparbnent status in their institutions with a secondary goal of improving that status outside their institution. 'lhe distribution of goals is of particular interest. Approx:ilnately SO% selected the aforementioned primal:y and sec:omary goals of improving deparbnent status either in or outside the institution. Relatively equal percentages (approx:ilnately 13% each) selected "improve the curriculum," "improve departmental teaching", or "better distribute resources/workload". ('lhe "other'' goals shown in Table 4. 4 were write-ins.) Table 4. 4 -Individual Clairperson s Goals for the Deparbnent Primary Secondaey 62 Ilrprove dept's status in the inst. 37 23 Ilrprove the curriculum 28 21 Ilrprove deparbnental teaching 16 19 Better distribute resources/workload 22 19 Inprove dept's status outside inst. 49 22 other (including: Build program, 14 improve working erwiromnent, improve reputationjquality, improve product, and none) 166 166 89

PAGE 98

'Ihe foregoing has summarized findings related to the characteristics of deparbnent chairpersons in institutions of higher education in Colorado relative to their job-related stress levels (measured by their degrees of burnout); their perceptions of their power relative to their faculty, peers, and deans; and their institutionaljdeparbnentaljpersonal situations. '!he next section will review the results of analyses conducted to identify associations among these three sets of data. Bivariate Analyses Power Perceptions and stress Chi -square tests were run on all combinations of power perceptions am burnout components. Using a significance level of .os, evidence was found to be sufficient to reject the hypothesis of irrleperrlence between burnout components am power perceptions in ten instances. Six of these related to faculty, two to peers am two to deans. The data constituting this evidence are reported in Appendix B. To illustrate these data, the test data for Depersonalization am Referent Power relative to Faculty are shown in Table 4. 5 (Table B.l in Appendix B) 90

PAGE 99

Table 4.5.-Perception of Referent Power Relative to Faculty level of l?ower Existence of Power Depemon-Perceived Uncertain or None Perceived alization Total IDt1 40/32.84 7/14.16 47 Average 62/63.59 29/27.41 91 High 14/19.57 14/ 8.43 28 All 116 50 116 arl-square = 10.567 p < .01 df = 2 In Table 4. 5, the first value in each cell represents the rnnnber of the 166 respon:ients observed in that category while the second value {following the slash) represents the rnnnber expected to be in that category if there were independence between the two variables. For exaiTple, the number of respondents who reported High Depersonalization ani were {statistically) expected either to be uncertain as to the existence of referent power relative to their faculty or to perceive none to exist was 8.43. In actuality, 14 of the respondents were observed to be in that category. '!he value of the chi -square statistic in this table was 10. 567. Using a chi -square distribution table for the two degrees of freedom represented by these data, the chances of getting this statistic with independence existing between the variables studied would be less than .01 (exactly .0051) .Accordingly, the hypothesis of independence in the case of these 91

PAGE 100

two variables is rejected at a significance level of 01. It was also noted that the group uncertain about or perceiving no referent :power to exist and reporting a High level of Depersonalization made the largest contribution to the chi -square statistic. Faculty and Power. 'lhe instances where independence between burnout components and :power perceptions was rejected in the case of faculty included all :power bases except coercive. (Tables B.l-B. 7, Appendix B.) Table 4. 6. -Imeperxience Rejected Between Power Type and Burnout Ccmp:>nent (Faculty) Power Type Referent Referent Referent Expert Expert Reward Legitimate Burnout Component EnotionalExhaustion Depersonalization Personal Acc:onplishment Depersonalization Personal Acc:onplishment Personal .Acccmplishment Emotional Exhaustion Significance level a. Although independence was not rejected in this instance, the significance level was sufficiently close to .05 to be of interest. b. Among those dla.irpersons who were uncertain as to the existence of the power or perceived that it did not exist, more reported high stress than would have been predicted to do so in the absence of a dependency relationship. c. Among those dla.irpersons who perceived that the :power did not exist, more reported high stress than would have been predicted to do so in the absence of a dependency relationship. 92

PAGE 101

d. In the two cases of expert power, the chi-square test may not have been appropriate since in more than 20% of the cells the predicted values were less than 5. '!he distribution of actual versus expected instances of stress, however, were consistent with the results in the other cases shown. '!he above results concerning the cllairperson -faculty relationship support the hypothesis that dependence may exist between stress arrl either an uncertainty regarding the presence of power or a perception that the power does not exist. Further, in all significant tests the observed frequencies of those scoring high on burnout and who perceived no power to exist or, at best, were uncertain as to its existence, were greater than would have been expected on the basis of chance alone. This fact would support a firrling regarding causality direction from the absence or uncertainty of power perception to the level of bumout. l?eers and Power. In the case of peers, independence was rejected in only two cases, one involving legitimate power and one reward power. (Tables B. 8-B. 9, Append.ix B. ) Table 4. 7. -Irrlependence Rejected Between Power Type and Bumout Cc:arponents (l?eers) Pc:Mer Type Legitimate Reward Burnout Component Emotional Exhaustion Personal Accomplishment 93 Significance level

PAGE 102

a. Among those chail:persons who perceived that the power did not exist, more reported high stress than would have been predicted to do so in the absence of a dependency relationship. For the chail:person -peer relationship, there were i.mications of dependence between power and stress in only 2 out of the 15 cases. Similarly to the chail:person -faculty relationship, however, there were strong indications of dependence between Emotional Exhaustion and legitilnate power. In both significant tests also, similarly to those related to perceptions regarding faculty, the absezved frequencies of those scoring high on burnout and who perceived no power to exist or, at best, were uncertain as to its existence, were greater than would have been expected on the basis of chance alone. Deans and Power. Similar to the situation with peers, independenCe between power perceptions and burnout components were rejected in only 2 of 15 instances: for legitilnate power and coercive power (Tables B.lO-B.ll, Appendix B) Table 4.8.-Irxiependence Rejected Between Power Type and Burnout (Deans) Power Type I.egitilnate coercive Burnout Component Personal Accamplishment Emotional Exhaustion Significance Level <.osa <.05a a. Among those chail:persons who perceived that the power did not exist, more reported high stress than would have been predicted to do so in the absence of a dependency relationship. 94

PAGE 103

'!he results are smprising for the chail:person -dean relationship since the power perceptions in the two significant cases related to coercive and legitimate. 01ail:persons are, at least nominally, subordinate to their academic deans. Accordin:jly, dlairpersons would not be expected to perceive those Jdnjs of power over their deans in the first place. Why then should there be a dependence between a perception that these particular types of power did not exist and the occurrence of higher stress? A possible answer relates to the agglutinative characteristic of power (see page 23). Since the dlail:persons perceive the existence of the other three bases, there may be an expectation that all fonns of power should exist. When they do not appear to exist, stress can resu1 t. OJ:ganizational/Personal Olaracteristics ari -square tests were also run on the various combinations of organizational/personal characteristicS and bumout components. Using a level of 05 for statistical significance, evidence "WaS found to be sufficient to reject a hypothesis of independence between these characteristics and bumout components in six instances (Tables C.l-c. 6, Appendix C) 95

PAGE 104

Table 4. 9. -organizational Characteristics and Bumout Colrp)nents Blll:nout eomoonent EnDtional Exhaustion I:lepersonalization Personal Accamplishment Explanatory Variables Characteristic Sign. level Influence as c::c.:mpared to faculty in the allocation of resources <.01 BS degrees offered? <.05 Secondary reason for taking position <.05 Percent of faculty paid more than chairperson <. 05 Number of faculty in <.01 deparbnent Percent of faculty paid more than chairperson <. 01 An analysis of variance was then conducted on each power or variable and burnout carrp:ment pair for which the chi -square tests indicated the existence of a deperxlency relationship. '!he results are shown in Table 4 .11. '!he only cases shown are thoSe in which the "F-ratio" for the burnout component indicated was sufficiently large to reject the null hypothesis that the means of the scores reported by each group for the different levels of the burnout component were equal. ('!he F-ratio is the ratio of the variances among the burnout component values of the chairpersons responding in 96

PAGE 105

the same way on any given power or organizational/personal characteristic variable to the variances between the means of the reported burnout component values in those groups. The variances are expressed in sums of squares.) The last column in the table imicates the "P -value, 11 the percent of the time that the ratio would be exceeded if the null hypothesis were true. Where that percentage is 05 or less, the null hypothesis that there is no difference in the group means can be rejected and the conclusion drawn that there are differences between the groups in stress levels. 'lb illustrate these data, the analysis of variance for EnDtional Exhaustion and Referent Power relative to Faculty is shown in Table 4.10 (Table 0.1, Appendix D). Table 4.10.-:Ana.lysis of Variance: Emotional Exhaustion vs Referent Power Relative to Faculty Deer. of SUm of SOURCE Freedom Sauares Ref Power 1 3.283 Error 164 113.299 Total 165 116.512 Power Perception N Existence 116 Uncert. jNon exist. 50 Mean of sauares 3.283 0.690 Mean 1.8534 2.1600 F-ratio 4.76 Std.Dev. 0.8157 0.8657 l;? 0.031 '!he "power perceptions" shown above represent the two groups previously referred to in the chi -square discussion: 1 -those dlail:persons perceiving referent power relative to their faculty 97

PAGE 106

(N=116) am 2 -those either uncertain about it or perceiving none to exist (N=50) 'lhe means indicated are the results of whether they reported a High level of Emotional Exhaustion (3), an Average level (2) or a I.J:M level (1). '!he F and p values are as previously explained, based on the sum of the squares shown. 'lhe data shown in Table 4.11 below include only those pairs of organizational/personal characteristic burnout component or power bunlout component for which the means of the scores reported by each group for the different levels of the component were not equal (Tables 0.1-0.9 for power and Tables 0.10-0.13 for characteristics, Appendix D). Table 4 .11. -Results of ANOVA on Pairs with Unequal Means Referent Power Referent Referent Expert Expert Legitimate Reward Legitimate Coercive Relative To Faculty Faculty Faculty Faculty Faculty Faculty Faculty Peers Deans Ol:ganizational/Personal Characteristic Secondary reason for taking position Percent of faculty wjmore pay Percent of faculty wjmore pay Number of faculty in dept. Burnout Component P-Value ERK:>tional Exhaustion <. 05 Depersonalization <. 01 Personal .Acccmplishrnent <. 01 Depersonalization <.01 Personal Accomplishment <. 05 Emotional Exhaustion <. 01 Personal Accomplishment <. 05 Emotional Exhaustion <.05 Emotional Exhaustion <. 05 Burnout Component Depersonalization P-value <.05 Depersonalization <.05 Personal Accomplishment <. 05 Depersonalization <.01 98

PAGE 107

In the above cases, the probability is high that the groups ( detennined by their categorization by, for example, the m.nnber of faculty in the deparbnent) have reported significantly different levels of stress. In these instances then, since the chi -square tests have indicated depeOO.ency, the evidence is stroiXJ that there is a significant relationship between the b.u:nout carrp:>nent and the organizational/personal characteristic or power perception indicated. In the power perceptions instances, there is similar evidence of direction of causality as that indicated in the chi -square tests. In each case, the mean score of the bmnout component level for those perceiving the power to exist was always less than the mean score for those perceiving it to be nonexistent or for those either uncertain as to its existence or perceiving none. 'Dle ANOVA results for organizational/personal characteristics and bw:nout in the above table (from Tables D.l0-0.13, .Appemix D) are also interesting. In the case of the secondary reason for taking the position, stress (Depersonalization) was highest when the reason was either to represent the faculty to the administration or to prevent a less capable person from taking the position. It was lowest when the reason was to respond to faculty encouragement. 99

PAGE 108

In regard tO percent of faculty receiving m::>re pay, the highest st:l:ess levels (lowered feelings of Personal Accomplistnnent) occurred when the percent fell in the highest ran]e: 31% or m::>re receiving m::>re pay. '!he allcx:ation of institutional resources was the only one of the control issues analyzed for which a stress relationship was in:lica:ted. '!he relationship was in the direction expected, with greater stress (Depersonalization) associated with less relative control. stress (Depersonalization) increased in the smaller departments as the rn.nnber of faculty increased; where, however, there were over 18 faculty in the deparbnent, the stress decreased as the size increased. No statistically significant dependence was fourrl between sex and any of the bumout camponents, supporting other research which has shown that llE1 and women are similar in their experience of burnout (Maslach 1982a, 58) However, Maslach fourrl that men show slightly m::>re of one camponent of bun1out (Depersonalization) while women show slightly m::>re of another (Em:>tional Exhaustion) In this research m::>re wcmen than men were at the High level in every burnout camponent on a percentage basis (vaeying from 43% to 19% for women and from 37% to 16% for llEl) 0 100

PAGE 109

Hypotheses Testing '!he data preceding this section provide evidence that addresses the hypotheses stated in Chapter 1. 'lhese hypotheses have been repeated here together with that evidence. Ol.apter 5 will discuss conclusions drawn from the evidence and the findings reganling the hypotheses. Hypotheses on stress m. 'lhat bJrrDit is exhibited by a statistically significant percentage of the .im.:ividnals per.fcmn:in] the role of departmerit dlair:persan in frur-year instibrt:ians of higher edt"Ulticn in Cblarado. Partially supported. '!he data shCMn in Figure 4.2 (p.69) indicate that 32% of the chaiJ:persons reported high levels of Emotional Exhaustion and 38% reported high levels of a reduced sense of Personal Accomplishment. Depersonalization however, was not identified as making a major contribution to an overall state of burnout, with only 17% of the chaiJ:persons reporting high levels of that component. ma: 'Ibe average bmncut CCIIpJI'lellt scores of these imi.vidnals exceed those displayed by post-sea:mdal:y educatm:s in the :results of :resean:n by Maslach am Jackson (1981, 1986; J). 101

PAGE 110

Partially SllplX)rted. Maslach and Jackson's research on post-secorx:lary educators indicated mean scores of 18.57, 5.57 and 39.17 for Emotional Exhaustion, Depersonalization and Personal .Acx:xllrplishment, respectively. '!he means of the CCIIl'Q;X)nent scores reported by these deparbnent chairpersons were 19 .19 for Emotional Exhaustion, 5. 22 for Depersonalization and 36. 22 for Personal .Acx:xllrplishment (Figure 4. 3, p. 70) While the mean of the Emotional Exhaustion scores reported by these chairpersons was slightly higher than their group and that of Personal Accomplishment slightly lower (both representing greater bumout) the mean of these chairpersons' Depersonalization scores was slightly lower than that of the Maslach and Jackson group. Hlb: M:n:e of these .imividnals will sc:m:e High rather than I.ai en the varioos CXIIpllleiit:s of bm:ncut. Partially SllplX)rted. '!Wice the proportion of chairpersons reported High levels of reduced feelings of Personal Accomplishment as reported row levels (38% c:anpired to 19%), however more chairpersons reported !J::M rather than High levels of Emotional Exhaustion and Depersonalization -38% compared to 32% and 29% compared to 17%, respectively (Figure 4.2, p.69). 102

PAGE 111

Hypotheses on Stress versus Power Perceptions H2. 'lbat bmuit :is :a:u:e prevalent aiOOl'J those dlahpe.rsans 'Who perceive they have inadequate power to px:operly can:y art: the responsibilities of their positions. Partially supported. In 22% (10 of 45) relationships analyzed between power perceptions and burnout components, chail:persons uncertain as to the existence of the power or perceiving that it did not exist inlicated greater levels of stress than would have been predicted under circmnstances of i.mependence (Tables 4.6, 4. 7, 4.8; pp.92-94). ma. SUbjects peroeivinj laver levels of power relative to their depart:Dental farulty have higher bmuit cxmpment scores than those who do not. Partially supported. In 40% (6 of 15) of the relationships analyzed between power perceptions and burnout components, chairpersons uncertain as to the existence of power relative to their faculty amjor perceiving that it did not exist inlicated greater levels of burnout than would have been predicted under circumstances of independence (Table 4.6, p.92). H2b. SUbjects peroeivinj laver levels of power relative to their peers have higher bumcut scores 103

PAGE 112

than these who do net. Partially supported. In 13% (2 of 15) of the relationships analyzed between power perceptions am bumout components, chaixpersons perceiving that power did not exist in relation to their peers indicated greater levels of burnout than would have been predicted under circumstances of independence (Table 4. 7, p. 93) H2c. SUbjects :perceiviDJ lower levels of pc7i11er :r::elative to their administrative superiors have higher bimcut scm:es than these who do net. Partially supported. In 13% (2 of 15) of the relationships analyzed between power perceptions and burnout components, chaixpersons perceiving that power did not exist in relation to their deans indicated greater levels of burnout than would have been predicted under circumstances of i.rrlependence. (Figure 4 8, p. 94) Hypotheses on Power Perceptions lD: 'lhat subjects perceive they have st:t:a:ger pc7i11er bases of sane types than others. SUpported. Relative to their faculty, chaixpersons perceived that they possessed referent and expert power but were uncertain as to their reward, coercive and 104

PAGE 113

legitilllate power. While, relative to their peers, they maintained the expert power perception they had in regaJ:d to their faculty, they were uncertain as to their referent and reward power and perceived no coercive or legitilllate power. The power perceptions they had in regaJ:d to their peers also held for their deans except that they were not sure of their legitilllate power relative to the latter (Table 4.1, p.79). lOa. SUbjects perceive that they have m:>re power due to their expertise than to the authority associated with their position. SUpported. Chairpersons perceived that they possessed expert power relative to all groups but were either uncertain regarding or perceived that no legitimate power existed (Table 4.1, p. 79). lOb. SUbjects perceive that they have mxe power due to their ability to :rewam than to their ability to pm:ish. Partially SUpported. Chairpersons were uncertain as to their reward power relative to all groups. They perceived no coercive power over peers and deans and were uncertain as to its existence in regaJ:d to their faculty (Table 4.1, p. 79). 105

PAGE 114

Hypotheses on Organizational/Personal Characteristics and Stress H4. the a:rganizatianal am perscna1 dlaracteristics of the imividnal s occupyin} these positialS do mt significantly ccntrib:rt:e to the vari.arKE in 1:m:nait
PAGE 115

Exhaustion, Depersonalization, and Personal Accomplishment. '!heir perceptions of expert power related to Depersonalization and Personal Accomplishment, a1 though the relationship was less precise. A depen:lency relationship was also indicated between reward power and Personal Accomplishment and between legitimate power and EnDtional Exhaustion. In regard to their peers, perceptions of the existence of legitimate and :rewan:i power were found to be related to Emotional Exhaustion and a lowered feelinge of Personal Accomplishment, respectively. In the case of their deans, relationships were imi.cated to exist between their perceptions of legitimate power and Personal Accomplishment and between their perceptions of coercive power and Elrotional Exhaustion. '!he only organizational/personal characteristics indicated by chi -square tests as being related to stress were seco:ndal:y reasons for taking the position, the percent of their faculty paid more than they were, their influence as compared to their faculty's in the allocation of institutional resources, the number of faculty in their department and whether or not their department awarded Bachelor of Science/Arts degrees. When analyses of variances were conducted. on these relationships, however, the differences of means in influence over resources and the issue of BS degrees were not sufficiently large to warrant rejection of the null hypothesis of equality. 107

PAGE 116

'lhe hypotheses studied were all supported to varying degrees. A high level of one bul:nout component or another was reported by from 17% to 38% of the chairpersons, the mean reported levels of two of the bun1out components were higher than those reported by a referent group of post-secondal:y educators, am twice as many chairpersons reported high levels of reduced feelings of Personal Accomplishment than reported low levels. In general, uncertainty as to the existence of power andjor a perception that the power did not exist were associated with greater levels of stress. 'lhe chairpersons reported a belief that they possessed expert and referent power relative to all groups and reward power relative to their deans. Finally, only five characteristics of an organizational or personal nature were found to be in deperrlency relationships with bul:nout components. 'lhree of these had significant mean squared variances between the categories of those characteristics. 108

PAGE 117

aJAPrER 5 OONClliSION Introduction In general, this research has discovered three things about department chairpersons in four-year institutions of higher education in Colorado. First, those deparbnent chairpersons responding to the survey, on the average, did not exhibit the high levels of stress that might be expected in consideration of the paradoxical nature of their managerial positions. '!he stress they reported was essentially similar to that discovered among 635 post-secomary educators studied by Maslach and Jackson (1981, 3). Among these particular individuals, however, many did report high levels of stress. '!hey, and their situations, should be of interest to deans and vice-presidents since, as previously mentioned, an entire institution can be adversely affected if its deparbnent chairpersons do not accomplish their roles effectively. Second, this research discovered additional support for other research (Birnbaum 1988, 14) indicating that expert and referent power are the primary bases for deparbnent chairpersons' influence on their faculty. Finally, an absence of independence was fot.md to exist between certain organizational/personal characteristics and the stress reported by these deparbnent 109

PAGE 118

dlail:persons. '!his discovery identifies areas in which deans and vice-presidents can take immediate actions to reduce bw:nout in their deparbnent dlail:persons as well as areas for further research on the causal relationships involved. 'lhe following disrnsses specific conclusions drawn from the data and f.inli.ngs presented in Olapter 4, strengths and weaknesses of the research, contributions of the research to a field of knowledge, and recommendations both for actions by deans and vice presidents and for further research. Conclusions General 'lhe data presented in Olapter 4 regarding department dlail:persons in four-year institutions of higher education in Colorado can be summarized into four categories. 1. 'lhe typical organizational/personal characteristics of these irx:lividuals. 2. 'lhe levels of stress they experienced, as indicated by their reported degrees of the three components of bw:nout. 3. 'lhe power perceptions of these individuals relative to their faculty, peers and deans. 4. Relationships between the stress they reported and their power perceptions and organizational/personal characteristics. 110

PAGE 119

'!he data were used to establish the presence or absence of support for several hypotheses reganling these categories. '!he general ard specific conclusions which follow are drawn from both the data ard the findings in regani to those hypotheses. Four general conclusions were derived from the study. 1. Deparbnent dlaiipersons at four-year institutions of higher education in COlorado, in general, did not exhibit high levels of job-related stress, operationalized as burnout. Among them, however, were a significant number who did exhibit such levels. 2. '!he subject individuals perceived that their ability to influence their faculty, peers, ard deans was based primarily on their expert ard referent power. 3. '!hose individuals who were uncertain as to the existence of those power bases were more likely to experience high stress levels. 4. '!here was an absence of inde:pendence shown between chairperson stress ard a select group of organizational characteristics. Typical
PAGE 120

1. In regard to etlmicity and gender, major strides have not yet been made in appointment of etlmic minorities and females to these positions. Whether that is due to the non-availability (or non-willingness) of qualified candidates or to a bias in the selection process was not an objective of this research. 2. Maturity (represented by the age group), tenure, a tenninal degree and full-time faculty status are all major considerations in department chairperson selection. 3. '!he position is not seen as long-tenn, or at least chairpersons do not appear to be making it a career o over one-half of those responding to the survey had spent only from 1 to 4 years in the position. '!his result is sllni.lar to results obtained by Kinpton, cited in Corson (1959, 94), indicating (as he noted then) that there is still no indication of a return to the concept of long tenn department "heads. 11 4. Part of the deparbnent chairperson's expert power may relate to his having, typically, full professor rank. 5o '!he typical salacy received places those department chairpersons receiving it in the "upper income" bracket but salaries are widely distributed among department chairpersons. one quarter can be expected to be paid less than $40,000 and one quarter more than $65, 000 annually. 112

PAGE 121

6. While the typical deparbnent chairperson has less than 10% of her faculty receiving higher salaries, over 60% of the chairpersons reported that some percentage of their faculty did receive higher salaries. Another way of looking at this would be to note that a majority of the department chairpersons were not the highest paid persons in their deparbnents. '!his could logically lead to same frustration and stress and pertlaps a power problem with those individuals receiving higher salaries. '!he "equity" theo:cy of ll'Dtivation (Adams 1963) would have application here. 7. '!he number of full-tilne deparbnental faculty typically exceeds that which woUld be desirable for an effective span of control in a non-collegial organization, with 40% of the departments exceeding 12 full-tilne faculty members -again similar to Kimpton's study. '!hat size is, however, relatively static with few having varied by more than 20% in either direction over the past three years. 8. '!he adequacy of institutional resources and the difficulty in obtaining noninstitutional resources are perceiyed as presenting significant problems for deparbnent chairpersons, much as it did to the chairpersons in Norton's 1978 study of deparbnent chairpersons in colleges of education (Norton 1978, 20). In that study, "the highest degree of frustration was reported in the area of inadequacies of deparbnent resources". 113

PAGE 122

9. '!he allegiance of the department dlall:persons to their faculty was demonstrated by the fact that over 50% stated their prilnazy reason for taking the position was either to represent the faculty to the administration or to respond to faculty encouragement. In Norton's study (1978, 5) a similar percentage (56.5%) gave their prilnazy reason as "an opportunity for providing leadership to the academic program." His listing of alternative choices for the prilnary reason, however, did not include the two choices in this study receiving the majority of responses. '!his study included questions which differentiated between a c.hallperson's personal reasons for taking the position am her prinm:y goals for the department; over 50% of the respondents indicated either "improving internal status" am "improving the curriculum" as the prilnazy goal. '!he similarity between these goals and the prilnazy reason for taking the position indicated by Norton's study is evident. stress levels '!he data do not support a conclUsion that these or similar department dlall:persons experience particularly high levels of stress. '!he majority can be expected to experience bmnout in the average range of that reported in Maslach and Jackson's research as being typically displayed by post-secondary education educators (1981, 1986; 3). '!he data do support, however, a conclusion that, within any group of such individuals, a 114

PAGE 123

significant nmnber (17-38%} will be experiencing burnout resulting from high levels of one or more of the components constituting burnout: Emotional Exhaustion, Depersonalization ani (a reduced feeling of} Personal Accomplishment. Also, a similar proportion (20-40%) can be expected to be at a l01r1 burnout level. Power Perceptions Expert and referent pc:Mer fonn the prilnary bases for the influence exerted by deparbnent chail::persons over their faculty, peers and deans. '!he expertise that the chail::person perceives gives hint that influence is not, however, clearly defined. '!he statements on the Maslach Bumout Invento:ry (the responses to which generate the conclusion that this power base exists} consist of "respect rcry competence about things in which I have more experience, 11 "defers to rcry judgment in areas with which I am more familiar, 11 and "believes that following rcry advice will result in better decisions. 11 Unfortunately, perllaps, the statements are not explicit as to what those "things, 11 "areas, 11 ani "decisions" are. Accordingly, the author questioned several department chairpersons to obtain their opinions on the matter. '!hey were unanimous as to believing it was not expertise in the field of kn01r1ledge represented by the deparbnent. Rather, it was their opinion that it was expertise in managing the deparbnent: in obtaining the resources the department needed; influencing higher administration as to program offerings, award of tenure, 115

PAGE 124

etc. ; achieving institutional objectives am conducting the myriad of other duties required by the position. Since a perception of having expert power would logically correlate with a high level of self-confidence, the presence of this perception alone may aCCO\mt for the overall low stress levels of department chairpersons. '!here is also a sort of "political correctness" to the concept that influence is best (ani most apPropriately) effected in educational institutions through the exercise of expert power, particularly as compared to coercive and legitimate. '!his concept may well have biased the chairpersons' responses. Cllail:persons also believe that their referent power plays a major role in their ability to influence. '!his, too, would be expected, considering the envirornnent in which they ftmction. Department chairpersons certainly would logically want to think that others respected and desired to identify with them rather than being primarily influenced by the rewards, coercive abilities or the authority associated with their position. In general, this research substantiates the claim that "normative organizations, such as colleges and universities, rely on referent ani expert power that is less likely to cause alienation and that produces conunitted participants who are influenced through the manipulation of symbols" (Bimbaum 1988, 14). 116

PAGE 125

stress Relationships It is apparent from this research that those deparbnent chail:persons 1.D'lCertain as to their power bases or perceiving the absence of power will be trore likely to experience feelings of Elrotional Exhaustion ani Depersonalization and a reduced feeling of Personal Accomplislnnent. Although a definite causal relationship has not been proven in this study, the logical direction for this oorrelation, supported by these findings as well as on the many studies done of bun1.out, would appear to be from the power perception to the stress. 'Iha.t is, if deparbnent chail:persons perceive an inadequacy of power in dealing with their faculty, peers and deans, they can also be expected to be experiencing job stress at same -level. FUrther, and of particular note, those deparbnent chai.z];.lersons pen::eiving problems particularly with their referent ani expert power bases relative to their faculty will have the highest probability of experiencing such stress and the results manifested as bun1.out. An absence of independence can also be expected to exist between chail:persons 1 stress and their power perceptions relative to their peers arrl their deans and stress -but to a lesser degree than it does in regard to their faculty. Similarly, only a small number of a chai.z];.lerson 1 s organizational or personal characteristics should have a dependency relationship with the 117

PAGE 126

stress which chail:persons experience. 'Ihese are :ilnportant characteristics, however, and for the most part they are controllable by the institution. Specifically, dependency exists between their job-related allocation of institutional resources, the percentage of faculty in the deparbnent paid higher salaries, the number of faculty in the deparbnent, and whether or not the department offers a bachelor's degree. ('!he observed :llnportance of their perceived influence on the allocation of institutional resources is consistent with the findings of Norton's 1978 study. ) '!he remaining characteristic having a dependency relationship to stress, a chail:person' s secondal:y reason for taking the position, while not controllable by the institution, might well be worthy of consideration in the process of chail:person selection. 'lhese specific conclusions on stress relationships support a number of the propositions put fo:rward in 1993 by Cordes and I:bugherty (refer to pp. 42-43,
PAGE 127

organization. Finally, as Comes and Dougherty proposed, employees enter the workplace with certain expectations regarding, among other things, what they will be able to accomplish professionally. '!he contrast between those expectations and daily realities will influence their perceptions of personal accomplishment (1993, 647). This last situation was clearly demonstrated by the employees smveyed, department chail::persons, who did report lowered feelings of personal accomplishment when they perceived an absence of the attributes necessary to accomplish those expectations, particularly in an academic envirornnent: referent, expert, and reward power. Strengths and Weaknesses A major strength of this research is in the fact that it used two instnnnents which had been previously used in other studies and thoroughly validated by their developers: the Maslach Burnout Inventory and a power typology questionnaire developed by Spekman. Reliability coefficients that have been demonstrated for these instruments were discussed in Chapter 3, 'Methodology. 11 A second strength results from the uniqueness of the population studied and the relating of the results of job-related stress {burnout) to power perceptions. Much study has been done on the occurrence of burnout among health care providers, teachers, post-secondary educators, scx::ial and human seJ:Vices 119

PAGE 128

workers, prison am law enforcement officials and even 269 members of the oregon Province of the Society of Jesus, an order of the catholic Church (Grimm 1986). No studies to date, however, have specifically focused on individuals occupying this "first-line" superviso:r::y role in higher education. Also, the sb.ldies referenced have shown relationships between bunlout am a host of variables' including gender' ethnicity, marital am family status, education, age, and workplace environment, including leadership style and role characteristics such as ambiguity, overload, and conflict (Clarke, 1991, 34-54 and Maslach 1982a, 58-62). Only in the case of leadership style, however, have these sb.ldies approached the focus of this study on power perceptions, and then only tangentially. In those studies the leadership style of nurses in wham bumout was found appeared to have been developed by them in response to a work situation which itself may have been the cause of the bumout. '!here are weaknesses am limitations to this research. For one, the questiomaires were only administered to chail:persons in higher education (not including community colleges) in Colorado. And, even though the literature tends to consider general homogeneity to exist in the functions of the role across the broad spectrum of higher education institutions nationally, the Colorado population may not accurately represent 120

PAGE 129

department chail:persons as a whole. Also, only 166 of a possible 444 department chail:persons responded. These were, however, well distributed across the state's colleges and mriversities, both public and private, am across the departments within those institutions. In fact, responses were received from all but one of COlorado's 16 four-year institutions of higher education. 'Ihe percentage of departments responding from an institution ranged from 21.5 to 84.6, with a median response percentage of 32.3. Additionally, a rn.nnber of the characteristics of the department chail:persons responding correlated with those of chail:persons sw:veyed in Norton's 1978 study. Specificaily, 55% of both groups were full professors at the time of their appoinbnents, their prinmy reasons for taking the positions were similar (to provide leadership to the program), their deparbnental sizes were similar (6-10 or 7-12) and the adequacy of deparbnental resources was identified as a significant problem by both groups. As with any self-administered questionnaire, there is always the possibility of a self-selection bias -but this could have affected the results either way. That is, those feeling more stress and wishing to vent their frustrations could have been more likely to respon:l in order to do so but, conversely, those feeling less stress could feel they had more time to devote to filling out the questionnaire. '!he range of responses, however, in:ticates that urrloubtedly both kinds responded. 121

PAGE 130

Finally, since the responses to many of the questions resulted in categorical instead of measurement data, regressions could not be run to detennine the explanatory power of the dlaracteristics ani power perceptions appearing to be related to burnout oomponents. Contribution '!he sine gye non of any dissertation of this sort is that it has added to a field of knowledge. In this instance, the research adds material to several fields of knowledge. '!his material not only contributes to what is known about those fields but also provides a basis and for even more fruitful study. First, the research adds significantly to the data existing on the effects of job-related stress (burnout) by having studied a particular group of managers not previously isolated for such study. It contributes data to that already existing on the power of managers, but which is focused on individuals occupying unique managerial positions. It adds descriptive data to the field of knowledge regarding department chairpersons in higher education, particularly in regard to their power perceptions and the results of the stress experienced by those individuals. In that way it helps to build a theory for this field of endeavor. '!he theoretical base existing on department chairpersons is p:x>rly 122

PAGE 131

defined with, as was previously mentioned, most of the literature on the area having concentrated on prescribing functions, descriliing characteristics and detennining the satisfaction levels of particular populations of '!he JOOSt i.n'portant contributions of this research may be its fllxling that high levels of stress do exist among a relatively large segment of the group surveyed (and, probably, among deparbnent chail:persons in other states as well) and its support of the hypothesis that stress and power perceptions are not independent. Recormnendations From the data gathered, it would appear that individuals lll'rlertake to serve as mainly in order to do a number of things that, as faculty, they see as needing to be done. Once they start acting as one of two things happens. 'Dley are able to make acceptable progress (in their own view as well as that of their faculty) toward doing these things and are then satisfied with their perfonnance and do not experience high levels of stress. Or they are not able to make such progress, t:h.ereby losing the respect and admiration of their faculty and perceiving that they are unable adequately to influence them, and then suffer from job-related stress. 123

PAGE 132

To nmdlnize a chairperson's productivity then, the possibility of such stress occurring should be minllnized. Two areas for action are recommended for achieving this. Actions by Vice-Presidents and Deans Since the reasons chairpersons have for taking on the position and their personalities may affect their potential for burnout, vice-presidents and deans should take steps in all aspects of the deparbnent chairperson lifecycle (recruitment, selection, training, employment and tennination) to avoid andjor alleviate the occurrence of bumout in these individuals. Specifically, they should give particular attention in the recruitment arrl selection processes to the level of self-confidence displayed by the carxtidates and the reasons they have for seeki.rg the position. In particular, since the research has shown its importance in the burnout equation, they should carefully evaluate those whose secondal:y reason for taking the position is either to represent the faculty to the administration or to prevent someone less capable from taking the position. '!his research, as well as that by others such as Chemis et al. (1985), has shown the importance of power perceptions in avoiding stress. Accordingly, vice-presidents and deans should make new deparbnent chairpersons urrlerstand that their work will be significantly "political" and that one of their challenges will be to acquire arrl use a variety of types of power. '!hey should 124

PAGE 133

therefore insure that newly selected chairpersons receive training to make them knowledgeable on the entire issue of power, particularly hO'tl to develop it and hO'tl to use it effectively. Olrrently this is not a topic included in preparato:cy courses for department chairpersons. '!he closest that two recent Department leadership Programs conducted by the American Council on Fducation carre to doing so was to include in one a session on "'!he Department Cllairperson as Cllange Agent." '!he acquisition and the use of power per se, hO'tlever, were not issues included in the workshop. FUrther, based on this study as well as Norton's (1978), vice-presidents and deans should prepare their department chairpersons for the reality of facing resource constraints. But they should also analyze the relative influence on institutional resource allocation either inherent to or traditionally exerted by the various departments in their institutions. Where feasible and appropriate, they should take actions necessa:cy to change the influence of departments so as to achieve greater equity among them in their influence over decisions on this critical area. One possible technique here would be to fonn groupings of less powerful departments into divisions where, through collaborative efforts, they could be better able to affect resource allocations as well as tenure decisions, etc. 125

PAGE 134

An additional rec:ammendation arises both from this study ani from Warnath ani Shelton's research (1976) which showed the contribution of situational factors to bmnout. That is, that vice-presidents ani deans should also consider using organization development consultants to help create greater trust between deparbnent chail:persons and their faculty, peers and deans and to open problem-solving avenues within the stru.cture of the institution. '1hese actions would serve not only to reduce the stress among deparbnent chail:persons resulting from gaps between ideals and job realities and a lack of collegial support but also could contribute significantly to the decision-making process in IOOSt institutions. As far as dealing directly with burnout, however, fet direct organization development interventions exist ani almost none provide data sufficient to assess effects. A large number of OD designs stani ready, though, to guide data-based efforts to create trust and reduce bmnout (Golembiewski and MUnzenrider 1988, 218-19). Finally, if symptoms of bmnout ocx:::ur, vice-presidents and deans should: (1) Plan for structured counseling sessions with the chail:persons concerned, (2) Consider alternatives such as sabbaticals, and (3) Offer the opportunity to retmn to teaching and research. 126

PAGE 135

An alternative operative at many schools, which experience has shown may be helpful in avoiding andjor relieving bumout, is fixed periods of tenure defined in advance. For exanple, at the university of Waterloo in canada, deparbnent heads and deans are allowed a maximum of seven years in a position. '!hey have to state, when they become candidates for the positions, how many years they plan to serve in this term (Woolsey 1993) F\lrth.er Research Qualitative studies would be most helpful in determining directions of causality among stress, power perceptions and organizational/personal characteristics as well as for determining the reasons for any dependencies existing between them. In particular, those studies should be focused on the power bases and organizational/personal characteristics identified by this study as not being independent from the existence of burnout in deparbnent chairpersons. 'Ihese studies could then lead to :rrore specific preventive or corrective actions involving those power bases andjor characteristics. Another area for further research involves perceptions of legitimate and coercive power. A majority of the deparbnent chairpersons in this study reported perceptions that they either were uncertain about the existence of these two fonns of power relative to their faculty or perceived none to exist. Is this because they neither want to acknowledge their existence nor to 127

PAGE 136

use them? or is it because they truly do not understand what they are? Similarly, relative to their deans, why were they W'lCertain as to whether or not they have legitiinate power? '!here should lCXJically have been a strong perception that no such power existed. calvin Lee, Cllancellor of the University of Maryland, Baltimore County, has stated "The relationship of the department chairman to the academic dean .. is the most ilnportant relationship within the administrative structure of the university" (in Brann and Emmet 1972, 54). FUrther research regarding the perception of power by department chairpersons relative to their deans (and vice versa) has the potential of developing same very interesting and useful infonnation regarding this relationship. 128

PAGE 137

APPENDIX A Department Ola:Uperson Questionnaire University of Colorado at Denver Graduate School of Public Affairs l44S Markel Srreer. Suire 3SO Denver. Colorado 80202 Phone : (303) 820 Fax: (303) 534-8774 February 1, 1992 Dear Department Cllair/}lead/Cllairperson: I am writin:J to request yoor participation in a research study conc:erni.rq persons directly responsible for departments/divisions in institutions of higher education in COlorado. Your response is critical to the success of this effort and would be lOClSt appreciated. 'Ihe study is part of rrrt doctoral dissertation at the Graduate School of PUblic Affairs, University of COlorado at Denver. It investigates your feelings about the position yoo hold and your perceptions of the influence you have upon those in yoor institution with whan you work. In this folder are five short questionnaires. It should take only 20 -25 minutes of yoor time to fill them out and return the folder to me in the envelope provided. For the results to be representative, it is inpJrtant that as nany folders as possible be CCI!pleted and retUl:ned. I am hopin:J for a good return rate and I believe that oor uutua1 interest in higher education will help to fulfill that forecast. I can assure you that all info:rmation provided will be kept CCitpletely confidential. 'lhe number shown on the folder is only for rrrt use in which ones have nat yet been returned so that I can follow up on them. Please be mtte to read the questions carefully and answer them as objectively as you can. 'nlere are no right or wron:1 answers. If you have arrt questions regard.j..n; the questionnaires or problems with their c::cmpletion, please call m at 1-SOQ-245-1060, ext. 3204. 129 o.J \ Dcx::toral camidate, UCD and Re:Jistrar, COlorado Sdlool of Mines

PAGE 138

("Deparbrent" an:i "Deparbrent Clair'' shall.d be translated as appropriate to yoor organizational titles. ) ) General: 1. personal reason for becanirq a depaitlrent d1air was to: a. Have a break fran an:i research. b. Prepare for a higher position in academic administration. c. Represent the faculty to the administration. d. Represent the administration to the faculty. e. Prevent a less capable person the position. f. Respon:l to faculty E!llC:nlr'agement to seek the position. g. other (specify) ------------------2. secord 1!DSt ilrp:lrtant personal reason for bec:anirq a department chair was to: (check gm) a _Have a break fran an:i research. b. Prepare for a higher position in academic administration. c. Represent the faculty to the administration. d. Represent the administration to thS faculty. e. Prevent a less capable person the position. f. Respon:l to faculty ena:aJragement to seek the position. g. other (specify) ------------------3. goal for the department is to: .(check _gm) a. Ill;n:'cve the curriculum. b. Inprcve departmental teadWq. c. Better distril:lute resources an:i wrkload within the department. d. Inprcve the department's status in the institution. e. Ill;n:'cve department's status outside the institution. f. other (specify) ------------------4. seocnd mcst ill;lortant goal for the department is to: a. Ill;n:'cve the curriculum. b. Illprcve clepartmental teadlin;J. c. Better distril::lUte resources ani wrkload within the department. d. Illprcve the department's status in the institution. e. Ilrprcve the department's status outside the institution. f. other (specify) ------------------5. 'lbe precess by w.i.ch I became the department chair involved: (check ill applicable choices) a. Appoint:!Ent by a higher level administrator (with faculty c::xmsul.tation). b. Appoinbrent by a higher level administrator (wit:halt faculty c::xmsul.tation). c. Rotation (amcn:J various faculty IIElli::lers in the department). d. Election by the department's faculty. e. other (specify) 130

PAGE 139

organizational: 6. Institution is: (check one choice in each row) a. (1) ?..ll:llic (2) Private (3) other ) b. (1) College (2) Professional School (3) University (4) other (Specify if (4) is checked: ) 7 Classification of departl!ent: Arts &lsiness/Mgt carm.mications Education En3'ineerin:] Frqn I..anq !Uth SVcs _Humanities_ Letters_ Life/Sciences Math/Q:IIp Sci Sci Soc Sci other '--------8. Depart:lDent has been in existence for: a. :tess than 5 years b. At least 5 I:M: less than 15 years c. At least 15 but less than 25 years d. At least 25 years 9. Cllrrent number of full-time faall.ty in the deparbnent: a._ 1-6 b._ 7-12 c._ 13-18 d._ 19-25 e._ 26-31 f._ IIX:)re than 31 10. Im"in:] the past 3 years the number of full-time faall. ty in the deparbnent has: a. Remained within 20\ of its current number. b. ExrPeied its current number by IIX:)re than 20%. c. :Been smaller than its current number by IIX:)re than 20%. 11. Within institution the department is primarily a:msidered as perfol111in;J: a. a SURJOrtin:J role b. a central role c. other than the above (Please elaborate on the choice selected, usin;r the reverse side if necessary: -------------------------------> 12. Degree programs offered by the department incllde (check all applicable): a. None b. BS/B.Vsimilar c. MS/MA,/similar d. 13. Availability to the department of institutiorial resau:oes necessary to achieve departmental c;joals is: (In:ticate by whole number based on scale below) Inadequate------Barely than adequate 1 5 9 14. a;m-instituticmal :rescurces for the department is: (In:ticate by whole number based on scale below) Exb:emely diffic:ult:----fUcderately diffiall.t-------Rel.atively easy 1 5 9 15. catpared to the influence which can be exerted by mcst of the department's faall.ty, l!rl influence in each of the areas below is as shown: (Enter 1 for llllCh IIX:)re, 2 for sc:m!What IIX:)re, 3 fer the salle arxi 4 for less). a. Faall.ty prcm:X.icn b. _, Faculty telUlre c. Faall.ty salaries d. Faall.ty teachinq assigrments e. Class meetin:] t.iJDes f. Cllrriall.um g. _Allocation of instituticmal/CX)llege resources 131 ....

PAGE 140

16. Persalal leadership style: (Irdic:ate by lolhole l'l\liOOer based en scale below) Autoc::rati.c--------Demccrati.o--------I.aissez-faire 1 5 9 17. !tj position is directly acx:cuntable to a: (check one) a._ Division Head b._ Dean c._ Vice Pres. d._ other (specify) -----18. Position title: -----------------------Bioqra;tlical data: (check only one choice for each question) 19. Aqe: t.Jn:l.er 25 26-30 -31-35 36-40 -41-45 46-50 51-55 _56-60 over 60 20. Sex: Feml.e Male 21. Tenured? Yes No 22. Ethnicity: _Native Al!erican _Hispanic Caucasian Asian Al!erican _._ Black -other 23. 1svel. of highest degree held: a. Na'le b. c. KS/MlVsimilar d. Fh.D/Ed.D/similar 24. l\c.ademic field of highest degree held: ___________ 25. Years as a full-time faculty IIElltler: -1-4 5-7 8-11 12-15 -16-19 -OVer 19 26. Olrrently: F\lll-time -Actin; 27. Years in position: 1-2 3-4 5-6 7-8 8-9 10 or IIIJre 28. .Acadenic rank when to department chair: Professor Assoc. Professor Asst. Professor Instructor other (specify:, _______ 29. OJrrent academic rank: Professor Assoc. Professor Asst. Professor Instructor other (specify: _______ 30. QJrrent a:lltract U'n:ier $30,000 $30,001 $35,000 $35,001 $40,000 $40,001 $45,000 $45,001 $50,000 $50,001 $55,000 $55,001 $60,000 $60,001 $65,000 Al:xllle $65,000 31. percentage of departmental faculty paid a greater amcamt than in:W::ateci in question 29 abave: 0% 1-10% 11-20% 21-30% 31-40% OVer 40t 132

PAGE 141

Christina Mastach Susan E. Jackson Richard L. Schwab Educators Survey The purpose of this survey is to discover how educators view their job and the people with whom they work closely. On the following page there are 22 statements of job-related feelings. Please read each statement carefully and decide if you ever feel this way about your job. If you have never had this feeling, write a "o (zero) in the space before the statement. If you have had this feeling, indicate how often you feel it by writing the number (from 1 to 6) that best describes how frequently you feel that way. An example is shown below. Example: HOW OFTEN: 0 2 Never A few times Once a ayear month or less or less HOW OFTEN 0 6 Statement: I feel depressed at wort<. 3 A few times a month 4 Once a week 5 A few times a week 6 Every day If you never feel depressed at work, you would write the number o" (zero) under the heading HOW OFTEN. If you rarely feel depressed at work (a few times a year or less), you would write the number If your feelings of depression are fairly frequent (a few times a week, but not daily) you would write a s: Consu/1/ng Psychologists Press, Inc. 3803 E. Bayshore Road Palo Alto, CA 94303 'Reproduced by special permission of the Publisher. CoDSulling Psychologists Inc . Palo Alto. CA 94303 from EduciJlors Survey by Christina Mulach, SUSID E. Jackson. and Richard L. Schwab. Copyright 1986 by CoiiSillting P1ychologbts Press, Inc. All rights reserved. Funher reproduction is prohibited without the Publisher's consent. 133

PAGE 142

Educators Survey HOW OFTEN: HOW OfTEN 0 l. __ 2. __ 3. __ 4. __ 5. __ 6. __ 7. __ a. __ 9. __ 10. __ 11. __ 12. __ 13. __ 14. __ lS. __ 16. __ 17. __ 18. __ 19. __ 20. __ 21. __ 22. __ 0 Never A few times a year or less Statements: 2 Once a month or less 3 A few times a month I feel emotionally drained fran rrrt work. I feel used up at the en:i of the workday. 4 Once a week 5 A few times a week 6 Every day I feel when I get up in the an:i have to face another day on the Job. I can easily understani hew rrrt depa.rt:nEntal faculty feel about th.i.ngs. I I treat sane departmental faculty as if they were i.n;:ersonal. obJects. Worki.n; with peq>le all day is really a strain for I deal very effectively with the problems of rrrt departmental faculty. I feel l::iurned out fran rrrt work. I feel I'm positively influencirg other people's lives through my work I've Cec:aDe more calloos towartl people sin::e I took this job. I worry that this job is hardeni.rg 1te emotionally. I feel very eneigetic. I feel frustrated by Jtrf job. I feel I'm worki.n; teo hard on rrrt job. I don't really care W..t tuq:pens to sane departmental faculty. Worki.n; with peq>le directly puts teo 1lllCh stress on I can easily create a relaxed atmosphere with rrrt departlnental faculty. I feel exhilarated after workirg closely with rrrt departmental faculty. I have ac:xx:mplishecl many worthWhile thin;s in this job. I feel like I'm at the en:i of rtrf rope. In rrrt work, I deal with emotional problems very calmly. I feel departmental faculty blaue 1te for sane of their problems. (Admi nistralive use only) cat cat. EE: __ OP : __ PA: __ 134

PAGE 143

RESEARai OOES'I'IQNNAIRE 1 Please resporxi to each statement bel.CM acx:ordirq to wat degree you concur with the statement. Faculty in my department: 1. Respect me arxi want to act in ways that merit my respect arxi admiration. st:ron:]ly agree ( ) agree ( ) neutral ( ) disagree ( ) st:ron:]ly disagree ( ) 2. Respect my competence about t:hi.rl3s in which I have more experience. strorgly agree ( ) agree ( ) disagree ( ) ( ) st:ron:]ly disagree ( ) 3. Believe I can give special help to those who cooperate with me. strorgly agree neutral disagree strorgly agree disagree ( ) ( ) ( ) ( ) ( ) 4_. Believe I can apply pressure on those who do not cooperate with me. st:ron:]ly agree neutral disagree st:ron:]ly agree disagree ( ) ( ) ( ) ( ) ( ) 5. Perceive me as havirg a legitimate right to expect that my suggestions will be carried out. st:ron:]ly agree neutral disagree st:ron:Jly agree disagree ( ) ( ) ( ) ( ) ( ) 6. Defer to my judgment in areas with which I am toore familiar. st:rcrgly agree neutral disagree stron::Jly agree disagree ( ) ( ) ( ) ( ) ( ) 7. Believe I am able to make t:lli.n3s difficult for them if they fail to follCM my advice. st.l:'orgly agree ( ) agree ( ) neutral ( ) 135 disagree ( ) stron::Jly disagree ( )

PAGE 144

8. Feel obligated to follow my suggestions. stron;ly agree ( ) agree ( ) neutral ( ) disagree ( ) 9. Believe they can personally benefit by CXXJperating with me. stron;ly agree ( ) agree ( ) neutral disagree ( ) ( ) stron;ly disagree ( ) stron;ly disagree ( ) 10. Believe that following my advice will result in better decisions. stron;ly agree ( ) agree ( ) neutral ( ) disagree ( ) strongly disagree ( ) 11. Cooperate with me because they have high regard for me as an irx:lividual. stron;ly agree neutral disagree strongly agree disagree ( ) ( ) ( ) ( ) ( ) 12. Believe I can penalize those who do not follow my suggestions. stron;ly agree neutral disagree stron;ly agree disagree ( ) ( ) ( ) ( ) ( ) 13. Feel they haVe to oooperate with me. strongly agree neutral disagree strongly agree disagree ( ) ( ) ( ) ( ) ( ) 14. Cooperate with me because they wish to be identified with me. stron;ly agree ( ) agree ( ) neutral ( ) disagree ( ) strongly disagree ( ) 15. Perceive that oooperating with me can positively ilrpact on their performance. sb:cn:jly agree ( ) agree ( ) neutral ( ) 136 disagree ( ) stron;ly disagree ( )

PAGE 145

RESFARai CUES'I'IONNAIRE 2 Please respom to each statement below accoi.'d..in;J to what degree you concur with the statement. other departrrent (division> chairs in my college (or similar unit>: 1. Respect me am want to act in ways that 100rit my respect am admiration. stroJ'x1ly agree ( ) agree ( ) neutral ( ) disagree ( ) strongly disagree ( ) 2. Respect my c:orrpetence about thin;Js in which I have 100re experience. stroiXJlY agree neutral disagree stroJ'x1ly agree disagree ( ) ( ) ( ) ( ) ( ) 3 Believe I can give special help to those who cooperate with me. stroJ'x1ly agree neutral disagree stroJ'x1ly agree disagree ( ) ( ) ( ) ( ) ( ) 4. Believe I can apply pressure on those who do not cooperate with me. stroJ'x1ly agree ( ) agree ( ) neutral ( ) disagree ( ) stroJ'x1ly disagree ( ) 5. Perceive me as havil'XJ a legitimate right to eJCPeCI: their acceptance of suggestions in which I have a stroJ'x1 interest. stl:on;ly agree neutral disagree stroiXJly agree disagree ( ) ( ) ( ) ( ) ( ) 6. Defer to my jOOgment in areas with which I am JIX)re familiar. stl:on;ly agree neutral disagree stroiXJly agree disagree ( ) ( ) ( ) ( ) ( ) 7. Believe I am able to make thin;Js difficult for them if they fail to follow my advice. stroiXJly agree ( ) agree ( ) neutral. ( ) 137 disagree ( ) stroiXJly disagree ( )

PAGE 146

8. Feel obligated to follow rrrt suggestions in IOOSt cases. strcn:Jly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 9. Believe they can personally benefit by cooperatirq with me. strorqly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 10. Believe that followirq rrrt advice will result in better decisions. strorqly agree ( ) agree ( ) disagree ( ) ( ) strorqly disagree ( ) 11. COoperate with me because they have high regard for me as an i.rx:li. vidual. strorqly agree ( ) agree ( ) neutral ( ) disagree ( ) stroi'l;}ly disagree ( ) 12. Believe I can penalize those who do not follow rrrt suggestions. strorqly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 13. Feel they have to cooperate with me. strcn:Jly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 14. Cooperate with me because they wish to be identified with me. strcn:Jly agree ( ) agree ( ) disagree ( ) ( ) strorqly disagree ( ) 15. Perceive that oooperatirq with me can positively on their performance. strcn:Jly agree ( ) agree ( ) neutral ( ) 138 disagree ( ) strorqly disagree ( )

PAGE 147

Please respord to each statement below aa::ordin;J to what degree ycu concur with the statement. 1he Dean/Vice President/other imivtdual to whom I am responsible; 1. Respects me and wants to act in ways that merit my respect and admiration. stron:;rly agree ( ) agree ( ) ( ) disagree ( ) stron:;rly disagree ( ) 2. Respects my c:anpetence about thin;Js in which I have mre experience. stron:;rly agree ( ) agree ( ) neutral ( ) disagree ( ) stron:;rly disagree ( ) 3. Believes I can give special help to those who cooperate with me. strongly agree ( ) agree ( ) neutral ( ) disagree ( ) stron:;rly disagree ( ) 4. Believes I can aJ;:PlY pressure on those who do not cooperate with me. stron:;rly agree ( ) agree ( ) neutral ( ) disagree ( ) stron:;rly disagree ( ) 5. Perceives me as hav.in;r a le;itimate right to expect acceptance of suggestions in which I have a strong interest. strongly agree neutral disagree strongly agree disagree ( ) ( ) ( ) ( ) ( ) 6. Defers to my judgment in areas with which I am mre familiar. strongly agree neutral disagree strcrgly agree disagree ( ) ( ) ( ) ( ) ( ) 7. Believes I am able to make thin;Js difficult for hi.uy'her if my advice is not follCM:!d. strongly agree ( ) agree ( ) neutral ( ) 139 disagree ( ) strongly disagree ( )

PAGE 148

B. Feels obligated to follow Irrf suggestions in most cases because of the department I chair. stroD;ly agree ( ) agree ( ) neutral ( ) disagree ( ) stroD;lY disagree ( ) 9. Believes that cooperatiD; with rre will result to his/her personal benefit. stroD;lY agree ( ) agree ( ) neutral ( ) disagree ( ) stroD;ly disagree ( ) Believes that followiD; Irrf advice will result in better decisions. stroD;ly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 11. Cooperates with me because of a high regard for me as an individual. stroD;ly . neutral disagree StroD;lY agree . disagree agree ( ) ( ) ( ) ( ) ( ) 12. Believes I can penalize those who do not follow Irrf suggestions. stroD;ly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 13. Feels obligated to cooperate with me. stroD;ly agree neutral disagree strorqly agree disagree ( ) ( ) ( ) ( ) ( ) 14. Cooperates with me because of a desire to be identified with me. agree ( ) agree ( ) neutral ( ) disagree ( ) strorqly disagree ( ) 15. Perceives that cooperatiD; with me can positively inpact on the perfonnance of the college/sChool as a whole. stroD;ly agree ( ) agree ( ) neutral ( ) 140 disagree ( ) stroD;lY disagree ( )

PAGE 149

APPENDIX B Chi -square Test Results Power Perceptions arx1 Burnout CO!tp:>nents Table 8.1. -Perception of Referent Power Relative to Faculty Does Exist Uncertain or NonDe personexistent alization Obs. Exp. Obs. Exp. Totals Level LOW 40 32.84 7 14.16 47 MEDIUM 62 63.59 29 27.41 91 HIGH 14 19.57 14 8.43 28 ALL 116 116 50 50 166 Chi-square = 1 0.567, df = 2, p <.01 Table 8.2.-Perception of Referent Power Relative to Faculty Does Exist Uncertain or NonEmotional existent Exhaustion Obs. Exp. Obs. Exp. Totals Level LOW 48 44.02 15 18.98 63 MEDIUM 37 34.24 12 14.76 49 HIGH 31 37.74 23 16.26 54 ALL 116 116 50 50 166 Chi-square = 5.921, df = 2, p <.1 0 141

PAGE 150

Table 8.3. Perception of Referent Power Relative to Faculty Does Exist Uncertain or NonPersonal existent AccomplishObs. Exp. Obs. E:xp. Totals ment Level LOW 30 22.36 2 9.64 32 MEDIUM 52 49.62 19 21.38 71 HIGH 34 44.02 29 18.98 63 ALL 116 116 50 50 166 Chi-square= 16.621, df = 2, p <.01 Table 8.4. Perception of Expert Power Relative to Faculty Does Exist Uncertain or NonDepersonexistent alization Obs. Exp. Obs. Exp. Totals Level LOW 46 43.6 1 3.4 47 MEDIUM 88 84.42 3 6.58 91 HIGH 20 25.98 8 2.02 28 ALL 154 154 12 12 166 Chi-square = 22.940, df = 2, p <.01 142

PAGE 151

Table 8.5. Perception of Expert Power Relative to Faculty Does Exist Uncertain or NonPersonal existent AccomplishObs. Exp. Obs. Exp. Totals ment Level LOW 32 29.69 0 2.31 32 MEDIUM 68 65.87 3 5.13 71 HIGH 54 58.44 9 4.56 63 ALL 154 154 12 12 166 Chi-square= 8.127, df = 2, p <.02 Table 8.6. -Perception of Reward Power Relative to Faculty Does Exist Uncertain or NonPersonal existent AccomplishObs. Exp. Obs. Exp. Totals ment Level LOW 26 20.82 6 11.18 32 MEDIUM 48 46.19 23 24.81 71 HIGH 34 40.99 29 22.01 63 ALL 108 108 58 58 166 Chi-square= 7.302, df = 2, p <.03 143

PAGE 152

Table 8.7. --Perception of Legitimate Power Relative to Faculty Emotional Exh sf au 1on Level LOW MEDIUM HIGH ALL Does Exist Ob Ex s. CQ. 18 18.22 18 14.17 12 15.61 48 48 Uncertain Non-Existent Ob Ex s. CQ. Ob Ex s. CQ. 40 34.91 5 9.87 28 27.16 3 7.67 24 29.93 18 8.46 92 92 26 26 Chi-square= 19.830, df = 4, p <.01 Table 8.8. -Perception of Legitimate Power Relative to Peers Does Exist Uncertain Non-Existent T ota s 63 49 54 166 Emotional Exhaustion Obs Calc Obs Calc Obs Calc Totals Level LOW 10 6.83 34 34.54 19 21.63 63 MEDIUM 7 5.31 30 26.86 12 16.83 49 HIGH 1 5.86 27 29.6 26 18.54 54 ALL 18 18 91 91 57 57 166 Chi-square = 11.339, df = 4, p <.05 Table 8 9 -Perception of Reward Power Relative to Peers -------------------------------------------------i Personal Does Exist Uncertain Non-Existent A r h ccomp 1s Ob E s. xp. Ob E 5. xp. Ob Ex 5. cp. Ttl oas ment Level LOW 14 9.06 12 16.96 6 5.98 32 MEDIUM 18 20.1 45 37.64 8 13.26 71 HIGH 15 17.84 31 33.4 17 11.76 63 ALL 47 47 88 88 31 31 166 Chi-square = 1 0.844, df = 4, p <.05 144

PAGE 153

Table 8.10 .-Perception of Legitimate Power Relative to Deans Personal Does Exist Uncertain Non-Existent A rh ccomp1sOb E s. xp. Ob E s. xp. Ob E T s. xp. otas ment Level LOW 11 6.36 14 20.05 7 5 59 MEDIUM 10 14.12 53 44.48 8 12.4 HIGH 12 12.52 37 39.47 14 11.01 ALL 33 33 104 104 29 29 Chi-square = 1 0.947, df = 4, p <.05 Table 8.11. -Perception of Coercive Power Relative to Deans Emotional Exh sf au 1on Level LOW MEDIUM HIGH ALL Does Exist Ob E s. xp. 7 8.73 11 6.79 5 7.48 23 23 Uncertain Non-Existent Ob E s. xp. Ob E s. xp. 37 29.6 19 24.67 21 23.02 17 19.19 20 25.38 29 21.14 78 78 65 65 Chi-square= 11.412, df = 4, p <.05 145 32 71 63 166 T t I oas 63 49 54 166

PAGE 154

APPENDIX C Chi-Square Test Results Organizational/Personal Olaracteristics and Burnout components Table C.1. Influence in Resource Allocation as Compared to Faculty Emotional Much More Somewhat More Same or Less Exhaustion Obs. Exp. Obs. Exp Obs. Exp. Totals Level LOW 28 27.33 28 23.53 7 12.14 63 MEDIUM 25 21.25 9 18.3 15 9.45 49 HIGH 19 23.42 25 20.17 10 10.41 54 ALL 72 72 62 62 32 32 166 Chi-square= 13.707, df = 4, p <.01 Table C.2. Percent of Faculty Paid More Than Chairperson De person0-10% 11-30% Over30% alization Obs. Exp. Obs. Exp Obs. Exp. Totals Level LOW 32 29 .16 5 9 .91 10 7.93 47 MEDIUM 60 56.46 21 19.19 10 15.35 91 HIGH 11 17.38 9 5.9 8 4.72 28 ALL 103 103 35 35 28 28 166 Chi-square = 11.7 43, df = 4, p <.05 146

PAGE 155

Table C.3.-Percent of Faculty Paid More Than Chairperson Personal 0-10% 11-30% Over30% AccomplishObs. Exp. Obs. Exp. Obs. Exp. Totals ment Level LOW 26 19.86 4 6.75 2 5.39 32 MEDIUM 48 44.05 13 14.97 10 11.98 71 HIGH 29 39.09 18 13 28 16 10.63 63 ALL 103 103 35 35 28 28 166 Chi-square = 13.094, df = 4, p <.05 Table C.4. Bachelors Degrees Offered by Department ? Emotional NO YES Exhaustion Obs. Exp. Obs. Exp. Totals Level LOW 8 9.11 55 53.89 63 MEDIUM 12 7.08 37 41. 92 49 HIGH 4 7.81 50 46.19 54 ALL 24 24 142 142 166 Chi-square = 6.315, df = 2, p <.05 147

PAGE 156

Table C.5. Second Most Important Personal Reason for Becoming Department Chairperson Break or Rep. Fac. to Rep. Adm. Fac. EncDepersonHigher Psn. Adm. to Fac. ouragement Other alization Obs. EXQ. Obs. Ex_Q. Obs. Exp. Obs. Exp. Obs. Exp. Totals Level LOW 5 5.1 12 9.91 6 10.8 11 7.93 13 13 47 MEDIUM 11 9.87 20 19.2 19 .. 20.8 16 15.4 25 26 91 HIGH 2 3.04 3 5.9 13 6.41 1 4.72 9 7.9 28 ALL 18 18 35 35 38 38 28 28 47 47 166 Chi-square = 15.759, df = 8, p <.05 Table C.6. Number of Faculty in Department Deperson1 to 6 7 to 12 13 to 18 Over 18 alization Obs. Exp. Obs Exp. Obs Exp Obs. Exp Totals Level LOW 20 12.17 18 16.14 5 9.63 4 9.06 47 MEDIUM 17 23.57 33 31.25 22 18.64 19 17.54 91 HIGH 6 7.26 6 9.61 7 5.73 9 5.4 28 ALL 43 43 57 57 34 34 32 32 166 Chi-square = 17.211, df = 6, p <.01 148

PAGE 157

APPENDIX D Analyses of Variance on Dependency Pairs Table D.1. Analysis of Variance On Emotional Exhaustion (Referent Power Relative to Faculty) SOURCE DF SUMSQ MEANS F p Ref Power 1 3.283 3.283 4.76 0.031 Error 164 113.229 0.69 Total 165 116.512 LEVEL N MEAN STDEV POOLED ST DEV Exists 116 1.8534 0.8157 Uncertain or so 2.16 0.8657 None 0.8309 Table D.2. Analysis of Variance On Depersonalization (Expert Power Relative to Faculty) SOURCE DF SUMSQ MEANS F p Ref Power 1 4.633 4.633 11.14 0.001 Error 164 68.192 0.416 Total 165 72.825 LEVEL N MEAN STDEV POOLED ST DEV Exists 116 1.7759 0.6472 Uncertain or 50 2.14 0.6392 None 0.6448 149

PAGE 158

Table 0.3. -Analysis of Variance On Oepersonalization (Referent Power Relative to Faculty) SOURCE OF SUMSQ MEANS F p Ref Power 1 6.298 6.298 15.53 0 000 Error 164 66.527 0.406 Total 165 72.825 LEVEL N MEAN STDEV POOLED ST DEV Exists 116 1.8312 0 6346 un certain or 50 2.5833 0.6686 None 0.6369 Table 0.4. -Analysis. of Variance On Personal Accomplishment (Referent Power Relative to Faculty) SOURCE OF SUMSQ MEANS F p Ref Power 1 8.929 8 929 18.24 0.000 Error 164 80.282 0.49 Total 165 89.211 LEVEL N MEAN STDEV POOLED ST DEV Exists 116 2.0345 0.7452 Uncertain or 50 2.54 0.5789 None 0.6997 Table 0.5. Analysis of Variance On Personal Accomplishment (Expert Power Relative to Faculty) SOURCE OF SUMSQ MEANS F p Ref Power 1 4.104 4.104 7.91 0.006 Error 164 85.107 0.519 Total 165 89.211 LEVEL N MEAN STDEV POOLED ST DEV Exists 116 2.1249 0.7359 Uncertain or 50 2.75 0.4523 None 0.7204 150

PAGE 159

Table D.6. Analysis of Variance On Personal Accomplishment (Reward Power Relative to Faculty) SOURCE DF SUMSQ MEANS F p Ref Power 1 3.924 3.924 7.55 0.007 Error 164 85.287 0.52 Total 165 89.211 LEVEL N MEAN STDEV POOLED ST DEV Exists 116 2.0741 0.7451 Uncertain or 50 2.3966 0.6738 None 0.7211 Table D.7. Analysis of Variance On Emotional Exhaustion (Legitimate Power Relative to Faculty) SOURCE DF SUMSQ MEANS F p Leg Power 2 9.545. 4.772 7.27 0.001 Error 163 106.967 o:656 Total 165 116.512 LEVEL N MEAN STDEV POOLED ST DEV Exists 48 1.875 0.7889 Uncertain 92 1.8261 0.8202 Non Existent 26 2.5 0.8124 0.8101 Table D.8. Analysis of Variance On Emotional Exhaustion (legitimate Power Relative to Peers) SOURCE DF SUM .SQ MEANS F p Coercive Pow 2 5.41 2.705 3.97 0.021 Error 163 111.102 0.682 Total 165 116.512 LEVEL N MEAN STDEV POOLED ST DEV Exists 18 1 5 0.6183 Uncertain 91 1.9231 0.8196 Non Existent 57 2.1228 0.8878 0.8256 151

PAGE 160

Table D.9. Analysis of Variance On Emotional Exhaustion (Coercive Power Relative to Dean) SOURCE DF SUMSQ MEANS F p Leg Power 2 4.93 2.465 3.6 0.03 Error 163 111.582 0.685 Total 165 116.512 LEVEL N MEAN STDEV POOLED ST DEV Exists 23 1.913 0.7332 Uncertain 78 1.7821 0.832 Non Existent 65 2.1538 0.852 0.8274 Table D.1 0. Analysis of Variance On Depersonalization (Second Most Important Reason for Taking Position) SOURCE DF SUMSQ MEANS F p 2nd Reason 4 5.841 1.46 3.51 0.009 Error 161 66.984 0.416 Total 165 72.825 LEVEL N MEAN STDEV POOLED ST DEV Take Break 18 1.833 0.6183 Rep. Fac. 35 2.7429 0.6108 Rep. Admin. 38 2.1842 0.6919 Fac. Encrgmnt 28 1.6429 0.5587 Other 47 1.9149 0.6862 0.645 Table D.11. Analysis of Variance On Depersonalization (Percent of Faculty Paid More Than Chairperson) SOURCE DF SUMSQ MEANS F p Faculty% 2 2.707 1.353 3.15 0.046 Error 163 70.118 0.43 Total 165 72.825 LEVEL N MEAN STDEV POOLED ST DEV 0 10% 103 2.707 0.6161 11 30% 35 70.118 0.6311 Over 30% 28 72.825 0.8133 0.6559 152

PAGE 161

Table 0.12. Analysis of Variance On Personal Accomplishment (Percent of Faculty Paid More Than Chairperson) SOURCE OF SUMSQ MEANS F p Faculty o/o 2 6.898 3.449 6.83 0.001 Error 163 82.313 0.505 Total 165 89.211 LEVEL N MEAN STDEV POOLED ST DEV 0 w 10% 103 2.0291 0.7337 11 w 30% 35 2.4 0.6945 Over 30% 28 2.5 0.6383 0.7106 Table 0.13. Analysis of Variance On Depersonalization (Number of Faculty in Department) SOURCE DF SUMSQ MEANS F p Faculty No. 3 5.809 1.936 4.68 0.004 Error 162 67.017 0.414 Total 165 72.825 LEVEL N MEAN STDEV POOLED ST DEV 1 to 6 43 1.6744 0.7145 7 to 12 57 1.7895 0.6192 13 to 18 34 2.0588 0.6001 Over 18 32 2.1562 0.6278 0.6432 153

PAGE 162

BIBLIOORARIY Adams, J. stacy. 1963. Toward an understanding of inequity. JOUD'lal. of Abnonnal. and Social Psychology, 67, 422-24. Arrlerson, M., and E. Iwanicki. 1984. Teacher motivation and its relationship to buznout. Educational Administration Quarterly, 20: 109-32. Baldridge, J. Victor. 1971. Power and conflict in the university. New York: Wiley. Baron, Robert A. 1986. Behavior in o:r:ganizations. Newton, Mass.: Allyn and Bacon, Inc. Beehr, T.A., and J.E. Newman. 1978. Jab stress, employee health, and organizational effectiveness. Personnel Psychology, 31, 665-99. Bennett, John B. 1983. Managing the academic deparbnent. New York: Mac::millan Publishing Company. Bennis, Warren, and Burt Nanus. 1985. leaders: The strategies for taking charge. New York: Hazper and ReM. Bimbaum, R. 1988. How colleges work: The cybernetics of academic governance and leadership. San Francisco, calif. : Jossey..;.Bass, Inc. Blackburn, R., s. Ho:rc:Mitz, D. F.dington and D. IO.os. 1986. University faculty and administrator responses to jab strain. Research in Higher Education, 25(1): 31-41. Blau, Peter M. 1973. The organization of academic work. New York: Wiley. Booth, D.B. 1982. '!he deparbnent chair: Professional development and role conflict. Washington, D.C.: American Association of Higher Education. Brann, James, and Thomas Ennnet (eds.). 1972. The academic department or division chainnan: A complex role. Detroit: MI Balamp Publishing. 154

PAGE 163

Burke, R.J., and E. Deszca. 1986. Correlates of psychological bumout phases among police officers. Human Relations,-39: 487-502. Bums, James MacGregor. 1978. leadership. New York: Harper & Row. cahoon, A.R. and Rowney, J. 1984. Managerial bmnout: A camparison by sex ancl level of responsibUity. Journal of Health and Human Resource Administration, 7: 249-63. ___ 1989. A comparison of bmnout across public/private sector managers. Joumal of Health and Human Resources Administration, 12: 79-94. carroll, J. F. X. 1980. staff bmnout as a fonn of ecological dysfunction. Contemporary Drug Problems, 8: 207-25. cartwright, (Ed.). 1959. studies in social power. University of Michigan. Cllemis, C., E.S. Egnatios, S. Wacker, and B. O'Dc:IWd. 1985. '!he professional mystique ancl bmnout in public sector professionals. Social Policy. ariannonte, A. J. 1983. Psychological correlates of burnout in clergymen. Dissertation Abst:ra.cts International, 44: 433. Cialdini, Robert B. 1988. Influence: Science and practice, 2d e:i. Glenview, Ill: Scott, Foresman. Clarke, Betty M. 1991. IJhe burnout maze: An exploration of the interrelationships of burnout, hardiness, depression, response 100de to burnout, and demographic variables. Ph.D diss., University of Colorado-Denver. Cohen, M., and J. March. 1974. leadership and ambiguity: '!he American college president. New York: McGraw Hill. Cook, 'lhamas D., ancl J::bnald T. Campbell. 1979. Quasi-Experimentation. Boston: Houghton Miflin Co. Cordes, c., and T.W. Dougherty. 1993. A review and an integration of research on job burnout. Academy of Management Review, 38(4): 621-56. Corson, J. 1959. Governance of colleges and universities. New York: McGraw Hill Book Company, Inc 155

PAGE 164

cresswell, J. 1986. '!he academic deparbnent: Faculty growth and development practices of excellent chairs. Lincoln, Neb: University of Nebraska. Demerath, Nicholas J., Richard w. stephens, and R. Robb Taylor. 1967. Power, presidents and professors. New York: Basic Books, Inc. Dilley, Frank B. 1972. 'nle department chainnan as academic plarmer. In James Brann and 'lhamas Ennnet, eds. '!he academic deparbnent or division chainnan: A complex role. Detroit, Mich.: Balamp PUblishing. Doyle, Edward A. 1953. '!he status and functions of the deparbnental chainnan. washington, D.C.: '!he catholic. University of America Press. Dunham, J. 1978. staff stress in residential wark. Social Work Tbday, 9(45): 18-20. Dyer, Wayne. 1981a. When I say no. I feel guilty. New York: Avon Books. ____ 198lb. Your erroneous zones. New York: Avon Books. Emener, W.G., R.S. Illck, and F.X. Gobs. 1982. 'Iheoretical investigation of the construct burnout. Journal of Rehabilitation Administration, 6: 188-96. F'al::ber, B.A. 1983. Introduction: A critical perspective on burnout. In B.A.F'al::ber. Stress and burnout in the human services profession. New York: Pergamon Press. Fiorelli, Joseph. 1988. Power in work groups: Team members' perspectives. Human Relations, 41(1): 1-12. Firth, H., and Britton. 1989. Burnout, absence and turnover among British nursing staff. Journal of Occupational Psychology, 62: 55-59. Fletcher, David. 1969. Administrative functions and o:rganizational networks utilized by academic deparbnent heads in selected instiutions of higher education. Unpublished doctoral diss.: University of Denver. French, John R.P. ,Jr., and Bertram Raven. 1959. '!he bases of social power. In Studies in Social Power, ed. D. cartwright, Arm Al:Dor, Mich.: Institute for Social Research. 156

PAGE 165

Freudenberger, H.J. 1977. :au:n-out: the organizational menace. Training and Development Journal, 31(7): 26-27. Gaines, J.R., and S.E. Jenniar. 1983. Emotional exhaustion in a high stress organization. Academy of Management Review, 26: 567-86. Gardner, Jolm w. 1990. On leadership. New York: Free Press. Ginsburg, S.G. 1974. '!he problem of the burned out executive. Personnel Journal, 53: 598-600. Gmelc:h, Walter H. Paying the Price for Academic leadership: Department Olair Tradeoffs. Educational Record, 72(3): 45. Gmelc:h, W.H., and J.S. Bun'ls. 1993. '!he cost of academic leadership: Department chair stress. Innovative Higher Education, 17(4): 259-69. Golembiewski, Robert T. 1988. Perspectives on Psychological :au:nout, IV: Empirical Research, Part 1. Journal of Health and Human Resources Administration, 13 (2): 63-75. ___ 1989. A note on Leiter's study: Highlighting the two m:xiels of burnout. Group and Organization studies, 14: 5-13. Golembiewski, R. T. and R. Mlmzenrider. 1981. Efficacy of three versions of one burnout Itieasure: MBI as a total score, sub-scale score or phases? Journal of Health and Human Resources Administration, 7: 228-46. ___ 1984. Phases of psychological burnout and organizational co-variants: A replication using nonns from a large p:>pUlation. Jow:nal of Health and Human Resources Administration, 7: 290-323. ---1988. Phases of burnout. New York: Praeger PUblications. 1990. Alternative combinations of phases of bumout: An illustrative exercise. Journal of Health and Human Resoriroes Administration, 3(4): 489-506. ____ 1991. Health related covariants of bumout: A pilot study. Journal of Health and Human Resources Administration (Fall): 204-24. 157

PAGE 166

Golembiewski, R.T., R. MLmzenrider, and D. carter. 1983. Phases of progressive burnout and their worksite covariants: Critical issues in CD research and praxis. '!he Journal of Applied Science, 19(4): 461-81. Grinun, Robert B. 1986. '!he contribution of personality type and selected individual, role/task, and organizational factors to the experience of burnout. Ph.D. diss. University of Colorado, Boulder. Hackman, Judith. 1985. Power and centrality in the allocation of resources in colleges and universities. Administrative Science Quarterly, 30: 61-77. Hackman, J.R., and G.R. Oldham. 1974. '!he Job Diagnostic SW:Vey: An i.nstrunlent for the diagnosis of jobs and the evaluation ofjob redesign projects. JSAS catalcxr of Selected Documents in Psychology, 4: 148. Hal.berstam, O:tvid. 1986. '!he reckoning. New York. William Morrow. Harxiy, Charles B. 1983. Understanding organizations. Baltilnore, M:l. Penguin Books, Ltd. Helmich, Donald L. and Warren B. Brown. 1972. successor type and organizational change in the corporate enterprise. Administrative Science Quarterly, 17: 371-81. Hill, Winston, and Wendell French. 1967. Perceptions of the power of department chainnen by professors. Administrative Science Quarterly, 11(4): 548-567. Hodgkinson, Harold L., and Richcu'd L.Meeth (eds.). 1971. Power and authority: Transfonnation of campus governance. San Francisco: Jessey-Bass, Inc. Howard, J.H., E.M. scott, R.L. Wiener, K.S. Boynton, and W.M. Rooney. 1991. Is a behavioral measure the best estimate of behavioral parameters? Perllaps not. Applied Psvcholcxrical Measurement, 4 (3) : 293-311, as reported in Betty M. Clarke. '!he burnout maze. Ivancevich, J .A. and M. T. Matteson. 1980. Stress and work: a managerial perspective. Glenview, Ill.: Scott, Foresman. Iwanicki, E., and R.L. Schwab. 1981. A cross-validation of the Maslach Bumout Inventory. Educational and Psycholcxrical Measurement, 41: 1167-74. 158

PAGE 167

Jayaratne, s. arx1 W.A. Chess. 1984. '!he effects of emotional support on perceived job stress arx1 strain. '!he Journal of lg)lied Behavioral Science, 20(2): 141-53. Kahill, s. 1988. Synptams of professional burnout: A review of the enpa-ical evidence. canadian Psychology, 29: 284-97. Kahn, R. 1978. Job bumout: Prevention arx1 remedies. Public Welfare, 36: 61-63. Kanter, Rosabeth M. 1977. Men and women of the corporation. New York. Basic Books. Katz, Daniel, arx1 Robert L. Kahn. 1966. '!he social psychology of o:rganizations. New York: John Wiley and Sons, Inc. Kinneyer, S.L., arx1 T.W. Ik>ugherty. 1988. Work load, tension and coping: Moderating effects of supervisor support. Personnel Psychology, 41: 125-39. Koeske, G.F., arx1 R.D Koeske. 1989. Construct validity of the Maslach Bumout Inventoey: A critical review arx1 reconceptualization. Journal of lg)lied Behavioral Science, 25(2): 131-32. Kotter, John P. 1985. Power and influence: Beyond fonnal authority. New York: Free Press. Krackhanlt, David. 1990. Assessing the political landscape: structure, cognition arx1 poWer in organizations. Administrative Science Quarterly, 35: 342-69. Iasswell, H. and A. Kaplan. 1950. Power and society: A framework for oolitical inquiry. New Haven: Yale University Press. I.ee, R.T. arx1 B.E. Ashforth. 1993. A further examination of managerial burnout: Toward an integrated model. Journal of Opganizational Behavior, 14: 3-20. leiter, M.P., arx1 c. Maslach. 1988. The impact of inteJ:personal enviromnent on burnout and organizational commitment. Journal of Organizational Behavior, 9: 297-308. leiter, M.P., arx1 K.A. Meechan. 1986. Role structure and burnout in the field of business services. Journal of Applied Behavioral Science, 22: 47-52. 159

PAGE 168

Likert, Rensis. 1961. New patterns of management. New York: Ma;raw Hill. Martin, Katln:yn A. 1966. Role perceptions of management and leadership responsibilities of academic chairs in colleges/schools of fine arts as perceived by deans and department chairs. Ph.D. diss., University of San Francisco. Maslach, Christina. 1976. Burned-out. Human Behavior, 5(9): 16-22. ___ 1978a. Rlmout: How people cope. Public Welfare, 36: 56-58. ___ 1978b. '!he client role in staff burnout. Journal of Social Issues, 34(4): 111-24. ___ 1979. '!he bumout syndrome and patient care. In c. Garfield, ed., Stress and surviVcu.: 'Ihe emotional realities of life-threatening illness. St. I.Duis: Mosby. ___ 1981. Bumout: A social psychological analysis. In J.W.Jo:nes, ed., 'Ihe burnout syndrome. Park Ridge, Ill.: I.Dmon House Press. ___ 1982a. Rlmout: '!he cost of caring. Englewood Cliffs, N.J. : Prentice Hall. ----.1982b. Urrlerstanding burnout: Definitional issues in analyzing a complex phenomena. In W.S.Paine, ed. 1 JOb stress and burnout. Beverly Hills, calif.: Sage. Maslach, c., and A. Pines. 1979. Bumout: 'Ihe loss of htnnan caring. In A. Pines and c. Maslach, eds., Experiencing social psychol
PAGE 169

McClelland, David c. 1974. 'lhe two faces of power. Organizational psychology: A book of readings. Englewood Cliffs, N.J.: Prentice Hall, Inc. ____ 1975. Power: 'Ihe inner experience. New York: II.Vington. Mechanic, David. 1962. Sources of power of lower participants in complex organizations. Administrative Science Quarterly, 7: 349-64. Merton, Robert K. 1957. Social theory and social structure. Toronto, canada: '!he Free Press of Glencoe. MonsOn, Olarles H., Jr. 1972. '!he University of Utah's department chainnen training program. In James Brann and '1homas Emmet, eds. 'lhe academic department or division chainna.n: A complex role. Detroit, Mich.: Balamp Publishing. Moore, William L., and Jeffrey Pfeffer. 1980. The relationship between departmental power and faculty careers on two canpuses: 'Ihe case for structural effects on faculty salaries. Research in Higher Education, 13: 291-306. Norton, M. Scott. 1978. A study of the department chairperson in colleges of education. Tempe, Ariz. :Arizona State University. ---1980. The academic chair: Tasks and resoonsibilities. Tempe, Ariz. :Arizona state university. Pfeffer, Jeffrey. 1992. Power in ornanizations. Boston, Mass.: Harvard Business School Press. Pfeffer, Jeffrey, and William L. Moore. 1980. Power in university budgeting: A replication and extension. Administrative Science Quarterly, 23: 637-53. Pines, A., and D. Kafry. 1981. Tedimn in the life of professional women as COITpared with men. Sex Roles, 7:10, 963-77. Pretty, G.M.H., M.E. McCarthy, and V.M. Catano. 1992. Psychological environments and bumout: Gender considerations within the corporation. Journal of Organizational Behavior, 13: 701-11. Quick, J. and J. Quick. 1979. Reducing stress through preventative management. Human Resource Management, 18. 161

PAGE 170

Reading, Hugo F. 1971. A dictionary of the social sciences. lorrlon: Routledge ani Kegan Paul. Ridnnan, Barry M. I ani Richard M. Fanner. 197 4. leadership r goals. ani power in higher education. San Francisco: Jossey-Bass. Rue, leslie w. and IJ.oyd L. Byars. 1989. Management. Homewood, Ill. : Richard D. Il:win, Inc. Russell, D.W., E. Allmaier, ani D. Van Velsen. 1987. Jab-related stress, social support ani bmnout among classroom teachers. Joumal of Applied Psychology, 72: 269-74. Ryan, B. B. Joiner, arx1 T. Ryan. 1985. MINITAB handbook. 2m ed. Boston: IWS-KENr Publishing. Salancik, Gerald R., ani Pfeffer, Jeffrey. 1974. '!he bases ani use of power in organizational decision making: '!he case of a university. Administrative Science Quarterly, 19: 453-73. 5ayles, .leonard R. 1979. Ieadership. New York: McGraw Hill. Schwab, R. L. ani E. F. Iwanicki. 1982. Perceived role conflict, role ambiguity, ani teacher bUrnout. Educational Administration Quarterly, 18: 60-74. Selye, H. 1983. 'Ihe stress concept: Past, present and future. In stress research, C.L. Cooper, ed., New York: John Wiley & sons. Silrpson, J.A., and E.S.C. Weiner. 1989. The Oxfo:rd English dictionary. Oxfo:rd: Clarerrlon Press. Singleton, B. 1987. Sources and consequences of role conflict and role ambiguity among department chairs. capstone Journal of Education, 7(2): 39-50. Sites, Paul. 1973. Control: The basis of social order. New York: nmel.len Publishi.n;J Co. Inc. Siu, R.G.H. '!he craft of power. New York: John Wiley & Sons. Spekman, R.E. 1979. Influence and infonnation: An explorato:cy investigation of the bounda:cy role person's basis of power. Academy of Management Journal, 24: 104-17. 162

PAGE 171

Spiegel, Murray R. 1975. Probability and statistics. New York: McGraw Hill Book Cc:mpany, Inc. staton-Spicer, A., arrl c. Spicer. 1987. Socialization of the academic chail:person: A typology of cammunication dilrensions. Fducational Administration Quarterly, 23 {1): 41-64. stogdill, Ralph M. 1974. Handbook of leadership: A sw:vey of theocy and research. New York: '!he Free Press. Tennis, c. E. 1984. An exploration of the relationships between jab bumout arrl selected worl