t. 3. z. .
..; I134. .. . 1..

 

 

 

VJ.
:3 .r... . .\..
. .1. «.

 

2.. gr? ;
i: . .12....2

 

I. 411.! it:
3:11: 5:11.

3.. 5, I
«{5}. :(.:.o.}.5¢.::r:r. 1‘
3‘1, 3

x . 2.... 9.3.1:)..3‘53;
. .5. I. .33).. I! .. In, (an-(1r.
~A¢C€<¢tnipzia .62! ) I
4) ).).ry(r;r.p 11:... re}
.i)).l:(.r.:: (I: i»

 

 

 

n ’" .. f...
0) Kc." ( (jun/9’ /

lIllllijiljlilfllljflliillllll

2488

 

”entire? W
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

Communication, Stress and Burnout in Education

presented by

Sandra M. Starnaman

has been accepted towards fulfillment
of the requirements for

Masters degree in Communication

 

 

Major professor

[hue February 22, 1990

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

. "T
V .
'2. . “1* (W “ '1"
r m" m - ‘ '
- “ I

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

DATE DUE DATE DUE DATE DUE

 

I

-fi,;_\i N

4

 

 

 

 

 

4'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Inditution
cmma-m

 

COMMUNICATION, STRESS AND BURNOUT IN EDUCATION

BY

Sandra M. Starnaman

A THESIS

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

MASTER OF ARTS

Department of Communication

1989

loos 40173

ABSTRACT
COMMUNICATION, STRESS AND BURNOUT IN EDUCATION
BY

Sandra M. Starnaman

Past studies of burnout have argued that teachers, as
caregivers, have a greater risk of burnout due to their
constant involvement with students. These studies have
shown repeatedly that the variables of overload, role
conflict and role ambiguity are associated with either the
burnout variables of emotional exhaustion, depersonalization
and reduced personal accomplishment and/or job satisfaction,
and occupational commitment.

There are, however, variables related to communication

within schools that are perceived as mediating the burnout

process. These are supervisory support and participation in

decision making.

This paper addresses the interactive relationship that
exists between all of these variables through the development

of a causal model of the burnout process.

To Craig, with love

iii

ACKNOWLEDGMENTS

I would like to express my thanks to some of the people
who helped make this work possible.

Thank you

Dr. Katherine Miller - for your guidance and support.

Dr. James Stiff and Dr. Cass Book — for your patience
and insights.

Craig — for your love and belief in me.

Mom - for believing her children could do anything
they set out to do.

Sabrina, Eric and Perry — for constantly worrying about
me.
Maureen — for making sure I kept both feet on the
ground.

iv

 

List of
List of
Chapter

1.

2.

TABLE OF CONTENTS

Tables . . . . . . . . . . . . . . . . . . .

Figures . . . . . . . . . . . . . . . . . .

THEORY AND LITERATURE REVIEW . . . . . . .
Burnout . . . . . . . . . . . . . . . .
Organizational Stressors . . . . . . . .

Workload . . . . . . . . . . . . . . .
Role Stress . . . . . . . . . . . . .
Communication Variables . . . . . . . .
Social Support . . . . . . . . . . . .

Participation in Decision Making

Outcome Variables . . . . . . . . . . .
METHODS . . . . . . . . . . . . . . . . .
Sample Description . . . . . . . . . . .
Administration Procedure . . . . . . . .
Instrumentation . . . . . . . . . . . .
Burnout . . . . . . . . . . . . . . .
Participation in Decision Making . . .

Work Satisfaction and Occupational
Commitment . . . . . . . . . . . . .

Social Support . . . . . . . . . . . .

Role Overload, Role Conflict and Role
Ambiguity . . . . . . . . . . . . .

V

Page

viii

21
27
28
29
29
30

30

30

31

31

Chapter

Analysis . . . . . . . . . . .
Confirmatory Factor Analysis

Path Analysis . . . . . . .

3. RESULTS . . . . . . . . . . . .
Confirmatory Factor Analysis .
Preliminary Path Analysis . .

Path Analysis . . . . . . . .

4. DISCUSSION . . . . . . . . . . .
Measurement Model . . . . . .

Path Model . . . . . . . . . .

Bibliography . . . . . . . . . . . . . . .

vi

Page

31

31

32

34

34

42

42

48

48

50

59

LIST OF TABLES

Table Page

1. Scale Items, Factor Loadings, and
RGliabilitieS o o o o o o o o o o o o o o 38"41

2. Correlation Matrix . . . . . . . . . . . . . 4S

vii

LIST OF FIGURES

Figure Page
1. Hypothesized Model of Communication, Stress
and Burnout . . . . . . . . . . . . . . . 26
2. Hypothesized Model of Communication, Stress
and Burnout With Path Coefficients . . . . 46
3. Final Model of Communication, Stress and

Burnout . . .

O O O O O O O O O O O O O O 47

viii

CHAPTER ONE
Theory and Literature Review

Without work all life goes rotten. But when
work is soulless, life stifles and dies. (Camus)

Soul....essence, life-blood, marrow, embodiment, vital
core. These are the ideas we associate with the soul of
something. While Camus writes that work is essential to
life, he qualifies his statement by writing that it must be
work which has a soul. As individuals and members of
organizations we search for this work. We attempt to find
and keep this vital core alive. How is this accomplished?
This question lies at the center of the research on burnout.

The research on the role of stress in the workplace is
perhaps more vital now then it was when initially begun.
Organizations are changing rapidly as a result of increasing
societal and technological change. These changes affect not
only the structure and outcomes of organizations but also
the interaction between employees within organizations.

Both internal and external change can lead to an increase in
the level of individual uncertainty about the organization,
and occupational and organizational role requirements.
Uncertainty associated with change has been shown to lead to
increased levels of stress and ultimately burnout (Tetrick &
La Rocco, 1987). Burnout has been associated with turnover

and absenteeism. Both absenteeism and turnover are

2
organizational problems that, as recently as 1983, have been
estimated to cost organizations in the United States between
$50 billion and $75 billion a year (Wallis, 1983).

Human service organizations and certain "helping"
professions have been central to the research on stress and
burnout. The educational organization and the profession of
teaching are prime examples. Stress in teaching induced by
the educational structure and process is widely documented.
Teachers work in limited contact with other staff, and yet
almost constant contact with their students. Teachers work
within a very structured organization and report limited
input into school level or policy making decisions. They
feel that their profession is given little community
recognition and that they are hindered in their work by
inadequate funding and overcrowded classrooms (Goodlad,
1983; Farber, 1984). Additionally, schools within the
United States have increasingly, in recent years, become
the target for major reform by governmental, educational,
parental and business groups (Timar & Kirp, 1989). In
addressing the changing nature of the school and,
consequently, the role of the teacher, the reform movement
has emphasized improvement in the quality of teachers
entering the profession. Conley, Bacharach, and Bauer
(1989) point out that the reformers have focused on changes
in teacher education to enhance the quality of teachers and
on the development of reward systems linked to professional

accomplishment. They have not, until recently, however,

 

 

3
seriously considered the necessity of changing organizational
structure and teacher activities to allow highly qualified
teachers to achieve rewards based on accomplishment. The
role of the teacher, as with other helping professionals, is
a very complex one. It is one which leads, in some instances,
to such elevated levels of stress that teachers burn out and
either leave the profession or continue to teach at a minimal
level of involvement (Farber, 1984; Schwab, Jackson & Schuler,
1986).

Awareness of the above becomes both an organizational
and a societal issue. If, due to certain variables-—
societal, organizational, situational or personal—-good
teachers are leaving the profession, not performing at their
maximum or performing without enthusiasm, then children and,
eventually, society are being negatively impacted. Goodlad
writes in his 1984 study of schools:

In general the practicing teacher—~to the degree
we can generalize from our findings-—functions in

a context where the beliefs and expectations are

those of a profession but where the realities tend

to constrain, likening actual practice to a trade

. . . a question arises as to whether the

circumstances can be made conducive to developing

in all teachers the behavior a profession entails.

By its very nature a profession involves both

considerable autonomy in decision making and

knowledge and skills developed before entry and

then honed in practice. The teachers in our sample,

on the whole, went into teaching because of those

inherent professional values. However, they

encountered in schools many realities not conducive

to professional growth. (Goodlad, 1984)

Identifying the variables that can intervene to halt or

reverse the erosion of skilled professionals is critical to

the continued improvement of the educational process.

4
The underlying argument presented in this paper is
that there is a model of the process of burnout within
educational settings that can be identified for teachers.

This paper addresses the relationship among organizational

 

stressors (work load, role conflict, role ambiguity),

 

communication variables (participation in decision making

 

and supervisory support), three dimensions of burnout

 

(emotional exhaustion, depersonalization, reduced personal

 

accomplishment), and the outcome variables of occupational

 

commitment and job satisfaction.

 

Burnout

Burnout is a complex, multisymptomatic concept that
has engendered a great deal of debate as to its definition
(Schuler, 1980). While, as Maslach points out, the evocative
quality of the term has fueled the development of multiple
definitions, it has also given rise to a large body of
research. This research on burnout has led to agreement on
some basic definitional issues. It is generally agreed by

researchers that burnout is an individualL psychological

 

process that is perceived as having negative consequences

 

by the individual (Maslach, 1982a). While burnout is viewed

as occurring within the individual in response to their needs,

values, abilities, experience and personality characteristics

(Schuler, 1980), its sources and interventions can occur at

the individual, dyadic, group, and the organizational levels.
From research done mainly within the human services

field, three dimensions of burnout are agreed to exist.

5

These are: (1) exhaustion, which is conceived of as

 

primarily psychological or emotional exhaustion. Emotional
exhaustion is characterized by: "a loss of feeling and
concern, a loss of trust, a loss of interest, a loss of

spirit" (Maslach, 1982a, p. 32), (2) depersonalization, or

 

the negative shift over time in responses to others and

(3) a sense of reduced personal accomplishment in terms of

 

morale and productivity. This is represented by an
individual's loss of a sense of efficacy on their job
(Pines, Aronson & Kafry, 1981; Maslach, 1982a; Jackson,
Schwab & Schuler, 1986; Miller, Ellis, Zook, & Lyles, 1990).

Changes in these dimensions have been shown to be
affected by situational variables perceived as individually
problematic such as workload, role conflict and role
ambiguity. Research also indicates, however, that these
burnout dimensions can be mediated or lessened by the
presence of communication variables such as social support
from either supervisors or co—workers (McGee, Cashman &
Goodson, 1987; Marcelissen et al., 1988; Miller, Ellis,
Zook & Lyles, 1990) and participation in decision making
(Jackson, 1983; Miller et al., 1990).

In an effort to conceptually construct the model
proposed in this study, I will discuss the three levels
of relationships, (Organizational stressors—->Burnout
dimensions and Outcome indicators; Communication/intervening
variables—->Burnout dimensions and Outcomes indicators; and

Burnout dimensions-—>Outcome indicators). The first of these

 

6

relationships to be examined is that of the organizational

variables, workload, role ambiguity and role conflict, which

 

are viewed as sources of stress, with the burnout dimensions

of emotional exhaustion, depersonalization, and personal

 

accomplishment and the outcome indicators of job satisfaction

and occupational commitment.

 

Organizational Stressors

 

Workload

Workload is defined as the amount and level of work an
individual is expected to accomplish in a specified amount
of time. Stress arises when the work or "burden . . .
exceeds the person's ability to handle it" (Maslach, 1982a,
p. 38). This is referred to in the burnout literature as
overload. Workload is sometimes viewed as a perceptual
variable. For example, two individuals may have identical
workloads and one may perceive him/herself in a state of
overload while the other perceives him/herself in a state of
underload. Workload can also be considered in terms of the
objective requirements of the position. This is illustrated
by the actual units (e.g., patients, students, cases, files)
an individual is responsible for (French & Caplan, 1972;
Pines, 1982). Objective and perceived workload have both
quantitative and qualitative aspects. Quantitative overload
refers to too much work assigned for a specific time.
Qualitative overload implies that the work is too difficult
(Cooper & Marshall, 1976). These concepts are interrelated

in that the quantity of work can have an effect on the

 

 

7
quality and vice versa (French & Caplan, 1972). To simplify
the process of addressing these issues, however, they are
presented separately here.

Pines and Maslach (1978), in research with psychiatric
hospital staff, found positive relations between quantitative
workload and level of stress reported. Interviews indicated
that (1) when patient load was too great staff members
attempted to separate themselves from their jobs and didn't
regard their jobs as sources of personal accomplishment and
(2) the longer the hours worked the more likely staff members
were to report stress and negative feelings.

Similar findings about quantitative workload have been
found in educational settings. Simply having more to do in
a day's time then they feel they can accomplish and too many
students to teach are often cited by teachers as major
sources of stress (Needle, Griffin & Svendsen, 1981). These
findings are also supported in a five year study of the
individual concepts of teacher stress held by-"experts"
conducted by Fimian (1987). The subject pool of experts was
comprised of those who had conducted research on teacher
stress during a given year. The sources of stress related
to the amount of work were too much work, overcrowded
classrooms or too large a caseload, too much paperwork,
having to do more than one thing, thinking of other things
at once, becoming easily overcommitted and not having enough
time to complete things. This stress was manifested

emotionally by feeling unable to cope, anxious, depressed,

 

8
vulnerable, and insecure and physiologically by exhaustion,
tiredness and related health problems. Levitov and Wangberg
(1983) cited too much to do, and paperwork as causing not
only negative feelings for teachers about their work but
also negative feelings about their own competence. The
teachers in this study also indicated that their workload
had a negative effect on their interactions with their
students. This leads us to consider the qualitative impact
of overload on teachers.

In schools, teachers spend the majority of each working
day in almost constant contact with their students.
Interactions with disruptive students are often stressful
(Cichon & Koff, 1980; Dedrick, Hawkes, & Smith, 1981; Feitler
& Tokar, 1982; Smilansky, 1984) perhaps because discipline
problems are perceived as indicative of a rejection of the
teacher's authority (Fimian, 1987). In summary, there are
qualitative aspects of teacher/student interactions that
lead teachers to feel a heightened sense of overload and
consequently, a greater level of stress.

The concept of load also often encompasses the types
of responsibilities one has. French and Cooper (1972)
differentiate between responsibility for people and
responsibility for things (budgets, equipment, etc.). They
write that responsibility for people results in greater
levels of physically related stress than does responsibility
for things. Teachers have direct and continual contact with

students. This is anticipated to be a significant distinction.

 

9

Within a caregiving organization there are various
groups that are crucial to the caregiving process. In a
school there are administrative, educational support,
clerical and general support functions in addition to those
of the teacher. These individuals do not, however, have the
continual student contact that is central to a teacher's
position. It is heuristically valuable, therefore, when
addressing the impact of workload on stress and burnout, to
separate groups within caregiving institutions based on
their level of client contact (Pines & Maslach, 1978; Miller,
Ellis, Zook & Lyles, 1990).

While the research presents a clear picture of the
general impact of increased workload, the goal of this paper
is to define specifically its links to various stressors and
burnout dimensions. This specification, as discussed in
Jackson et al. (1986) is not a simple matter. The interaction
of the qualitative and quantitative aspects of workload tend
to color the relationships. The hypothesized links presented
here are assumed, however, to be general enough to apply to
both the elementary and high school teacher and both
qualitative and quantitative aspects of the classroom
situation. These hypotheses and the logic underlying them
will now be discussed. In the model, workload is an exogenous
variable predicted to be positively related to role ambiguity,
role conflict and emotional exhaustion. This addresses both
the issues of quantity and quality of client contact. As the

sheer quantity of their work increases teachers will become

 

 

10

more physically and emotionally exhausted. This quantity may
refer to simply having a large number students in the
classroom or additional responsibilities (reports, meetings)
that they feel are tangential to the process of teaching.

As the quantity of their work increases the quality of their
interactions with students may diminish. There will be less
time to focus on the individual needs of students. Because
of this teachers may perceive a heightened sense of
uncertainty about their expected role within the school
(role ambiguity) and may find themselves receiving what they
may perceive or what actually are conflicting directions

concerning their role (role conflict).

Role Stress

The next variables of interest are role ambiguity and
role conflict. When teachers arrive at school in the
morning, they have certain expectations about how they will
spend the day. These expectations are associated with the
role they fill in the organization. Roles, as defined by
Katz and Kahn (1978) ”are standardized patterns of behavior
required of all persons playing a part in a given functional
relationship" (p. 43). Roles carry with them both internal
and external expectations. The internal expectations may be
derived from perceptions of and attitudes about the occupation
or organization. External expectations may be those
associated with the objective criteria of the job as defined
by the organization, such as a job description, or from the

more subjective expectations associated with supervisor or

 

“VI. 7;,, - -~ . .9.-

11
co-worker perceptions of the role (Kahn, Wolfe, Quinn &
Snoek, 1964).

Whatever the source, perceived environmental uncertainty
has been shown in research to be linked to increased levels
of stress (Van Sell, Brief, & Schuler, 1981). Uncertainty
about one's role within an organization can fall into two
categories——role ambiguity and role conflict.

Role ambiguity occurs when the person holding a
particular position isn't sure of what the role actually
consists of and how it is measured. Katz and Kahn (1978)
write that early researchers assumed that "ambiguity
frustrates the human need for clarity and structure in the
environment” (p. 206) and, therefore, leads to increased
stress.

Role conflict, on the other hand, is derived from
conflicting job or role demands. Conflict can arise from
an individual's identification with a role and an opposing
demand from a supervisor or organization; from actually
receiving multiple, conflicting sets of instructions or from
conflicting demands inherent in the requirements of the
position itself (Kahn et al., 1964). Within the school
setting, for example, teachers may find themselves attempting
to balance the goals of maximizing educational outcomes with
the reality of fulfilling their administrative functions.

Both role ambiguity and role conflict have been shown
to be positively related to burnout indicators. In their

review of the role conflict and ambiguity literature,

 

12
Van Sell, Brief, and Schuler (1981) report that, across a
variety of occupational groups, the "best-documented outcomes
of role conflict are job dissatisfaction and job—related
tension" (p. 48). Outcomes for role ambiguity are less
consistent. These writers do indicate, however, that most
studies show that those with high levels of ambiguity
perceive themselves as anxious, depressed, having a lower
level of job involvement and organizational commitment and
lowered perceptions of their own, their supervisors and the
organization's performance.

Research on teacher stress by Schwab and Iwanicki
(1982) indicated that role conflict explained more of the
variance associated with both emotional exhaustion and
depersonalization than did role ambiguity. Role ambiguity
alone, however, accounted for most of the variance in
personal accomplishment. Similarly, Schwab, Jackson and
Schuler (1986) found that role conflict accounted for the
largest amount of variance for emotional exhaustion and
depersonalization among a sample of teachers. They also
found, however, that role ambiguity, while not as great a
predictor as role conflict, was still significantly related
to emotional exhaustion. Neither role conflict nor role
ambiguity were significant predictors of personal
accomplishment.

Finally, research indicates that these job related
stressors are not only precursors to job burnout but

negatively related to job satisfaction and occupational

 

13
commitment as well (Golembiewski, 1982; Pines, 1982;
Miller et al., 1990; Schwab, Jackson & Schuler, 1986).

In the model of the process of burnout for teachers,
role conflict acts as an antecedent to depersonalization
and emotional exhaustion. As the level of role conflict
increases so the quality of teachers' interactions with
others will decrease and so will their sense of concern and
interest. This can be expected to lead to an increased need
to withdraw from ones students, co-workers and supervisor.
It is hypothesized, therefore, that a positive relationship
will exist between role conflict and depersonalization and
role conflict and emotional exhaustion. Role ambiguity is
proposed to be positively associated with emotional
exhaustion. The greater the role uncertainty teachers
experience the more apt they will be to express the loss of
spirit and interest associated with emotional exhaustion.
Role ambiguity is also indicated by research as the sole
antecedent to reduced personal accomplishment. Uncertainty
leads to teachers questioning the level of accomplishment
they derive from their work. These two variables are,

consequently, hypothesized as being positively related.

Communication Variables

 

While certain organizational variables may increase the
level of stress, there are communication variables that can
intervene in the burnout process and reduce the level
of stress. Research has indicated that two variables--

participation in decision making and support (both co-worker

 

14

and supervisor)——lessen the impact of stress. In this
section I will explore the relationship that exists between
these communication variables and the stressors of role
ambiguity, role conflict; the burnout variables of emotional
exhaustion, depersonalization and reduced personal
accomplishment and the outcome variables of job satisfaction

and organizational commitment.

Social Support

Support in the workplace——both supervisory and
co—worker—-has been viewed in research as a key factor to
reducing the stress induced by occupational requirements and
organizational structure and demands. Social support is
defined by Albrecht and Adelman as ”verbal and nonverbal
communication between recipients and providers that reduces
uncertainty about the situation, the self, the other, or the
relationship, and functions to enhance a perception of
personal control in one's life experience" (1987, p. 19).

A question central to support in the workplace is
whether two types of support (supervisory and co—worker)
have equal impact on stress. Research indicates that while
co—worker support may be received more frequently, supervisor
support is perceived as more important to employees in both
the educational and non-educational setting (Marcellissen et
al., 1988; Conley et al., 1989). Brissie, Hoover—Dempsey and
Bassler (1988) found that principal support accounted for
25 percent of the variance in burnout reported by 1,213

elementary teachers. Lack of administrative support is

 

15
reported by teachers as evidence of increased stress and
job dissatisfaction (Levitov & Wangberg, 1983).

Jackson et al. (1986) note that "feelings of personal
accomplishment are highest for teachers in supportive
environments, with support from one's principal appearing to
be particularly important" (p. 636). They also found that
the lack of support from a teacher's principal was the only
condition leading to a sense of depersonalization. These
findings indicate that support from ones supervisor is not
only important to teachers but also has broad ranging effects.

The opposite argument is presented by Schwab et al.
(1986) who did not find a significant relationship between
principal social support and various aspects of burnout but
did find that co—worker social support acted as a small but
still significant moderating factor for emotional exhaustion,
depersonalization and personal accomplishment.

In addressing the issue of co—worker support, Pines
writes that support from co—workers is one of the components
of the organizational environment that is an important
intervening factor in the process of burnout (Pines, 1982).
While this has been suggested theoretically, it has not been
empirically supported. Contrary to the theoretically
predicted connection co—worker support is not often cited as
a significant intervening variable in the lives of teachers
(Bacharach, Bauer & Conley, 1986). Lack of support from

colleagues is cited as having less impact on teachers' stress

 

16
level than lack of positive feedback from administrators
(Dedrick, Hawkes & Smith, 1981).

Lack of effect of co-worker support can, perhaps, be
explained by the professional status of teaching and the
rather self-contained nature of the work. Team activities
or sharing of work, which would allow for the diffusion of
stress associated with one's work, is not a key element in
teacher success in most school settings. In an overview of
the research done on teacher participation in schools,
Conley, Schmidle and Shedd (1988) write about the social
norms and structural features within schools that interfere
with peer interaction. "The solitary nature of most teaching
assignments, the physical layout of school facilities, and
restrictive time schedules usually preclude such interaction,
as do organizational norms that discourage advice giving (or
seeking) and treat work as something necessarily and
exclusively done in the classroom" (p. 26).

Based on these findings, therefore, supervisory support
will be the sole form of social support considered in this
study. The impact of supervisory support itself, however,
can be a rather complex issue. Ray and Miller (1989) write
that the impact of supervisory support within the school
setting may be limited by the inability of supervisors to
make broad ranging structural changes. In the school setting
a principal may be limited to less concrete types of support.
These researchers found that perceived supervisory support

mediated teacher levels of role ambiguity but had no impact

17

on their sense of role conflict. This may come about because
issues or situations that cause increases in conflicting role
demands for teachers may be more student and curriculum
oriented. Decisions about wide ranging changes on these
issues are not totally within the realm of the principal but
must be coordinated with district, local and state agencies.

Given the above research findings the following
relationships are predicted: Supervisory support is
hypothesized as negatively related to role ambiguity because
of a reduction in the level of uncertainty. It is also
predicted to be negatively correlated with depersonalization
and reduced personal accomplishment. The greater the sense
of a supportive supervisor, the more positive teacher
reactions should be to their supervisors, co—workers and
students and the less apt they should be to withdraw from
them. The more supervisory support teachers receive, the
greater the sense of personal accomplishment. As principals
indicate support for teachers' efforts the more apt teachers
are to feel that they are productive and, subsequently,
experience an increase in their level of personal
accomplishment. Supervisory support and reduced personal
accomplishment are, therefore, predicted to be negatively

correlated.

Participation in Decision Making

 

Participation in decision making is a widely studied
construct related to organizational communication. The

interest in it comes from a variety of directions. In a

18

metaéanalysis of the research, Miller and Monge (1986) found
that there is strong support for the idea that participation
effects satisfaction and somewhat lesser support for its
effect on productivity. Their findings also indicate that
satisfaction is enhanced more by a supportive, participative
climate than by the opportunity to participate in specific
decisions within an organization. This addresses a theme
that is at the core of participation research—-the
opportunity to participate versus the perceived or actual
influence associated with participation.

Two assumptions could be made about the supportive,
participative climate mentioned by Miller and Monge (1986).
First, the lack of opportunity for communicative interaction
between an individual and his or her supervisor would limit
the amount and type of support that could be offered, hinder
the level of participation and, consequently, increase the
level of stress. The opportunity for teachers to participate
in decision making within their schools_has been shown to be
negatively correlated with stress, in general, (Brissie,
Hoover-Dempsey & Bassler, 1988) and, specifically with
job dissatisfaction (Levitov & Wangberg, 1983) and
depersonalization (Schwab, Jackson, & Schuler, 1986). The
relationship to depersonalization indicates that the less
input teachers have in the decision making process within
their school the more negative their responses to others

become over time.

19

This lack of opportunity to participate could lead to
an increase in individual's uncertainty about where and how
he or she fits into the organization. Uncertainty about the
organization or one's role within the organization could be
ameliorated by increased participation in decision making.
Certainty about the requirements of a position are enhanced
if one is involved in determining some of the factors
that affect that position. Jackson (1983) found that,
participation reduced role conflict and role ambiguity.
Increases in these were positively correlated with emotional 4
stress. Participation was also positively correlated with
the individual's sense of perceived influence. In Jackson's
study there we ::rong correlation (.81) between a staff
member's sense a supportive social environment within the
organization and their level of perceived influence within
the organization and with their supervisor. Since Jackson
saw perceived influence as more central to the process of
decision making she dropped social support. This raises the
second point to be made about the supportive, participative
climate that Miller and Monge found in the literature——
opportunity for participation must be linked to influence
to impact satisfaction.

This idea is supported by Benson and Malone (1987) who
note that teachers in their study perceived themselves as
being deprived of influence in both managerial and technical
decisions made in their schools. These authors stress that

their study focuses on the ability of the teachers to have

 

20
influence as opposed to the opportunity to have involvement
in decision making activities. They write that teachers
report a greater opportunity for involvement than for
influence.

Perceived influence is, therefore, a concept that is
central to both participation and supervisory support. Using
a longitudinal design, Jackson (1983) found that there were
differences between the levels of role conflict and role
ambiguity reported by subjects in the control groups (no
increases in the level of participation) and those reported
by the intervention group (increased participation). Jackson
suggests, however, "that the benefits of participation in
decision making are not manifested immediately" (p. 14). For
example, when staff meetings occur regularly and frequently,
they may become more meaningful and, therefore, run more
effectively. Consequently, she speculates that while there
are quantitative differences, there may also be qualitative
differences such as changing perceptions of supportive
relationships in the workplace.

Thus participation is hypothesized to be an antecedent
to the perception of a supportive supervisor. This
prediction is supported by the work of Ray and Miller (1989)
in which they found that participation in decision making is
an antecedent to perceived supervisory support. It is through
participation's link to supervisory support that uncertainty
reduction associated with role ambiguity is postulated.

Participation is, however, predicted to be positively related

 

21
to the outcome variable job satisfaction. The more
opportunities teachers have for participation in decision
making and the greater their level of perceived influence,
the more control they will perceive that they have over
their jobs and, consequently, the more apt they are to

report a greater level of job satisfaction.

Outcome Variables

 

This last segment of this chapter will address the
relationship that exists among the burnout dimensions of
emotional exhaustion, depersonalization and reduced
personal accomplishment and with the outcome variables
of job satisfaction and occupational commitment.

Changes in the burnout variables have been shown to be
affected by environmental or situational variables perceived
as individually problematic. These burnout variables have
been shown, in more recent research, to have an effect on
each other (Miller et al., 1990). Miller et al., in their
research with caregivers and support staff in a hospital
setting, found that depersonalization was an antecedent
to both emotional exhaustion and decreased personal
accomplishment for caregivers. They suggest that this
relationship occurs because of the importance of the client/
caregiver interaction in the health care setting. This same
relationship can be inferred to exist for teacher burnout
due to the centrality of the teacher/student interaction.
The further teachers withdraw from those with whom they have

the opportunity to interact at school and the more negative

 

22
their responses become, the greater their level of emotional
exhaustion will be. This same withdrawal can be
hypothesized to lead to an increased level of reduced
personal accomplishment. This last hypothesis is based on
the premise that central to teachers' sense of personal
accomplishment are the successful——both productive and
pleasant-—interactions with their students. If they
withdraw from their students, the fewer successful
encounters and, therefore, a reduction in their sense of

personal accomplishment.

Much of the research on burnout has focused on the
relationship between the dimensions of burnout and the
outcome variables of employee job satisfaction and commitment
to the occupation. Increased levels of emotional stress or
exhaustion have been widely indicated as an antecedent to
decreased job satisfaction and occupational commitment
(Jackson, 1983; Kahn, Wolfe, Quinn & Snoek, 1963; Maslach,
1982b; Miller et al., 1990).

Katz and Kahn (1966) write that ". . . job satisfaction
is used loosely to cover overall liking for the job
situation as well as intrinsic job satisfaction deriving
from the content of the work process." Pines reports that
in research with mental health professionals the greater the
degree of burnout the more apt they were to stop looking for
self—fulfillment in their work and that their "good days"
become less and less frequent (p. 20). Consequently, the

more burned out a person is the less apt they are to find

 

23

satisfaction in their work. For those in caregiving
positions emotional exhaustion and depersonalization have
been found to be negatively correlated with job satisfaction
(Miller et al., 1990).

When individuals enter an occupation they have certain
expectations about what it means to them in terms of rewards,
both intrinsic and extrinsic. Matteson (1987) writes that
the gap between these individual expectations and situational
reality is one of degree-—the degree to which one's
expectations are met or not met by the occupation. If the
gap is too broad it can lead to burnout and a lessening of
commitment to the occupation. It has been proposed that this
reality gap may be broader for those who choose caregiving
occupations (Matteson, 1987; Goodlad, 1984). Miller et al.
(1990) found that for caregivers emotional exhaustion was
negatively correlated with occupational commitment.

These findings have generally been shown to occur for
teachers as well. Goodlad (1984) writes that in his research
"a large portion (57%) of the prime reasons for entering
teaching chosen by these teachers clustered around the nature
of teaching itself: the desire to teach in general or teach
a subject in particular (22%); the idea of teaching as a good
and worthy profession (18%); and a desire to be of service
to others (17%)" (p. 171). In addressing why teachers leave
the profession, he reports that "If one goes into teaching
with expectations of being able to teach and be of service

and then is frustrated in realizing these expectations,

24
dissatisfaction sets in and quitting becomes an alternative"
(p. 172). There seems to be a conflict in the values
teachers initially associated with the profession and the
frustration of those values by the reality of the teaching
situation. Schwab et al. (1986) found that teachers who
experienced an elevated state of emotional exhaustion
reported greater intentions of leaving their job and were
shown to have a higher absentee rate. Depersonalization
and reduced personal accomplishment were associated with a
decrease in effort on the job. In research conducted by S
Susan Jackson (1983), participation was demonstrated to
affect the variables of role conflict, role ambiguity,
perceived influence and emotional strain. These, in turn,
were related to both perceived employee job satisfaction and
intention to leave a job.

In the model presented in this paper, therefore,
emotionally exhausted teachers are expected to report a
lowered level of job satisfaction and occupational commitment.
Additionally, once the teacher has withdrawn from others it
is also feasible to assume a negative relationship between
depersonalization and job satisfaction as well.

In summary, the model (Figure 1) hypothesized in this
thesis illustrates the burnout process in teachers as having
its origins in increased workload, role conflict and role
ambiguity and being mediated by increases in supervisory
support and participation in decision making. This model

addresses the impact these variables have on the burnout

25
indicators of emotional exhaustion, depersonalization,
reduced personal accomplishment and subsequently, the
outcome variables of occupational commitment and job
satisfaction. The next chapter will describe how this
model was tested with a sample of elementary and secondary

school teachers.

26

 

 

Workload (+) Emotional (') Occupational
(+) Exhaustion ' Commitment
<+/
Role Conflict ()
(+) (+) (1‘)
(+

Sugervisiry__lLi____¢ Depersonalization_ili____* Sat'::bct'

uppor _ \\ /// 1 a ion

\K). (+> <+>

Role Ambiguity
H (+) /

Participation—z” Reduced Personal
in Decision Making Accomplishment

 
    
   

 

Figure 1

Hypothesized Model of Communication, Stress and Burnout

 

CHAPTER TWO

Methods

This chapter explains the methods used to assess the
relationship between organizational stressors, mediating
communication variables, dimensions of burnout and the
outcome variables of job satisfaction and occupational
commitment. It includes: (1) a description of the sample
of those from whom data were drawn, (2) the process of data
collection that was used, (3) a description of the survey
instrument with a breakdown of the items and (4) analysis

of the data.

Sample Description

 

This study was part of a larger research project. The
sample consisted of teachers, administrators and support
staff in a mid—size, urban school district in the Midwest.
The subjects were employed in elementary, middle, and
secondary schools and a center for severely mentally
impaired and severely, multiply impaired students. A
random, stratified sample of 880 employees was selected
as representative of the composition of the district by
research and evaluation personnel. This sample had 645
(73%) females and 235 (27%) males. Of these 538 (61%) were
teachers. For the purposes of this paper it is this group

(teachers) that was of interest.

27

28

Two hundred and ninety-eight completed surveys or 34%
of the 880 mailed were returned. Of the 538 teachers in the
original sample, 182 (34%) responded. The sample of teachers
(182) analyzed represents 61% of the final sample of 298. ‘In
the original sample of 880, teachers (538) represented 61% of
the total. The distribution of teachers in the final sample
is representative of the distribution of teachers in the
initial sample.

In terms of distribution by grade the original sample
was composed of 50% elementary school teachers (271); 42%
middle/high school teachers and 2% (11) special education
teachers. Of the sample of 182 analyzed 51% were elementary
school teachers (92) and 41% were employed in high schools
(74). Eight percent indicated they were teachers but did
not indicate grade level. Distribution by grade in the
final sample, therefore, is also representative of the
original sample.

The final sample of administrators, teachers and support
staff (298) was composed of 73% females and 26% males and 1%
unidentified. Teachers in the final sample were 67% female
and 30% male. Five respondents did not indicate their sex.
In the original sample of 880 teachers, administrators and
support staff, 73% were females and 27% were males. This
indicates that the breakdown by sex in the final sample was
representative of the original sample.

The total group (298) of individuals who returned their

surveys had a mean job tenure of 17.7 years and mean tenure

29

of 8.6 years at their current school. The subsample of
teachers (182) analyzed has a mean job tenure of 19 years
and a mean tenure of 9.6 years at their current school.
Tenure information was not available for the original sample
(880). Teachers in the final sample have, therefore, a

somewhat greater level of tenure than the total final sample.

Administration Procedure

 

The survey was mailed directly to respondents at their
homes at the end of the 1988-89 school year. The timing of
this mailing may explain the relatively low response rate.
Surveys were returned directly to the researchers at Michigan
State University. This was done to protect the anonymity of
the respondents and to help foster the fact that this was an
independent research effort rather than a school district

effort.

Instrumentation

 

The presence of individual job—related stress was
measured using self-report data from the subjects. The use

of self-report data assesses the individual's perception of

 

the stress versus an objective measure of stress. This is
an important distinction given the criticism of self-report
data. In the case of stress the argument has been made that
the subject's determination of their level of stress is more
important than a third party measure (Bacharach, Bauer &

Conley, 1986).

30

The measurement instrument used in this study was a 95
item survey. The variables related to individual perceptions
of workplace stressors, interventions (supervisory support,
communication) and the burnout process were assessed using

the following items:

Burnout

This was measured using the Maslach Burnout Inventory
(Maslach & Jackson, 1981). This well-validated measure is
comprised of three subscales which ask individuals for their
perceptions of their degree of emotional exhaustion, their
depersonalization of others and their sense of job—related

personal accomplishment.

Participation in Decision Making

 

The individual's perceptions of their opportunities for
participation and influence in the decision making process
was assessed using items from the Survey of Organizations
developed at the Institute for Social Research at the

University of Michigan.

Work Satisfaction and Occupational Commitment

These outcome measures were assessed using the Job
Descriptive Index (Smith et al., 1969) for work satisfaction
and the adaptation of relevant items on the Organizational
Commitment Questionnaire (Mowday, Steers, & Porter, 1979)

for occupational commitment.

 

31

Social Support

Items from the Caplan, Cobb, French, Harrison and
Pinneau Survey (1975), revised to reflect the educational
setting, were used to measure social support from three
sources: supervisory/principal, co-workers, and family and

friends.

Role Overload, Role Conflict and Role Ambiguipy

 

These areas were assessed using a survey instrument
developed by Pettegrew and Wolf (1982) to measure the

individual perceptions of educators.

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

model was estimated.

Confirmatory Factor Analysis

 

The measurement models were analyzed using the
confirmatory factor analysis subroutine of the PACKAGE
computer program (Hunter & Lim, 1987). The measurement
models were defined by an a priori analysis of the item
content. The fit of the specified measurement model to the
data was then evaluated by comparing the observed correlations
between the variables with the correlations predicted by the

measurement model. The items should "share common meaning

 

32

and conform to the product rules of internal and external
consistency" (Hunter & Gerbing, 1982). The criteria,
therefore, for confirming the factor structure of each scale
is as follows:

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

2. The scale should be internally consistent. The

 

items should satisfy the Spearman Product Rule. "That is,
the correlation between two items in the same cluster should
be the product of their correlations with the underlying
trait" (Hunter & Gerbing, 1982).

3. It should satisfy the parallelism requirement.
Parallelism refers to the extent to which items in a factor

are related in a similar way to other factors.

Path Analysis

A second subroutine of the PACKAGE computer program was
used to evaluate the proposed path model. "Path analysis is
a procedure for systematically combining the use of partial
and multiple correlations to study the causal relations
among a set of variables" (Hunter & Gerbing, 1982). This
program is a least squares path analysis program. Its output
consists of the original correlations, the path coefficients,
the reproduced path coefficients, the errors between the
original and reproduced correlations and the sum of squared

errors. The fit of the model is evaluated by examination of

 

33

the errors. Revisions to the model are then done to develop
a closer fitting model.

Once a well—fitted model has been developed through the
use of the confirmatory factor analysis and path analysis
subroutines of the PACKAGE computer program, a final
confirmation of the model will be done using the LISREL
computer program. LISREL uses full information, maximum
likelihood techniques. The output used to assess the fit of
the model is the Chi—square estimation of the "goodness of
fit" of the model. An insignificant Chi—square indicates a
good fitting model. Since Chi-square is sensitive to sample
size, the Chi—square, degrees of freedom ratio is also
considered along with the Adjusted and Unadjusted Goodness
of Fit Indices. At the more micro level of analysis LISREL
indicates whether the links specified in the model should be
retained and links not specified added. This is assessed by
looking at the T—values and the Modification Indices. The
T-values look at the significance of the path loadings and
the correlation between factors. The Modifications Indices
looks at unspecified parameters and indicates the degree to
which Chi—square would drop if a parameter was estimated.
The amount of variance accounted for by the model is
determined by looking at the R-squared values and the

Coefficient of Determination.

 

 

CHAPTER THREE

Results

This chapter includes the results of the confirmatory
factor analysis and the path analysis. Since this research
is part of a larger research project the CFA includes all of
the subjects in the sample (298). Because the model under
analysis deals only with the process of teacher burnout, the
path analysis was run on only the teachers (N = 182). The
use of the larger sample size for the CFA, however,
accomplishes two things: (1) lower levels of sampling error
should provide more stable estimates of parameters within
the measurement model and (2) it allows for comparison

across groups at a later date.

Confirmatory Factor Analysis

 

Confirmatory factor analysis was used to determine the
dimensionality of the measurement scales used to assess work
load, role conflict, role ambiguity, participation in
decision making, supervisory support, emotional exhaustion,
depersonalization, reduced personal accomplishment, job
satisfaction and occupational commitment. A 12 item scale
from an instrument developed by Pettegrew and Wolf (1982) to
assess teacher workload was initially analyzed. Due to lack
of internal consistency and low factor loadings 6 items were

dropped. A 6 item solution was achieved which proved to

34

35

have strong internal consistency and parallelism. These
remaining 6 items dealt primarily with issues that are part
of the 19p of being a teacher but not necessarily with the
process of teaching. For example, "Trying to complete
reports and paperwork on time causes me a lot of stress" and
"I find that dealing with student discipline problems puts

a lot of stress on me.” Scale items, factor loadings and
reliabilities are presented on Table 1.

Of a 5 item scale measuring role conflict (Pettegrew
& Wolf, 1982) 1 item was dropped. This item did not
discriminate between conflict and workload and loaded almost
equally on both factors. Role ambiguity was evaluated using
a 5 item scale (Pettegrew & Wolf, 1982). A 4 item solution
which was internally consistent and parallel was achieved by
dropping 1 item. This item assessed the teacher's ability
to predict what will be expected of him/her on the next day.
This is somewhat different from the other 4 items which focus
on the teacher's understanding of: (1) the scope if his/her
responsibilities and (2) the criteria used to evaluate their
performance. Items, factor loadings and reliabilities for
these scales are provided in Table 1.

Two communication scales were predicted——participation
in decision making and supervisory support. These, however,
were found to be strongly correlated (.89) and, therefore a
1 factor solution was evaluated. The combined factor had 4
items from the participation in decision making scale and 3

from the supervisory support scale. All of these items

 

36

focused heavily on communication with the supervisor and
influence in the decision making process. The resulting
scale was internally consistent, parallel with outside
factors and had a reliability of .93. The factor loadings
for the items in this factor are presented in Table 1.

Items from the Maslach Burnout Inventory were used to
assess emotional exhaustion (9 items); depersonalization (5
items); and reduced personal accomplishment (8 items). Four
items were dropped from the emotional exhaustion scale
resulting in a 5 item solution.

Difficulties arose in obtaining a unidimensional scale
for depersonalization. Three of the items related to
depersonalization of students and one of the others to
depersonalization of others (not students) and the last one
to the depersonalizing impact of the job on teachers. A
scale that was internally consistent and parallel was
obtained by dropping the items not related to students.
This final scale is consistent with the idea that client
interaction pays a critical part in the teacher burnout
process. The resulting scale proved to be internally
consistent and parallel with a reliability of .65.

Three items were dropped from the reduced personal
accomplishment scale leaving a 5 item scale. Examination
of the remaining items indicated that contextually they were
similar. They focus on the academic accomplishments of
teachers with their students and their sense of positive

influence. This scale was internally consistent with the

 

37

exception of a .11 (p > .05) correlation between the 2
strongest items. The items, factor loadings and
reliabilities for these scales are included on Table 1.

Job satisfaction was measured using the 18 item Job
Descriptive Index developed by Smith, Kendally and Hunter
(1969). A 5 item scale was finally arrived at which was
internally consistent and parallel. The items dropped were
determined by low factor loadings and the impact of the
items on the internal consistency of the scale. Reviewing
the items dropped, it is interesting to note that items such
as fascinating, creative, respected and challenging as well
as routine, boring, simple and tiresome dropped out. The
remaining items were satisfying, good, pleasant, useful and
gives a sense of accomplishment. Factor loadings and the
reliability for this factor are on Table 1.

The original occupational commitment scale consisted

 

of 6 items adapted from Mowday, Steers and Porter's
Organizational Commitment Questionnaire. Three items were
dropped resulting in a 3 item scale which was internally
consistent and parallel. These 3 items addressed the value
teachers' placed on their occupation; the similarities
between their values and those of their occupation and
ability of their occupation to inspire exceptional job
performance. Factor loadings and the reliability for this

factor are on Table 1.

38
Table 1

Scale Items, Factor Loadings, and Reliabilities

Item

Role Conflict (Alpha = .78)

Factor Loading

 

 

1. I receive conflict demands from two or

more people or groups in the school

setting. .75
2. I often have to buck a rule or policy to

carry out an assignment. .79
3. I am given school—related duties without

adequate resources and materials to

carry them out. .61
4. There is a difference between the way my

administrative head thinks things should

be done and the way I think they should

be done. .59
Role Ambigpity (Alpha = .77)
1. I am unclear on what the scope and

responsibilities of my job are. .70
2. I am uncertain what the criteria for

evaluating my performance actually are. .64
3. I receive enough information to carry

out my job effectively. .68
4. When asked, I am able to tell someone

exactly what the demands of my job are. .66
Work Load (Alpha = .74)
1. I feel constant pressure from others to

improve the quality of my work. .47
2. Trying to complete reports and paperwork

on time causes me a lot of stress. .58
3. I find that dealing with student

discipline problems puts a lot of stress

on me. .57
4. Complying with state, federal and school

rules and policies is very stressful. .61

39
Table 1 (cont'd.)

Item Factor Loading
5. I experience a lot of stress trying to

resolve conflicts between parents and

the school. .63
6. Having to participate in school

activities outside of the normal working

hours is very stressful to me. .57

 

Emotional Exhaustion (Alpha = .89)

 

 

 

1. I feel emotionally drained from my work. .87
2. I feel used up at the end of a workday. .80
3. I feel fatigued when I get up in the

morning and have to face another day on

the job. .73
4. I feel burned out from my work. .85
5. I feel frustrated by my job. .68
Depersonalization (Alpha = .65)
1. I feel I treat some students as if they

were impersonal "objects". .60
2. I really don't care what happens to some

students. .71
3. I feel the students blame me for some of

their problems. .54

 

Personal Accomplishment (Lack of) (Alpha = .69)
1. I deal very effectively with the

problems of students. .55
2. I feel I'm positively influencing other

people's lives through my work. .61
3. I can create an academically focused

atmosphere with the students. .51
4. I feel exhilarated after working closely

with the students. .39

40
Table 1 (cont'd.)

 

 

 

 

 

 

Item Factor Loading
5. I have accomplished many worthwhile
things in my job. .67
Work Satisfaction (Alpha = .80)
Do the following words describe your work:
1. Satisfying .77
2. Good .68
9. Pleasant .62
10. Useful .60
18. Gives a Sense of Accomplishment .69
Occupational Commitment (Alpha = .72)
1. I believe I've chosen the best of all
possible occupations to work in. .58
2. I find that my values and the values of
my occupation are very similar. .69
3. My occupation really inspires the very
best from me in terms of job performance. .76
Supervisory Support (Alpha = .93)
1. It is easy to get my ideas across to my
supervisor/principal if I have a
suggestion. .76
2. When decisions are being made at this
school, the persons affected are asked
for their ideas. .79
3. Information is widely shared at this
school so that those who make decisions
have access to such knowledge. .82
4. At this school, decisions are made at
those levels where the most adequate
information is available. .82

 

 

41
Table 1 (cont'd.)

My principal goes out of his/her way to
make my life easier for me.

It is easy to talk with my principal.

My principal can be relied on when
things get tough at work.

Factor Loading

.82

.82

.83

 

 

42

Preliminary Path Analysis

 

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

the fit of the model are included in Table 2.

Path Analysis

The original model (Figure 2) proved to be a poor fit
to the data. There were two nonsignificant path coefficients;
several residuals of questionable proportions and a sum of
squared error of 2.327. The Chi—square analysis did not
indicate a good fit of the model (computed X2 = 568.38;
required X2 = 28.87/df = 18, p > .05). In examining the
residuals it became apparent that the following paths were
underrepresented in this model: (1) Supervisory support to
Reduced Personal Accomplishment (2 to 7) = path coefficient
of .09; residual error, —.12; (2) Depersonalization to
Emotional Exhaustion (5 to 6) = path coefficent of -.02;
residual error .12. These links were dropped.

The residual errors were then analyzed and based on the
amount of error associated with a linkage, decisions were
made on whether to drop that link. A final model (Figure 3)

was arrived at in which the path coefficients were

43
significant, the squared error was decreased to .165 and
the average squared error was .007. The Chi-square analysis
indicated a good fit of the model (computed x2 = 30.;
required X2 = 30./df = 19, p > .05).

The following links were dropped from the original
model: (1) Work Load was not associated with Role Ambiguity,
(2) Role Ambiguity did not prove to be an antecedent to
Emotional Exhaustion, (3) Depersonalization was not found to
be a cause of Emotional Exhaustion nor was it linked to Job
Satisfaction and (4) The link from Participation in Decision
Making to Job Satisfaction dropped out when it was combined
with Supervisory Support. The following links were added:
(1) Workload to Depersonalization, (2) Role Conflict to Role
Ambiguity, (3) Role Ambiguity to Depersonalization,

(4) Reduced Personal Accomplishment to Job Satisfaction,
and (5) Job Satisfaction to Occupational Commitment.

The analysis of the final model using LISREL indicated
the following. The goodness—of—fit index is .871 (adjusted =
.747). The Chi—square to degrees of freedom ratio is 7.6.
This is somewhat larger than the recommended limit of 5.
Only 4% of the normalized residuals were greater than 2.0.
The coefficient of determination and R—squared values
indicate that a substantial amount of the variance is
accounted for by the final model—-role conflict (R-squared =
.586), role ambiguity (R-squared = .530), emotional
exhaustion (R—squared = .425), depersonalization (R-squared

= .356), reduced personal accomplishment (R—squared = .679),

 

44
job satisfaction (R-squared = .717), occupational commitment
(R—squared = .694). The modification indices indicate that
the Chi-square estimate would be improved by an addition of
a link between role conflict and reduced personal
accomplishment. A revision with this link was computed. It,
however, produced modifications calling for feedback loops
from job satisfaction to reduced personal accomplishment and
emotional exhaustion to reduced personal accomplishment.
While the idea that there could conceptually be feedback
loops is feasible, the current model is accepted with the
suggestion that in the future researchers may want to

explore this area.

 

45
Table 2

Correlation Matrix @

 

 

Items 1 2 3 4 5 6 7 8 9
l -— —.25 .59 .36 .60 .45 .45 —.40 -.47
2 —.21 —- —.62 -.59 —.35 -.21 -.34 .46 .41
3 .45 -.53 -- .70 .56 .49 .43 —.63 -.56
4 .27 —.50 .54 —- .45 .49 .73 —.67 -.65
5 .49 -.32 .47 .37 —— .35 .49 —.66 -.66
p .31 -.16 .35 .35 .27 —— .69 -.47 —.49
Z .32 —.27 .31 .53 .38 .46 —— -.69 —.80
8 -.29 .38 —.47 —.50 —.53 —.32 —.48 —— .83
9 —.36 .35 -.44 —.51 —.56 —.35 -.59 .63 ——

Means 18.09 20.88 11.80 8.65 16.04 6.70 10.76 10.48 11.91

S.D. 4.5 6.9 3.8 3.1 5.2 2.6 2.6 2.6 3.9

Key:

1. Workload 6. Depersonalization of Students
2. Supervisory Support 7. Reduced Personal

3. Role Conflict Accomplishment

4. Role Ambiguity 8. Occupational Commitment

5. Emotional Exhaustion 9. Job Satisfaction

@ Correlations in the lower half of the matrix are
uncorrelated for attenuation; corrected correlations
appear in the upper half.

46

    
  

Workload ('43) Emotional Exhaustion
.59) ('-65)
(21)
Role Conflict\ 02) Occupational
Commitment
.22)
(J6 Depersonalization
Role Ambiguity 43)
(-. 53 (.16) (57

Supervisory (0 92 Reduced Personal

Support Accomplishment
(.17) Job
Satisfaction

Figure 2

Hypothesized Model of Communication, Stress
and Burnout With Path Coefficients

*Coefficient significant, p < .05

47

Workload ('4') Emotional (--20) Occupational
Exhaustion_———_————_* Commitment
4|

    
   

(.46)

   
     

Depersonalization

 

('54) (.3él///’ (.70)
QnSO) Role Ambiguity 2 (~44)
(-.25 ('5)
(.20 Reduced Personal
Accomplishment
Supervisory —.63)
Support
Job
Satisfaction

Figure 3
Final Model of Communication, Stress and Burnout

All Coefficients Significant, p < .05

 

CHAPTER FOUR

Discussion

This chapter includes a discussion of the findings from
the measurement model and the path model. In the measurement
model, while most of the scales were confirmed in a fairly
standard fashion, two scales are of particular interest--
supervisory support and depersonalization. These will be
discussed in the measurement section. In the path analysis
section the total interaction of the final path model will

be discussed.

Measurement Model

 

The high correlation between the items in the
participation in decision making scale and the supervisory
support scale resulted in testing a unidimensional scale.
The findings suggest that the level of perceived influence
in decision making may play a key role in an employee's
sense of supervisory support. As discussed earlier there
is a major difference between opportunity for involvement
in the decision making process and the level of influence
this opportunity provides. Supportive supervisors would
seem to be those who are not only interested in a person's
work; concerned enough to show affective supportive but are
also receptive to the ideas of subordinates and able to act

upon those ideas. In essence, these supervisors allow

48

 

49

teachers to become influential in the educational process.
This indicates that supervisory support and participation
in decision making are not as distinct as has been assumed.
The second measure of particular interest was the
depersonalization scale. While the majority of the items in
this scale have a client focus, two are not directly aimed
at the teacher's concern for depersonalizing students. These
two items were dropped from the measurement model. The items
that remain are, as noted earlier, consistent with the idea
that student interaction is critical to the burnout process
of teachers. Maslach (1982b) suggests that while the
"development of this detached, callous and even dehumanized
response" (p. 4) known as depersonalization may start with
one's clients it can progress until it effects one's
relationships with others who are important in one's life as
well. When thinking about the order in which burnout in
teachers occurs, however, it seems logical that the workplace
depersonalization process for teachers may happen in two
stages with the final stage being the depersonalization of
students. This would suggest that this process starts with
teachers isolating themselves first and, additionally, from
their supervisor and co-workers and, as a last desperate
attempt to shield themselves from burnout, from their
students. Teachers work, from their first day on the
job, in relative isolation from each other with limited
opportunities for professional or social interaction and

in almost constant contact with their students. The job

 

 

50

requirements and limitations that come from outside of the
classroom are, however, a source of stress. As noted by
Goodlad (1984), teachers enter the profession because of
inherent professional values. These values are central of
the role of being a teacher and working with students. Two
important issues that need to be addressed in association
with the process of depersonalization as a way to control
burnout are:

1. How would a teacher go about isolating him/herself
as way to control burnout?

2. Which causes the initial burnout process,
disillusionment with the system and the others in it or the

children?

Path Model

The revised model indicates that overload is, indeed, a
major source of stress. It is most strongly correlated with
role conflict (.46) and emotional exhaustion (.41), but also
accounted for increases in depersonalization of students
(.24). These are rather logical connections. As workload
increases, either in quantity or complexity, the probability
that a teacher will receive conflicting demands from
supervisors and clients also increases. This heightened
workload will be a source of exhaustion, physical as well
as emotional. Of special interest is the impact on
depersonalization of students. As their responsibilities

increase or their student load increases, teachers will

 

51

either spend less time with individual students or start to
distance themselves psychologically from them.

Both role conflict and role ambiguity have an impact on
the burnout variables (emotional exhaustion, depersonalization,
and reduced personal accomplishment). Of the role related
stressors, role conflict is sole predictor of emotional
exhaustion (.32). Role ambiguity, on the other hand, is the
only predictor of reduced personal accomplishment (.52). Both
role stressors are antecedents to depersonalization with role
conflict having a path coefficient of .21 and role ambiguity
.38. These findings are consistent with previous research
(Schwab & Iwanicki, 1982). Of particular interest is the link
that predicts that the greater the role conflict experienced
by a teacher the greater the level of role ambiguity will
become. The more demands a teacher receives or perceives he
or she receives the less able they become at sorting out what
their responsibilities are and, concurrently, what the
criteria are for measuring their successful accomplishment
of those responsibilities. This reflects the difficulties
that arise for teachers because of the multiple and sometimes
conflicting expectations that exist for schools. Goodlad
(1984) concludes, in assessing the expectations that exist
for schools beyond academics, that "those associated with
schools are not united in and consumed with a common,
pervasive, educational mission" (p. 71).

As a way to counterbalance the effect of increased

workload, increases in supervisory support are shown in this

 

 

52
model to lessen the effects of both role ambiguity (-.25)

and role conflict (—.50). The findings are not consistent
with the initial hypothesized lack of relationship between
supervisory support and role conflict. Supervisory support
was not predicted to mediate role conflict because of the
assumption that the principal would have limited impact on
the more substantial educational changes that might be
associated with role conflict. The final model indicates,
however, that supervisory support has a stronger effect on
role conflict than on role ambiguity. A teacher's perception
of role conflict is lessened by support from their supervisor.
Stress reduction may be linked to teachers' influence with
their supervisors and the supervisor's ability to influence
more concrete issues rather than provide just affective
support. The link hypothesized between supervisory support
and role ambiguity did not, however, drop out. Support that
can help teachers more clearly define their role in the
educational setting lessens their sense of role ambiguity.
The assumption that principals would be limited in the
instrumental support they could provide teachers; the
finding that supervisory support lessens role conflict and
role ambiguity all point to the multidimensional nature of
support. Social support has been purported to break out
into three separate categories——emotional, cognitive or
informational and instrumental. The multidimensionality of
this concept and the difficulty that comes about in trying

to separate these dimensions increases the difficulty of

53

clearly and concisely studying it. When a principal's
support reduces a teacher's role conflict is that
instrumental support, informational support, emotional
support or, perhaps, all three? Is the type of support
contingent on the situation? Are principals limited in the
type of support they can offer? If so, what effect does
this limitation have on the burnout of teachers? Is one
type of support more important than another in the burnout
process? These are the types of questions that muddy the
research in this area and call for more research to sort
out the specific relationships.

Something to be kept in mind, however, is that while
supervisory support is shown to act as a buffer to stress,
the literature indicates that the educational level and
professional autonomy of this population may lessen the
impact of supervisory support. Marcellissen et al. (1988)
in their research report that supervisory support is less
important to higher occupational groups than lower
occupational groups.

Supervisory support is also positively related to
depersonalization (.21). In the original correlations these
variables are negatively correlated (—.21). Supervisory
support was perceived as lessening depersonalization. When
the model was developed, however, the negative correlation
was suppressed by the impact of the load and role variables
(workload, .45; role conflict, .49; and role ambiguity, .49)

on depersonalization. The relationship proposed by the

54

model is that as teachers experience increases in the level
of supervisory support they increasingly withdraw from
students.

Another possibility to be considered, however, is that
when teachers reach the stage at which they have become the
object of supervisory concern for whatever reason and,
therefore, the object of a great deal of attention from
their supervisors they are more apt to be depersonalizing
their students. The question that must be asked, however,
is why this support leads to further depersonalization.

One possibility is that principals, in an attempt to be
supportive, are commiserating with teachers by sharing "war
stories". Rather than addressing the issues associated with
why the teacher feels the need to depersonalize their
students and how these issues may be remedied, this support
may only serve to further confirm for the teacher that life
in the classroom is bad and what they are experiencing is
normal. Consequently, supervisory support may increase in
quantity but be of a type that further exacerbates the
burnout process.

Of the burnout variables depersonalization seems to be
at the core of the burnout process. It is, as mentioned above,
positively linked to increases in role conflict, workload,
role ambiguity and supervisory support. Depersonalization
is characterized by negative responses to others, increases
in irritability and a lessening of idealism (Maslach, 1982b).

Jackson et al. write that there are not only quantitative

 

 

55

differences in client contact (the number of client contacts)

but also qualitative.

Depersonalization,

they suggest may

occur in jobs at either extreme more than for jobs

characterized by intermediate amounts of interpersonal

distance" (Jackson et al.,

1986,

p. 638). Since the

teacher's position falls within the extreme area mentioned

by Jackson,

reduced personal accomplishment (.44) is to be expected.

the connection between depersonalization and

As

teachers feel distanced from their students their sense of

personal accomplishment is diminished.

What was not confirmed by
link between depersonalization
While emotional exhaustion and
antecedents (workload and role

predictors of each other. The

this model was the predicted
and emotional exhaustion.
depersonalization have common
they are not

conflict)

greater the workload (both in

quantity and quality) and the greater the perceived role

conflict, the more frustration

fatigue a teacher feels.

As hypothesized,

and emotional and physical

emotional

exhaustion seems to separate out as a distinct predictor of

lessened degrees of job satisfaction (—.35) and occupational

commitment (—.20).

Of the other burnout measures, depersonalization was

predicted to directly lead to decreased job satisfaction.

This was not confirmed in the analysis.

Depersonalization

is indirectly linked to job satisfaction through reduced'

personal accomplishment which was found to be causally

antecedent to job satisfaction.

Logically as teachers

 

 

56

become more distant from their students and their sense of
personal accomplishment decreases so does their sense of
job satisfaction.

One additional link that emerged from the analysis is a
rather substantial (.70) correlation between job satisfaction
and occupational commitment. This connection certainly makes
sense. A lessened sense of job satisfaction leads teachers
to assume that the values they associate with teaching and
education are not consistent with those required for the
actual occupation.

In general, a number of themes emerge from this model.
The first of these is the idea that depersonalization of
students is a key link in the burnout process. The impact
of depersonalization and role ambiguity on reduced personal
accomplishment and indirectly on job satisfaction presents
a picture of teachers who have become uncertain about their
roles, have started to distance themselves from their
students and, therefore, feel that they are not accomplishing
what they want to and are, consequently, not satisfied with
their jobs. As their job satisfaction is lessened so is
their commitment to the occupation of teaching.

Secondly, overload, whether quantitative or qualitative,
is not only directly related to two of the burnout dimensions
(emotional exhaustion and depersonalization) but is also a
contributor to the conflict teachers associate with their
roles. Workload and conflict imply, in one sense, the more

concrete aspects of the jobs such as the number of students

 

57

and the number of conflicting assignments. Through increases
in these more concrete aspects teachers become emotionally
exhausted and indirectly experience a reduction in their
satisfaction and their commitment to the values of their
occupation.

Last, the role of support communicated by the supervisor
is confirmed as a moderating force in the burnout process.

As has been discussed, however, the issue of support is not
clearly defined and the interaction of participation in
decision making and supervisory support has not been either
successfully redefined or differentiated.

In closing, issues related to these research findings
that have implications for future research and by extension,
practical application will be briefly addressed.

The collapse of the participation and supervisory
support scales into a unidimensional scale emphasizes the
exchange of information and the degree of influence the
teacher has with the supervisor. It points to the need to
focus in burnout research on participation in decision making
as a level of influence rather than just as an opportunity
to participate. Rather than being a mediating factor to
the burnout process, the opportunity to participate in
organizational decision making without the assumption of
influence may actually lead to a greater level of burnout.
Teachers indicate that they are reluctant to become involved
in the decision making process because involvement did not

bring with it any meaningful influence (Benson & Malone,

 

 

 

58
1987). It would be interesting to test scales that attempt

to break out these separate and, perhaps, contradictory
pieces.

Another aspect of this relationship that would be
interesting to investigate is the relationship that exists
between the supervisor's sense of influence on substantial
issues that effect the school and the teacher's perception
of their supervisor as a mediator in the stress process.
Having influence with a supervisor who doesn't perceive that
they have the opportunity or ability to effect change may
also be another avenue of increased stress.

Of both research and practical interest is the
sequencing of the depersonalization process. Practically,
if one is interested in developing an intervention strategy
it would seem heuristically valuable to be able to measure
depersonalization of supervisors and co—workers and
depersonalization of students as separate indicators of
burnout. The intervention process may be dramatically
different for individuals in the initial stages than for
those who have reached the point of withdrawing from their
students. Of research interest is the further clarification
of the causal sequence that this indicates in the burnout

process.

BIBLIOGRAPHY

 

BIBLIOGRAPHY

Albrecht, T. L., & Adelman, M. B. (1987). Communicating
social support. Newbury Park: Sage.

Bacharach, S. B., Bauer, S. C., & Conley, S., (1986).
Organizational analysis of stress: the case of
elementary and secondary schools. Work and
Occupations, l5 (1), 7—32.

Benson, N., & Malone, P. (1987). Teachers' beliefs about
shared decision making and work alienation. Education,
107 (3), 244—251.

Camus, A. The Myth of Sisyphus (1955). New York: Vantage
Books.

 

Cichon, D. J., & Koff, R. H. (1980). Stress and teaching.
NASSP Bulletin, 54, 91—104.

Conley, S. C., Bacharach, S. B., & Bauer, S. (1989).
The school work environment and teacher career
dissatisfaction. Educational Administration Quarterly,
55 (1), 58-81.

 

Cooper, C. L., & Marshall, J. (1976). Occupational sources
of stress: a review of the literature relating to
coronary heart disease and mental ill health. Journal
of Occupational Psychology, 45, 11—28.

 

Dedrick, C. V., Hawkes, R. R., & Smith, J. K. (1981).
Teacher stress: a descriptive study of the concerns.
NASSP Bulletin, 55, 31-35.

Evertson, C. M., & Emmer, E. T. (1982). Preventive classroom
management. In D. L. Duke (Ed.), Helping teachers
manage classrooms (DP. 2—31). Alexandria: Association
for Supervision and Curriculum Development.

Farber, B. A. (1984). Teacher burnout: assumptions, myths,
and issues. Teachers College Record, 55, 321—338.

 

Feitler, F., & Tokar, E. (1982). Getting a handle on teacher
stress: how bad is the problem. Educational Leadership,
55, 379—388.

 

59

 

 

60

Fimian, M. (1987). Teacher stress: an expert appraisal.
Psychology in the Schools, 25, 5-13.

 

Glickman, C. D. (1985). Supervision of instruction. Boston:
Allyn and Bacon, Inc. '

Golembiewski, R. T. (1982). Organizational development
interventions: changing interaction, structures, and
policies. In W. S. Paine (Ed.), Job stress and burnout
(pp. 229-253). Beverly Hills: Sage Publications.

 

Goodlad, J. I. (1984). A place called school: pgospects for
the future. New York: McGraw—Hill.

Hackman, R. J., & Oldham, G. R. (1976). Motivation through
the design of work: test of a theory. Organizational
Behavior and Human Performance, 15, 250—279.

 

Jackson, S. E. (1983). Participation in decision making as
a strategy for reducing job—related strain. Journal of
Applied Psychology, 55, 3—19.

 

Jackson, S. E., Schuler, R. S., & Schwab, R. L. (1986).
Toward an understanding of the burnout phenomenon.
Journal of Applied Psychology, 11, 630—640.

 

Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek, J. D., &
Rosenthal, R. A. (1964). Organizational stress.
New York: Wiley & Sons, Inc.

Koff, R. H., Laffey, J. M., Olson, G. E., & Cichon, D. J.
(1981). Executive stress and the school administrator.
NASSP Bulletin, 55, 1—9.

Marcelissen, F. H. G., Winnubst, J. A. M., Buunk, B., &
De Wolff, C. J. (1988). Social support and
occupational stress: a causal analysis. Social
Science and Medicine, 55 (3), 365—373.

 

Maslach, C. (1982a). Understanding burnout: definitional
issues in analyzing a complex phenomenon. In
W. S. Paine (Ed.), Job stress and burnout (Pp. 29—40).
Beverly Hills: Sage Publications.

 

Maslach, C. (1982b). Burnout: the cost of caring.
Englewood Cliffs, NJ: Prentice—Hall, Inc.

 

Matteson, M. T. (1987). Individual~organizational
relationships: implications for preventing job stress
and burnout. In J. C. Quick, R. S. Bhagat, J. E. Dalton,
& J. D. Quick (Eds.), Work stress; health care systems
in the workplace (pp. 156—170). New York: Praeger.

61

McGee, G. W., Goodson, J. R., & Cashman, J. F. (1987). Job
stress and job dissatisfaction: influence of contextual
factors. Psychological Reports, 51, 367-375.

Miller, K. I., Ellis, B. H., Zook, B., & Lyles, J. (1990).
An integrated model of stress, burnout, and
communication in the workplace. Michigan State
University, E. Lansing, MI. (unpublished manuscript).

Miller, K. I., & Monge, P. R. (1985). Social information
and employee anxiety about organizational change.
Human Communication Research, 11, 365-386.

 

Needle, R. H., Griffin, T., & Svendsen, R. (1981).
Occupational stress: coping and health problems of
teachers. The Journal of School Health, 51 (3), 175-181.

 

Pettegrew, L. S., & Wolf G. E. (1982). Validating measures
of teacher stress. American Educational Research
Journal, 15 (3), 373-396.

 

Pines, A. (1982). Changing organizations: is a work
environment without burnout an impossible goal? In
W. S. Paine (Ed.), Job Stress and Burnout (Pp. 189—211).
Beverly Hills: Sage Publications.

 

Pines, A., & Maslach, C. (1978). Characteristics of staff
burnout in mental health settings. Hospital and
Community Psychiatry, 15, 233—237.

 

Pines, A. M., Aronson, E., & Kafry, D. (1981). Burnout: From
tedium to personal growth. New York: The Free Press.

 

Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970). Role
conflict and ambiguity in complex organizations.
Administrative Science Quarterly, 15, 150-163.

 

Schuler, R. S. (1980). Definition and conceptuatlization of
stress in organizations. Organizational Behavior and
Human Performance, 15, 184—215.

 

 

Schwab, R. L., Jackson, S. B., & Schuler, R. S. (1986).
Educator burnout: sources and consequences.
Educational Research Quarterly, 15 (3), 14—30.

 

Similansky, J. (1984). External and internal correlates of
teacher satisfaction and willing to report stress.
British Journal of Educational Psychology, 51, 84—92.

 

U.S. Department of Education. (1983). A nation at risk: The
imperative for educational reform. Washington, D.C.:
U.S. Department of Education, April.

 

 

62

Van Sell, M., Brief, A. P., & Schuler, R. S. (1981). Role
conflict and role ambiguity: integration of the
literature and directions for future research. Human
Relations, 55 (1), 43-71.

Wallis, C. (1983). Stress: can we cope?, Time, June 6, 1983.

Weiman, C. (1977). A study of occupational stressors and
the incidence of disease/risk. Journal of Occupational
Medicine, 15, 119-122.

Wiggins, T. (1988). Stress and administrative role in
educational organizations. Journal of Educational
Research, 55 (2), 120-125.

 

 

 

 

 

 

MICHIGAN STATE UNIV. LIBRARIES
iliilWINilliNIHWWHllWillHiiililiiliilllliil
. 31293307532488

 

 

   

 

V \. :