A comparison of the supervision of principals in two school districts

Material Information

A comparison of the supervision of principals in two school districts
Brady, Ann Mathewson
Publication Date:
Physical Description:
ix, 255 leaves : forms ; 29 cm

Thesis/Dissertation Information

Doctorate ( Doctor of philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Education and Human Development, CU Denver
Degree Disciplines:
Administration, curriculum, and supervision


Subjects / Keywords:
School principals -- Rating of -- United States ( lcsh )
School principals -- Rating of ( fast )
United States ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Doctor of Philosophy, School of Education and Human Development, Administration, Curriculum, and Supervision.
Statement of Responsibility:
by Ann Mathewson Brady.

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:
28863658 ( OCLC )
LD1190.E3 1993d .B73 ( lcc )

Full Text
Ann Mathewson Brady
B. A., University of Colorado, 1958
M. S., University of Colorado, 1973
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
Administration, Curriculum, and Supervision

This Thesis proposal for the Doctor of Philosophy
degree by
Ann Mathewson Brady
has been approved for the
School of Education
Laiira D. Goodwin
lll flj, fi I9cib
fj D^te
Kari H. Flaming

Brady, Ann Mathewson (Ph.D., Administration, Curriculum, and Supervision)
A Comparison of the Supervision of Principals in Two School Districts
Thesis directed by Professor Richard P. Koeppe
With the shift of decision making from the central office to the local
school level, supervisors of principals must adjust their supervisory
practices. An urban and a suburban school district, in the process of
changing from central office to school-centered decisions under different
circumstances, were used in this comparative study. Comparisons of the
supervision of principals in these two districts were made by coding and
analyzing 52 interviews, observations of principals and supervisors working
together, and documents pertaining to the supervisory process.
Descriptions of the supervision of principals in the two districts were
completed under categories and superordinate dimensions of supervisory
interactions that emerged from the analysis. There was congruence
between the principals and supervisors descriptions of supervision within
districts, but there were striking differences between the two districts on
seven dimensions: Support, Supervision, Sharing, Decision, Time,
Relationships, and Control.
Similarities in the supervision of principals in the two districts were
found in the way both sets of principals described ideal supervision and in
the limited importance of the principals performance appraisals. In both
districts, the culture of the district influenced the supervision of principals.
Key elements in the supervision of principals emerged from the data.
Principals want a collaborative relationship with their supervisors based on
a common focus. Collaborative supervision is developed by devoting
substantial time to gaining and sharing information about current

educational practices and specific school situations. Shared information
leads to shared decisions and mutual professional growth.
The new fiscal reality of reduced central office budgets as well as
greater leadership responsibilities at the local school level require new
methods of supervision. Collaborative supervision, encouraging shared
leadership, as well as policy changes to reduce time spent on legislated '
accountability requirements, are recommended. Time and organizational
structures that allow school administrators opportunities to work
collaboratively were found to be necessary for successful supervision.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Richard R Koeppe

1. INTRODUCTION...................................................1
Purpose....................................................... 7
Importance of Study............................................8
Theoretical Framework..........................................9
Definition of Terms...........................................10
2. REVIEW OF LITERATURE........................................ 13
Evaluation of Principals......................................14
Learnings from the Theorists Writing on Management Behavior... 16
School Organizations as Bureaucracies.........................20
Dilemmas and Choices Facing School Administrators.............22
The Use of Control and Influence..............................24
General Supervision in Education..............................27
Final Words.................................................. 29
3. METHODOLOGY...................................................31
The Qualitative Research Design...............................31
Coding the Pilot Data.........................................33
Site Selection................................................41
Selection of Participants.....................................44

Principal Interviews............................................51
Supervisor Interviews.......................................... 53
Collection of Documents..............................................57
Methods Used to Analyze Data.........................................57
Outside Verification of Interviews and Categories...............60
Principal Verification of Description of Supervision............62
Additional Verification From Observations and Documents. .63
Reliability and Validity.............................................64
Internal reliability.....................................64
External reliability.....................................65
Internal validity........................................67
External validity........................................68
4. RESULTS.............................................................. 69
The Description of Supervision of Principal
in the Suburban School District........................................71
Personalized Supervision........................................71
Personal support.........................................75
Substantive support......................................76
Time Commitment.................................................77
Sharing Information.............................................79

Shared Decisions...............................................80
School Centered................................................85
A Story Line Description of Supervision in the
Suburban District..............................................88
Suburban Principals Ideal Supervision.........................91
Similarities Between the Supervisors
and Principals Interviews.....................................92
The Effect of the Suburban District Culture on
the Supervision of Principals..................................93
The Description of the Supervision of Principals in the
Urban School District................................................94
Bulk Supervision...............................................94
Cursory support........................................101
Personal support.......................................102
Lack of Time................................................. 103
Psychological Distance........................................105
Multiple Bosses...............................................113
A Story Line Description of Supervision
in the Urban District.........................................118
Urban Principals Ideal Supervision...........................122
Similarities Between Supervisors and
Principals Interviews........................................123

The Effect of the Urban District Culture
on the Supervision of Principals.........................124
Comparison of the Supervision of Principals
in an Urban and a Suburban District.........................126
Supervision Dimension................................... 129
Support Dimension........................................131
Time Dimension......................................... 134
Sharing Dimension...................................... 135
Decision Dimension.......................................138
Relationship Dimension.................................. 140
Control Dimension..................................... 142
5. DISCUSSION OF FINDINGS......................................149
Characteristics of the Supervision of Principals............149
Support................................................. 152
Time.................................................... 153
Sharing................................................. 154
Control................................................. 156
The Effect of District Culture on the Supervision of Principals. .158
The Change to School-Centered Decisions..158
The Selection and Training of Supervisors................160
Organizational Structure.................................163
Supervisors and Principals Responses to District Culture. 164

Similarity of Supervisory Practices in the Two Districts
to Previous Studies...........................................167
Urban District............................................167
Suburban District.........................................170
Similarities Between the Urban and Suburban Principals........172
Information about Schools.................................173
Feelings of Loneliness....................................173
Ideal Supervision.........................................175
Time to Talk--With an Important Purpose.................. 175
Application of the Results to Supervisory Practices...........176
Principal Led Collaborative Supervision...................178
Recommendations for Future Research.......................180
Qualitative research................................181
Quantitative research...............................181
Policy Implications.......................................183
A. Pilot Data Analysis..........................................192
B. Schedule of Interviews and Document Collection...............202
C. Transcripts of Interviews....................................206
D. Categories, Observation, and Verification of Data............224
E. Documentation................................................245

With only a few studies as exceptions, the supervision of principals has
been either overlooked or considered as part of the aggregated, general
material on the supervision of educational personnel. Most of this general
information is focused on the supervision of teachers by principals or central
office supervisors. The lack of interest in the supervision of principals may
be attributed to the limited contact principals have with their supervisors.
One ethnographic study of 16 elementary principals in Chicago, focused on
the discretionary areas of administrative decision-making at the school
building level, found that interaction pairings with the district
superintendent accounted for 7% of the principals time (Morris, Crowson,
Hurwitz, & Porter-Gehrie,1981, p. 32). The same study gave an example of
district level personnel undercutting an instructional innovation made by a
principal at the school level by not supporting the principals evaluation of
teachers. Other examples of minimal mutual support between principals
and central office supervisors were described in the study under a chapter
heading entitled creative insubordination and civilized disobedience. What
appears to be limited interaction may, in reality, be avoidance in a continual
confrontation between principals and their supervisors. Principals may
interact with their supervisors in a manner similar to the way principals and
teachers interact, described by Roland Barth (1990) as parallel play or
adversarial and competitive relationships. More information is needed to

describe what happens between principals and the people who supervise
them. The relationship between principals and their supervisors was the
subject of this research.
Control was one topic that emerged from the limited literature about the
interaction between principals and district level managers. The types of
controls used by the district organization to direct the daily work of principals
was the principal finding in a study by Peterson (1984) that examined six
organizational control mechanisms used over elementary principals. The
need for connection between principals and the central office was
established early in the study report:
Functioning as the linkage between central office and classrooms as
well as between parents and teachers, principals must keep
resources, personnel, and students working efficiently toward
organizational goals and objectives. To do this, they must neither
be so tightly constrained that they cannot respond to changing
conditions, nor so loosely controlled that they seek personal rather
than organizational goals... In short, superiors seek an appropriate
balance of control and autonomy that will maximize organizational
effectiveness, (p. 573)
Petersoni (1984) confirmed the limited interaction between principals
and their supervisors noted in the Chicago study: The scarcity of research
into supervision as one form of control in school systems may be due to the
fact that principals spend so little time interacting with central office
personnel (p, 578). His study found that supervision resulted in only partial
control over principals: however, it was complemented by other controls,
such as output control and environmental control which act in place of or in

addition to direct supervision" (p. 584). In addition to supervision, he
studied five other administrative-level control mechanisms: input control,
behavior control, output control, selection-socialization control, and
environmental control. Each of the six organizational control mechanisms
adds some influence and increases the potency of the overall system of
control (p. 594). After presenting the findings regarding the six control
mechanisms used with principals, Peterson suggested that a balance
between organizational control and managerial prerogatives of school
principals may be the direction school districts should take.
The relative balance of control and autonomy may affect the
motivation of principals. There should be greater motivation in
districts in which principals are afforded greater autonomy and are
less confined by various mechanisms of control. In districts in which
one finds constraints over principals and less autonomy, motivation
among administrators should be low, depleted by the binding web of
control, (p. 594)
Control of school principals is the subject of an article by E. Mark
Hanson (1975). He discussed two systems of control over school principals:
the superordinate structure of superintendent and the subordinate structure
of the teachers. The growth and development of the modern bureaucracy
as a balance of power has come at considerable cost to the public school
middle manager. More than ever he has become the man caught between
two colliding systems of control (p. 28).
Five years later, a similar statement came from another source: Caught
between the power of teachers unions and the directives of supervisors, the
principal must assert authority in order to exercise leadership (Crowson &

Porter-Gehrie, 1980, p. 47). These authors went on to reinforce the
importance of the principalship;
More than any other single position in the American school
hierarchy, the principalship represents the pivotal exchange point,
the most important point of connection between teachers, students
and parents on one hand and the educational policy-making
structure-superintendent, Board of Education, taxpayer-on the
other, (p. 65)
The aforementioned studies about the interaction between school
principals and others in the school district organization have described
control mechanisms, limited interaction, and manipulation of directives.
These qualities are not what the authors of organizational management
books would suggest as the preferred approach. Organizational literature
includes ample evidence that participative management, open
communication, and inclusive decision making procedures result in greater
productivity and personal satisfaction. For example, McGregor (1960)
showed that with a choice between the belief that personnel must be
directed in Theory X and the belief that personnel can be self-directed in
Theory Y, Theory Y management behavior was far more successful. Likert
(1961,1967; Likert & Likert, 1976) put years of research behind System 4,
the participative group approach to management and found it an effective
approach. Combs, Avila, and Purkey (1978) described the successful open
system used in helping professions. Sergiovanni and Starratt (1983)
conceptualized productive people interaction in a human resources
supervision theory. Clark and Astuto (1988) presented seven choice options
in organizations with a clear preference for those options that validate
employees. According to Clark and Astuto,

Those who fear the process of moving from a centralized autocracy
that assumes accountability, exercises control, ensures stability and
regularity, and designs interventions and intentions are those who
fear the efficacy of people. People in a humane organization will
protect! their opportunities for self-actualization by resisting the
negative confusion and formlessness of anarchy; they will exhibit
self-control. Such an organization, in the final analysis, will be
wiser, more effective, and more humane than that which could be
invented and implemented by most contemporary designated
leaders, (p. 128)
Have school districts missed these recommendations, or are the
interactions that occur between managers and others in school districts not
yet understood? How can the discrepancy between the information gained
from studies about school district management and the research about
productive management techniques be reconciled? The idea of greater
mutual support among people within school districts is appearing in current
journal articles. Pajak and Glickman (1989) expressed in the following
statement their views on what is needed for school improvement: What is
important is to! create district expectations of professional dialogue and
support so that educators in all positions in a school system can share in that
inventiveness and express that commitment ( p. 64). A recent issue of
Education Leadership was about Transforming Leadership (February,
1992). One quote from among many statements in this issue demonstrates
that educational writers are thinking about how we work together:
Transformational leadership ... arises when leaders are more concerned
about gaining overall cooperation and energetic participation from
organization members than they are in getting particular tasks performed
(Mitchell & Tucker, 1992, p. 32). The term transforming leadership
originated from the well-known leadership theorist, James Burns (1978).
Perhaps reconciliation of the dichotomous descriptions between what is and

what should be at the managerial level of educational organizations is in the
process of being addressed.
In an attempt to get information about one part of the puzzle of
management practices in school districts, this study focused on a fraction of
the many interactions that are part of the complex human interchanges in
public education. It concentrated on the interaction between principals
and the people who supervise them. Heretofore, studies done with
principals and superintendents have been done with a focus on one role or
the other. In other words, principals are studied or superintendents are
studied separately, not together. In discovering how principals and their
supervisors work together, it is appropriate to seek information from both
parties to the interaction and to observe what occurs when the two work
We have learned from the literature and our own experience that
principalships are key positions in school districts (Dwyer, Barnett, & Lee,
1987; Greenfield, 1987; Purkey & Smith, 1983). We also know that linkages
between principals and the central office are critical for both the schools and
the larger organization (Crowson & Porter-Gehrie, 1980; Peterson, 1984;
Wimpelberg, 1987), although some suggest that those linkages are another
means for control (Peterson, Murphy, & Hallinger, 1987). To learn how those
critical positions can be supported and how positive linkages can be
strengthened, we need to discover what happens between principals and
their supervisors.
The information we have about leadership styles and the adjustment of

leadership approaches under changing situations was hard to translate to
what actually happens when principals and their supervisors interact. This
researcher could find no description in the literature of the relationship
between supervisors and principals in the context of leadership or
supervision. In pilot interviews with principals about the supervision they
had experienced, the relationship with their supervisor emerged as an
important factor. Interviews with both principals and their supervisors about
their supervisory interactions as well as observations of principals and their
supervisors working together in meetings were the methods used to obtain
data in the present study. Artifacts germane to those reported interactions
and observations were collected as corroboration and verification. By
analyzing the characteristics of the interaction between principals and their
supervisors, the researcher hoped to formulate a description of the
relationship th'at exists between them.
As the downward thrust of decision making from the district level to the
local school becomes a reality, the role of the principals as well as the
people who supervise principals will need to adjust to the shifting locus of
control. Before we know how best to accommodate to local decision making
and to support the changing role of the principal, we must know how
principals and their supervisors work together. If supervisors more often
make choicesjthat recognize the efficacy of people (Clark & Astuto, 1988),
we should be able to move with the new direction without difficulty. If, on the
other hand, centralized autocracy lingers as the mode of operation used by
supervisors of! principals, we may experience a more difficult change

process. In either case, we want to know how these roles work together so
we can determine how best to help and support people in key positions in
school districts. Thus, the purpose of this study was to provide a description
of the interactions that occur between school principals and their
supervisors and the relationship that develops between the people who
serve in those managerial positions in two districts.
The following questions provided a focus for the study. These were
used to start the discovery process about the interactions between principals
and the people to whom they report and to direct the study of principals and
their supervisors in a work environment.
The general question is: What characterizes the supervision of
principals in two school districts?
Supporting questions are:
1. How do principals and their supervisors describe their work together?
2. How do the situations in the two school districts affect the supervision of
3. To what degree are the principals perceptions of their interactions with
their supervisors congruent with the supervisors perceptions of the same
interactions? j
4. Are there; commonalities across pairings and across school districts?
Importance of the Study
Educational managers are likely to have read in a current journal that
transforming leadership is the wave of the future (Brandt & Scherer, 1992).

Perhaps it is time to recognize that leadership is less a matter of aggressive
action than a:way of thinking and feeling--about ourselves, about our jobs,
and about the nature of the educational process (Mitchell & Tucker, 1992,
p. 30). This study asked principals and their supervisors what they think
and feel about the way they work with each other. Currently, we are faced
with a shift in the locus of control in decision making in public schools, a shift
that will bring principals and the local school staffs into policy level decision
making. Is the way principals and their supervisors presently work together
supportive of that concept? In a recent article, Glickman (1992) expressed
concern that the new movement to site-based control in schools will not
succeed. One ingredient for its success is for superintendents, school
boards, and district personnel to understand, approve, work with, and
support the schools vision and operating plan, as well as provide technical
assistance and resources to work ahead" (p. 25). How is that done?
Before answering that question we need first to discover how they work
together. Than we may have a base from which we can try to figure out how
we can develop a stickier management and leadership practice, one that
touches people and stays with them (Sergiovanni & Brandt, 1992, p. 49).
Theoretical Framework
Goetz and LeCompte (1984) and Miles and Huberman (1984)
recommended that every research design include the theoretical framework
assumed by the researcher. The assumptions, approach, and biases of the
researcher can thus be known by the reader. Herbert Blumers Symbolic
Interactionism influenced the present researchers thinking about the way
people attribute meaning to the actions of those with whom they interact.

Based on the theories of George Herbert Mead, Blumer (1969) gave this
definition of Symbolic interaction:
The term symbolic interaction refers to...the peculiar and
distinctive character of interaction as it takes place between human
beings1. The peculiarity consists in the fact that human beings
interpret or define each others actions instead of merely reacting to
each others actions. Their response is not made directly to the
actions of one another but instead is based on meaning which they
attach to such actions. Human interaction is mediated by the use of
symbols, by interpretation, or by ascertaining the meaning of one
anothers actions. ( pp. 78-79)
The definition of each others actions as perceived by principals and
their supervisors is what the researcher wanted to discover in this study. It
was assumed that the meaning each one gives to the other would influence
the nature of the supervisory process in which they are involved.
Definition of Terms
Principals and the people who supervise them are the persons
appointed by the Board of Education as school administrators responsible to
lead a school or to supervise school principals.
Supervision is the act, process, or occupation of supervising: direction,
inspection, arid critical evaluation: OVERSIGHT... (Webster, 1981, p.
2296). The use of oversight as a synonym for supervision leads to the
operational use of supervision for this study: the way supervisors of
principals oversee the work of school-based administrators who report
directly to them.
Interaction, by dictionary definition, is mutual or reciprocal action or
influence (Webster, 1981, p. 1176). In this study, it is every action that takes
place between principals and the people who supervise them. For example,

all communications, written and spoken; meeting times, formal and
informal; social interactions; gossip; feelings; reactions; body language.
Relationship is a state of affairs existing between those having
relations or dealings (Webster, 1981, p. 1916). The relationship between
principals and their supervisors was determined by analyzing the
characteristics of the interaction between the two parties and how both
parties felt about those interactions. Definitions of the terms used by
principals and supervisors, such as openness, trust, collegiality, support,
communication, were defined by them through interviews.
The interactions between principals and their supervisors within the
organizational structure of two school districts was studied. The importance
of organizational expectations in determining the characteristics of the
phenomena Linder study helped the researcher understand the reasons why
supervisors and principals acted as they did. Learning about all the
organizational expectations and how these factors affected the supervision
of principals was beyond the scope of the study. The ability of one person to
study fully two management levels in two school districts was a limitation.
Important features of the two organizations may have been omitted due to
time restrictions and the complexities of large public school districts.
The goal of qualitative studies is to develop a theory or explanation of
phenomena that has enough variation to be applicable to different
phenomena in other, related situations (Strauss & Corbin, 1990). Two
school districts were used in this study to minimize the effect of a single
school system on the data collected as well as to allow additional

dimensions of the supervision of principals to emerge. The difference in the
number of students enrolled in the two school districts as well as the
different demographics of the communities they serve may be dominant
factors in how principals are supervised. These factors limited the study to a
description and comparison of the supervision of principals in two school
districts at a particular time period.

Strauss and Corbin (1990) direct the qualitative researcher to avoid
extensive technical literature review because you want to explain
phenomena in light of the theoretical framework that evolves during the
research itself ; thus, you do not want to be constrained by having to adhere
to a previously developed theory that may or may not apply to the area
under investigation (p. 49). This researcher had already examined some
literature in an attempt to discover a theory and a measurement tool to study
the relationship between principals and their supervisors. The literature
search was enlightening: very little pertained to the study of the interaction
between those two management levels in a complex, school district
organization. The exceptions were two articles written by a school
superintendent about mutual administrative support and how to evaluate
that support (Koeppe, 1977, 1981). The readings also showed there was a
discrepancy between what theorists said was the preferred managerial
approach and what the school district management practices were. The
review supported the researchers belief that research studies described
only part of the interaction that occurs between management people in
public school districts (Rammer, 1991; Sullivan, 1982). Another author
agreed with the researchers perceptions that authoritative relationships did
not meet the current demands of good relationships within organizations:

Clearly, the way in which people view themselves and their relations
to others in organizations has changed to the point that the theories of
administrative and organizational behavior we have been using are
no longer applicable. They do not describe organizational behavior,
nor do they predict such behavior. They ignore the basic change in
authority relationships that has been underway since World War II
and that accelerated in the early 1960s. (Griffiths, 1977, p. 4)
The readings described herein are thematic as well as sequential. The
evaluation of principals, which is often mistaken for what principals actually
do or should do in schools, are the first literature samples to be reported.
An article that pleas for a return to the wisdom of McGregors Theory X and Y,
extends the literature review to those writings supporting participation and
openness in organizations. In contrast to what theorists tell us are the best
ways for people to relate to each other in organizations, the next group of
articles included in this chapter are the realities of school district
bureaucracies along with the dilemmas school administrators face. An
exploration into the ever-present theme of control and influence will be
followed by a limited review of writings about general supervision in
education. A final quotation completes the chapter.
Evaluation of Principals
Much of the material on evaluation of principals addressed how
principals could improve as school leaders. (Bolton, 1980; Croghan & Lake,
1984; Holdzkc?m, 1985; Zappulla, 1983) This information was helpful,
however, it focused on how-to rather than what-happens and presented a
narrow, prescriptive approach to working with school principals.
Bolton (1980) presented skills needed to be a successful school
administrator and used the management-by-objectives approach to help

administrators improve their skills. The Educational Research Service, inc.
report on Evaluation Administrative Performance (Carnes. 1985) contained
interesting facts on the recent adoption of formal evaluation procedures for
school administrators throughout the country. The most commonly used
systems were described and various administrative evaluation forms were
included. None of those forms referred to interactions or relationships
between principals and their supervisors.
In Phi Ddlta Kappas Hot Topics Series on Administrator Evaluation.
(1985) there Was a good literature synthesis on educational managers.
Included in the Hot Topics Series was an article by Holdzkom (1985) that
presented a summary of writings on managerial proficiencies from
Mintzberg, Peters and Waterman, and Heller. It closed with Boyatzis twelve
competencies. Another article in the same publication gave the results of
two researchers who analyzed competencies from three Florida principal
assessment Renters, checked them against Boyatzis list, matched them with
competencies from one other investigation, added a review of two other
studies, and finally emerged with nineteen high-performing and basic
competencies under six clusters (Croghan & Lake, 1984). This
comprehensive but overwhelming list of competencies led to an
apprehensive! thought: evaluation for principals was a check list of
effective behaviors to be achieved. There were no words of wisdom in the
endless lists of competencies on how to help people who worked in complex
organizations; called schools and school districts. The researcher
questioned if lists were the best way to motivate principals to improve their
skills and gladly turned to more thoughtful, theoretical studies.

Learnings from the Theorists Writing on Management Behavior
When an article from the business world was included in the latter part
of Evaluating Administrative Performance: Current Trends and Techniques.
(Zappulla, 1983) something more interesting and applicable to managers
than lists of competencies was promoted. Rieder (1973) stated that he was
suffering from managerial performance review battle fatigue (p. 401) and
pointed out six flaws in performance review programs. He suggested a
return to fundamentals: partner relationship; individual accountability; build
on strengths; managerial support; flexible systems quantitative feedback;
and action commitment (pp. 404-408). Rieder based his ideas on
McGregors Theory Y: Douglas McGregors assumptions still cast the
longest shadow (p. 409).
In a bookwritten in 1960 that remains a classic, McGregor established a
framework to explain the differences in how managers approached their
managerial tasks. The Human Side of Enterprise used examples from
industry to explain Theory X and Theory Y, two diverse approaches to
managerial behavior. Theory X described managerial behavior that was
based on the belief that workers must be controlled and directed. Theory Y,
on the other hand, was rooted in the belief that people would do a good job
if given the opportunity to be self directed and integrate their own goals with
those of the organization. McGregor made it clear that Theory X led to job
dissatisfaction! and poor performance from employees working under that
managerial approach. Theory Y, with the belief that people were basically
good and trustworthy, helped employees reach higher levels of performance
and greater job satisfaction.
Other worthwhile books and articles were written in the 1960s about

working successfully with people in organizations. Two of these books were
most germane to this study: The Organizational Climate of Schools (Halpin
& Croft, 1963) and The Human Organization: Its Management and Value
(Likert, 1967). Halpin and Croft developed an Organizational Climate
Description Questionnaire which they used with the staffs of 71 schools.
They identified six Organizational Climates that fell on a continuum defined
at one end as Open, at the other, as Closed (1963, p.2). An Open Climate
was characterized by esprit, low hindrance, high intimacy, and high
consideration! A Closed Climate had disengaged teachers, high hindrance,
high production emphasis, little trust, low consideration, and highly aloof
principals. Furthermore, they inferred three parameters which can be used
to conceptualize the social interactions that take place within an
organization: authenticity (openness); satisfaction; and leadership initiation
from any group member (1963, p. 76). This study has been criticized for its
narrow focus on relationships (Miskel & Ogawa, 1988); however, it gave a
picture of the way people responded to different personal attitudes used by
people in schools.
Likerts (1967) labels for management systems were on a four-part
scale that ranged from System 1, exploitive-authoritative, and System 2,
benevolent-authoritative, to System 3, consultative, and System 4,
participative group. His studies showed that even when confronted with
evidence that System 4 resulted in better productivity, many managers
retained the practices of System 1. In one case, a System 2 vice president
deliberately destroyed a System 4 department under him because it made
other departments look bad. Likerts principal of supportive relationships
explains the basic concept behind System 4 management practices:

The leadership and other processes of the organization must be such
as to ensure a maximum probability that in all interactions and in all
relationships within the organization, each member, in the light of his
background, values, desires, and expectations, will view the
experience as supportive and one which builds and maintains his
sense of personal worth and importance. (Likert, 1961, p. 103)
In a later book, he wrote that the relationship between the superior and
subordinate is crucial (1967, p.47). The Table of Organizational Variables
included in Likerts works contained the same components found in the
human resources supervision theory found in a text by Sergiovanni and
Starratt (1983). Likerts work is still applicable today and is found in current
literature on principalships (Sergiovanni, 1987, pp. 263-269).
More wisdom from the 1960s was found in an article written in 1968 by
F. Herzberg explaining his hygiene vs. motivators theory of job satisfaction
which supported the knowledge that people were motivated by their
achievements not by KITA (kick in the pants). Hygiene and motivation
factors were not the opposite of each other but were separate factors
determined by two different needs of people.
The growth or motivator factors that are intrinsic to the job are:
achievement, recognition for achievement, the work itself,
responsibility, and growth or advancement. The
dissatisfaction-avoidance or hygiene (KITA) factors that are
extrinsic to the job include: company policy and administration,
supervision, interpersonal relationships, working
conditions, salary, status, and security. (1968, p.174)
It appeared there were two points of view between Likert and Herzberg
about the importance of relationships to job satisfaction. There was
similarity in their thinking, however, in that you motivate people through
greater job responsibilities. Herzbergs job enrichment approach was the
vertical loading of more job responsibilities that resulted in greater

achievement thus greater job satisfaction. With this approach, more and
more job responsibilities would come under subordinates, and supervisors
might begin to feel useless unless they did more training and evaluation with
their subordinates. What has been called an employee-centered style of
supervision will come about not through education of supervisors, but by
changing the jobs that they do (1968, p. 183).
Further support for the positive relationships between managers and
their subordinates that were proposed in McGregors Theory Y, Halpin and
Crofts Open Climate, and Likerts System 4, came from authors in another
field of study-i-perceptual psychology. Helping Relationships, bv Combs,
Avila, and Purkey, presented two frames of reference for working with
people. The, authors described two fundamental choices available to
helpers for dealing with human problems: closed or open systems. In
closed systems, responsibility rests with the directors, managers, or
administrators. These systems were based on stimulus-response theory.
The closed system proceeds by defining a final objective in the
clearest possible terms, then establishes the machinery to reach that
objective.. .Closed systems are project-oriented. As a consequence
persons involved tend to be regarded as part of the machinery by
which the product is produced. (Combs, Avila, & Prukey, 1978,
pp. 101-103)
in open systems, responsibilities for outcomes was shared by all who
confronted the problem. The emphasis in open systems is on participation
by all with shared power and decision making (Combs, Avila, & Purkey,
1978, p. 105).
An open system...may begin without a manifest object. It proceeds to
confront a problem, then searches for solutions...People are regarded
as dynamic rather than static processes. Problems confronted and
solutions accepted are judged in terms of the persons involved.
(Combs, Avila, & Purkey, 1978, pp. 101-104)

The research and rationale for positive, supportive relationships
between managers and subordinates in these theories was very persuasive.
The assumption from those readings was that all managers of people,
particularly in organizations that relied on personal interactions, would
follow an open climate, Theory Y, System 4 approach. Experience tells us
that was not the case in most school districts. Their bureaucratic structures
permeated personal interactions.
School Organizations as Bureaucracies
An author from the same era of McGregor and Likert, gave another
point of view about management relationships--those affected by a
bureaucratic organization. Suggesting that principals behavior was molded
by their bureaucratic role, Bridges (1965) pointed out that principals were
expected to minimize personal relations and to ignore the peculiarities of
individual cases...If he substitutes personalized relationships for the
structurally required impersonal relationships, the principal is likely to
encounter conflict within the bureaucratic setting (p. 25). Another paper
written by Bridges deals with administrative man as a pawn of subordinates
rather than the originator of decisions. He discusses different sets of
conditions under which the subordinates goals intrude into and influence
administrative decisions (Bridges, 1970, p. 12). Here is one author
who did not agree with the concept of open, supportive relationships
between managers and subordinates, at least in a bureaucratic view of
school organizations.
Using Chester Barnards concept that subordinates had the authority to
support or deny the outcomes of centralized decision making in an

organization,!Hanson (1975) suggested that there were two authority
structures, one in the hands of the superordinates and the other in the
hands of the subordinate...Modern bureaucracy concept suggests that the
authority structure of the school is best described as a balance of power
(p.26). He perceived principals having formal authority and teachers having
informal authority. There was tension between accountability requirements
that were the responsibility of the principal and teacher power.
Michael Lipskys theory of street-level bureaucracy was used by
Crowson and ;Porter-Gehrie to describe the complex and uncertain work
environment of principals. Caught between the power of teachers unions
and the directives of supervisors, the principal must assert authority in order
to exercise leadership (1980, p. 47). They suggested that one coping
mechanism that principals uses was to blame downtown for mistakes
made by the central office. It appeared that each group was trying to
undercut the other in this balance of power.
In a discussion of the concepts of power, authority, and bureaucracy to
be used to enhance our understanding of schools, Abbott and Caracheo
(1988) used research findings to call into question the monistic conception
of hierarchy.. iAuthority can be, and is, centralized or decentralized
depending on the type of decision to be made (p.252). In a bureaucratic
structure decision making can be either centralized or decentralized...The
important question is what the consequences are of centralizing decision
making (p.255). Their suggestion was to stop studying bureaucracy per se
and start using other concepts to understand the organizational structure of

Dilemmas and Choices Facing School Administrators
The classical dilemma of organizational control and personal efficacy
had not been forgotten in the readings for this study. A current article by
Clark and Astuto (1988) entitled Paradoxical Choice Options in
Organizations presented seven choice options of paired elements found in
literature on organizational theory that confront leaders. The authors were
clear about which choices were made by people in effective organizations.
...leaders in effective organizations adopt strategies that result in a
consistent pattern of choice options across the seven paired
elements. The contemporary literature on effective organizations
supports the position that any such strategic pattern would include
activity,1 distinction, variability, efficacy, facilitation, empowerment, and
disaggregation. (1988, p. 126)
Corwin and Borman (1988) described the structural constraints that
limit the capacity of administrators to control the work of teachers and
students in schools [as] structural incompatibilities (p. 209). The six types
of structural incompatibilities they identified as applicable to school districts
were the source of the way power and authority was distributed. The first
structural incompatibility was the:
Dilemma of Control. Central-office administrators are legally
and politically responsible for the actions of schools, but total
centralization is neither administratively nor technically feasible.
Chain-of-command protocols help preserve administrative
control, but they can interfere with the ability of schools to solve
local problems, whereas decentralization and slippage can
interfere with coordination and with administrators official
responsibilities, (p. 209)
The resolution!of this and the other five dilemmas may be found only
through continuing analysis addressed to revamping the archaic structure of
American school systems (p. 234).

Wimpelberg (1987) proposed linkages between supervisors and
principals to resolve the dilemma of helping noneffective schools improve
instructional!^. By developing linkages and using intermediate supervisors,
...the thrust for instructional leadership can come from the intermediate
administrators who supervise and evaluate the work of schooi principals
(p. 106).
He presented a clear viewpoint that teachers and principals could not
accomplish instructional improvement goals alone, even though the center
of change activity remained at the school level. His five propositions were:
1. Instruction in most schools is not likely to improve unless a
leadership consciousness at the district level develops in such a
way as to forge linkages between school and central office,
among schools, and among teachers within schools, (p.106)
2. The best lineages are forged, not through centralized
instructional prescriptions but through an exchange process in
which the central office and school administrators simultaneously
challenge and support each other, (p. 107)
3. The central office personnel with the highest potential for
exercising instructional leadership are intermediate
administrators who have the organizational authority to
supervise and evaluate principals and the expert and referent
authority to support them, (p.108)
4. The primary responsibility of the intermediate administrator is
to see that every school principal develops both a technical and
cultural consciousness of the school, (p. 110)
5. The instructional leadership role of the central office administrator
requires la new kind of intimacy with schools, (p. 111)
The way these five propositions could be accomplished was to have
accurate knowledge (knowing), exchange of information (communication),
and time. Amid all the rhetoric on control, power, authority, and decision

making, here: were straightforward proposals on how to establish
connections among the different roles and structures in the school districts
complex organization. For those school employees who believed in the
efficacy of people, those five proposals were the most satisfying ideas
presented in the literature.
The Use of Control and Influence
Studies about how principals and their supervisors worked together
were infused by the words control, influence, and power. In one study,
control mechanisms were described as the methods used by central
administrators [to] constrain the work of school principals through a complex
web of controls" (Peterson, 1984, p.573). Organizational controls may be
hierarchical, collegial, or nonhierarchical (p. 574). The supervision of
principals was one of the six control mechanisms in Petersons study. His
findings indicated that principals were infrequently supervised, however,
other control rinechanisms were used in addition to supervision. For
example, input control that determined the amount and flow of resources to
schools, and behavior control that used rules, procedures, directives, and
required activities. The central office also used outputs as a control
mechanism, such as: student performance, public reaction, teacher
performance and attitudes, adherence to district rules and procedures, not
making waves, etc. (p. 589). Principal evaluations were included under
environmental |:control with 50% of the principals reporting that the
community was used as a source of information for their evaluations. The
top four sources of information for principals evaluations, according to
principal responses in Petersons study, were: 1) Community and parents;

2) Superintendent; 3) Teachers; and, 4) Central office personnel. Evaluation
information taken directly from the principal was in a distant fifth position.
This perception of how evaluations were written could not lead to trustful
relationships!- Furthermore, public and teacher reaction, compliance to rules
and procedures, not making waves, and student performance were listed as
additional top criteria for principal evaluations. Instructional programs were
a distant sixth.
...the pattern of control in these organizations may increase
managerial job stress. These systems of control employ
multiple sources of information; principals are monitored by
virtually everyone around them. This should increase their
stress. Furthermore, the results of their work, which bring
rewards! or sanctions, are often not clearly delineated by
supervisors, nor are the way superiors determine effective
performance always evident, (p. 595)
The influence of the superintendent on student performance was the
focus of two California studies. To study the effect of the school
superintendents on twelve effective school school districts, Murphy and
Hallinger (1986) developed nine control functions used by superintendents
to coordinate, and control the work activities of site level administrators. The
results of the study showed that superintendents:
...controlled the development of goals both at the district and
school level; they were influential in establishing procedures for
the selection of staff; they took personal responsibility for the
supervision and evaluations of principals; and they established
and regularly monitored a district wide instructional and
curricular focus, (p. 220)
The authors drew similarities between principals as instructional leaders
and superintendents as instructional leaders. Superintendents also had
more active involvement in establishing district direction in curriculum and

instruction, coordinating technical core operations, monitoring internal
processes, and checking outcomes.
A similar study about the influence of superintendents on the academic
achievement of school districts was published in 1987 by Hart and Ogawa.
They used leadership, environment, organization, and district academic
performance as variables. The results were similar to the Murphy and
Hallinger results. Hart and Ogawa note that they thought the estimates of
the influence; of superintendents in the studys results were too conservative.
A third study in the same time frame determined that superintendents
spent most of their time interacting with subordinates and if they changed
the rules it was administrators who were affected not teachers.
Superintendents must use indirect influence if they wanted to make changes
at the teacher level. (Crowson, 1987)
Peterson! joined Murphy and Hallinger to write a companion article to
the one alreaidy reviewed on superintendents and student achievement. In
this article, the authors showed ways that superintendents could control
classroom activities. Mintzbergs concept of professional bureaucracy was
discussed and the definition of technical core (the classroom in school
organizations) was given. Although, in professional bureaucracies,
professionals had discretion over the technical core, constraints were
possible through specifications, standardizations, and monitoring.
Superiors are increasingly using student test scores to assess
educational output. Under these conditions, educators might
develop control and coordination systems that specify the
content of curriculum, methods of instruction, and textbooks to
be used. Increases also may be occurring in the monitoring of
teachers and in the use of texts to assess student achievement.
(Peterson, Murphy, & Hallinger, 1987, p. 81)

The supervision and evaluation of principals and the use of a preferred
model of teaching were additional methods superintendents used to
coordinate and control the technical core.
General Supervision in Education
Textbooks on supervision in education were focused primarily on the
supervision of teachers. Supervisory Behavior in Education bv Ben M.
Harris (1975)1 was most thorough in his discussion of the change process,
studies that related to supervisory practice, and the part that
communications, power, and influence play in interactions with
subordinates.. Wiles and Bondi (1986) were equally thorough in their
development of the concept of supervision as well as the human
development and human relations skills needed for successful supervision.
Another text on supervision by Sergiovanni and Starratt (1983)
i .
emphasized human resources in supervisory theory. As with McGregor
(1960), Sergiovanni and Starratt connected the assumptions and beliefs of
the supervisor to the pattern of supervision used by that supervisor. They
introduced the factor of interaction in their human resources supervision
theory. It had three interacting variables: initiating, mediating, and
effectiveness. These were similar to the causal, intervening, and end-result
variables used by Likert (1961). Initiating variables were the assumptions
and behaviors supervisors had toward supervision. Mediating variables
were the attitudes that employees had toward their jobs and the people with
whom they worked. Mediating variables produced three types of reactions
in response to the pattern of supervision used. Type 1 reaction was
dissatisfaction, prompted by rational-bureaucratic supervision. Human

resources supervisory patterns resulted in type 3 reaction, high commitment
and satisfaction. The Type 3 reaction was similar to the reaction to Theory Y
management patterns described by McGregor.
A recent study on supervision, sponsored by ASCD, presented an
indication of the importance of the people factor in supervision. Identification
of Supervisory Proficiencies Project was completed by a research team at
the University of Georgia (Pajek, 1989). The study identified twelve
categories or dimensions of supervision from an extensive examination of
the literature from the past fifteen years. These categories were verified as
valid by a selected group of exemplary practitioners throughout the country
through the use of two written surveys. The report identified the knowledge,
attitudes, and; skills which were characteristic of twelve dimensions of
especially effective supervisory practices. The attitude statements under all
twelve dimensions of effective supervisory practice used the same language
that we found1 in McGregors Theory Y, the open system of Combs, Avila and
Purkey, Halpih and Crofts school climates, System 4 of Likert, and the
human resources supervision theory of Sergiovanni and Starratt. The
research team reviewed literature on supervision from the last fifteen years,
therefore (except for Sergiovanni & Starratt), this language and related
concepts continued through time from the writings and research of earlier
theorists through the current literature. The attitude statements relevant to
effective communication were especially parallel to earlier works:
encouraging mutual trust; open and approachable; collegial; committed to
open channels of communication; responsive to concerns and aspirations of
others; accepting of diverse viewpoints (Pajak,1989, p.82).

The last section of the ASCD report presented the results of the
telephone survey done with twelve participants who had already returned
the written surveys. In the telephone survey, the twelve were asked to
identify the knowledge, attitudes, and skills that they personally considered
to be important to supervisory practice (Pajek,1989, p. 143). This question
elicited the response which supports the importance of people in the
practice of supervision. One of the clearest themes that appeared in the
telephone interview data was an attitude reflecting a strong belief in the
importance of people (Pajak, 1989, p. 143). The telephone survey section
concluded with a statement about the importance of the human element in
supervision: ,
...the attitudes that were expressed by the outstanding practitioners
in the telephone interviews suggest that supervision in education is a
very personal activity. It requires knowledge and skills, to be sure, but
the human element is paramount. ( p.146)
Final Words
The literature review concludes with a quotation from a textbook used in
college courses for aspiring school administrators and which reflects this
researchers bias:
The real question concerns how the school can be a democratic
community allowing for the participation of its members in its
decision,;making while still being responsible to the larger
democratic community and to its elected representatives...An
administrator is more than a link in the chain of command. He or
she shodld also be a channel of communication, a facilitator of
dialogue, and a repository of educational wisdom. (Haller &
Strike, 1986 p. 328)

Practitioners in public school systems often complain that the public schools
must be all things to all people. What this study hoped to discover was what
supervisors and principals needed to be to each other. If managers in
school districts worked with each other to achieve a democratic community
rather than be locked in a bureaucratic system, the move toward site-based
decision making might prove to be an effective management process. In
the readings referenced above, none gave a total picture of the supervision
of principals. If the most common themes found in the research studies,
control and influence, are discovered to be the most common managerial
practices in school districts, positive relationships between principals and
their supervisors might remain at the conceptual level of discussion.

The Qualitative Research Design
A qualitative research design was developed for this study because little
has been done to examine the supervision of school principals. Although
the supervision of middle managers in industry has received considerably
more attention in the research literature, survey instruments like those used
in industry would not have allowed this researcher to examine the
complexity and subtlety of the way school principals and their supervisors
work together. The researchers experience as an administrator in a large
public schoohdistrict led to the realization that interactions between the two
management: levels are personalized and varied. This was confirmed by
reviewing the contents of interviews conducted in 1989 with seven
principals and two supervisors about the supervision of principals. These
interviews were conducted when the researcher began her long-established
interest in the supervision of principals. A review of these taped interviews,
transcribed shortly after they were completed, reaffirmed that verbal
testimony revealed the personal needs and feelings of people in both
administrative roles. It also confirmed the influence that unique school
situations give to the interactions. This led to the conclusion that listening to
people and watching their behavior would be more effective in capturing the
complexities and subtleties of their interactions with each other than
responses on- a questionnaire developed in industries. In addition,

established management theory from industry may or may not fit all the
interactions and relationships between school principals and the people
who supervise them. Norman J. Boyan (1988) supported this premise:
Another problem is that overdependence on conceptual perspectives
developed outside the schools underestimates the unique characteristics of
educational organizations ( p. 93).
Further study into the works of Glaser and Strauss (1967), Goetz and
LeCompte (1984), Miles and Huberman (1984), Schatzman and Strauss
(1973), and Strauss and Corbin (1990) reinforced the decision to pursue a
qualitative rather than a quantitative research design. Their well-developed
positions on qualitative research and procedures on how to analyze data
obtained through interviews and observations provided
the background needed to initiate this type of research. The vision that
sociologists Glaser and Strauss (1967) presented about this method of data
collection provided the researcher with a sense of rightness. Their
commitment to the development of theory that emerges from careful,
ongoing analysis of the data collected in naturalistic settings was strong
encouragement to use this approach. The detailed techniques on how to
organize, analyze, and present data presented by Strauss and Corbin
(1990) and Miles and Huberman (1984) provided examples of the
systematic steps of coding, matrices, and displays needed to give structure
to qualitative data. Goetz and LeCompte (1984) explained in a thorough,
scholarly manner the full range of preparation and precision required for
comparability and translatability of the research data. The work of these
authors was used in designing and conducting this study.
Strauss and Corbin (1990) described the three components of

qualitative research: (a) data, usually collected from interviews and
observations! (b) analytic procedures, techniques for conceptualizing data;
and (c) written reports, the final product. Qualitative research is any kind of
research that produces findings not arrived at by means of statistical
procedures... It can refer to research about persons lives, stories,
behavior, but also about organizational functioning, social movements, or
interactional relationships" (p. 17). Rather than starting with a theory and
testing it through quantitative statistical methods, the aim of the constant
comparative method of qualitative research is to develop theory by carefully
comparing and contrasting behavior described and recorded as it happens.
Theory development was too extensive a goal for the present study. An
accurate description of the supervision of principals, as reported by
principals and supervisors and observed by one investigator in two school
districts, wasithe aim of this research. Before applying the methods
developed by the authors cited above to the process of selecting a site and
participants fpr the study, coding and analysis had to be completed on pilot
data to develop categories for use in the initial phases of the study.
Coding the Pilot Data
The coding procedures used by Strauss and Corbin (1990) and Miles
and Huberman (1984) were applied to the taped and transcribed pilot
interviews of seven principals and two supervisors conducted 1989. This
activity follows the advice that, creating codes prior to fieldwork is helpful; it
forces the analyst to tie research questions or conceptual interests directly to
the data (Miles & Huberman, 1984, p. 64). Coding procedures are the
bedrock on which grounded theory rests. Coding represents the operations

by which data are broken down, conceptualized, and put back together in
new ways (S>trauss & Corbin, 1990, p. 57).
The nine administrators in the pilot sample were two of the researchers
associates and seven successful elementary, middle school, and high
school principals who reported to them. All nine were asked to talk about
the supervision of principals. The content of these interviews was the first
data to be coded, the results of which served as a starting point for the
interviews and observations to be conducted later in the study. The use of
I ' '
these procedures with the pilot data accomplished three things: (a) provided
practice in the use of coding and categorizing procedures, (b) gave some
understanding of the difficulties inherent in the use of qualitative research
techniques, aind (c) produced a start list of codes to use during the first
stages of datk collection for the larger study. The 60 open codes or
descriptive codes, developed from the pilot data, are listed in Appendix A-1.
These descriptive words were drawn from both principals and supervisors in
their description of what the supervision of principals meant to them. It is
interesting to;note that there was agreement between the seven principals
and two supervisors on 19 of the 60 codes (32%).
The process of breaking down, examining, comparing, conceptualizing,
and categorizing data (Strauss & Corbin, 1990, p. 61) from the pilot
interviews brought a recognition of the difficulties in understanding and
recording thejreal meaning of what people say with single words and
phrases. Listening again to the tapes was as important as rereading the
transcriptions because it gave feeling and meaning to the respondents
words. In this activity, it was discovered that the use of the word
supervision iin the interview questions was somewhat restrictive because,

at first, the interviewees formed their responses in terms of what supervision
should be rather than what it is. Fortunately, the information that emerged
later in the interview became specific to their situations with illustrations and
descriptions of what really happens.
A statement from one of the supervisors helped the researcher move
beyond the word supervision to a more interactive concept, interaction.
He said, I dont think we can effectively supervise someone unless you first
establish a relationship with them that allows a very comfortable interaction
and information flow (Brady, 1989) Interaction and relationship are more
descriptive of the way principals and supervisors describe how they
work together than supervision. From this discovery, the researcher found
further references to the same concept and identified interactions that
described how the two administrative levels work together. Thus, the first
words developed during open coding are listed under interactions,
relationships,! and communications.
The researchers first attempt at axial and pattern coding naturally
evolved from the open coding exercise. After listening to the tapes again,
she found the words listed under open coding were just a small part of what
was described. For example, the nine interviews contained references to
contacts between principals and supervisors (both positive and negative) as
well as non-contacts (both acceptable and unacceptable). The need to look
for what does not happen between principals and supervisors is as
important as what does happen and why.
The code list was a starting point, but the researcher had to look for the
interactions between the descriptors to get a clearer definition of what the
principals and their supervisors in the pilot interviews truly thought about

how supervision of principals was conducted. One example of an
interaction was a contrasting comparison between a principals and her
supervisors description of an interaction labeled quarterly reports. The
two had very different reactions to the same interaction. The principal
indicated that writing quarterly reports was a required task that had no
benefit to hen but she completed it with good grace. The supervisor, who
asked for the; reports, described them as important sources of information for
his written performance appraisals of principals. Thus, there is one
interaction with two diverse reactions. The principal wanted conversations
with her supervisor; instead, she had to produce written reports. The
supervisor gqt material for his evaluations and felt justified in his supervisory
role for requiring the reports. Thus, quarterly reports can be seen as an
i ' . . .
interaction by which principals and supervisors share information with two,
diverse motivations and with two reactions. As a result, the researcher
added reaction to the action/interaction method proposed by Strauss and
Corbin (1990), identifying an action/interaction/reacf/on schema to
accommodate differing reactions of supervisors and principals to
the same interactions, ranging from positive to negative. A series of
descriptions of how principals and supervisors share information and the
reactions to the sharing could form a sequence or series of events that
describes an interaction between principals and supervisors.
This analysis fit under the axial coding process explained by Strauss
and Corbin (1990): a set of procedures whereby data are put back together
in new ways after open coding, by making connections between categories.
This is done by utilizing a coding paradigm involving conditions, context,
action/interactional strategies and consequences ( p. 96). To test the

comparative value of axial coding to pattern coding described by Miles and
Huberman (1984) in the analysis of data, the researcher first applied the
axial coding paradigm model from Strauss and Corbin to the phenomenon,
Scheduled School Visitation," an interaction the principals and supervisors
in the pilot interviews reported as positive. The result of this application is
found in Appendix A-2. Although the process was cumbersome, discoveries
were made after the first list of conditions and properties was completed.
The researcher changed the causal conditions for the school visit from
Supervisor wants to have first-hand knowledge about the school..." to:
The supervisor visits the school to resolve a conflict." Rather than a
scheduled school visit the descriptor now have a problem-solving school
visit." It was apparent immediately that the dimensions, properties,
strategies, arid conditions in the rest of the model would change as a result
of this modification. The reaction to a school visit could be either positive or
negative. For example, one pilot study principal said he wanted to make the
decisions in his school. Using the framework of a problem-solving school
visit, examination of the pilot interviews indicated that principals prefer to
discuss a problem with their supervisor and solve it without a school visit by
the supervisor. A school visit in that instance would be interference and
have a low support dimension. It would appear that the causal conditions
play an anticipatory role in the interaction properties and could be
categorized on a continuum from positive causal conditions to negative
causal conditions that, as already discovered, can differ with the same
interaction depending on the administrative role.
The experience of working with the pilot data using the open and axial
coding procedures described by Strauss and Corbin (1990) indicated that

their processiwas helpful but restrictive when applied to data from
interviews. The importance of relating contrasting information from
interviews was recognized, but the use of pattern code procedures from
Miles and Huberman (1984) would produce the same result. They
summarized data under themes, causes/explanations, relationships among
people, and more-theoretical constructs (p. 68). This straightforward
approach seemed to be more applicable to the comparisons found with
interactions between principals and supervisors. What was emphasized by
both sets of authors is the requirement that the data be analyzed as it is
collected to allow for ongoing verification.
Memoing was suggested by both Miles and Huberman (1984) and
Strauss and Corbin (1990) as a method for conceptualizing data: [Memos]
do not just report data, but they tie different pieces of data together in a
cluster, or they show that a particular piece of data is an instance of a
general concept (Miles & Huberman, 1984, p. 69). The information written
about School Visits" used in the axial coding process could be written up as
a dated memo, noting the effect of contrasting casual properties. Memoing
was given equal importance by both sets of authors, but was most
pragmatically described by Miles and Huberman (1984): When an idea
strikes you, write it down. Memoing should be started early in the process of
data gathering, making sure the memos are accessible. Memos are used for
ideas and for fun.
By testing the pilot data further through the application of Strauss and
Corbins (1990) selective coding as well as Miles and Hubermans (1984)
developing propositions, the researcher was able to analyze the evolving
nature of events in the next step of coding procedures. Selective coding is

the process of selecting the core category, systematically relating it to other
categories, validating those relationships, and filling in categories that need
further refinement and development (Strauss & Corbin, 1990, p.116). The
core category becomes the story line, a conceptualization of a descriptive
narrative about the central phenomenon of the study. In a similar manner,
Miles and Huberman (1984) used "developing propositions as a way to
connect statements that reflect the findings and conclusions of the study.
In reviewing and analyzing the pilot data this story line emerged:
The interactions between principals and
supervisors, over time, form the basis of the relationship
between the two persons serving in those administrative
roles. The frequency and properties of the interactions as
well as the reactions to them by both parties will
determine the quality of the relationship.
Through the interactions, principals and supervisors
learn about each other and develop ways to work together in
a mutually supportive, trustful way. These learnings lead
to an understanding of what the other person will do in
different interactional situations. Events in the school
district may change the way principals and supervisors
interact, thus changing the relationship between the people
in the two administrative roles. The more trustful the
relationship, the less affect the district level changes
will have on that relationship.
This core category, story line, or developed proposition might or might
not have been verified by future interviews. However, by using coding
procedures the researcher began to understand the significance of

emerging categories and the connections among them. The concern was
that the selection of a firm story line at this point in the research might invite a
lack of objectivity and flexibility needed to analyze the data of the next
interviews and observations to be made in the two school districts. This was
only a proposed story line or proposition of how principals and supervisors
work together and parts or all of it will be either confirmed or refuted with
additional data. It was to be used at this point in the research to structure the
questions for the initial interviews of the main study with reservations and the
realization that new questions would be developed from new data.
The core category story line was given more detail and explanation by
clustering information from the pilot interview data under six categories. The
identified six interactions (categories) with dimensions and properties taken
from the pilot data are presented in Appendix A-3. The six categories are:
1. Sharing Information/Communicating can be written or verbal,
mutually interactive or directional, required or volunteered. The dimensions
range from helpful to unhelpful, open to closed, supportive to punitive.
2. Making Contact can be face-to-face (school visits, meetings,
conferences) interactions that are scheduled or impromptu or can be a
phone call that asks for help or made in response to a parental complaint.
3. Problem Solving/Conflict Resolution can be dictatorial or shared and
is often directed by policy.
4. Evaluating/Accountability is generally formalized and unidirectional.
5. Decision Making can be cooperative or directive.
6. Staff Development, when it occurs, is generally interactive.
These six categories are not dissimilar to the systems characteristics in
Likerts profile of an organization: leadership, motivation, communication,

interaction, decision making, goal setting, control, feedback (Likert & Likert,
The coding of pilot data in the A Appendixes was changed and
reworked as the descriptions of supervision emerged from subsequent
interviews. Strauss and Corbin (1990) warned the beginning researcher
about the need to ask questions about the data to verify conclusions as well
as the requirement to redefine or reject categories when additional data
refutes earlier findings. While the pilot categories were not refuted by later
data, they were constantly evaluated, refined, and consolidated. Although
somewhat cumbersome, they were an essential step in the study, serving as
the framework to guide the first interviews and to assist in the early analysis
of new data.
Site Selection
After the decision was reached to use a qualitative research approach
and the coding of the pilot data was completed, the next task in the research
was the selection of sites for the study. Because qualitative research is
accomplished in the context of the working environment where the
action/interaction to be studied takes place, careful selection of the locations
as well as recognition of the events that might influence the
action/interaction in those locations was essential. Thoughtful consideration
of conditions in local school districts under which administrators are
working was necessary. The primary concern, however, must be to develop
a set of attributes based on the research problem and questions so that the
researcher could match the site and participant selections to that set of
attributes (Goetz & LeCompte, 1984).

The most critical attribute for site selection in this study was to have
school districts in which local school decision making was an expectation.
This was an important factor because the perceptions of both principals and
their supervisors toward the supervision of principals needed to be explored
within a district structure that expects school administrators to have a
decision-making process at his/her school. The geographical area in which
the researcher lived and worked had several school districts that were
working on local school collaborative decision making, however, there were
variations among them on how it was to be accomplished. Because of those
variations, the decision to use two school districts rather than one was
required to provide a greater opportunity to discover commonalities as well
as differences in supervision of principals. Glaser and Strauss (1967)
recommended the use of comparison groups because consideration of
similarities and differences forces a more extensive and better analysis of
data. Miles and Huberman (1984) stated that by comparing sites or cases,
one can establish the range of generality of a finding or explanation, and at
the same time, pin down the conditions under which that finding will occur
(P- 151).
The two districts selected as locations for the study were organizations
working toward an emphasis on local school decision making, but in
different ways. The first was an urban district still working under a court
order to integrate its schools and faced with another forced change: a
directive for collaborative decision making at the school level through a
process established by political forces at the district and state level. The
second district was a nearby suburban district where restructuring
and decision making at the local school level was established in a mission

statement developed with extensive community input and adopted by the
local Board of Education. The two organizations presented exciting
prospects for interviews and observations about how school managers work
together under varying conditions of change with the same rubric:
collaborative decision making at the local school level. Both school districts
were moving toward school-based decision making: one through strategic
planning and self-developed collaborative decision making procedures, the
other under an imposed state and district political intervention.
Upper level administrators in both districts were willing to
accommodate a graduate student interviewing and observing principals and
their supervisors. The entry into the two districts varied slightly. In the urban
school district, a supervisor of principals was interested in learning about the
subject and gave the researcher permission to interview some of the
principals she supervised. The suburban district gave the researcher
introductions to the three educational level directors with an open invitation
to precede. Thus, the researcher started at the principal level in the urban
district and at the supervisor level in the suburban district. The order of the
initial interviews did not affect the outcome because after the entry level
interviews were made, the frequency and overlap of principal and
supervisor interviews soon blended the process into a continuous series of
interviews and observations.
Demographic differences between the two districts affected the way
each district was perceived. The numbers and ethnic breakdown of the
student enrollment was one primary factor. In 1991, the urban district had
approximately 60,000 students, with an ethnic distribution of 1.4% American
Indian, 3.6% Asian, 21.6% Black, 40.4% Hispanic, and 33.0% White. The

suburban district had approximately 16,000 students with a distribution of
1% American Indian, 3% Asian, 1% Black, 4% Hispanic, and 92% White.
(The source of figures were from telephone conversations with the public
relations offices in both school districts and were based on 1991 data.)
The numbers of principals in the two districts varied based on these
numbers. The urban school district had 106 school principals excluding
alternative schools: 78 elementary, 18 middle, and 10 senior high. It had
four executive directors who directly supervised the 106 principals as well as
a budget director who handled all budgetary matters. An associate
superintendent had responsibility for the executive directors and all schools.
The suburban school district has 22 school principals: 15 elementary, 4
middle, and 3 senior high. Three educational level directors directly
supervise the'22 principals. There was an assistant superintendent who has
supervisory responsibilities for instruction and the three level directors.
The number of principals supervised by directors was a factor that must
be noted and recognized as an influence on the interactions between the
two levels of administrative responsibilities. The supervisor to principal ratio
in the urban district was approximately 26:1. In the suburban district, it was
approximately 7:1, but principals were assigned to supervisors based on the
grade levels of the schools. Because of this significant difference between
the two districts, commonalities of the supervision of principals across
districts have added depth and the differences needed to be examined in
the context of the demographics of the two districts.
Selection of Participants
As with the site selection, decisions on participant selection were based

on a set of criteria or list of attributes that related to the research question
(Goetz & LeGompte, 1984). The purpose of this study was to observe and
interview both principals and their supervisors about the supervision of
principals. In the two sites selected, all principals had responsibility for one
school and reported to an executive director or level director. The
supervisors of principals reported to one associate or assistant
Following criterion-based selection (Goetz & LeCompte, 1984) and the
purposeful sampling of participants (McMillan & Schumacher, 1989),
selection methods for this study included comprehensive selection, quota
selection, network selection, unique case selection, and reputational case
selection. In addition, the principals whose names were given to the
researcher by the executive director in the urban district who was the first to
express an interest in the study were included.
Comprehensive selection is the use of all cases in the group to be
studied. In this study, comprehensive selection was used with the
supervisors of principals. All direct supervisors of principals in both districts,
four in the urban district and three in the suburban district, were interviewed
for this study. In addition, the budget director and a retiring, direct supervisor
in the urban district were interviewed. The title of the direct supervisors of
principals in the urban district was Executive Director. The title of the direct
supervisors of principals in the suburban district was Director. The associate
superintendent in the urban district and assistant superintendent
in the suburban district who supervise the supervisors of principals were
interviewed for their description of the supervisory process.
Due to time constraints, it was not possible to include all principals from

the districts that participated in the study. Therefore, decisions based on
school level, gender, length of service, and school location were used to
select principals from the subsets of the principal populations. The selection
of representative groups from a larger population is called quota selection
(Goetz & LeCompte, 1984; McMillan & Schumacher, 1989). Elementary,
middle school, and high school principals were included. When both
genders were represented at a school level, both genders were included as
interviewees from that level. Because there were more elementary school
principals than secondary school principals, more elementary principals
were interviewed. Some balance between first year principals and
experienced principals was considered also. Length of service in the urban
district was important because of the large number of new principalship
assignments due to early retirements.
There were four areas in the urban school district comprised of a
group of schools either in the same part of the district or connected by
integration requirements from the court order. Urban principals from each
area were interviewed. The suburban district was not organized into areas,
however, the schools of the principals selected were located in a variety of
neighborhoods within that district.
Additional criteria were used to add participants within the principal
groups. These criteria were: leadership at a level (spokespersons for a
group of urban principals at one level), recommendations made by other
principals or network selection, unique case, and reputational case
selection. The unique case was in the urban school district, and it helped
with the selection of urban participants considerably. The Deputy
Superintendent was appointed 6 months prior to the interviews and,

according to principals, he had spent that time visiting, listening,
interviewing, and learning. Fortunately, this experienced front man was
the only person the researcher knew in the district through their graduate
school classes. His recommendations regarding principals that had the
strength and confidence to share accurate information were employed in the
principal selection process.
Two reputational cases emerged from the urban district. A first year
principal was identified in the local news press when a group of minority
parents tried to get her transferred to another school. Another, more
experienced principal, was noted in the newspapers with less than positive
statements from a central administrator. Because the researcher learned that
principals in that district were essentially left alone unless there was a
problem, these two principals were included in the study.
Three reputational cases in the suburban district also gained notice
through newsipaper accounts. The restructuring process that two high
schools and one elementary school in the suburban district had undergone
resulted in some community unrest. Questions from parents were noted in
the newspaper. These three principals were interviewed.
After selecting the two sites, contact was made to administrators in both
locations to get permission to conduct the study. In the urban district, as
noted earlier, the researcher was given the name of one supervisor of
principals. She gave the researcher the names of four principals: three
elementary and one middle school. The researcher took the opportunity to
interview those principals to start the research project. These four principals,
interviewed at the researchers entry into the school district, were included
as participants in the urban district. The suburban district presented a less

formidable selection process because it was smaller, has fewer principals,
and was not divided into organizational areas. Reputational selection was
used at the high school level and elementary level because three principals
were known for school restructuring through newspaper accounts.
Each district was considered as a separate site. The principal selection
in the urban and suburban districts, using quota, network, unique, and
reputational case selection as well as entry interviews, is summarized in
Table 1.
To enable the researcher and readers to know the source of the
information from interviews without divulging names, codes will be used to
identify participants. For example, U/F/M/2/7 means an urban school district,
female, middle school principal whose school is located in Area 2 with
seven years of experience.
Semi-structured interviews were the major source of data for this study.
They were audio taped to strengthen the internal reliability of the results
(Goetz & LeCompte, 1984). The object of the interviews was to find out what
principals and their supervisors thought about the supervision of principals.
Therefore, the interview questions were directed toward that objective. The
format of the interviews was somewhat determined by the respondents style
(Schatzman & Strauss, 1973) following the lead question about perceptions
of the supervision of principals in the district that opened each interview.
Some participants were able to talk with very little prompting; others needed
the guiding questions.
The first principals were interviewed more than twice because the initial

Table 1
Principals Interviewed for the Study
Suburban Principals
School level n Description
High school 2 2 males o selected on reputational basis o no female hiqh school principals in this district
Middle school 2 1 male 1 female
Elementary school 5 2 males (most senior) 3 females o 1 selected on a reputational basis
Total 9 Suburban Principals
Urban Principals*
School Level1 n Description
High school 3 2 males 1 female o 1 was spokesperson o Areas 1,2. & 3 represented
Middle school 6 3 males
3 females
o 1 was spokesperson
o Areas 1,3, & 4 represented
o including 1 entry interview
Elementary school 14 7 males 7 females o 1 was spokesperson o all 4 Areas represented o including 2 reputational cases o including 3 entry interviews o most senior
Total 23 Urban Principals
* Selections based on recommendations from other principals, the deputy
superintendent, and from one supervisor in the initial phase of the study.

intent was to;do a series of in-depth interviews with a few participants. From
the seven repeat principal interviews, it was learned that the information
provided at the first interview was repeated in the second interview with only
some more elaboration and illustration. Through these seven interviews, the
researcher progressively became satisfied that second interviews were not
needed for theoretical saturation to be reached (Glaser & Strauss, 1967;
Strauss & Corbin, 1990).
Spradley (1979) defined ethnographic interviews as a series of friendly
conversations into which the researcher slowly introduces new elements to
assist informants to respond as informants (p. 58). He warned against
introducing ethnographic elements so quickly that interviews become
interrogations. The desired outcome is to keep the person talking about
what the interviewer is interested in learning. The first part of the interviews
established a connection between the researcher and the interviewees as
well as defining the scope of the questions. Descriptive questions were
asked at the start of the interview, and structural/contrast questions were
asked at the conclusion of the interview. New, emerging concepts led to
questions that searched for clarification and verification in the later
Schatzman and Strauss (1973) recommend a more structured interview
approach with questions designed to elicit meaning from respondents under
specific topics taken from the research question. These follow a who, what,
when, how, and why format. These types of questions were included in the
interviews for both principals and supervisors. The authors further
recommended additional probing questions under reportorial, devils
advocate, hypothetical, ideal, and propositional headings. The question of

ideal supervision of principals enhanced the principal interviews and clearly
separated the responses about what they perceived as real from what they
thought would be ideal.
Goetz and LeCompte (1984) combine several interview strategies with
the advice that the goals and designs of the particular research project
should guide the construction of the interview. They further recommended
that interviews be organized and sequenced as well as include clear
statements as to the purpose of the study. Their recommendations were
followed at the beginning of each interview, but the sequence of the
interviews varied with each participant. Some needed probing questions;
others talked with only clarifying questions as prompts.
Principal Interviews
In all but four cases, principals were interviewed in their schools. The
four exceptions were in the urban school district. One female elementary
principal wanted to meet early for breakfast. (Three secondary principals
also met the researcher for an early breakfast interview, but because they
were also interviewed in their schools, they are not included here.) Two
urban middle school principals were interviewed together in a residence.
One elementary principal, who was briefly interviewed, was unwilling for me
to come to his/her school. The dates of principal interviews and
observations are listed in Appendix B-1.
In two cases, one in each district, the principals originally selected for an
interview were unable to schedule time with the researcher. The suburban
principal was replaced with another principal at the same level and gender.
An urban senior high principal and her potential replacement were equally

difficult to schedule. Therefore, only three urban high school principals were
interviewed.; A former urban high school principal who had recently left the
district to join the suburban district was also interviewed for this study.
Two urban principals were added to the list for interviews. Because of
the large number of new principals starting in the fall of 1992, the researcher
decided to interview a new elementary principal who had not worked in the
district before. The other was a principal who had received unfavorable
comments from a superior. Because two other principals who thought their
positions were in jeopardy declined to be interviewed, this principals
confidence and candor was especially appreciated. The portion of the
interview about the principals difficult situation was not taped; the rest of the
interview about supervision was taped.
The principal interviews ranged in length from 35 minutes to 2 hours.
The two 35-minute interviews were given by the two suburban high
school principals. They were unable to give more time due to scheduling
constraints. The 2-hour interview was with an elementary principal in the
suburban district who was interviewed during a non-student contact day.
Although many other interactions with principals were extended because
they wanted to show me around their schools, most interviews were
between 45 to 60 minutes in length. All agreed to have the interview taped,
except for one principal who was afraid of losing his/her principalship
position. This principal only gave partial information during the interview.
Six urban principals were interviewed more than once during the initial
stage of the study. Those three elementary and three secondary urban
principals were in the researchers district-entry group. It was in their
interviews that;suggestions of other principals to interview were made.

As the interviews progressed from May to December, 1992 and the
researcher became more familiar with the situations under which the
principals were supervised, she was able to verify statements made during
the early interviews with principals in the same district. Principals in both
districts openly discussed supervisory practices in their districts. This
was due partially to the researchers ability to ask relevant questions and
partially because principals had either heard about her or met her in a
principals meeting. It was apparent that the principals generally felt it was
sage to talk with this researcher. The only time the researcher met
apprehension (Spradley, 1979) was with the principal who was under close
scrutiny because community unrest had drawn attention to that principals
school. I was pleased with the frequent comments at the end of interviews
that, It was fun. The goal to have a conversation (Spradley, 1979) was met
in all but the apprehension case. Because two of the seven repeat
interviews were held with two or three principals together, a total of 28
interviews were held with 23 urban principals and a total of 9 interviews
were held with 9 suburban principals.
Supervisor Interviews
Appendix B-2 reports the schedule of supervisor interviews. Because
early contact was made with principals in the urban district and supervisors
in the suburban district, the order of initial interviews with supervisors was
reversed between the two districts. Suburban supervisors were interviewed
in June, 1992; urban supervisors were interviewed at the end of August and
in September, 1992. The interviews of administrators in the two
management levels were independent of each other; therefore, the different

sequence for the two districts did not affect the content of the interviews.
Why were some supervisors readily available for interviews when others
were not? After analyzing the situation, the researcher determined that
personnel changes in the urban district caused the delay, not a reluctance
on the part of the supervisors to be interviewed.
Due to early retirements, there was a major turnover in urban district
personnel at all levels during the interview phase of this study. Three of five
urban directors were new to their positions in July, and the only remaining
supervisor was reassigned to a new group of principals. (The researcher
was able to interview one retiring supervisor before he left in August.)
Two of the urban supervisors were uncomfortable with the tape
recorder so the researcher took notes of those interviews. Another urban
supervisor was interviewed at a restaurant where taping was difficult to
accomplish. Fortunately, the researcher was able to have an incidental
interview with her at a later time. There was another incidental, supervisor
interview in the urban district with a supervisor not originally selected for an
interview. A conversation with the supervisor of high school alternative
programs (also new to her position) gave some interesting insights into the
way principals respond to supervisory directions. A total of eight interviews
were held with eight urban supervisors (and one incidental interview with a
ninth). A total of five interviews were held with four suburban supervisors.
Observations of principals and their supervisors working together
occurred concurrently with the interview phase of the study. The objective of
the observations was to confirm what was said in the interviews and to watch

the supervisors and principals in action-together. As a nonparticipant and
as unobtrusively as possible in regularly scheduled meetings, the
researcher noted the interactions of principals and their supervisors working
together under normal conditions. Attendance at meetings was
supplemented by observing goal setting conferences with principals and
supervisors. Observations were not taped. Goetz and LeComptes
example of noninteractive observation notes (1984), descriptions of what
happened in 5-minute intervals, was the structure used to record events
during meetings. A coding system was devised prior to gathering data
through observations. For example, S indicated the supervisor was talking
to the principals. Notes of the content of the interaction described the type of
interaction it is, such as, decision making, directive, or supportive. P
means principals making a statement, and this notion also was
accompanied by explanatory notes. All agendas and handouts at meetings
were collected.
Scheduling attendance in these meetings was the most difficult
arrangement the researcher had to make. Supervisors were reluctant to let
outsiders into meetings, concerned that the content of the discussion would
be altered. Negotiations were needed to accomplish this essential part of
the investigation. The researcher was excluded from the all-principals
meetings in the suburban district, and she observed the urban district area
meetings only because she was at the place of the meetings at the
scheduled time. Her request to attend the area meetings had not been
explicitly denied, but neither had it been affirmed. The difficulty in the urban
district was one of uncertainty and an inability of the budget director who
was in charge of the meetings to make a decision about the appropriateness

of the researchers attendance. Nevertheless, she was welcomed when she
appeared at the door. The difficulty in the suburban district was the concern
that her presence would limit the participants freedom to express feelings
and opinions in front of a stranger.
Attendance at each of the four area meetings in the urban school district
and two level level meetings in the suburban district were adequate to
assure that the description given by both principals and supervisors of these
meetings was accurate. An elementary principals meeting and a middle
school principals meeting were the two observed in the suburban district.
The senior high director in the same district thought the researchers
presence would disrupt the senior high principals meetings so she was
unable to observe them. The list of observations is in Appendix B-3.
Observations of two goal setting sessions with principals in their
offices and their supervisors were made in October, 1992. Both were with
elementary principals; one in each district. Although the principals in these
meetings did not appear to be concerned, these sessions were awkward for
the supervisors. In one case the supervisor repeatedly turned to the
researcher for confirmation or explanation and asked for feedback about her
performance. In the other, the supervisor limited his responses to brief
statements and questions. Neither session was taped; the researcher took
notes from a position away from the two administrators. After these two
difficult observations, no more were scheduled. These did not reach the
researchers goal of observing unobtrusively and did not add new
information to what she had learned from interviews.
Three incidental observations were made while waiting for appointments
in the supervisors outer offices. I observed how a supervisor handled
ininrw im

appraisal forms submitted by principals, talked to a supervisor about a
school visit with an errant principal, and waited while a principal completed
an unscheduled visit with his supervisor. These observations gave
confirming insights into how the supervisors operated
Collection of Documents
Copies of documents were collected were in conjunction with principal
and supervisor interviews or in principals meetings. If a principal referred to
a written document received from his/her supervisor or other
administrators from the administration building, the researcher asked for a
copy. For example, the urban district Board of Education passed a
resolution about the assessment of students that shifted the responsibility of
testing procedures from the district to the individual schools under certain
guidelines. A copy of that document was obtained from a principal. Both
districts provided documents on the administrators performance appraisal
process. These are illustrative of the way the principals in the two district
relate to their performance appraisals. District mission statements also were
available from both districts. Detailed lists of goals and outcome statements
were not collected because they were too lengthy and rarely were
referenced by the principals. An example of another document which that
the researcher read and returned is a book written by a supervisor in the
suburban district. The list of artifacts appears in Appendix B-3.

Methods Used to Analyze Data
The work of several researchers framed the data analysis. Goetz and
LeCompte (1984) provided a matrix on the relative characteristics of

selected analytic strategies (p. 180). The labeled dimensions on the four
sides are: inductive/deductive, verification/generative,
constructive/enumerative, and objective/subjective. Examination of this
matrix gives a clear explanation of the continua on which different types of
analyses are placed and the rationale for methodology decisions that need
to be made. The use of constant comparisons from Glaser and Strauss
(1967) and Strauss and Corbin (1990) as well as Hutchinsons (1988)
similar comparative analysis process, implies an inductive, generative,
constructive, subjective approach. Using enumeration implies a deductive,
verification, enumerative, objective approach. Aggregation of these
methods is not only possible but recommended and, as with other methods
in qualitative research, depends on what is needed for the research
question (Goetz & LeCompte, 1984).
Shortly after the first urban principal interviews and the suburban
supervisor interview were completed, the researcher transcribed and coded
the interviews using the codes from the pilot interviews (See Appendix C-1
for an example). Two things were immediately evident: The codes from the
pilot study were not adequate for these interviews and patterns emerged
immediately. Pattern coding is the process of looking for themes,
causes/explanations, relationships among people, and theoretical
constructs. One will often note recurring patterns, themes or Gestalts,
which pull together a lot of separate pieces of data. Something jumps out
at you, suddenly makes sense (Miles & Huberman, 1984, p. 216). Others
have used a similar description of the process: While coding and analyzing
the data, the researcher looks for patterns. He compares incident with
incident, incident with category, and, finally, category with category or

incident, incident with category, and, finally, category with category or
construct with construct (Hutchinson, 1988, p. 135).
Categories of meanings began to emerge in the first urban principal
interview, for example, multiple bosses (see Appendix C-2). Urban
principals reiterated the need to relate to many different people in different
departments downtown. After several transcriptions were completed, the
researcher found that making comparisons among the principals
perceptions and combining those perceptions into common meaning was a
natural thing to do. Coded and transcribed interviews (see Appendixes C-3
through C-7 for examples) led to a more efficient process of assigning
descriptions and illustrations to the emerging categories from the taped
interviews. The process became one of expanding the descriptions under
each category as more and more interviews revealed similar perceptions.
When similar perceptions were repeatedly found among persons from a
particular subsample, categories were added. For example, every urban
principal talked about the lack of time their supervisors had because of the
number of schools they were assigned to supervise. This led to a category
called lack of time. The consistency among the principals' and
supervisors interviews led to the development of a list of Categories
Describing Principals and Supervisors Interactions. (Appendix D-1) Many
categories in the list are also found in the pilot data open codes: support,
time, relationships, and shared information. The new interviews added
richer descriptions and illustrations to the pilot data, although, there is a
degree of congruence between the two.

Outside Verification of Interviews and Categories
The danger of becoming satisfied with the results of the data analysis
process without outside verification of the accuracy of the categories was
overcome by the use a group of fellow graduate students in a research
seminar. As constructs became clear and theoretical saturation (Glaser &
Strauss, 1967; Strauss & Corbin, 1990) occurred, these students were used
as outside consultants to verify the accuracy of the codes and categories.
To demonstrate the interview process during one class session, a portion of
a taped interview was played while students read copies of the transcript.
The principal who had been interviewed was present to respond to
questions and to verify the accuracy of the transcription. After reading the
transcript and; listening to a segment of the taped interview, a transcript of a
supervisor/principal observation with the same principal was read and
discussed (see Appendix D-2). The transcription was taken from
observation notes so the presence of the principal to respond to questions
was particularly valuable.
At a later class session with the same group of graduate students,
explanations of the categories that had been developed from interviews was
given during the first hour of class. Following a brief break, four groups of
three graduate students used the categories to rate 5 of the 20 excerpts
taken from both districts principals and supervisors interviews ( see
Appendix D-3). The excerpts were selected at random from the interviews
completed between May and November, 1992. The groups of three were
given one page of five excerpts to rate. Group consensus was reached on
which categories to assign to an excerpt. When the four groups had
completed the task, the researcher collected their category selections and

compared them with the ones she had already assigned to the excerpts.
The details of the process and results are recorded in Appendix D-4.
There was 83% agreement between the groups category
assignments and the researchers: The poorest agreement was with the
category Interaction, which the researcher subsequently deleted from the
categories list. Upon reflection, the researcher saw it as a theoretical
concept that could be assigned to all the interactions between principals and
supervisors. It did not distinguish among the supervisory interactions. The
disagreements with the Relationships category were due to an inadequate
definition, which was strengthened with additional review of the interview
tapes. The recommendation from some members of the class to change
Fragmentation" to Compartmentalization was considered and rejected
because fragmentation was actually used by urban principals as a term to
define their district. Another useful recommendation was to look for
differences between supervisors and principals perceptions of rescuing.
Principals generally perceived it as non-support; supervisors perceived it as
being supportive.
The verification process used with the graduate students to confirm the
accuracy of the interview transcripts, and to verify the categories that
describe the content of the interviews was an essential part of this study.
Quotations from the interviews, used to illustrate the categories of
supervisory interactions in Chapter 4, may now be used with increased

Principal Verification of Descriptions of Supervision
Abstracting salient points from fifty three interviews and eleven
observations is tedious and time consuming. The whole picture can get
blurred in the plethora of details. A story line of the supervision of principals
for each district was developed to describe the supervisory interactions and
to tie the categories together into a contextual framework (Strauss & Corbin,
1990). The descriptions of supervision were sent to all the principals
interviewed in both districts with a cover letter and response form (see
Appendix D-5 and D-6) asking for verification or correction.
Fourteen urban principals returned the response form and the
researcher contacted four other principals to obtain a 78% response rate.
Seventeen reported that the description was correct (four with suggested
clarifications) and one checked that it was incorrect. The principal who
thought the description was incorrect, plus two other principals who said the
description was correct, identified the level of the support they have
received from their immediate supervisor as the one issue they thought was
misrepresented. This information helped to clarify the varying levels of
support from supervisors as perceived by principals. One principal who
returned her form late wrote, When I first read this document several weeks
ago, I worried that is was too general and too generally negative. Now I
believe the report is accurate. I apologize for the delay.
The same process was used with the suburban principals and resulted
in a response rate of 100%. All the suburban principals returned the
response form or called with confirmation that the description was correct, a
few with minor corrections.

The verification of the data analysis by graduate students and principals
was essential for the reliability and validity of this study. The process of
refining the categories of supervisory interactions extracted from the
interviews was integrated throughout the data analysis with the development
of the descriptions of supervision in the two school districts. It was the
interactive, ongoing comparison of the interview data that led to the
refinement of the categories and to the descriptions of supervision (story
lines). The integration of the analysis procedures was essential to this
study; one would not be complete without the other. A complete explanation
of the categories and the descriptions of supervision are given in Chapter 4.
Additional Verification From Observations and Documents
The principal and supervisor interviews provided the data for analysis in
this study. Observations of principal meetings and documents used in the
supervisory process were used to verify the descriptions of supervision
given in the interviews. Although verification of principals descriptions and
supervisors descriptions was aided by comparing and contrasting the
interviews from both sets of administrators, it was necessary to observe the
administrators in action during regularly scheduled meetings to confirm what
was said in the interviews. Thus, this aspect of the study was used to verify
the participants perceptions of behavior against observations of actual
Documents used in supervision and management interactions gave
additional credence to interview statements. Directives from the Board of
Education, memos from central office administrators, performance appraisal

forms, mission statements, and other important documents that impact
principals provided important verification of how principals are supervised.
Reliability and Validity
Every research design must address issues of reliability and validity.
This is particularly true in qualitative research that does not use an
instrument previously tested for reliability and validity.
Quantitative researchers use numbers and measurement to
document consistencies in behavior, and they use definitions of
validity and reliability that relate to instrumentation employed to
show consistency. Qualitative researchers, however, assume that
the meaningfulness of human actions depends on the contents of
situations in which these action, feelings, and perceptions occur.
(McMillan & Schumacher, 1989, p. 187)
The extent to which studies can be replicated and similar results can be
obtained by other researchers is problematic in qualitative research,
particularly when the study is done by a single investigator. The use of
participants to verify the results, the careful selection of participants, and the
detailed description of data collection methods were some of the strategies
used in this study to minimize questions of reliability.
Internal reliability. Because only one researcher conducted this study,
the problem of internal reliability was a major concern. The problem of solo
researcher bias was minimized by using a group graduate students to verify
the accuracy of the interview transcriptions and the emerging coding
categories. Asking the interviewed principals to confirm the descriptions of
supervision for their districts also was used to guard against internal

reliability problems. Principal participants were used in other ways to
address the question of reliability: A few randomly selected principals were
asked to verify the transcriptions of their interviews, and principals
interviewed later in the data collection process were asked to verify the
accuracy of conclusions drawn from earlier interviews. By tape recording all
but four interviews, it was possible to review the exact statements made by
the participants during ongoing analysis of the data.
External Reliability. External reliability is the probability of another
researcher getting similar results in the same or a similar situation.
Ethnographers enhance the external reliability of their data by recognizing
and handling five major problems: researcher status position, informant
choices, social situations and conditions, analytic constructs and premises,
and methods of data collection and analysis (LeCompte & Goetz, 1982, p.
The greatest attention to external reliability problems in this study was
given to the selection of participants. The principal informants were chosen
with equal gender representation from all three school levels and from
school locations that were widely spread throughout the districts. Providing
a careful description of the informants and the reasons for their inclusion in
the study allows another researcher to contact individuals similar to those
who were informants in the prior study (McMillan & Schumacher, 1989, p.
189). All supervisors of principals in both districts were interviewed and
observed to increase the likelihood of getting a more complete picture of
supervision of principals in those two districts.
The context in which the interviews and observations were made in both
districts was an integral part of this study and an important part of the

analysis. Principals and their supervisors work in a public organization with
a hierarchical structure that does impact their work. Their perceptions of
supervision could change with changes in the organizational structure, and
because both districts were undergoing significant change due to budget
reductions, this could become an external reliability problem. To counteract
the problem, the participants explanations of what was happening in their
districts when the interviews and observations were made for this study was
used to provide the social context of the supervision of principals in these
two districts.
The theoretical framework for this study was discussed in Chapter 1
and a detailed account of how the initial constructs about the supervision of
principals emerged from analysis of the pilot data was given in Chapter 3.
Descriptions of the analytic procedures used to discover the categories were
also are presented. The researchers premise that relationships develop
between principals and their supervisors from frequent positive interactions
has been made explicit as well. These detailed accounts were used to aid
the external reliability of this study in response to a statement made by
McMillan and Schumacher (1989):
Simply asserting that formal data analysis was done carefully is
insufficient for establishing reliability and validity. The researcher
must provide retrospective accounts of how data were synthesized
and identify the general strategies of data analysis and interpretation,
(p. 189)
In the latter part of Chapter 4, the dimensions of supervision will be
refined to reflect the analysis of the action/interaction process (Strauss
& Corbin, 1990) between principals and supervisors as described in the
interviews and confirmed in the observations. Perhaps the best analysis

strategy is the use of quotes from principals and supervisors that truly tell the
story. Many of these are formally reported in the following chapters and
Validity does not present the same threat to qualitative research as
reliability: Reliability is a serious threat to much qualitative research.
Validity, however, may be its major strength (McMillan & Schumacher, 1989,
p. 191). This strength is derived from the methods used by ethnographic
researchers: the collection of data for long periods of time; the reality
reflected in participants interviews; the observations of participants in
natural settings; and, the self-monitoring done by the researcher (Goetz &
LeCompte, 1984).
Internal Validity. Internal validity is the degree of accuracy between the
researchers observations and reality. By carefully describing the changing
situations that occurred during the study, the effect of those changes on the
participants responses become part of the data and give added perspective
to the conclusions. Internal validity problems were addressed in this study
by paying close attention to the evolving situations in both districts.
Because she was the sole investigator, the researcher had to be
watchful for any bias she might have brought to the process and carefully
examine her own reactions to informant behaviors. One of the benefits of
living in the greater community but outside the districts under study was the
knowledge she had gained of the districts without being involved directly.
This knowledge helped her understand the history of the events that were

part of the interviews and observations. The use of unique cases and a wide
selection of principals helped guard against selection bias.
External Validity. External validity depends on the identification and
description of those characteristics of phenomena salient for comparison
with other, similar types (Goetz & LeCompte, 1984, p. 229). Although this
was a descriptive study without pretense of generalizing the results to other
school districts, the researcher hoped that the conclusions she was able to
make from the data collected would be translatable and comparable to other
supervisory situations involving school principals and their supervisors. By
using two contrasting districts, whose structures and histories were well
known to the researcher, and by selecting informants from different groups of
principals in each district as well as all supervisors of principals, the
researcher carefully addressed the selection, history, and setting problems
of those two districts that might affect external validity. Any applicability to
other districts; can only be made if administrators in those districts perceive
the information as being helpful to their own situation. The results of this
investigation are reported in the following chapter.

Organizing rich, qualitative interview data into precise results was a
challenging task made even more difficult by the interactive nature of the
analysis process. The description of principals supervision in this chapter
by categories and through story lines answers the question: How do
principals and their supervisors describe their work together? Due to the
contrasting nature of the interview data from the two districts, the results from
the districts are presented separately. First the suburban school district is
described arid then the urban district is described. Because the coding
categories emerged first from the analysis of the data in this study, followed
by the development of the story line description of the supervision of
principals in the two school districts, that sequence determined the order of
presentation within the descriptions of each district. The supervision
categories were given meaning by the use of selected quotations from
principals and supervisors interviews. The story line descriptions
integrated the categories into a conceptual picture or framework of the
supervision pf principals in the two districts and were the descriptions of
supervision sent to all the principal participants in both districts for validation.
A brief explanation of what the principals wanted their supervision to be is
found under their descriptions of ideal supervision. This provides an
understanding of the comparability between the real and the ideal

supervisory practices in the two districts. Then, the principals and
supervisors perceptions of the supervision of principals in the district are
compared. Additional insights into the conditions that influenced the
supervisory situations in the two districts are then addressed based on the
perceptions of what the principals called the culture of their districts and the
effect of that culture on the supervision of principals.
Following the descriptive analyses of the two districts, a comparison
between the districts is presented by contrasting the established categories
on seven supervisory dimensions. The category continuums combine the
descriptions of the supervision of principals from both districts. The
comparison demonstrates a distinct difference between the districts by
identifying on which end of the dimensions the districts descriptions were
primarily located. A summary of the overall perceptions of the supervision of
principals in :the two districts closes this chapter.
The researcher began this the study with only a vague perception of the
two school districts that had been formed from her work in an administrative
position in a neighboring district. The results in this chapter are based on
the content and analysis of the interviews, observations, and documents
obtained during the study. The distinctly contrasting nature of the
supervision of principals in the two districts was an unanticipated surprise to
the researcher. After 10 years as a supervisor of principals, she found the
testimony of: principals and supervisors from both districts presented new
issues and ideas she had not experienced or had overlooked in her work as
a supervisor To be more confident of the accuracy of the her descriptions of
categories and dimensions of supervision, the researcher listened several
times to the interview tapes to reassure herself that her analysis was correct.

This was done especially with the urban principals interviews for an
additional fairness check. In addition, she verified the accuracy of the story
lines with those who participated in the study and had graduate students
verify the appropriateness of her interpretations and coding of the interview
data (See Chapter 3). Thus, it is assumed that what follows is a fairly
accurate description of the supervision of principals in two school districts
during the Idst half of the 1991-1992 academic year and first half of the
1992-1993 academic year.
The Description of the Supervision of Principals
in the Suburban District
Personalized Supervision
When the supervisor and principal shared leadership and took time to
understand each others professional responsibilities, mutual respect and
trust were developed. Personalized supervision was realized when there
was fairly complete, in-depth knowledge about the school and school
j ' .
community as well as a true understanding of the principals preferred style
of leadership. Supervisors with that knowledge as well as an understanding
of recent educational research could provide specific encouragement for
principals to explore new ways to educate students. Personalized
supervision was a partnership of educators that fostered shared information,
shared input] and shared feedback: a mutuality between two school
administrators. It was often described under ideal supervision.
The unique styles of the principals and supervisors were shared and
respected in the suburban district. Both sets of administrators benefited from

the value placed on individual skills and styles. A supervisor of principals in
the suburban district said:
It is wonderful to be in an organization that lets you do what you do best.
I use artistic supervision: I experience the persons environment and write a
metaphor describing it. Good people get bored with evaluations. They are
unique. The cookie cutter kind of mentality can be offset by using the artistic
A principal described his perceptions of the supervision in the suburban
Supervision is a two-way street. Principals must communicate needs and wants
to supervisors. There must be congruency between principals and supervisors.
A whole belief system has to be shared. Supervisory behavior and principal
attitudes must be similar.
The personalized nature of the supervision was illustrated in a principals
description Of her directors statement:
The director said, I dont have a set format. I would like you to do a self
assessment and tell me how youve done your goals and objectives.
There was a great deal of consistency in the supervision of principals in
this school district. Both the supervisors and principals perceived
supervision in a very similar way. The consistency was attributed to a
commitment to the districts Mission Statement, Outcome Statements, and
Adopted Priorities that are included in the Strategic Plan. The meaning
behind the districts four priorities, remembered by the acronym CARS, gave
a common purpose and understanding to the direction the district was
taking: changing needs of students; assessment practices; restructuring
schools; school-centered decision making. Principals also attributed the
congruency to the outstanding skills and abilities of the top management

team. All were liked and respected; the superintendent was revered. As
one principal put it,
I dont know where there is a better team. [Superintendent, Board of
Education, Assistant Superintendent, Director]: there is not a weak one in the
bunch. With that kind of leadership at the top, it is not hard to be successful.
We dont have to second guess.
A supervisor described her need to always be in tune with the district
priorities and with principals:
There ;is a level of trust, mutual respect, about each other. We are held together
by the priorities. There is an understanding that we are all part of the district,
but can be unique. That is a paradox. When principals have the freedom to go
out and do different things, the Board of Education can get beat up, principals
will stub their toes. The ultimate risk is the removal of unity. The thing that has
enabled them to be free. We must be in constant touch with each other. If you
disarm those people who have enabled you to be free then you can lose your
Personalized supervision in the suburban district was the byproduct of
closeness between the people involved and honest feedback. One
supervisor defined supervision as:
Intimaby, physical proximity, closeness, frequency of telephone conversations,
understanding without intruding in personal events in principals life. The
principals and I developed norms together: Principals Principles. How we
will behave with each other. Honesty. Give honest feedback, opportunity to
vent, to process their feelings. A supervisor must be a good listener, not just
listener, not to fix it, but to listen to listen.
The performance appraisal was not as important to principals as the
daily interaction they had with their supervisors; however, the formal

validation of ; what they had accomplished was rewarding. Based on
leadership behaviors, the position description, and performance objectives,
appraisals of principals were completed annually by the supervisors after
getting input from the principals. (Elementary principals now have an
alternate-year appraisal system.) One principal thought the collection of
material for the appraisal was too time consuming, but others found that the
self-reflection helped them to grow. All preferred the conversations and
dialogue with their supervisor throughout the year more than the time
consuming process of writing an appraisal. One principal commented,
A lot of what the supervisor knows about principals is from inference. The
important things are not done in front of him. The meaningful part of the
appraisal is the discussion I have with [my supervisor] about self evaluation.
What he does on a day-by-day and weekly basis and the kind of support that is
real informal is more important to me, much more useful to me. Summation is
not as accurate as the week-by-week interactions. The daily redirecting is much
more important A good evaluation for me is an indication of the success he has
had. The stronger I am, the more it means that he is a good supervisor.
The assistant superintendent indicated that it took commitment to learn
about schools and the role of the principal:
I shadow a principal at each level each year. Just follow the person around for a
day. It replenishes my knowledge of what it is like to be on the firing line.
Supervision was used to maintain continuity among the districts
schools. Although the expectation was that principals, staff, and community
would determine their own approach to implementing curriculum and
restructuring; their schools, the supervisory process helped to keep
principals from straying too far from the districts norms and culture. The
priorities were the guidelines that helped principals make decisions that

were in line with the Strategic Plan. A principals term was for this was
We dont all have to do it the same way and at the same time. The congruency
comes in the direction. We are all going in the direction, using the same
Two kinds of support were found in the data from the suburban district,
personal support and substantive support. Personal support involves a
genuine caring for the individual. Substantive support is built on personal
support and goes beyond it to the provision of new ideas and suggestions
based on the needs of the recipient.
Personal support. The genuine interpersonal caring between principals
and supervisors in the suburban district fostered mutual respect beciause it
implied knowledge of the person and the role that person must perform in
the district. It was more than lip service; it was a demonstration of regard. A
principal reported,
When youve had a bad day, [the supervisor] carries around a little thing like a
bird swallowing up a person and saying, having fun yet? He gave it to me. It
lightens up the whole thing. That is an example of the collegial relationship we
A supervisor talked about the closeness needed for supervision:
Intimacy is necessary for supervision, that is physical proximity, closeness,
frequency of telephone conversations, understanding without intruding in
personal events in their lives. I called a principal to remind him of a report that
was due because he was very busy with other things. I did it to help him

Substantive Support. Substantive support necessitated knowing the
other persons role responsibilities and thinking in terms of those
responsibilities rather than ones own. Substantive support in this district
was focused on the principals needs and enhanced his/her leadership and
professional; growth. It required a true understanding of how the principal
thought and led his/her school. A principal reported,
The idea of the restructuring came from the school, but it was made in a district
that fostered that kind of change. We developed along parallel tracks. The
school made a presentation to central staff. Central administration had
confidence in what the school was doing.
The supervisors in this district came close to achieving ideal supervision
by understanding the individual leadership styles of each principal and by
using that knowledge while interacting with them. Substantive support was
given through suggestions, offers of assistance, and time for conversations.
The category name, Substantive Support, came to the researchers mind
after leaving the suburban central office building where she had heard a
director plan a way to help a principal. The thought was that the exchange
went beyond personal support because it had substance as well as
personal regard. A principal expressed what substantive supervision meant
to her :
Directors need to make clear to principals that these are tricky waters. If
principals are not sure of legal positions, the director needs to be available to
answer questions. I cant stand really close supervision, but I have to have
room to make mistakes. He doesnt tell me what to do, but tells me what he
recommends... He gives direction, but never judges.

Two good examples of substantive supervision were noted by the
researcher during observations of principals and supervisors in the
suburban district. During a goal setting session, a supervisor suggested that
he and the principal get together to brainstorm some ideas about
assessment after he gave her warm praise for her performance and her
goals. In two principals meetings, the researcher observed principals
participating not only as members of a team; in one case they directed the
meeting themselves with support and facilitation from the assistant
A principal described a situation in which her director attended a
sensitive parent meeting at her school:
He didnt assert himself, but sat in the back of the room. He didnt want people
to talk to him and didnt want to be in the way. He was there in case we had
any problem; he just keeps himself informed. I was very comfortable about
him being there.
The directors in the suburban district did not rescue principals; they gave
them substantive support that enhanced their leadership skills.
Time Commitment
This category was developed to demonstrate the amount of time
provided for the supervision of principals in the suburban district. Much time
was given to supervision to ensure continuity of educational programs as
well as to learn what was happening in the schools. Collaboration, time to
develop relationships, team building, unity, and mutual trust were terms used
by supervisors to describe the supervision of principals in their district. Time
was provided to develop a common commitment to the districts priorities

(CARS). That common commitment allowed the principals and supervisors
to develop common bonds. The issues were clear, so interactions could be
based on shared goals. Mutual support and mutual respect developed from
those positive interactions.
Supervisors valued the time they spent with principals. Meeting
schedules ranged from weekly to bimonthly, and school visits were frequent.
Time commitment in the suburban district meant giving time for principals in
lieu of doing other things. Supervisors believed it was necessary to give
time to develop relationships. Relationships among administrators were
essential to share the culture and ensure that the district priorities were met.
A principal stated,
Time together to build those relationships is where it all grows from. If you
dont have time to build relationships and if you dont understand peoples
belief systems and what they are trying to do, you dont do anything. Weve
got to spend that kind of time together. Our teaming beliefs are a result of a
retreat we did together. The retreat was set up by our supervisor. We put it on
our calendars. We would never have come up with that if we hadnt gone away
to the retreat.
Another example was the time needed to solve a competitive problem that
arose within the senior high school ranks. The associate superintendent
We had to pull back on student transfers and give that responsibility to the
director. Not to rescue. Just buying time so the group can work out norms and
solve their own issue.
A supervisor acknowledged that time was given to help each other.
Supervisory time has high value in this district. We spend time talking.

Sharing Information
Sharing information was characterized by openness to different
points of view and to giving and receiving feedback. Collaborative behavior
was enhanced because administrators discovered what the other person
was thinking. This was shared knowledge. In the suburban district, all
administrators took a personal style inventory that described each persons
preferred style. This information was shared and used to foster
understanding and open communication.
Principals appreciated the number of visits made to schools by
directors. Although there was no way supervisors could understand fully the
daily work of the principals, those frequent visits permitted supervisors to
gain an understanding of the school, its successes and challenges.
Intimacy was nurtured because principals and supervisors learned about
each other through frequent interactions as they worked together toward the
same goals.
I keep my supervisor informed; he keeps his supervisor informed. It is a good
model. If the central office is not informed, they cant help.
The personal, interactive style of sharing information was fostered by
frequent meetings that were scheduled for all three levels of school
administrators, 1) elementary 2) middle 3) high school.
Communication does not occur via memo. Weekly informal breakfasts,
monthly half-day meetings, and calls give an information-rich environment.
A senior high principal described the importance of meeting together:
We meet twice a month. It is very important to share information from the
school district. Our director goes around the table and makes sure everyone

gives an opinion. If our director is going to represent us, he must get input
from the principals.
Shared Decisions
Directives were not the way supervisors communicated with principals
in the suburban district. Supervisors and principals were described as a
team, a closely knit unit that knows each others warts, and a sharing, self-
actualizing group. Principals were encouraged to call on any person who
could best help them. The assistant superintendent was contacted if she
was the one with the knowledge or skill needed to assist the principal. The
superintendent as well as Board of Education members were viewed as
supporters for the schools.
The suburban supervisors descriptions of supervision were full of
references to teaming, relationships, mutual respect, and intimacy. They
talked about the Strategic Plan as the glue that formed ties within the district
and helped everyone work in the same direction. The directions given to
schools were within the expectations of the Strategic Plan and did not have
to come from supervisors as directives. Everyone understood the
framework, and relationships were developed between people, unimpeded
by reminders of what to do. One supervisor described her task this way:
The supervisors job is to help principals to take time to make sure the
innovation fits in the districts priorities and that everyone understands it You
are far less apt to make decisions based on the moment. Decisions are based on
Shared Decisions were made through an inclusive process that
necessitated time to include stakeholders in decisions. Both suburban

principals and supervisors reported a similar commitment to shared decision
making. It was a district norm. For example, the agendas for suburban
principal meetings had a great deal of input from principals and were often
designed by the principals.
One incident that described the practice of asking principals for their
opinions was reported by a suburban principal. When the final decisions
were made on the administrators evaluation process, the assistant
superintendent visited with this principal to get his opinion. His comments
indicated that evaluation was not his favorite activity; she was checking with
him to get feedback from a negative voice. The process used in the district
to revise the administrators performance appraisal system was developed
through shared leadership:
We developed the leadership behaviors and appraisal system through a shared
process. The committee working on it asked principals what they wanted; came
back to principals meetings for more input. We feel like a team; when
director is not there, we miss him.
Another principal expressed the togetherness that was developed among
principals and with their supervisors.
We do things together. We develop shared decisions as a level, for example,
assessment of students. Our teaming belief is that we do things together.
Our director believes in that too. We are in this all together. We wrote the
proposal for site-based management; it did not come from [the] Board of
Education. We did that together.
Budget decisions were made for the district by groups made up of
representative school personnel, parents, and community members. The
decisions were communicated frequently in all the schools for verification

and validation. The same process was used in local school budgets. The
superintendent of schools explained the process the whole district went
through to make budget decisions. A committee made up of teachers,
administrators, classified employees, and community members solicited
input from all their groups and formed recommendations. The committee
members then went back to schools to explain the recommendations to them
and get feedback before taking the final recommendations to the Board of
We put together a team to discuss funding, not budget cuts, called the
Strategic Funding Team. The Board took 44 of the 56 recommendations to
reduce 2.7 million [dollars] from the budget
The superintendent talked openly about how much trust had been lost
when she made an unilateral decision. Following that incident, she met with
the principals to ask how she could regain the trust that was lost. The district
was committed to shared, school centered decisions rather than central
office decisions.
Principals expressed their commitment to the same value:
We make decisions together that affect our schools. We do it together. We
developed a staffing process to share staff among the four middle schools. The
director facilitated it.
Relationships were the heart of the supervision in the suburban district,
frequently mentioned in the interviews and given a lot of attention by both
principals and supervisors. Relationships were developed over time
through frequent interactions. Both sets of administrators understood each

others roles and behavior patterns. Relationships were influenced
significantly by district expectations, culture, and norms. It was through
interpersonal relationships that trust and mutual respect were developed.
Supervisors: and principals valued the frequent contact they had with each
other and recognized the importance of open, honest communication.
Supervisors frequently referred to the need for positive relationships and the
need to schedule time to maintain relationships:
You must have time together to develop relationships. You schedule time
together because you value relationships. Trust develops through collaboration,
follow-through (do what you say you are going to do), and problem solving to
assist the schools. Mutual respect and honesty are part of relationships.
Another statement from a supervisor addressed the need for relationships
between principals and supervisors so principals would know it was all right
to try and fail:
It is important for relationships to be solid and positive between supervisors and
principals. Principals must be able to tell that, because if they have to ask a
question, they may be put in a position where they may fail. We are dealing
with complex situations. No one person can do it all themselves. Principals are
responsible, so [they] must have good relationships with the supervisor so
the supervisor can give help, talk things over.
Principals echoed that perception:
A positive relationship is necessary for supervisors to help principals. A sense
of we as we solve problem together. Our director listened to principals and
built a level of trust because he valued principals. He avoided being all
knowing or fixing things.
Taking time together was the way relationships in the suburban district
were developed. Meetings, once a week or twice a month, were perceived

by the principals and supervisors as valuable time together. Principals had
a lot to contribute during the meetings. One meeting the researcher
observed was planned and implemented by the principals. Principals
thought that meetings gave the structure needed for relationships with their
supervisors and with their fellow principals.
I am most appreciative of the way the four middle school principals do things
together. Middle level principals wrote a proposal for site-based management.
It came from principals, not the Board of Education or superintendent.
Meetings are used to maintain relationships, used to develop relationships.
As another example, articulation among levels was addressed in
discussions about assessment of students. Level directors were facilitating
that process.
Level directors have set up monthly meetings so middle school and high school
people could work together to resolve articulation problems and assessment
In all three educational levels, principals were closest to their own peers
and directors, taking pride in the work they did together. There was a
difference in the relationships between principals and supervisors within the
three levels. Because of the newness of the elementary director to the
district and because the number of principals at that level made interactions
less frequent, mutuality was still being developed. Elementary principals felt
free to give their director feedback, and then reported that he responded to
feedback in a very positive way. Supervisors expected and valued direct
feedback from their principals. A principal stated,
Our supervisor wants feedback. We have confrontations so we can understand
each other, share ideas, our point of view. We get our ideas out on the table. It
is a team process. I credit my director for allowing that to happen. We have

open communication. The [Director] is right with us. I have a respect for a
supervisor who encourages principals, who challenges principals to express
beliefs. My director is not threatened and continues to strengthen me.
A supervisor was pleased that all but two principals had returned a recent
request for feedback:
It takes a lot of time, but we work collaboratively. They give me feedback. I
sent out an evaluation request to principals. All but two came back.
The mutuality between principals and supervisors was stated by a
supervisor like this:
The principals and I negotiate what is reasonable to expect me to do for them.
They set goals for me. We have an off-site retreat every year to build our
relationship. It is a real mutuality; a building principalship is an impossible job
to do.
There was a closeness among the principals and between the
supervisors and principals. If competition occurred, steps were taken to
reduce it. There also was a close relationship within the upper
management team. Each director and the assistant superintendent shared
information during individual weekly meetings and group weekly meetings.
Shared decisions were made after considerable dialogue and with an
understanding of each others roles.
School Centered
This category was developed from one of the four explicitly stated
priorities in the suburban district: school-centered decision making. The
expectation was that decisions were made at the school level and shared
decision making was the method used to make those decisions. Principals

were responsible for those decisions and were supervised under that
assumption. Service to schools was the modus operandi throughout the
School-centered decisions required a service orientation from people
downtown to the schools. Decisions were driven by school needs not by
downtown demands, it took a commitment from everyone in the
organization to make that happen. The suburban district was committed to
site-based management at all levels within the organization. When there
was a lapse in this commitment, it was quickly rectified. One principal
wanted more central office people available to help the schools, not less:
The role of people supervising schools is one of support and coordination and
resources. We need more people downtown to help the schools.
The predominance of the service commitment was demonstrated by district
people reacting to principals concerns about the load of paper coming to
schools from downtown. A decision was made to route all paperwork
through one office so principals received only the important, relevant
material. Principals felt very free to call a downtown office and complain
about a repeated request for information or about a meeting that called staff
away from the building.
I tell people downtown that I will not do the paperwork. The less I talk to
people downtown, the better I like it. The high school level people meet with
our director and thats all I need.
Two principals reported that there were some vestiges" of former top-
down directives remaining and a balance between schools and central
offices was still being established.

We have vestiges left over that are still top down. Finance, personnel,
maintenance-that bureaucracy is still there. Data processing think we work
for them.
A principal at another level talked about pockets of departments in the
district that had not shifted their roles:
The place where budgetary decisions are made needs more people who help us
do what we decide to do. We dont need more people if they are going to
decide what we are going to do. There is a role change that has not taken place
in this school district that supports school centered decision makingpockets of
district departments.
Generally, however, there was an expectation that principals were
responsible for their schools and that restructuring would be done with a
process that included staff and community. A principal said that supervision
follows that expectation.
He doesnt expect to have all of us do things in the same way. He is very open
with supervision. He says, You have to do this, but you can do it in your own
way. You each are different.
Supervisors were used as support persons, sounding boards, and listeners.
They provided options for principals to think about; there was sharing of
information rather than telling. A suburban supervisor described her role in
helping principals move forward without tripping over their enthusiasm:
It is hard for principals when they are ready to go and we have to put up a door
so the rest of the district has time to understand the new ideas. We dont put up
a wall, but a door with a knob. The supervisorss job is to help principals take
time to make sure the innovation fits into the districts priorities.
The school-centered category presents a good overall description of the
supervision of principals in the suburban district. It reflects the sharing of

decisions and information, the amount of time, and type of support given to
principals. School centered, as a district priority, is a pervasive value that
gives direction to how people in the suburban district work together. This
category is the core category for the suburban district that the following story
line description describes in more detail.
A Storv Line Description of Supervision
in the Suburban District
There was a great deal of congruency in the supervision
of principals in the suburban school district.
Collaboration, time to develop relationships, teams, unity,
and mutual trust were terms used by both supervisors and
principals to describe the supervision of principals in
their district. A common commitment to the district's four
priorities (CARS: changing needs of students; assessment
of students; restructuring schools; school-centered
decision making) provided a common bond for supervisors and
principals. Principals also attributed the congruency to
the outstanding skills and abilities of the top management
team. All were liked and respected; the superintendent was
Principals were responsible for the decisions made at
their schools and were supervised under that assumption.
Service to schools was the modus operandi throughout the
district. Although some vestiges of former top down
directives remained and a balance between schools and
central offices was still being established, principals

reported the service mode was there. The prevalence of the
service commitment was demonstrated by district people
reacting to principals' concerns about the load of paper
sent to schools from "downtown." A decision was made to
route all paperwork through one office so principals
received only the important, relevant material. Principals
felt very free to call a downtown office and complain about
repeated requests for information or a meeting that called
staff away from their school.
Supervision was used to maintain continuity among the
district's schools. Although the expectation was that
principals, staff, and community would determine their own
approach to implementing curriculum and restructuring their
schools, the supervisory process kept them from straying
too far from the district's norms, culture, and priorities.
Supervisors were used as support persons, sounding boards,
and listeners. They provided options for principals to
think about; there was sharing rather than sharing of
information. Norms for principal/supervisor interactions
were developed mutually among the principals within a
Good relationships between principals and supervisors
were the heart of the supervisory process in the suburban
district. Taking time together was the way those
relationships were developed. Meetings, once a week or
twice a month, were perceived by the principals and
supervisors as valuable time together. Principals had a
lot to contribute during the meetings. Supervisors

expected and valued direct feedback from their principals.
Articulation between educational levels was not a
natural by-product of the close association among people
working at the same school level. Supervisors worked with
principals from the different levels to overcome
differences and assure continuity.
Principals appreciated the number of visits made to
schools by directors. Although there was no way
supervisors could understand fully what the principals'
daily work was, those frequent visits permitted supervisors
to gain an understanding of the school. Principals learned
from trying new educational approaches and were supported
by the presence of their supervisors in their schools.
Performance appraisals were not as important to
principals as the daily interactions they had with their
supervisors; however, the formal validation of what they
had accomplished was rewarding. Based on leadership
behaviors, position descriptions, and performance
objectives, appraisals were completed annually by the
supervisors after getting input from the principals. One
principal thought it too time consuming, but others found
that the self-reflection helped them to grow. All
preferred the conversations and dialogue throughout the
year rather than the time consuming process of writing an
This district came close to ideal supervision through
positive interactions. These interactions were improved
with the knowledge gained from the personal style inventory

taken by each administrator. Substantive support was given
through suggestions, offers of assistance, and time for
conversations. Directives were not the way supervisors
communicated with principals. Supervisors and principals
were "a team," a "closely knit unit that knows each other's
warts," and a sharing, self-actualizing group. Principals
were encouraged to call on any person who could best help
them. The assistant superintendent was contacted if she
was the one with the knowledge or skill needed to assist
the principal. The superintendent as well as the Board of
Education members were viewed as supporters for the
schools. The supervisors were selected for their experience
and commitment to the district values. The interview
process for directorships was thorough and comprehensive.
The motto used by supervisors was: Lead, guide, and
Suburban Principals Ideal Supervision
The suburban principals told the researcher that they thought the
supervision they were experiencing was close to ideal. There were
comments from principals who wanted a perfect balance between clear
directions and shared decisions, between complete support and full
autonomy, and between site-based decisions and central personnel ready to
help. Given the balance of options in the suburban district, however, the
principals believed they were very well supervised. Every suburban
principal interviewed expressed satisfaction for the supervision they were
given and described with pleasure their interactions with their supervisors.