|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 LEADERSHIP FRAME PREFERENCES OF ELECTED AND APPOINTED SCHOOL DISTRICT SUPERINTENDENTS IN ALABAMA, FLORIDA, AND MISSISSIPPI By CHERI LYNNE LANDRY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Cheri Lynne Landry
3 T o my husband Carl for all of your support
4 ACKNOWLEDGMENTS It is with much gratitud e that I express heartfelt thanks to members of my dissertation committee. I especially want to thank Dr. James Doud for all of his mentoring, his guidance, and most of all for his patience and understanding. He kept me going during difficult times when I just wanted to give up. His detailed feedback and his encouragement enabled me to complete this arduous task. Dr. Douds dedication to his students even after his retirement has been phenomenal. I am most grateful that he never seemed to lose faith in me.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...................................................................................................... 4 LIST OF TABLES ................................................................................................................ 8 ABSTRACT ........................................................................................................................ 10 CHAPTER 1 INTRODUCTION ........................................................................................................ 12 Historical Roles for Superintendents .......................................................................... 13 Statement of th e Problem ........................................................................................... 17 Purpose of the Study .................................................................................................. 18 Research Questions ................................................................................................... 18 Delim itations ......................................................................................................... 18 Limitations ............................................................................................................ 19 Definition of Terms ...................................................................................................... 19 Justification for the Study ........................................................................................... 20 Organization of the Study ........................................................................................... 22 2 REVIEW OF RELATED LITERATURE ...................................................................... 23 Organizational Leadership Theory ............................................................................. 24 Superintendent Leadership ........................................................................................ 29 Role of the Superintendent .................................................................................. 30 Characteristics of Superintendents ..................................................................... 32 Appointed or Elected Superintendents ...................................................................... 33 Leadership Orientation Frame Studies ...................................................................... 43 Frame Choices ..................................................................................................... 43 Frames and Effectiveness ................................................................................... 44 College Presidents/Deans ................................................................................... 45 Superintendents ................................................................................................... 47 Foreign Cultures ................................................................................................... 49 School Administrators .......................................................................................... 51 Studies Addressing Specific Demographic Characteristics ................................ 54 Gender .................................................................................................................. 54 Age, Race, Education, and Experience .............................................................. 55 Tenure and Size of District .................................................................................. 56 Summar y ..................................................................................................................... 56 3 METHODOLOGY ....................................................................................................... 59 Procedure .................................................................................................................... 59 Research Questions ................................................................................................... 59
6 Research Hypotheses ................................................................................................ 60 Type of Study .............................................................................................................. 60 Study Population ......................................................................................................... 61 Instrumentation ........................................................................................................... 61 Reliability of the Leadership Orientations (Self) Survey ..................................... 63 Validity of the Leadership Orientations (Self) Survey ......................................... 64 Data Collection ............................................................................................................ 64 Data Analysis .............................................................................................................. 65 Summary ..................................................................................................................... 68 4 RESULTS AND DATA ANALYSIS ............................................................................. 6 9 Descriptive Data.......................................................................................................... 71 Demographics ...................................................................................................... 71 Self -rated Effectiveness as Manager and Leader ............................................... 74 Participants Use of Cognitive Frames ................................................................ 75 What are the self -reported leadership orientation frame preferences of the elected and appointed school district superintendents in Alabama, Florida, and Mississippi? ........................................................................... 75 Is there a difference between self -reported leadership style preferences of elected and appointed school district superintendents? ....................... 76 Resear ch hypothesis: There are no differences between self -reported leadership frame preferences of elected and appointed school district superintendents in Alabama, Florida, and Mississippi. ............................. 76 Is there a difference in single or multi -frame preferences between elected and appointed school district superintendents? ........................... 78 Research hypothesis: There are no differences in single or multi -fra me preferences between elected and appointed school district superintendents in Alabama, Florida, and Mississippi. ............................. 79 Multiple Regression .................................................................................................... 80 Is there a difference in leadership style preferences of elected and appointed school district superintendents when considering selected demographic variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in educati on, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district)? ................................................................................... 81 Demographic variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district) have no statistically significant effect and/or predictive value on the self -reported lead ership frame preferences of elected and appointed school district superintendents. ............................ 81 5 SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ................................... 92 Summary ..................................................................................................................... 92 Discussion ................................................................................................................... 93
7 Demographics ...................................................................................................... 93 Superintendent Frame Use.................................................................................. 97 Multiple Regression Analysis ............................................................................. 100 Conclusions .............................................................................................................. 102 Impli cations of the Findings ...................................................................................... 103 Recommendations for Further Study ....................................................................... 104 APPENDIX A Request and Permission to Use Survey Instrument ................................................ 106 B Survey Letter, Endorsement, and Follow -up ........................................................... 109 C Leadership Orientation (Self) Survey ....................................................................... 111 D IRB FORMS .............................................................................................................. 116 LIST OF REFERENCES ................................................................................................. 119 BIOGRAPHICAL SKETCH .............................................................................................. 127
8 LI ST OF TABLES Table page 2 -1 Superintendent demographics from AASA superintendent studies 2000 and 2006 ........................................................................................................................ 33 3 -1 Descripti ons of dimensions of leadership of the LO(S)S ...................................... 62 3 -2 Reliability of the Leadership Orientations (Self) Survey ....................................... 63 4 -1 Age, gender and ethnicity of superintendents N = 145 ........................................ 72 4 -2 Highest degree, method of selection, years as superintendent, years in current district, experience in education, district size, state (N = 1 45) ................ 73 4 -3 Effectiveness as manager (M) and leader (L) ....................................................... 74 4 -4 Effectiveness self -rating: Manager (M) and leader (L) by selection method........ 74 4 -5 Superintendent frame preference (mean and standard deviation) (N = 145) ..... 76 4 -6 Superintendent frame preference ( *n = number of valid responses) ................... 76 4 -7 Superintendent frame preferences (M) by selection method ............................... 77 4 -8 Analysis of variance of f rame preferences by method of selection (N=136) ...... 77 4 -9 Single and multiple frame preferences N = 145 ................................................... 78 4 -10 Single and multiple frame preferences by method (N=145) ................................. 79 4 -11 Chi -square ( x 2) of no preference, single and multi -frame use by appointed and elected superintendents (N = 145, Missing = 14) .......................................... 80 4 -11 Summary of Multiple Regression Analyses for Variables Related to Superintendents Use of the Structural Frame ...................................................... 83 4 -12 Summary of mult iple regression coefficients for model 1 Structural (appointed) ............................................................................................................. 83 4 -13 Summary of multiple regression coefficients for model 2 Structural (elected) ..... 84 4 -14. Summary of multiple regression analyses for variables related to superintendents use of the Human Resource Frame .......................................... 84 4 -15 Summary of multiple regression coefficients for Model 1 Human Resource Frame (appointed) .................................................................................................. 85
9 4 -16 Summary of multiple regression coefficients for Model 2 Human Resource Frame (elected) ...................................................................................................... 86 4 -17 Summary of multiple regression analyses for variables related to superintendents use of the Political Frame .......................................................... 86 4 -18 Summary of multiple regression coefficients for Mod el 1 Political Frame (appointed) ............................................................................................................. 87 4 -19 Summary of multiple regression coefficients for Model 2 Political Frame (elected) .................................................................................................................. 87 4 -20 Summary of multiple regression analyses for variables related to superintendents use of the Symbolic Frame ........................................................ 88 4 -21 Summary of multiple regression coefficients for Model 1 Symbolic Frame (a ppointed) ............................................................................................................. 88 4 -22 Summary of multiple regression coefficients for Model 2 Symbolic Frame (elected) .................................................................................................................. 89 5 -1 African American popul ation percentage (N = 2008 population) .......................... 95 5 -2 Participating superintendents by gender, ethnicity and academic degree (N = 145) ......................................................................................................................... 96 5 -3 Appointed and elected superintendents by state .................................................. 97
10 Abstract of Dissertation Presented to the Graduate School of the University o f Florida in Partial Fulfillment of the Requirements for the Degree of D octor of Philosophy LEADERSHIP FRAME PREFERENCES OF ELECTED AND APPOINTED SCHOOL DISTRICT SUPERINTENDENTS IN ALABAMA, FLORIDA, AND MISSISSIPPI By Cheri Lynne Landry December 2009 Chair: Name James L. Doud Co -chair: David Quinn Major: Educational Leaders hip This study investigated the leadership orientation frames of school district superintendents in Alabama, Florida, and Mississippi, which are the only three states in the United States with both elected and appointed school district superintendents. The theoretical framework for this study was developed by Bolman and Deals (1997) four frame typology. The frames included in the typology are the Structural, Human Resource, Political, and Symbolic. The purpose of the s tudy was to determine whether th ere was a difference in frame use between elected and appointed school district superintendents based on their method of selection. The study also investigated the effects of demographic characteristics (age, gender, ethnicity, highest academic degree, ye ars of experience in education, years of experience as superintendent, years of experience as superintendent in current district, and size of district) on the superintendents cognitive frame use Superintendent use of single and multiple frames was also investigated.
11 Of the 348 Leadership Orientation (Self) Surveys distributed to the population of school district superintendents 145 (42%) were returned. Surveys were completed by 89 appointed superintendents and 56 elected superintendents. The demographic characteristics of the participating superintendents mirrored the demographics of the larger population of superintendents in these states. The pattern of frame use (Human Resource, Structural, Symbolic, and Political) proved to be the same for elected and appointed superintendents ; however, elected superintendents had a statistically significant mean score for the Human Resource Frame. Frame use implied that the school district superintendents in these three states were effective school managers. Regre ssion analysis results suggested that method of selection as well as demographics (age, gender, ethnicity, highest academic degree, and district size) had a statistically significant effect in the use of Structural, Political, and Symbolic frames when perf ormed with appointed as the selection variable. The findings have implication for professional development of school district superintendents as well as those aspiring to the position. Further studies on superintendent leadership effectiveness and frame c hoice are indicated, especially those that include feedback from peers and colleagues.
12 CHAPTER 1 INTRODUCTION Lifes daily challenges rarely arrive neatly categorized or clearly labeled. Instead, they flow over us in a murky and turbulent stream of exper ience. The art of reframing, and of leadership, uses knowledge and intuition to make sense of the flow and to find sensible and effective ways to channel the current in productive directions (Bolman & Deal, 1997, p. 354). The challenges of leadership in various types of organizations have long been a topic for serious study. Just as theories of leadership and organizations have evolved over the years, so has the role of school district superintendent changed since the first local superintendencies were e stablished in 1837 (Konnert & Augenstein, 1995; Norton, Webb, Dlugosh, & Sybouts, 1996). The superintendent of schools in America have one of the most responsible and complex roles in modern society. The superintendents role mak es them central to the w elfare of their communities, and they are often very visible in their communities, but their job is rarely understood or fully appreciated (Houston, 2007, p. ix). Leadership has only grown in complexity requiring us to reassess our vision of how leaders f unction in our contemporary, complex, technological organizations. In 1962 the American Association of School Administrators wrote that the success of schools and students is strongly related to the caliber of school district leadership (Carter & Cunningh am, 1997). There are major gaps in the research regarding the superintendency and its impact and even larger gaps in the research regarding any differences in the leadership of appointed and elected superintendents. This topic merits reflection as well as empirical examination since only a few studies have analyzed the effects of the chief executive (Bridges, 1982). The literature is also lacking on educational qualifications of school district superintendents (Schuh &
13 Herrington, 1990). The superintendency encompasses responsibilities in instructional leadership, fiscal management, community relations, board relations, personnel management, and operations management. The role is one of both leadership and management within the district and the community (Glass & Franceschini, 2007, p. xiii). In the face of todays demands, further study of leadership of district school superintendents, how they are selected, and how they lead is warranted. Historical Roles for Superintendents The role of superinten dent has changed greatly throughout history. Superintendents play a unique and critical role being the connecting link between schools and communities represented by school boards (Glass & Franceschini, 2007, p. xiii). They have served as clerks for th e school board, scholarly educational leaders, business managers, educational statesmen, educational professionals, accountable public servants, and political strategists (Petersen & Barnett, 2003). The first superintendents had no special training for t he position and essentially performed the duties of a school inspector more clerical than administrative. The duties of the early superintendents were limited to serving as representatives of the school boards. They prepared annual reports, visited and su pervised schools, and had little influence on financial matters. As superintendents grew in experience, they began to acquire more duties and responsibilities. The business of administering school systems fell less and less to committees. The position ev olved into executive officer of the board with the additional responsibility of employing teachers, business management, finance and facilities (Konnert & Augenstein, 1995; Norton et al 1996). As states began requiring that teachers be certified, prog rams were also developed to provide additional training and course work for school administrators and
14 superintendents. Following the Great Depression administrative emphasis began to change from the Taylorist scientific management to a more democratic, par ticipatory style of leadership. With the teachings of John Dewey encouraging teachers to become involved in decisions related to the goals and operation of their schools, superintendents were urged to become human relations experts (Carter & Cunningham, 1 997; Chapman, 1997; Konnert & Augenstein, 1995; Norton et al 1996). Superintendents then became interested in behavioral theory in anticipation that reactions to change in district procedures and policies could better be predicted. The emergence of behavior science theory also gave credibility to the science of administration (Chapman, 1997; Norton et al 1996). The superintendency encountered many transformations in the realm of education, not the least of which was desegregation. Feder a l and state governments became more involved in the authority of school systems, and teachers unions grew in numbers and influence. More conflict in education arose during the 1980s and 1990s. Superintendents could no longer deny the political nature of the posi tion. School reform became a serious entity to be addressed by superintendents along with school based management, schools of choice, and teacher empowerment. Houston (2001) proposed that for superintendents to be successful, the traditional approach need s to be revised. He projected that, rather than the management issues of the Killer Bs (buildings, buses, books, budgets, and bonds) (p. 428), superintendents will have to deal with the leadership issues of the Crucial Cs (connection, communication, collaboration, community building, child advocacy, and curricular choices) (p. 428). A more recent transformation has occurred as a result of school accountability. Mandates for
15 increased student performance began to come from all directions with littl e or no increase in funding, prompting accountability and school finance to become the greatest challenges facing school districts and superintendents (Carter & Cunningham, 1997; Chapman, 1997; Konnert & Augenstein, 1995; Norton et al 1996). Hoyle (19 99) stated that the responsibilities and stressors for school superintendents intensify each year creating the need to visualize the solutions to problems through nontraditional lenses. In the January 2003 newsletter, the American Educational Research Ass ociation (AERA) proposed that educational leaders have the responsibility to lead their schools through the tests presented by an environment which grows increasingly problematical. Indeed, the position of superintendent has evolved into a much more comp lex and demanding role. Many have recognized the difficulties leaders face in approaching new problems with old ideas relying on old thinking rather than using imagination and metaphor. A new approach which fosters critical thinking and promotes understanding of multiple meanings of situations must be learned to enable leaders and managers to face and control new situations effectively (Carlson, 1996; Morgan, 1997). Hersey and Blanchard (1977) stated that, in order to be effective, a leader must be able to analyze his environment and adapt his leadership style to fit the demands of that environment. Leaders are forced to simplify their situations by reframing what they see based on their own experiences. They must be able to create new viewpoints in o rder to develop new ways of addressing the concerns and issues that face most educational organizations. In order to accomplish this feat, leaders must filter their experience through their own frames or lenses developed through both experience and
16 educat ion. Unfortunately this may cause some situations to be ignored or be imperceptible as they are out of the leaders frame of cognizance. Leaders have not been successful when their perception of a situation is restricted. A misinterpretation of situations will lead to inappropriate action, just as medical misdiagnosis leads to mistreatment. Leaders must be able to envision organizations in flexible terms and view them from various perspectives in order to contend with the complex matters they will encoun ter (Bolman & Deal, 1997; Morgan, 1997; Norton et al 1996). Phil Schlecty looks upon the role of superintendent as that of CEO (Brandt, 1993). He stated that superintendents cause decisions, not make them. They are the ones that determine which problem s are worth solving and then establish an environment in which those problems are solved. Superintendents must not only be able to monitor and evaluate student performance, but also to explain how well their students compare with students in their state and the nation. They can no longer proclaim that their districts are doing well. They must be able to communicate this message through the interpretation of student performance scores on state and national standardized tests Political and societal demands for success on high stakes test are primarily the result of mandates for test based accountability and standardized curriculum created by the No Child Left Behind Act (NCLB). NCLB was enacted to ensure that all children would learn and contribute to th e economic vitality of the nation and preserve democratic traditions of the republic (Hoyle, Bjork, Collier & Glass, 2005, p. 42). The superintendent must go beyond the emphasis on test accountability to support their teachers and students by creating
17 learning environments that are comprehensive and inclusive. Superintendents must promote respect for all in an increasingly multicultural society. The need for strategic planning is also driven by legislative accountability systems. Strategic planning must be expanded from merely addressing curriculum alignment to including finance, facilities, staff development, adequate meals, safe transportation, disabled and bilingual students, athletic and music programs, as well as the physical and mental well -being of the students. All require adequate planning for effective implementation (Hoyle et al, 2005). Statement of the Problem Although the leadership orientations of school principals have been studied at all grade levels, few studies of the leadership orientat ions of district school superintendents have been conducted, in particular those directed at the method of selection of district school superintendents. There is very little research comparing leadership orientations of appointed and elected superintenden ts (Sumner, 1986). One reason for this dearth of research may be the fact that in 1990 there were six remaining states with elected and appointed superintendents (Schuh & Herrington, 1990). By 2006 that number had further declined to only three states: Alabama, Florida, and Mississippi (SREB, 2000). With an impending shortage of superintendents due to the graying of the superintendency, more qualified individuals will be needed to fill this critical position. In order to select (by whatever means) th e best individuals for the job, we must know more about the individuals occupying the office of the superintendent. It is imperative to understand their leadership styles as they relate to decisionmaking and effectiveness.
18 Purpose of the Study The purpos e of this study was to examine the leadership orientation preferences (as developed by Bolman & Deal 1984, 1991c, 1997, 2008) of elected and appointed school district superintendents in Alabama, Florida, and Mississippi. Demographic characteristics of the se superintendents, including age, gender, ethnicity, education, method of selection, years of experience in education, years of experience as a superintendent, length of tenure in current district, and size of district (number of students) were also analy zed. The Leadership Orientation (Self) Survey (LO(S)S) (Bolman & Deal, 1990) was employed to determine leadership orientation frame preferences of the superintendents inc luded in the study. Research Questions This study examined the following research questions. What are the self -reported leadership orientation frame preferences of the elected and appointed school district superintendents in Alabama, Florida, and M ississippi? Is there a difference between self -reported leadership style preferences of elect ed and appointed school district superintendents? Is there a difference in single or multi -frame pref erences between elected and appointed school district superintendents? Is there a difference in leadership style preferences of elected and appointed school district superintendents when considering selected demographic variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current dist rict, and size of district )? Delimitations T he following delimitations were observed while completing this study. The study was delimited to data gathered from selected school district s uperintendents in Alabama, Florida, and Mississippi who returned a co mpleted survey within the specified time period.
19 The study was delimited to responses obtained during the 2006-2007 school year. Data collection was delimited to responses to the LO(S)S and the characteristics demographic addendum. This study made no effor t to measure effectiveness of leadership perspectives Limitations I n this study, the following limitations were recognized. It was assumed that the superintendents included in the study and the researcher had a mutual comprehension of the terminology in B olman and Deals (1990) LO(S)S. It was assumed that the school district superintendent respondents accurately indicated their perceptions of their use of leadership orientation frames on the LO(S)S. Definition of Terms Frame is a tool or lens that brings a situation into focus providing individuals with a particular perspective through which to view a situation. The Bolman and Deal (1984, 1991c, 1997, 2008) frames are structural, human resource, political and symbolic. Multi-frame thinking is the ability to view situations within an organization through various perspectives that identify with combinations of three or four of Bolman and Deals (1984, 1991c, 1997, 2008) leadership orientation frames. The human resource frame emphasizes the needs of individ uals within an organization and adapts the needs of the organization to fit the people so that they might experience improved self esteem when performing their responsibilities (Bolman & Deal, 1984, 1991c, 1997, 2008). The political frame views organizations as groups of different interests vying for power and scant resources; conflict is the central theme of this perspective (Bolman & Deal, 1984, 1991c, 1997, 2008).
20 The structural frame based on a division of labor and the creation of policies, rules and procedures, is a more traditional approach rooted in the factory metaphor (Bolman & Deal, 1984, 1991c, 1997, 2008). The symbolic frame emphasizes the culture of organizations and is concerned with rebuilding the expressive or spiritual aspects of the organ ization through the use of stories, myths, metaphors, heroes, ceremonies and rituals (Bolman & Deal, 1984, 1991c, 1997, 2008). Leadership styles are behaviors that are representative of managers and leaders of organizations, and that have a powerful effect on morale and productivity (Bolman & Deal, 1997). Leadership behaviors refer to specific characteristics demonstrated by managers and leaders of organizations (Bolman & Deal, 1997). Justification for the Study The role of district school superintendent ha s become increasingly complicated, especially in todays modern technological society. The rapid increase in both number and diversity of students in our nations largest urban areas demands new skills of teachers and administrators (Houston, 2000, p.1) The widespread use of the internet complicates matters even more, creating a digital divide between mainstream society and the poor, in addition to our nationwide commitment to accountability and maintenance of high standards (Houston, 2000). As scho ol districts experience the retirements of the baby boomers, more qualified superintendents will be needed. Frames are mental models which enable leaders to comprehend and negotiate a particular territory. A good frame makes it easier to know what you are up against and, ultimately what you can do about it (Bolman & Deal, 2008, p.11) These mental
21 maps in effective serve as a global positioning system that leaders carry in their heads to enable them to navigate the twists and turns of leadership in org anizations (Bolman & Deal, 2008). Our basic premise is that a primary cause of managerial failure is faulty thinking rooted in inadequate ideas. Managers too often rely on constricted models that capture only part of organizational life. Learning multiple perspectives, or frames is a defense against thrashing around without a clue about what you are doing or why (Bolman & Deal, 2008, p. 21). Anna -Marie Cote (1999) revealed the importance of multi -frame thinking in school principals for enhancin g problem -solving skills and the ability to develop creative solutions. Leaders fail when they take too narrow a view. Unless they can think flexibly and see organizations from multiple angles, they will be unable to deal with the full range of issues t hey inevitably encounter (Bolman & Deal, 2008, p. 437) It is necessary for us to know and understand the decision making processes that superintendents employ. Due to the pressures on district school superintendents and the complexities of their role, mu lti-frame thinking is critical. Awareness of the frame preferences of school district superintendents both appointed and elected and how they view their leadership styles would expand the research knowledge and understanding of the use superintendents mak e of multi -frame thinking and how this relates to effectiveness. It is tempting to track familiar paths in a shifting terrain and to summon timeworn solutions, even when problems have changed (Bolman & Deal, 2008, p. 435) The issue of whether to elect or appoint school district superintendents is an important one. Do the qualifications and size of the candidate pool outweigh the issue of local control? The leadership of schools is critical to the development of American
22 society. The selection proces s should lead to the very best leaders available. The leadership orientations of superintendents have a tremendous impact on the school systems that they lead. The examination of leadership styles of appointed and elected superintendents would increase kn owledge as to the importance of the method of selecting school district superintendents. If there is a difference in leadership frame orientation based on method, then the method should be examined in order to ensure that the appropriate individual is sel ected for the position. The leadership frame orientation should fit the needs of the school district. Because of the essential role of the school district superintendent, there is a need for research to determine if appointed school superintendents lead ership orientations, and thus their behavior, differ from those of elected superintendents. Organization of the Study Chapter 1 provided background information, a statement of the problem, purpose of the study, research questions, delimitations and limitat ions, definition of terms and justification for the study. Chapter 2 contains a review of related literature and research. Chapter 3 presents the population, the data collection method, and the methodology used to respond to the research questions. Chapt er 4 contains a presentation and analysis of the data collected. Chapter 5 provides a summary of the study, conclusions drawn from the data analysis, major findings of the study, and recommendations for further research.
23 CHAPTER 2 REVIEW OF RELATED LI TER ATURE The history of the school superintendent has been a fitful journey from manager to leader. As conditions have changed over the last 200 years, the role has evolved from an ad hoc response to local needs for school management to leading a complex comm unity learning enterprise. Superintendents typically lead one of the largest institutions in the community and they have some of the greatest responsibilities in town, yet little is known about them. It is a position that is widely influential but narrow ly understood. (Houston, 2006) This study investigated the leadership frame orientations of school district superintendents in the states of Alabama, Florida, and Mississippi. The study also focused on whether the leadership frame orientation of the sup erintendents who were elected differed from those who were appointed to their positions. Was there a difference in their use of single, paired, and multi -frame use of leadership orientation frames? Demographic characteristics associated with these super intendents, including age, gender, race, education, method of selection, years of experience in education, years of experience as a superintendent, size of district (number of students), and length of tenure in current district were also analyzed to determ ine their influence on leadership frame choices of the superintendents. This review of the literature is organized into four sections. The first section addresses the alignment of organizational leadership theory and leadership orientation frame developm ent. Section two examines literature related to superintendent leadership (including appointed and elected superintendents) as well as gender and the superintendency. Section three examines studies employing the leadership orientation frame theory of Bol man and Deal. The final section is devoted to studies that relate to the specific demographic characteristics addressed in this study.
24 Organizational Leadership Theory Leaders and leader/managers are distinguished from ord inary managers by six aspects: t hinking in the long term; understanding the relationship of their organization to the larger organization; reaching constituents beyond the organization; emphasizing vision, values, motivation, and interaction; employing political skills to cope with confl ict; and emphasizing renewal and revision to meet requirements of a changing reality (Gardner, 2000). There are many kinds of leaders and styles of leadership. Diverse settings clearly affect leadership, but the most critical influence is the nature of t he group being led. In exploring the elements of a classical view of leadership, four characteristics of leaders using the classical view were identified: (a) identified by position as part of the hierarchy ; (b) focus on solutions and answers; (c) have vi sion and give direction; and (d) exhibit special characteristics that set them apart from their followers (Doyle & Smith, 2001). Organizations require five critical disciplines: (a) systems thinking which is a framework, a body of knowledge, and tools th at have been developed over time to make patterns clear in order to be able to change them effectively; (b) personal mastery which requires a commitment to lifelong learning about oneself and entails continuously clarifying and developing personal vision, energy, patience, and objectivity; (c) mental models which influence how one sees the world; (d) building shared vision rather than dictating a personal vision; and (e) team learning which emphasizes dialogue (Senge, 2000). Senge believes that a learning organization is where people learn how they create their own problems by their actions, and how they can change. Norton et al (1996) discussed Senges mental models and the importance of checking current ones
25 in order to develop new ones by which to func tion. They recognize that superintendents must examine their deep-rooted assumptions in order to be able to revise them as necessary. New ways of thinking must be created to meet new challenges. In order to be effective a leader must be able to analyze his environment and adapt his leadership to fit the demands of that environment (Hersey & Blanchard, 1977). Leaders are forced to simplify their situations by reframing what they see based on their own experiences. They must be able to create new viewpoi nts in order to develop new ways of addressing the concerns and issues that face most educational organizations. In order to accomplish this feat, leaders must filter their experience through their own frames or lenses developed through both experience an d education. Leaders must be able to envision organizations in flexible terms and view them from various perspectives in order to contend with the complex matters they will encounter (Bolman & Deal, 1997; Morgan, 1997; Norton et al, 1996). Leaders have difficulty when they approach new problems with old ideas (Carlson, 1996). The four frames of Bolman and Deal (1984, 1991c, 1997, 2008) provide leaders/managers with new and different views for approaching situations within organizations, thus increasing t heir problem -solving repertoire. Carlson noted concerns regarding the limitations of reliance on old thinking rather than using imagination and metaphor. Again, this new approach fosters critical thinking and promotes understanding of multiple meanings o f situations enabling leaders/managers to face and control new situations effectively. Thinking in a new way in effect jeopardizes a leaders security. Individuals grow accustomed to leaders behaving a certain way and come to expect the same patterns of
26 behavior. Although some would prefer to try alternate views, it would seem that intrapersonal and interpersonal forces prefer the status quo. Leader/managers that are able to create new viewpoints are capable of developing new ways of addressing the unen ding concerns and issues which face most organizations (Carlson, 1996) Administrators who try approach the problems in educational organizations base their methodology on how they view their options. Similar to Bolman and Deals frames, Owens (1991) exp lains that individuals are limited in their ability to comprehend problems and attack them by the number and variety of ways with which they are familiar. When Bolman and Deal attempted to teach an organizational leadership course together, they found tha t their different perspectives influenced their ability to work together and to further student understand. As a result, they developed a framework that aided in enhancing their own understanding by looking at issues through different lenses (Englert, 200 8). Bolman and Deal (1984, 1991b, 1991c, 1997, 2008) describe four frames: Structural, Human Resources, Political, and Symbolic in order to enhance mutual understanding for leaders and followers (Bolman & Deal, 1997). Each of the frames has a distinct p urpose and function in creating a clearer image of understanding. The Structural frame emphasizes details and goals. Individuals are categorized into roles while the various activities are managed by rules, policies, and chain of command. The leaders of these organizations value data and analysis of data. Their concern is for the bottom line. The Human Resource frame focuses on the needs of people rather than the needs of the organization. When the leader seeks alignment between the people
27 and the org anization, the organization is adjusted to meet the requirements of the people. Leadership is achieved through empowerment and facilitation. These leaders value feelings and relationships. Using the Political frame, leaders view organizations as arenas of continuous conflict and competition for scarce resources (Bolman and Deal, 1991b, p. 515). Leaders serve as advocates and intermediaries while valuing realism and common sense. Political leaders spend their time forming coalitions, networking, bui lding power bases, and negotiating compromises. The Symbolic frame is used to give meaning and predictability to a disordered world. Facts are interpretive rather than objective. The organization creates symbols to provide a shared sense of community an d human behavior. The leaders job is to inspire commitment and enthusiasm using rituals and drama. Charisma is a critical characteristic of the Symbolic leader. Leaders view experience through a set of preconditioned lenses and filters (Bo lman and Deal, 1991b, p. 510). They resist questioning their view of the manner in which the organization functions. Accordingly, if the frames fit, the leaders understand and can shape the human experience; if not, the frames freeze into distorted pictures trapping them in misconceptions. Circumstance is then blamed for failure. Just as a medical misdiagnosis rarely results in a cure, so misinterpretation of situations can weaken even an exceptional leader. The question remains as to whether or not patterns appear in the frames or lenses leaders utilize. If these frames represent organizing theories, then one must suppose that individuals view situations in different ways because of differing values and cultures. Frames of reference shape how individuals view si tuations and also establish the actions that they choose to take.
28 Leading with Soul an allegory written by Bolman and Deal (1994), draws from the Symbolic frame. Its purpose is to show that leaders actions serve to create meaning in the organization. L eaders must realize the importance of developing a philosophical framework by which they will act. Educational leaders need assistance in integrating the logical and artistic aspects of their position. Leaders must be bifocal in nature, having the abilit y to manage the technical while still maintaining the importance of the symbolic side of leadership (Deal & Peterson, 1994). Leadership is often mistakenly considered rational when actually it is basically political. Achieving power in order to get things accomplished is critical and makes conflict inevitable. The symbolic role including cultures, values and vision, is equally critical to leadership. Having a vision may not be as important as engaging people in a process of visioning (Bolman & Deal, 19 94b, p. 84). Leadership is primarily developed through experience. The authors believe that leaders should be exposed to difficult challenges early in their experience to enhance their learning. Another aspect critical to leadership is self -reflection including feedback from others. Leaders require their own personal frameworks in order to be able to interpret feedback. Leaders are being provided with adequate training in the areas of management; however, they are not often mastering the symbolic side of organizations. Leadership must be redefined in human, moral, and spiritual terms (p. 95). Bensimon (198 7 ) considered Bolman and Deals (1984) frame theory to be one of the most useful organizational theories from a leadership perspective. The idea of frames is useful as it suggests that different perspectives of followers and leaders will result in different interpretations of leadership. As perceptions change so the
29 expectations of leadership change. As leaders develop cognitive complexity, they will be more able to contend with the uncertain and volatile environments of organizations. In a leadership case study, Parry and Horton (1998) utilized the Bolman and Deal frames model and determined that leadership is crisis or change driven, but it a ddresses change through cooperation. Leadership is based on power through communication and implementation of ideas. While using vision and the existing culture, leadership is symbolic but still recognizes politics. It is learned and based on intangibl es (Parry & Horton, 1998). A leadership framework for educational change based on Bolman and Deals model as well as those of others was developed by Creasey (2002). Relationshipbuilding for trusting, caring relationships, passion and communication were all included in her model. Each of these components is also loosely based on the Bolman and Deal Human Resources frame. Superintendent Leadership Major gaps exist in the research regarding the superintendency (Bridges, 1982). Considering the importance of the position of the superintendent to education and society, only a small number of studies have analyzed the effect of the leadership of the chief executive. Bridges suggested that this topic merits reflection as well as empirical examination since nothing of consequence is known about the impact of the occupants of this role (p. 26). Further, Bridges believes that formidable theoretical, procedural, and political obstacles stand in the way of such a study.
30 Role of the Superintendent To outsiders, the role of the school superintendent has always been a little mystifying. Most people can explain that the superintendent is the ultimate person in charge, but what superintendents actually do remains vague (Lashway, 2002). Petersen and Barnett (2003) aggregated the development of the position of superintendent into seven eras: (1) 1820-1850 clerks for the school board; (2) 1850early 1900s scholarly educational leaders; (3) early 1900s -1930 business managers; (4) 1930mid 1950s educational s tatesmen; (5) mid 1950s -1970 educational professionals; (6) 1970-1980 accountable public servant; (7) 1980-1990 political strategist. Now it appears that district superintendents are required to provide powerful, authentic and rigorous learning f or all students (p. 10). Holloway (2001) developed nine job domain categories important to beginning superintendents identifying the most important as comprehension and reaction to political issues. The least important was developing school board relatio ns. Facilitating student learning was one of the least important, while providing professional development and maintaining group processes were considered of high value. Holloway also listed standards for effective leadership which reflect best practices in the role of superintendent. These included aiding in the creation of a shared learning vision; maintaining an instructional program that is conducive to student learning; providing a safe learning environment; collaborating with members of the communi ty; displaying ethics, fairness, and integrity; and finally, being knowledgeable of the political, social, economic, legal and cultural composition of the school district. Cuban (1998) describes three roles of superintendents: instructional, managerial, and political. They must improve student achievement, operate their
31 districts efficiently, and negotiate with numerous stakeholders for approval of programs and resources. In order for superintendents of the future to be successful, they must learn a comp letely new approach to their jobs. Excelling at managerial issues will no longer provide success. Superintendents will have to master the processes that support the work, i.e., connection, communication, collaboration, community building, child advocacy and curricular choices (Houston, 2001, p. 430). Houston believes that leadership of the future will necessarily focus on relationships between children, adults, school, and community. Communication skills will be critical for success. Superintendents will no longer need command skills as much as collaboration skills. The quality of the individual who is selected as superintendent is primarily responsible for the scope and quality of the educational program that will be developed in a school district (American Association of School Administrators, 1962). Schlecty envisioned the superintendent as CEO who is called on not to solve problems, but to decide which problems are worth solving, and to create the environment in which the problems get solved. H e sees the superintendent as a decision causer rather than a decision maker (Brandt, 1993). An additional role for the school district superintendent is that of change agent in school reform. Superintendents are critical to school reform as they make the majority of reform proposals in their districts (Bjork & Rodgers, 1999). Superintendents are key in creating organizational changes, thus causing a transformation in the ways that problems are viewed and solved (Ireh & Bailey, 1999). The major force pus hing superintendents to reframe problems and solutions to various issues is the ever -
32 changing world. The pressures and responsibilities of superintendents increase yearly because of the speed of information processing (Hoyle, 1999). The superintendent i s somewhat of a generalist The superintendent cannot be an expert in one area and ignore others (Sharp & Walter, 2004, p.15). The importance of the superintendent is to understand the big picture. Characteristics of Superintendents Superintendents from the largest school districts have the shortest tenure, 26 to 28 months (Snider, 2006). Regardless of their accomplishments in the school districts, superintendents have become political scapegoats. They are one of the few remaining district employees with little or no job security, making it difficult to properly implement effect ive reforms. When other school district employees know that they will more than likely outlast the tenure of the superintendent, it becomes more difficult for superintenden ts to be effective leaders. In a study of 50 superintendents, Willower and Fraser (1980) found that superintendents deal with a wide range of problems, are irked by the paperwork demands of state and federal agencies, regret not being closer to the clas sroom, and feel the pressure of the job but are ready to do it over again (p. 5). The superintendency is the most male dominated executive position of any profession in the United States (Table 2 -1) Bjork and Rodgers (1999) found that 96% of the superi ntendents in the United States are male. Females in the superintendency appear to have greater knowledge of instructional methods, provide more assistance to beginning teachers, supervise teachers directly, and provide school environments which are more s uited to learning. They also tend to employ more democratic methods that encourage participation thus achieving higher levels of participation and job satisfaction.
33 Table 2 1. Superintendent d emographics from AASA s uperintendent s tudies 2000 and 2006 Year Age % Female % Minority % Doctorate Tenure (yrs) District Size 2000 52.5 14 5.1 45 6.47 4,026 2007 54.5 21.7 6.2 51 5.5 2 ,750 Appointed or Elected Superintendents Using the Leadership Behavior Description Questionnaire developed by Halpin, Sumner (1986) studied how principals perceived the leadership behavior of elected and appointed school superintendents in Mississippi. He found very little literature comparing the leadership behaviors of appointed and elected superintendents The results of his study concluded that appointed superintendents held more advanced degrees and were rated higher in Initiation of Structure. Initiation of Structure is defined as leadership behavior endeavoring to establish well defined patterns of organization, channels of communication, and methods of procedure (p. 5). There was not enough diversity in the sample of superintendents to draw conclusions regarding gender or race. Sumner recommended that further studies using other instruments were needed to explore any leadership differences in elected and appointed superintendents. He also recommended further studies to investigate any factors that may have influenced superintendent behaviors. Further, Sumner suggested that state leaders review t his study and reconsider Mississippis method of selection for superintendents. In examining both elected and appointed school district superintendents, the Southern Regional Education Board (SREB, 2000) found that the number of elected superintendents is diminishing. The number has declined by one -half in the last decade. The practice of electing superintendents is now limited to three states: Alabama, Florida, and Georgia. There was still one district in South Carolina with an
34 elected superintendent, but that district also had an appointed superintendent and was in the process of eliminating the elected superintendent position. In 1992 the manner of selection for Georgias superintendents was changed to appointed, with the state of Tennessee making t he same change in 1999. Those who argue in favor of electing superintendents believe that the process is more democratic resulting in a superintendent who better represents the communitys and districts needs. Electing superintendents also creates a bal ance of power with the school board. Opponents of elected superintendents claim that an appointed system creates a clear line of accountability with the superintendent being accountable to the board. An appointive system also increased the pool of candid ates for the position. Superintendents can then spend their time running the schools and not running for office (SREB, 2000). In an attempt to determine why superintendents had been successful in maintaining their positions over an extended period of tim e (20 plus years), Waller and Richardson (1997) interviewed 13 long-term Georgia superintendents (3 appointed, 10 elected). Superintendents were also questioned regarding their perceptions of educational reform. All were from small towns, were middleaged, and held varying degrees. There was only one minority represented in the group. These superintendents expressed that an increase in political awareness aided in their ability to hold their positions for 20 or more years. In regard to school reform, t he superintendents expressed their concerns regarding the change from elected to appointed superintendents which they believed would remove the control of education from the political arena. They observed that the superintendency is a political office and
35 expressed the concern that elected school boards could be even more political. The superintendents predicted that the educational system would return to the previous lack of stability and continuity. McGriff, Bishop, and Rice (1997) examined how teacher leaders and school board chairmen perceive the work -related behaviors of local s c hool district superintendents in Alabama. The study identified 23 work -related behaviors that were frequently demonstrated by superintendents and examined these behaviors to determine whether they were demonstrated equally by elected and appointed superintendents. They found that there was a difference in the effectiveness of superintendents based on whether they were appointed or elected. The appointed superintendents demonstrated these work -related behaviors more often than elected superintendents, while the elected superintendents completed specific improvement actions more often. The process of electing superintendents gives control of schools to the public thus creating a superintendent who is more responsive to the needs of his/her constituents. It is less likely to have high turnover in the superintendency or abuse of power. Appointed systems, however, create specific personnel selection qualifications and a more eff icient process of decision making. This method does away with the political process and establishes a professional one. There is access to a greater number of candidates that are highly qualified. Appointed superintendents are more likely to hold advanc ed degrees. The authors recommended that the state of Alabama re evaluate continuing their system of electing district school superintendents. When Miller Whitehead (2000) studied school districts in Alabama that had received a grade of A from the state b oard of education, she was searching for a
36 common denominator for success. In Alabama the city school systems appoint their superintendents while county system superintendents are elected. Miller Whitehead found the superintendents of these successful sc hool districts were all appointed. While examining student performance in Oklahoma schools, it was found that elected superintendents were less likely to publicize poor academic performances by their schools for fear of losing votes. In 1989 the appointed superintendent printed 2500 copies of state test results so that as many individuals as possible could see them and compare their district and school results with others. In 1991, 1992, and 1993 the elected state superintendent had only 300 copies of th e state test results printed, which was not even enough copies for every school district superintendent to examine the results (Wood, 2004). The Georgia Governors Review Commission (1994) prepared a report defining quality basic education in the state of Georgia and made several recommendations for achieving this for all students. One of the recommendations was to change the system to all appointed school superintendents rather than elected. Their argument in favor of appointed superintendents was that appointed superintendents tend to be more responsive to the school board. They demonstrate greater support for taxing for education and spend more of their funds on items directly related to instruction. Appointed superintendents produce higher achievement scores. The qualifications for appointed superintendents would be more professional and not related to fund-raising abilities for a political campaign. Appointed superintendents are not as susceptible to political pressures.
37 A constitutional provision in Georgia requires that all school district school superintendents be appointed by the school board. Part of the justification for this change was to remove politics from the position. There is some question, however, whether this provision has served to remove politics or has simply changed them (Lindsay, 1996). Schuh and Herrington (1990) found that 97.8% of school district superintendents were appointed with only six states (Alabama, Florida, Georgia, Mississippi, South Carolina, and Tennessee) hav ing elected and appointed superintendents. Initiatives in each of these states failed to change the process to appointed superintendents. They suggest that the issue is basically local control versus professionalism. Although having schools governed by the public is a valid concern, the complex issues facing school districts today require a professional who holds greater authority and influence in policy making and implementation. Further, regardless of the method of selection, school boards and superin tendents are all involved in political activity. Although the literature is lacking on educational qualifications of school district superintendents, it appears that appointed superintendents tend to have more and higher degrees than those who are elected. The three states that are still electing superintendents are all southern and primarily rural. Legislatures in all of these states have considered changing their systems to all appointed superintendents, but the importance of local control (also called democratization) has prevailed. Alabama and Mississippi would only require some statutory revision to change, but for Florida to make the change to all appointed superintendents would require a constitutional amendment approved through a general statewid e election (Schuh & Herrington, 1990).
38 In Alabama the state superintendent is appointed and so are the city superintendents. County superintendents, however, are mostly elected. Legislation in Alabama has been filed, but has failed since 1984 for lack o f support from educational organizations. A significant factor in the failure to change to an appointed system is the fact that some of the local school boards are appointed. Voters would lose any control of the school system (Schuh & Herrington, 1990). In Florida after 1885, it was mandated that all district school superintendents be elected because of issues perpetrated by carpetbaggers during Reconstruction. Later (1955 and 1962) constitutional provisions were adopted to allow some large counties to vote for appointing or electing local school district superintendents. A special act by the legislature could also create this change in designated counties. Florida Statutes (Section 230.241 FS) require that the school boards of these counties adopt and present a formal resolution to the county commission in order to place the resolution on the general election ballot. Twenty -two (22) of the 67 counties in Florida now have appointed superintendents (Schuh & Herrington, 1990). Although the state of Geor gia recently changed to a system of appointed superintendents, this was achieved only after many failed efforts. Again, the primary reason for the prior repeated failure of this legislation was local control (Schuh & Herrington, 1990). Schuh and Herringt on (1990) further reported that although Mississippi changed the state superintendent of education from elected to appointed in 1982, many county superintendents are still elected. Superintendents in municipal, agricultural, and consolidated districts hav e appointed superintendents. In 1988 the Mississippi
39 legislature passed a bill which allowed school districts to decide through a vote to change the superintendent position from elected to appointed. No county took advantage of this opportunity even though the NAACP filed lawsuits claiming that the system of electing superintendents is discriminatory. There is, however, a requirement in Mississippi for all local school superintendents (elected and appointed) to have the equivalent of a masters degree. The constitution in South Carolina allows for both appointed and elected superintendents. Schuh and Herrington (1990) found that recent legislation was enacted requiring all superintendents to be appointed by the local board. The state of Tennessee had a system of appointed and elected superintendents with the majority of them elected. Districts were allowed to change the system through local referendum or private act. After many failed attempts, the Tennessee legislature finally passed a proposal for all appointed superintendents. Schuh and Herrington (1990) found that the major stumbling block to early attempts at change was again the issue of local control. Schuh and Herrington (1990) also discussed the pros and cons of appointing versus electing lo cal school district school superintendents. Those who favor electing superintendents reason that the people should have control of the schools. They believe that elected superintendents in essence are closer to the people and thus tend to be more respons ive to their needs and desires. This is more important to voters than the possibility of greater efficiency. The elected superintendent is not owned by the board and therefore is more independent. Appointed superintendents are often viewed as puppets of the board, reducing the power of the voters in decisionmaking and
40 strengthening the autocratic rule of the school board. The turnover rate of appointed superintendents is greater and there is a greater possibility of abuse of power. One of the strengths of an appointed system is that a qualified individual can be selected, as qualifications for the position can be established along with the salary; unqualified candidates can be eliminated, creating a smaller, quality group to choose from and making the selection process quicker and easier. The pool of applicants is generated from a larger geographic area. Schuh and Herrington (1990) also point out that in business the board of directors selects the CEO. Appointed superintendents can focus their time and efforts on their educational jobs and not on raising money and running for office. The position becomes professional rather than political. Also, when the school board can be held responsible, schools tend to be closer to the needs and desires of th e electorate. Elected superintendents and elected boards often have allegiances to different individuals creating conflict and confusion. When the superintendent is held responsible to the board, there is less opportunity for an abuse of power. Groups wh o are in favor of a system of appointed superintendents are more concerned with the qualifications of elected superintendents. They assume that appointed superintendents are more highly educated, better qualified, and more stable. In a focus report for t he Southern Regional Education Board (SREB, 2000) the issue of elected and appointed superintendents was investigated. It was found that those in favor of electing superintendents believe this to be more democratic and that elected superintendents will be more responsible to the people and the needs of the district. They also believe that elected superintendents are freer to disagree with the school boards. Proponents of appointed superintendents assert that this system
41 creates a more clear line of accou ntability and increases the pool of candidates providing a better opportunity to find a qualified individual who will meet the needs of the district. They say that appointed superintendents spend their time running schools, not running for office (p. 1) The prediction of this report was that by 2000 there would only be 154 elected superintendents left in the United States. The American Association of School Administrators (1962) proposed that elected superintendents should meet the same qualifications as those who are appointed. The operation and maintenance of a successful school system requires a leader who is knowledgeable and professional. Without qualifications, there are no safeguards against mismanagement; however, smaller counties often have considerable difficulty in finding qualified educational leaders who are willing to run for office. Situations where there are elected superintendents and elected school boards create a system of dual accountability which creates a dichotomy of authority and responsibility. Local politics can have a negative impact on schools, even though elected superintendents do have strong ties to the community (AASA, 1962). The AASA (1962) also found that groups who oppose changing to an appointed system include ele cted superintendents, state legislators, and in some instances (e.g. Florida) school superintendent associations. The superintendents in Florida argue that the voters of the school districts are the most qualified to make decisions about how the school sy stem should be run. Since the school system in many districts is the largest employer, the voters want greater control of the decisionmaking process. The remaining states with elected superintendents are primarily rural, Reconstructionist states. Loc al control in the South is considered a basic democratic
42 right and continues to be an important issue in many communities (Schuh & Herrington, 1990, p. 32). Two national associations performed an evaluation of Indianas education system and, as a result, encouraged the state to abolish the process of electing the chief state school officer and implement the office of appointed state superintendent. Although this refers to the superintendency at the state level, the same objections held true. Elected sup erintendents not only have to raise money in order to run a campaign, they also must carry out a political agenda once in office. This detracts from the goal of improving education (Zehr, 2001). The executive director of the National Association of State Boards of Education (NASBE) stated that state superintendents should be hired and evaluated based on what they can do to improve education. The argument against this is again that individuals elected by the people have allegiances only to those people an d not just to one individual or administration Investigating whether the mode of selection (appointed or elected) relates to leadership, Carol Brand (1993) asked principals in Georgia to rate their superintendents. Brand discovered that the majority of f emale superintendents were elected. She found that appointed superintendents were older; more experienced in administration, and held higher degrees. Elected superintendents had been involved in education longer and also held their positions longer. Appointed superintendents appeared to be concerned with school board relations, while elected superintendents were focused on budget and finance. Most importantly, appointed superintendents were rated as superior overall by their colleagues. Brand found tha t appointed superintendents received higher salaries and more perks with their positions. Larger
43 school systems frequently have appointed superintendents with more professional preparation. The majority of female superintendents are elected. In November 1992 Georgians voted to change to a system of appointed superintendents. Brand recommended that this study be repeated after this change was fully implemented. Leadership Orientation Frame Studies Frame C hoices Bolman and Deal (1991a) explained that leaders view situations based on their own experience, so if their perspectives (or frames) match the situation, they can be successful. If their frames do not fit, the situation can become vague and misdiagnosed thus leading to action that is not effective. Bolman and Deal used both qualitative and quantitative methods in their data gathering. The Leadership Orientations (Self) Survey (LO(S)S), 1990) was used including both the self and colleague ratings. Their hypothesis was that the capacity to refram e is a critical issue in success as both manager and leader (p. 519). They discovered that the respondents saw linkage between frames in the items and saw them as separate and distinct as well. The results of this study showed that American leaders scor ed higher in both the structural and human resources frames while international leaders scored highest in structural and lowest in symbolic. Leaders from Singapore had results similar to the Americans except for scoring unusually high in the symbolic fram e. Another study by Bolman and Deal (1991b) was used to operationalize the Bolman and Deal model (1984, 1991b, 1991c) for understanding leadership and organizations through the four frames: Structural, Human Resources, Political, and Symbolic. This study used critical incidents to assess which and how many frames were used. One or two frames were frequently used most often Structural and least
44 often Symbolic. The use of Human resources and Political frames varied. The frames as implemented by managers w ere the focus of this particular report. Frames and Effectiveness The findings of Bolman and Deal (1991a) study showed that frame orientations are related to success as manager and leader. Further, they showed that the effectiveness of managers and leaders was not the same thing. While managerial success emphasized the Structural frame, this frame was found to be ineffective in leaders. Leadership effectiveness was related to use of the Symbolic and Political frames. Frame orientations that are succe ssful depend on the work environment. In the corporate environment, the Structural frame proved to be more effective than the Symbolic. Most managerial training programs were found to focus on management rather than leadership while leadership programs fo cus on Structural and Human Resources frames. Use of the Political frame appears to be a better predictor of management and leadership effectiveness than the use of the Human Resource frame, possibly because those who are skilled in the Political frame ar e perceived as better leaders and managers. In regards to gender, there were little or no differences in frame orientation preference; however, females tended to rate themselves lower in the Political frame than they were rated by their colleagues. While it appears that managers frequently employ only one or two frames, they need to rely on all four frames in order to be fully effective as both managers and leaders (Bolman & Deal, 1991a). Further, leadership frames reflect significant elements of the lead ership approach and have a significant association with leader effectiveness. The frames orientation of an individual can predict effectiveness of an individual as a manager or a leader,
45 although the frames preferences are different. In three different s amples of educational administrators, Bolman and Deal (1991a) revealed that more than two frames were rarely used and that all four frames were almost never used. Leadership is significantly associated with the Symbolic frame and almost never with the Str uctural frame. In determining the effectiveness of managers, the results were the opposite with the Structural frame being a positive predictor and the Symbolic frame never a predictor. Bolman and Deal (1991b) assessed frame preferences of managers and t hen had colleagues rate the managers effectiveness as manager and leader based on their frame orientations. The effectiveness of managers was found to be linked to the Structural frame, while the effectiveness of leaders related more closely to the Symbolic frame. Gender was not significantly related to orientation or effectiveness. Leadership and management are two very different entities (Bolman & Deal, 1992a). T his research was based on the hypothesis that leaders and managers must both be able to r eframe, and that the ability to use multiple frames has a positive effect on the success of managers and leaders. Secondly, they hypothesized that leadership is based on context with different situations requiring different thinking patterns. The results suggested that having the ability to employ multiple frames is critical to success as leader or manager. College P residents/ D eans Several studies have been performed to assess the leadership orientation frames preferences of college presidents (Bensimon, 1987; Echols -To be, 2000; Jordan, 1999; Tedesco, 2004). Bensimon explored the extent that college presidents employed single or multiple frame perspectives in evaluating good leadership based on the Bolman and Deal model. She concluded that of the 32 pres idents interviewed, 13 relied on a single
46 frame, 11 used two frames, seven employed three frames and only one perceived using all four. The Structural or Human Resources frames were most used by single-frame presidents. Dual frame presidents were likely t o use a combination of Human Resources and Symbolic frames, while the majority of the three frame presidents identified Human Resources Political Symbolic combination. The ability to use multiple frames positively correlates to the length and range of experience of the presidents. As a result of this study, Bensimon (1987) recommended that the colleges create leadership teams composed of individuals with complementary frame orientations in instances where the president relied on only one or two frames in order to create a balanced approach to leadership. Another study of college presidents studied 13 female presidents in research universities. Experience did not appear to correlate to a use of multiple frames; however, it was found that more experienced presidents were more likely to prefer the Human Resources or Symbolic frame (Welch, 2002). Echols -Tobe ( 2000 ) studied leadership development among African American female college and university presidents collecting data through the LO(S)S (Bolman and Deal, 1990) and through personal interviews. It was determined that the vast majority of presidents had multiple frame orientations. Jordan (1999) also used the LO(S)S to examine the leadership frames of presidents of southern womens colleges. The pres idents in this study were all found to use a combination of single, paired, or multiple frames; however, due to the small population (3), results were inconclusive. Tedesco (2004) investigated the leadership frames orientations of Iowa community college p residents and k -12 superintendents. The differences that he found between
47 the leadership orientations of these two groups existed only in the Symbolic frame which was more likely to be used by the community college presidents. Cantu (1997) investigated t he leadership orientation frame preferences of public college and university deans. The deans had either been nominated because of their effectiveness or randomly selected. The most frequently preferred leadership frame used by the deans was the Human Re sources frame with the Symbolic frame being the least preferred. The difference between the two groups of deans chosen was in the use of the Political frame by the nominated deans. Experience was found to have a significant impact on the leadership orientation of the deans studied. The nominated deans used the Political frame more often than the randomly selected deans. Englert (2008) investigated the leadership orientations of rural community college presidents serving appointed or elected independent g overning boards using the LO(S)S. Significant differences were found between the presidents and board chairs as well as between presidents serving appointed and elected boards. Board chairs reported that the presidents used the Structural and Political f rames more often than the presidents reported in their survey responses. The self -perception of the presidents was that the Human Resources and Symbolic frames were used more. It was also reported that presidents serving appointed boards employed the Hum an Resources frame more than presidents serving elected boards. Superintendents Strickland (1992) also used the LO(S)S (Bolman and Deal, 1990) to survey superintendents and their colleagues in Tennessee regarding superintendent leadership and management s tyle. Her results were supplied to the school systems human resources department to assist in understanding expectations related to the four frames.
48 She found that the Tennessee superintendents generally employed multiple frames with the Human Resources and Structural frames being the most preferred and the Symbolic frame the least. In her study of leadership orientations of selected female school superintendents, Flak (1998) also found that these women employed a multi -frame approach to leadership. Ha rlow (1994) showed different results in her study of the leadership orientation frames of Washington state public school superintendents. She found that the superintendents rarely used more than one or two frames. More experienced superintendents used th e Political frame most often in describing critical incidents but used the Human Resources frame when defining leadership. Less experienced superintendents described critical incidents using the Structural frame, but defined leadership using either the St ructural or Human Resources frames. Superintendents rated themselves highest in Human Resources followed by Political, Structural, and Symbolic. Harlow indicates that employing all four frames produces a more complete understanding of organizations as well as the leadership perspective necessary for success. She recommended that leaders need to develop more sophisticated modes of thinking and concluded that frame orientation was not a predictor of leadership effectiveness. These findings were in contras t to previous findings by Bolman and Deal. Using the Bolman and Deal LO(S)S, Faverty (1997) surveyed a random sample of 55 of 539 California small district superintendents and the entire population of 14 superintendents trained by the ACSA Small School Di strict Superintendents Academy. He found no significant differences in the leadership orientations of the superintendents
49 surveyed; however the Human Resources frame was shown to be the primary leadership orientation. In another study of leadership orien tations of superintendents, Moss (2002) examined district superintendents in a mid western state. He found that while the superintendents perceived themselves as predominantly using the symbolic frame, their school board presidents saw the superintendents as being much more technical (structural frame). Clisbee (2004) investigated the leadership styles of school superintendents specifically to determine if gender impacted leadership style. After surveying 100 superintendents in Massachusetts (76 male and 24 female), the author found that there were no significant difference s in leadership based on gender. An examination of the leadership orientation of Indiana public school superintendents, utilizing the LO(S)S, identified the Human Resources frame of lead ership as their dominant frame while the Political frame was the least preferred. The study included demographic variables of gender, years of experience as a superintendent, school district setting, and initial level of training which were shown to have no significant effect on leadership frame choices (Ward, 2006). Foreign C ultures Redman (1991) compared the leadership orientations of higher education administrators in five private Japanese institutions with the leadership orientations of higher educati on administrators in five private American institutions using the LO(S)S. She found that the frame rankings of both groups were identical; however, the means of the Americans surveyed were much higher. The differences between the two groups
50 may be due to cultural influences which could indicate that the two groups were more alike than different. Using the LO(S)S (1990) Bolman and Bloch (1999) surveyed 788 managers and interviewed 37 managers in Venezuela in order to determine the validity of the Bolman an d Deal leadership orientation frames in another culture. Their findings upheld the validity of the frames research. It was shown that Venezuelan managers tend to use one or two frames, preferably structural and human resources. There was some variabilit y in the patterns of frame use. Suzuki (1994) studied leadership orientation of Asian and other K -12 school principals using the LO(S)S. His research showed that females used the human resources frame more often than males. Further, foreignborn princip als preferred the structural frame. Again, this study revealed a strong relationship between the use of multiple frames and leader effectiveness. Bolman and Deals LO(S)S and other instruments were used in this comparison study by Childress (1994). Japanese and American leaders in U.S. based automotive manufacturing industry were studied to determine leadership orientation frames. American participants viewed structural, political and symbolic frames as indicators of management effectiveness and the pol itical frame as indicators of leadership effectiveness. They saw themselves as primarily employing the human resources frame, while others saw them as more political. The Japanese participants viewed the structural frame as a management indicator, but di d not equate any frame with an indicator of leadership. The findings do not support previous Bolman and Deal studies, but do corroborate cultural differences between American and Japanese participants.
51 McCartney -Infelise (1999) studied the leadership ori entations of school directors and compared them with those of small school district superintendents in California. This researcher surveyed directors of overseas American curriculum schools in the Mediterranean region of Europe as well as superintendents of small school districts in California. The findings showed that there was a preference for the structural leadership frame in female overseas directors and in the California superintendents. Both groups also scored highly in the human resources orientation. These findings were similar to those of Bolman and Bloch (1990) and Suzuki (1994). Hodge (2003) investigated the cognitive frame orientations of school leaders in Belize and demographic factors that influenced their leadership. She determined that the pattern of frame use was the same for all principals in Belize. Human Resource frame was the first choice followed by the Structural and Symbolic frames. The Political frame was the least used. Demographic variables did not show a significant effec t on leadership orientation frames choices. School A dministrators Burks (1992) applied organizational frame theory to determine the need for leadership development training in the Tennessee Board of Regents system by examining leadership orientations of e ffective school administrators. His results showed a need for leadership development in all frames. Further, he found no significant difference between genders. Durocher (1995) also examined leadership orientations of effective school administrators usi ng the LO(S)S. Durocher reported significant differences in the number of frames used by male and female administrators. Her findings showed the Human Resource frame as the dominant choice with Structural being the least chosen.
52 She also found that nearl y half of the administrators surveyed used three to four of the Bolman and Deal frames. This use of multiple frames was concluded to be partially responsible for the success of these administrators. A study of elementary principals by Johnson (1995) determined that these principals tend to view their organizations from more than one frame of reference. He also found the females in the study exhibited higher levels of complex thinking; however, there were no significant differences in the number or types of leadership frames in males and females. A similar study of elementary principals by Meade (1992) found that females exhibited a stronger tendency toward the Human Resource and Symbolic frames. Principals gender was shown to have a significant impact on both of these frames. Leadership orientation was shown to be a stable characteristic which was not just a function of leadership training. Cote (1999) also performed an investigation of the leadership orientation frames of a stratified random sample of elementary school principals in Florida utilizing the LO(S)S. Socio economic status and school enrollment were used for stratification purposes. Cote found a significant relationship between enrollment and frame categorization; gender and Human Resource frame use; gender and tenure. Similar to the findings of Bolman and Deal (1991a, 1991b, 1992a), she also found a significant relationship between managerial effectiveness and the use of the Structural and Political frames. Use of the Symbolic frame and S tructural frame were significantly related to leadership effectiveness. Martinez (1996) compared leadership orientations of elementary school principals who completed the California School Leadership Academy (CSLA) with California elementary school princip als who had not participated in the training. Not only did the
53 CSLA training not make a significant difference in the leadership orientations of the principals, but gender age, ethnicity, location, size of district, and other demographic factors did not h ave a significant impact on the leadership orientations of these principals. Miro (1993) studied the leadership orientation frames of high school principals. Defining frame as a particular vantage point one takes to view the world which impacts leadershi p and management, Miro found the dominant frame of the principals studied was Human Resources with Symbolic being the least. He also found no significant difference in gender. These finding concur with previous findings of Bolman and Deal (1991b). Thomps on (2000) explored the differences in gender between leadership orientation, leadership characteristics, and the perceived effectiveness of educational leaders by examining the responses of subordinates using Bolman and Deals frames theory (1984, 1994, 19 97) as well as Quinns competing values model. He found that males and females were seen to be equally effective and their leadership characteristics were similar. In her study, Rivers (1996) employed the LO(S)S to study all levels of principals. She f ound that more than half of the principals surveyed employed multiple frames while threequarters of the high school principals used multiple frames. As in similar studies, Rivers found that the dominant frame used was Human Resources followed by Structur al, Symbolic, and Political. She concluded that the use of multiple frames enhanced leadership and management effectiveness. No significance was found related to gender, age, or school level and the principals leadership orientations
54 Studies A ddressing S pecific D emographic C haracteristics In reviewing the literature regarding the superintendency, many studies were found that address some of the demographic characteristic s addressed in this study ; other studies had few or none of these The demographic characteristics studied in this research were age, gender, race, highest academic degree earned, experience in education, length of experience as a superintendent, size of district (number of students), and length of tenure as superintendent in current di strict. Gender The demographic characteristic found to be most commonly addressed in the literature was gender. Ninety -six percent ( 96% ) of the superintendents in the United States are male with the superintendency being the most male-dominated profession in our country (Bjork & Rodgers (1999). Bjork and Rodgers further reported that s tudents see females as having higher morals. They are more oriented towards the instructional concerns of teachers and strive to provide environments conducive to learning. Females are also viewed as more democratic, demonstrating a more participative style of leadership (Bjork & Rodgers, 1999) Brand (1993) reported that the majority of female superintendents are elected. The bulk of the literature, including studies of business leaders, college presidents, superintendents, and school administrators, shows little or no differences in leadership orientation frame preference or effectiveness in regards to gender (Bolman & Deal, 1991a, 1991b; Burks, 1992; Clisbee, 1994; DeF rank -Cole, 2003; Guidry, 2007; Hodge, 2003; Johnson, 1995; Martinez, 1996; Miro, 1993; Rivers, 1996; Thompson, 2000; Ward, 2006). Hodge (2003) did note a difference in frame orientation by gender in the qualitative analysis of her study, but found no diff erences in the quantitative analysis. Davis (1996) did not find gender to be significant in
55 determining the primary orientation of principals; however, female principals were found to use all four frames more often while men used no more than two. There are, however, some studies that found gender to be a significant factor. These studies found that females were more likely to use multiple frames (Cote, 1999; Durocher, 1995; Meade, 1992; Suzuki, 1994). Meade (1992) found that females used the Human Resou rce Frame most frequently, while Cote (1999) determined that males were more likely to employ the Human Resources frame. Age, R ace, Education, and E xperience Studies which included age, race, education, and experience were not as prevalent in the research. None of these factors was found to be significant in the majority of the studies reviewed (Durocher, 1995; Kelly, 1997; Martinez, 1996; Salley et al, 1979; Ward, 2006). Since the superintendency is a white, male dominated profession, study populations m ay not have yielded enough racial diversity for race to have been a factor. Experience was found to be a significant factor, or at least to have an effect on individual frame use (Cantu, 1997; Harlow 1994; Welch, 2002). These studies determined that mor e experience had a significant impact on the use of the Political frame with more experienced leaders showing more political orientation than those with less experience. Many s tudies found that superintendents are older and hold varying degrees (Brand, 199 3; Shuh & Herrington, 1990; Waller & Richardson, 1997). A difference was found between appointed and elected superintendents with appointed superintendents generally having more and higher degrees than elected superintendents (Shuh & Herrington, 1990); how ever, elected superintendents had been in education and held their positions longer (Brand, 1993).
56 Cote (1999) determined that education was a significant factor in principal leadership orientation, finding that principals with less education were more li kely to use the Structural frame than those with more education. She also determined that use of the symbolic frame decreased as principals experience increased. Educational background was also found to be a significant influence on frame choice in a st udy of deans of women (Guidry, 2007). Tenure and Size of D istrict Snider (2006) found that superintendents from the largest school districts have the shortest tenure making it difficult for them to develop the loyalty of other district employees needed to achieve necessary goals. Waller and Richardson (1997) were also concerned with the tenure of superintendents but did not determine its significance as an influence on leadership orientation frame choices. Cote (1999) reported a significant difference among mean school enrollments of principals using different frames, but there was no indication in the research whether this result would carry over to superintendents. In a study of self -perceived leadership styles of 18 permanently appointed African Amer ican women presidents of historically black colleges and universities, years of service as president was found to be related to the number of leadership orientation frames utilized. Summary In this review of literature, research was presented to support i nvestigation of the problem statement and resulting research questions. The problem addressed in this study was to determine if there is a difference between the leadership orientation frames of appointed and elected school district superintendents in the states of Alabama, Florida, and Mississippi. Research regarding Bolman and Deal s frame
57 theory was presented. Examination of the use of structural, human resources, political, and symbolic frames by leaders in a variety of organizations revealed that leaders abilities to define, analyze, and address situations from multiple perspectives enhances their abilities to solve problems. Research addressing superintendent leadership was presented in section two. Research addressing the issue of appointed and elected superintendents was included in this section along with projections relating to modern leadership challenges confirmed the need for continuing to conduct investigations to support the superintendency. The third section reviewed studies specifically related to Bolman and Deals leadership orientation frame categorization based on use of the Leadership Orientation (Self) Survey The emphasis in this section was to include available research on the leadership orientation frames of school district super intendents and those that addressed other demographic characteristics that will be included in the study. Results of these studies were presented and discussed (Bensimon, 1987; Bolman & Deal, 1991a, 1991b, 1992a; Childress, 1994; Clisbee, 2004; Cote, 1999; Durocher, 1995; Harlow, 1994; Johnson, 1995; Martinez, 1996; Meade, 1992; Miro, 1993; Redman, 1991; Rivers, 1996; Strickland, 1992; Suzuki, 1994). In the final section studies which specifically address the demographic characteristics included in this st udy were reviewed. The following studies related to the demographics were presented and discussed: Gender: (Bjork & Rodgers, 1999; Bolman & Deal, 11991a, 1991b; Brand, 1993; Burks, 1992; Clisbee, 1994; Cote, 1999; DeFrank -Cole, 2003; Durocher, 1995, Guidry, 2007; Hodge, 2003; Johnson, 1995; Martinez, 1996; Meade, 1992; Miro, 1993; Rivers, 1996; Suzuki, 1994; Thompson,
58 2000; Ward, 2006): Age, Ethnicity, Education, and Experience: (Brand, 1993; Cantu, 1997; Cote, 1999; Durocher, 1995; Guidry, 2007; Harlow 1994; Kelly, 1997; Martinez, 1996; Salley et al, 1979; Shuh & Herrington, 1990; Welch, 2002; Ward, 2006); Tenure and Size of District: (Cote, 1999; Snider, 2006; Waller & Richardson, 1997). Review of the literature has shown that more research on this critical position is needed. Studies have shown that superintendents tend to use the Human Resource Frames more than the other three frames, but the studies also show that the use of Human Resource Frame is not necessarily associated with effective leader ship, only effective management. The second most commonly used frame is Structural, also related to management effectiveness. The use of multiple frames has also been shown to increase leadership effectiveness, but many of the studies have reported little self reported use of multiple frames. This could indicate a need for further training of superintendents and those wishing to be superintendents in the use of all four of the leadership orientation frames. The review of literature also found few studies which addressed the method by which superintendents were selected, but does show that most of the United States has moved to systems requiring that all superintendents be appointed. There is a need to determine if there is a difference in the process use d by superintendents in making decisions, and whether this process is linked to their method of selection.
59 CHAPTER 3 METHODOLOGY Procedure The purpose of this study was to examine the leadership orientation frame preferences (as developed by Bolman & Deal 1984, 1991c, 1997, 2008) of elected and appointed school district superintendents in the states of Alabama, Florida, and Mississippi. Demographic characteristics of these superintendents, including age, gender, ethnicity highest academic degree, method of selection, years of experience in education, years of experience as a superintendent, length of tenure in current district, and size of district (number of students ) were also analyzed. The Leadership Orientation (Self) Survey (LO(S)S) (Bolman & Deal, 1990) was utilized to determine leadership orientation frame preferences of the superintendents participating in the study Research Questions This study investigated the self -reported leadership orientation frame preferences in order to determine whether or not there were any d ifferences in leadershi p frame use by the appointed and elected superintendents in the states of Alabama, Florida, and Mississippi. This study also examined the effects of contextual factors (district size) and superintendents ind ividual characteristics (age, gender, ethnicity, highest academic degree, method of selection, years of experience in education, years of experience as superintendent, years of experience as superintendent in current district) on their frame use. This stu dy focused on answering four questions: What are the self -reported leadership orientation frame preferences of the elected and appointed school district superintendents in Alabama, Florida, and Mississippi?
60 Is there a difference between self -reported leadership style preferences of elected and appointed school district superintendents? Is there a difference in single or multi -frame preferences between elected and appointed school district superintendents? Is there a difference in leadership style preference s of elected and appointed school district superintendents when considering selected demographic variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district)? Research Hypotheses The following hypotheses were created for the purpose of statistical analysis: There are no differences between self -reported leadership frame preferences of elected and appoi nted school district superintendents in Alabama, Florida, and Mississippi. There are no differences in single or multi -frame preferences between elected and appointed school district superintendents in Alabama, Florida, and Mississippi. Demographic variabl es (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district) have no statistically significant effect and/ or predictive value on the self -reported leadership frame preferences of elected and appointed school district superintendents. Type of Study Survey methodology was used to investigate the research questions. Reactions to Bolman and Deals (1990) LO(S)S Sections I, II, and III offered descriptive data. The survey included a demographic addendum, Section IV, which provided data concerning selected context and demographic variables (i.e., age, gender, ethnicity, highest academic degree, experience in education, length of experience as superintendent, length of tenure in current district, and size of district). The demographic addendum attached to the end of the survey did not influence responses to the previous sections.
61 Study Population The superintendents associations of Alabama and Mississippi reported that there were 131 superintendents in Alabama (39 elected, 92 appointed) and 150 superintendents in Mississippi (65 elected, 85 appointed). According to the Florida Department of Education online serv ices, there were 67 superintendents in Florida (44 elected, 23 appointed). This provided a population of 348 superintendents 148 elected and 200 appointed, in the three states. As these were the only states having elected and appointed superintendents, all 348 were included in the survey. Instrumentation The Leadership Orientations (Self) Survey (Bolman & Deal, 1990) instrument was used to identify superintendents self -reported leadership style preferences in terms of the four frames introduced by Bolm an and Deal. The instrument included 40 items, as well as subscales to represent each frame, and was divided into three sections. Section I included 32 questions to be rated on a Likert type scale of 1 (Never) to 5 (Always) depending on how often each of the items applied to the respondent. The questions were divided equally among the Structural, Human Resource, Political, and Symbolic frames. There were two leadership dimensions associated with each of the four frames as shown in Table 31 (Bolman & Deal, 2001). There were four (4) subscales in this section with eight statements to assess each of the four frames. Items 1, 9, 17, and 25 were associated with the Analytic section of S tructural Frame, while items 5, 13, 21, and 29 assessed the Organized component of the Structural Frame. The Supportive piece of the Human Resource Frame was asse ssed by items 2, 10, 18, and 26; and items 6,14, 22, and 30 were associated with the Participative component of Human Resource. Political Frame was composed of Powerful (items 3, 11, 19, 27) and
62 Adroit (items 7, 15, 23, 31). Inspirational (items 4, 12, 20, and 28) and Charismatic (items 8, 16, 24, 32) made up the Symbolic Frame. Table 3 1. Descriptions of d imensions of l eadership of the LO(S)S Frame (Dimensions) D escription Structural Analytic Thinks clearly and logically; approaches problems with facts and attends to detail. (Items 1, 9, 17, 25) Organized Develops clear goals and policies; holds people accountable for results. (Items 5, 13, 21, 29) Human Resource Supportive Concerned about the feelings of others; supportive and responsive. (Items 2, 10, 18, 26) Participative Fosters participation and involvement; listens and is open to new ideas. (Items 6, 14, 22, 30) Political Powerful Persuasive, high level of ability to mobilize people and resources; effective at building alliances and support. (Items 3, 11, 19, 27) Adroit Politically sensitive and skillful; a skillful negotiator in face of conflict and opposition. (Items 7, 15, 23, 31) Symbolic Inspirational Ins pires others to loyalty and enthusiasm; communicates a strong sense of vision. (Items 4, 12, 20, 28) Charismatic Imaginative, emphasize s culture and values; is highly charismatic. (Items 8, 16, 24, 32) Section II of the LO(S)S consisted of 6 forced-choic e items which produces a sharper differentiation among the frames because it does not permit rating someone high on everything (Bolman & Deal, 1992a, p. 320). Section III, comprised of two items, asked the respondents to rate their own effectiveness as managers and as leaders. These two terms are purposely not defined. The ratings are compared with other leaders with comparable levels of experience and responsibility. The section employed an anchored scale with the lowest score (1) being the bottom 20% and the highest score (5) being the top 20%. Bolman and Deal (1992a) expected to see a high
63 correlation between the two measures, and found that they are typically between .75 and .85. Section IV is a demographic addendum including age, gender, ethnicit y, highest academic degree earned, experience in education, years of experience as a superintendent, years of experience as superintendent in current district, and size of district. The resulting data was used to determine whether these characteristics have a significant influence on the self -reported leadership frame preferences of the responding superintendents. In the current study, participants were asked to respond to all three sections of the LO(S)S; however, only the responses to Section I were used to identify frame use T he items in Section II were more appropriate for determining variance within people rather than between them (Hodge, 2003). Reliability of the Leadership Orientations (Self) Survey Bolman and Deal (2004) reported reliability stat istics for the LO(S)S based on 1309 colleague ratings of managers in education and business. It should be noted that the alpha coefficients for items in Section I were consistently higher than the reliability coefficients for Section II. Chronbach alpha coefficients for all items were reported in Table 3 2. Table 3 2. Reliability of the Leadership Orientations (Self) Survey Frame Coefficient a lpha (Section I) Coefficient a lpha (Section II) Structural .92 .84 Human Resource .93 .84 Political 91 .79 Symbolic .93 .84
64 Validity of the Leadership Orientations (Self) Survey Bolman and Deal (1991a, 1991b, 1992a) examined the validity of the LO(S)S using regression analysis They determined that the self -rating of managerial and leadership effectiveness (Section III) was predicted by the four frames. A minimum of 66% of the variance in managerial effectiveness and 74% in leadership effectiveness were predicted. Bolman and Deal also concluded that those variables that were associated with m anagerial effectiveness were almost the opposite of those associated with leadership effectiveness. The structural frame was determined the best predictor for management effectiveness and the worst for leadership effectiveness. The symbolic frame proved to be the best predictor for leadership effectiveness and the worst for managerial effectiveness. Bolman and Deal (1991a) employed factor analysis to indicate the internal consistency of frame scores of the LO(S)S. They deleted six of the 32 items that s howed loading factors of less than .50 and retained all of those with .50 and above. It should be noted that the original instrument with 32 questions has been used successfully in many other studies (Cote, 1999; Durocher, 1995; Harlow, 1994; Hodge, 2003; Johnson, 1995; Rivers, 1996; Suzuki, 1994). Data Collection Permission was requested from Dr. Lee Bolman and Dr. Terrence Deal to use the LO(S)S in this study. A copy of the findings of the study as well as the actual raw data will be made available to them. The Alabama, Mississippi and Florida superintendents associations were contacted and asked to endorse the study to encourage superintendents responses to the survey. No response was received from either the Alabama or Mississippi
65 associations. A phone call from the president of the Florida Association of School Superintendents was received indicating that he would not provide endorsement for any studies. The American Association of School Administrators (AASA) was also contacted for an endorsement, and an endorsement of the study was received from Dr. Paul Houston, Executive Director of the AASA. Packets including a cover letter with endorsement f rom AASA, a copy of the LO(S)S, and demographic addendum were mailed to the 348 school superintenden ts in Alabama, Florida, and Mississippi in July 2007 In addition to the paper surveys a website was included for those superintendents who preferred to respond electronically. Superintendents were also sent the survey via email when requested. Summari es of the completed study will be sent to any superintendent who expressed interest. Follow up emails were also sent in an effort to increase response rate. Responses were considered confidential. Data Analysis Each item response for Section I of the LO( S)S was scored based on the respondents rating of how often each behavior item was true, using a five-point scale. Scores for the eight items measuring each frame were added and then divided by 8 to provide a mean score that reflected the superintendent s leadership behavior on that frame. Leadership orientation studies that have used only the questions in Sections I and III to investigate leaders leadership orientation frames have used the mean of the eight items on each of the four subscales in Section I to identify frame use (Durocher, 1995; Hodge, 2005; Johnson, 1995; Rivers, 1996; Suzuki, 1994). The highest possible mean was five and the lowest was zero. For the purposes of this study, the scores for each of
66 the four frames were computed by finding the average rating of each of the subscales (eight items per subscale) in Section I. A mean score of 4.0 or greater was used to indicate frame use often or always; frames with scores lower than 4.0 were considered to be used less often (Durocher, 1995; H odge, 2003). The quantitative data were analyzed using the Statistical Package for the Social Sciences (SPSS), Version 16.0. Descriptive s tatistics, analysis of variance, and Chi square were computed, and operations of regression analysis were conducted. The self -reported leadership frame preferences of elected and appointed school district superintendents were compared on each of the four frames (Structural, Human Resource, Political, and Symbolic). Oneway analysis of variance (ANOVA), which focused on a comparison of the variability within groups and between groups, was computed on each of the four frames to determine whether statistically significant differences existed between the leadership orientation frame preferences of appointed and elected superintendents. The ANOVA yielded a value of Fobserved on each variable which was compared with Fcritical using a l evel of significance of p<.05. Another focus of this study was whether or not there was a difference between appointed and elected in the use o f single or multiple frames. Chi Square was used to determine if statistically significant differences were discovered between the groups of appointed and elected superintendents. The inquiry of this dissertation also focused on discovering what contextual and personal characteristics were related to superintendents frame use; therefore, multiple regression analyses were also used in this study. Multiple linear regressions were used to analyze the strength of the relationship between the independent vari ables and a
67 single dependent variable. For this analysis, the dependent variables were the four frames: Structural, Human Resources, Political, and Symbolic as measured by Section I of the Leadership Orientation (Self) Survey The independent variables were the demographic characteristics and their various categories after collinearity diagnostics. The methods of selection (appointed, elected) were used as selection variables. A general linear regression model was used to examine the relationship between superintendents scores on specific leadership orientation frames and individual contextual and demographic variables. The frame scores were used as a continuous dependent variable with the independent variables of age, gender, ethnicity, highest academ ic degree earned, experience in education, length of experience as a superintendent, length of tenure, and size of district while controlling for method of selection. Collinearity, or close correlations among independent variables is always a concern in Multiple Regression; consequently, collinearity diagnostics were run to determine collinearity among variables. The first regression was performed using the Structural Frame as the dependent variable and appointed as the method of selection. This was fol lowed by another regression using the elected method of selection. The other models followed the same pattern changing frame types and methods of selection. These eight models were estimated to address the fourth and final research question. Model fit w as judged on the basis of the magnitude of the model R2 and the statistical significance of the associated F value. All statistical tests were conducted at p = <.05. Multiple linear regression was used by the researcher to analyze whether superintendent frame preference could be predicted by one or more of the demographic variables included in the survey.
68 Summary Chapter 3 described the method used to investigate the leadership orientation frames used by school district superintendents in Alabama, Florida, and Mississippi. The Leadership Orientation (Self) Survey and demographic addendum were used to collect data. Population demographics, data collection techniques, as well data analysis procedures were presented in this chapter. Chapter 4 presents the results from the survey instrument as well as the data analysis and interpretation in view of the research questions and hypotheses. Chapter 5 presents conclusions, implications of the study, and suggestions for further research.
69 CHAPTER 4 RESULTS AND DATA ANALYSIS The purpose of this study was to examine the leadership orientation preferences (as developed by Bolman & Deal, 1984, 1991c, 1997) of elected and appointed school district superintendents in Alabama, Florida, and Mississippi. Demographic cha racteristics of these superintendents, including age, gender, race, education, method of selection, years of experience in education, years of experience as a superintendent, size of district (number of students), and length of tenure in the district were also analyzed. The Leadership Orientation (Self) Survey (LO(S)S) (Bolman & Deal, 1990) was employed to determine leadership orientation frame preferences of the superintendents included in the study. Four research questions framed and structured this stu dy: What are the self -reported leadership orientation frame preferences of the elected and appointed school district superintendents in Alabama, Florida, and Mississippi? Is there a difference between self -reported leadership style preferences of elected and appointed school district superintendents? Is there a difference in single or multi -frame preferences between elected and appointed school district superintendents? Is there a difference in leadership style preferences of elected and appointed school d istrict superintendents when considering selected demographic variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current distric t, and size of district)? Chapter 4 presents the results of the analysis of the quantitative data. The results are presented based on the research questions. First, descriptive data are presented. These include the demographic characteristics as well as descriptive statistics and comparisons of participants espoused use of the cognitive leadership frames. Next,
70 analyses and results of analysis of variance and linear regressions are presented that were used to examine relationships between variables and the self -reported leadership orientation frames. Respondent superintendents leadership orientation frames were determined by summarizing their responses to Section I of the LO(S)S (Bolman and Deal, 1990). Scores on leadership and management effectiveness were taken from superintendent responses to Section III, but were not used in determining preferences in leadership orientation frames. Demographic characteristics of superintendent respondents were reported from responses to Section IV the demographic addendum. Superintendent demographic data were grouped and reported using four (4) categories for age, ethnicity, years completed in education, total years as superintendent, years as superintendent in current district, and the FTE enrollment in the distric t; and three (3) categories for highest degree earned. A total of 348 Leadership Orientations (Self) Survey instruments were distributed to the entire population of school district superintendents in Alabama, Florida, and Mississippi. One hundred forty -five (145) questionnaires were returned either by U.S. mail or electronically, res ulting in a return rate of 41.7%. Of the returned questionnaires, 89 (61.4%) were from appointed school district superintendents and 56 (38.6%) from elected school district superintendents. Participation rate for appointed school district superintendents was 44.5%; participation rate for elected school district superintendents was 37.8%. Participants from each state represented approximately 40% of their respective states district superintendent population (Alabama 38.3%; Florida 44.7%; Mississippi 42.06%). The size of participating school districts ranged
71 from 170 to 128,000 student FTEs (full time equivalence) with a median district size of 3,327 student FTEs. Descrip tive Data Demographics Table 4 1 is a summary of the number and percentage of superintendent characteristic responses as well as a breakdown of appointed and elected superintendents according to the variables of age, gender, and ethnicity. Superintendents were asked to identify their ages according to the following ranges: (a) 2534, (b) 35-44, (c) 45 -54, and (d) 55 and over. Ethnicity was also identified according to four categories: (a) African American, (b) Caucasian, (c) Hispanic, and (d) Other. Data on age, gender and ethnicity revealed that 63.4% of the superintendents were age 55 and over with only just 4.1% within the ages of 3544. No superintendents were reported between 25-34 years. Most of the superintendents were male (77.2%) and Caucasian ( 82.8%). Superintendent answers to the question regarding highest academic degree were limited to (a) masters, (b) specialist, and (c) doctorate. Only two (2) responses, appointed and elected, identified the method of selection to the superintendency. Su rvey participants were asked to indicate both their total years as a superintendent as well as the total number of complete years as superintendent in their current district. Response categories for these items were: (a) less than one year, (b) 15 years (c) 6 15 years, and (d) 16 or more years. Superintendents were also asked to indicate their total number of complete years in education based on four (4) categories: (a) 15; (b) 615; (c) 16-20; or (d) 21 or more.
72 Table 4 1. Age, g ender, and e thni city of s uperintendents N = 145 Characteristic Frequency % Appointed (89) Elected (56) Age 2534 0 0 0 0 3544 6 4.1 6 0 4554 47 32.4 33 14 55 + 92 63.4 50 42 Gender Female 33 22.8 24 9 Male 112 7.2 65 50 Ethnicity African Am. 21 14.5 14 7 Caucasian 120 82.8 72 48 Hispanic 3 2.1 2 1 Other 1 .7 1 0 Responses to the item regarding nu mber of full time equivalency (FTE) students enrolled in a district were free responses which were placed into four (4) categories for statistical purposes. These data, reported in Table 42, indicate that 49.7% of the respondents held doctoral degrees, w hile 31% held masters degrees. Appointed superintendents comprised 61.4% of the respondents; 38.6% of the superintendents were elected to their positions. The vast majority of superintendents (93.1%) had spent 21 or more years in the field of education and 48% had served as superintendents for six or more years. School district sizes were divided into four categories to coincide with the AASAs 2000 study (Glass, Bjork, Brunner, 2000) for the purpose of reporting results. (Category 1 contains school di stricts with a student FTE of less than 300; category 2 districts serve a student population between 3002,999; category 3 includes FTE student enrollments of 3000-24,999; and category 4 districts have student FTE enrollments of 25,000 or more.) This stud y included superintendents from three different states: 51 from Alabama (20.7%), 30 from Florida (35.2%) and 51 from
73 Mississippi (42.1%). Three (3) respondents failed to report the state in which they currently served as superintendent. Table 4 2. Highes t d egree, m ethod of s election, y ears as s uperintendent, y ears in c urrent d istrict, e xperience in e ducation, d istrict s ize, s tate (N = 145) Superintendents Frequency % Appointed (89) Elected (56) Highest degree Masters 45 31.0 12 33 Specialist 28 19.3 16 12 Doctorate 72 49.7 61 11 Method of s election Appointed 89 61.4 Elected 56 38.6 Years in e ducation 1 -5 0 0 0 0 6 -15 0 0 0 0 1620 10 6.9 9 1 21+ 1 35 93.1 80 55 Years as s uperintendent <1 13 9.0 10 3 1 -5 59 40.7 38 21 6 -15 60 41.4 32 28 16+ 11 7.6 8 3 No r esponse 2 1.4 1 1 Years s uperintendent in c urrent d istrict Less tha n one 13 9.0 9 4 1 -5 72 49.7 53 19 6 -15 57 39.3 26 31 16+ 2 1.4 1 1 No r esponse 1 .6 0 1 Size of d istrict Group 1 (<300) 1 .7 1 0 Group 2 (300 2999) 63 44.1 42 21 Group 3 (3000 24999) 65 45.5 35 30 Group 4 (>24999) 14 9.8 10 4 No r esponse 2 1.4 1 1 State Alabama 51 35.9 40 11 Florida 30 21.1 12 18 Mississippi 61 42.9 35 26 Not i dentified 3 2.1 2 1
74 Self -rated Effectiveness as M anager and L eader This study included opportunities for respondents to indicate self -ratings in the areas of effectiveness both as managers and leaders (Table 4 3). There were five possible ratings, ranging from 1 (Bottom 20%) to 5 (Top 20%) in both areas. The mean self -rating of participating superintendents on management effectiveness was 4.38 (SD .68). The mean self -rating of appointed superintendents on management effectiveness was 4.36 (SD .71), while the mean self -rating of elected superintendents was 4.42 (SD .63). Self -ratings on effectiveness as a leader were even higher, with a mean of the total superintendents (4.52, SD.63). The mean of app ointed superintendents in leadership effectiveness was 4.56 ( SD .64) with elected superintendents reporting a mean of 4.45 (SD .60). There were no significant differences between the self -ratings of appointed and elected superintendents in either manager or leader effectiveness (Table 44). These self -ratings were not used in computing respondents leadership orientation frame preferences. Table 4 3. Effectiveness as m anager (M) and l eader (L) Self -Rating n (M) % (M) n (L) % (L) Below m iddle 0 0 1 .7 Middle 20% 16 11.0 7 4.8 Above m iddle 56 38.6 52 35.9 Top 20% 71 49.0 83 57.2 No r esponse 2 1.4 2 1.4 Table 4 4. Effectiveness s elf -r ating: M anager (M) and l eader (L) by s election m ethod Method n (M) Mean SD t p n (L) Mean SD t p All 144 4.38 .68 144 4.52 .63 Appointed 88 4.36 .71 88 4.56 .64 Elected 56 4.42 .63 56 4.45 .60 .95 .17 47 .32 p <.05
75 Participants U se of C ognitive F rames What are the self -reported leadership orientation frame preferences of the elected and appointed school district superintendents in Alabama, Florida, and Mississippi? Section I of the LO(S)S was used to determine the leadership orientation frame preferences of participating superintendents. Frame use was determined by adding the eight (8) item scores in pertinent sections and dividing by eight (8). This method of scoring was selected because the current study is specifically investigating the differences between elected and appointed superintendents. The method which includes Section II in determining frame use is preferred for determining differences within groups (Hodge, 2003). Structural Frame use was determined by adding item scores for Analytic (items 1, 9, 17, 25) and Organized (items 5, 13, 2 1, 26). Responses on Supportive (items 2, 10, 18, 26) and Participative (items 6, 14, 22, 30) determined Human Resource Frame preference. Scores on Powerful (items 3, 11 19, 27) and Adroit (items 7, 15, 23, 31) were used to determine Political Frame, and scores on Inspirational (items 4, 12, 20, 28) and Charismatic (items 8, 16, 24, and 32) were determiners for Symbolic Frame. Tables 4 -5 and 46 summarize descriptive data of the four frames used by elected, appointed and combined superintendents. Respon dents scores could range from 1.00 to 5.00 on each item. A mean score of 4.00 or above indicated that participants used that frame often or always. Scores of less than 4.00 indicated that the frame was used less frequently (Durocher, 1995; Hodge, 2003). Table 4 5 shows that t he mean score on the Human Resource Frame (3.78) was the highest frame score for all superintendents participating in this study. Scores ranged from 2.5 to 4.57. The Structural Frame received a mean score of 3.49 (Range
76 2.145.2 9), and the Symbolic Frame received a 3.39 (range 2.07-4.43). The Political Frame received the lowest mean score ( 3.23 ) with a range of 2.21-4.50. Is there a difference between self -reported leadership style preferences of elected and appointed school dis trict superintendents? Research hypothesis: There are no differences between self -reported leadership frame preferences of elected and appointed school district superintendents in Alabama, Florida, and Mississippi. This hypothesis was rejected as a signif icant difference was found in the self reported leadership frame preferences of elected and appointed superintendents in the use of the Human Resource Frame. The data below support the rejection of this hypothesis. Table 4 6 documents that 9 superintenden ts ( 6 appointed; 3 elected) showed a preference for the Structural Frame. The Human Resource Frame was preferred by 32 superintendents ( 15 appointed; 17 elected). The Political Frame was identified by 3 superintendents ( 1 appointed; 2 elected), while the Symbolic Frame was selected by 6 superintendents ( 4 appointed, 2 elected). Table 4 5. Superintendent f rame p reference ( m ean and s tandard d eviation) (N = 145) Frames M SD Minimum Maximum ST 3.49 .53 2.14 5.29* HR 3.78 .41 2.50 4.57 PO 3.23 .47 2.21 4.50 SY 3.39 .49 2.07 4.43 ST = Structural, HR = Human Resource, PO = Political, SY = Symbolic *Maximum >5.0 caused by invalid response Table 4 6. Superintendent f rame p reference (*n = number of valid responses) Frame *n Users % Appointed Elected ST 134 9 6 6 3 HR 134 32 22.1 15 17 PO 135 3 2 1 2 SY 136 6 4 4 2
77 The means and standard deviations for appointed and elected superintendents in the preference for e ach frame are listed in Table 47 as well as the totals and standard deviations for both. Results of the analysis of variance between groups test results are presented in Table 4-8 A significant difference, F (1, 132) = 4.11; p < .05, was revealed amon g appointed and elected superintendent self -reported preference s for the Human Resource Frame. The analysis of variance reported no significant differences between appointed and elected superintendents in self -reported preferences for the other three fram es. Table 4 7. Superintendent f rame p references (M) by s election m ethod Frame n (number of responses) Method Mean SD ST 82 Appointed 3.51 .50 52 Elected 3.46 .58 134 All 3.49 .53 HR 82 Appointed 3.73 .43 52 Elec ted 3.87 .36 134 All 3.78 .41 PO 83 Appointed 3.23 .44 52 Elected 3.33 .53 135 All 3.27 .47 SY 84 Appointed 3.40 .47 52 Elected 3.37 .52 136 All 3.39 .49 Data in Table 4 9 show that of th e supe rintendents responding, 60 (41%) showed no frame preference. This means that none of their self -reported frame scores reached the 4.0 mean required for reporting frame preference. Fifty (5 0 ) superintendents (35%) demonstrated a prefer ence for using one fr ame, and 32 (22 %) of those preferred the Human Resource Frame. The use of two frames was reported by 11 of the respondents (8%). The most common two -frame combination was that of the Structural and Human
78 Resource Frames. Six superintendents using a three-frame combination comprised 4% of the respondents, and four superintendents (3%) indicated that they used all four frames. A total of 21 (15%) superintendents reported using multiple frames. Table 4 8. Analysis of v ariance of f rame p references by m ethod of s election (N=136) Frame SS df MS F Sig. Structural .09 1 .09 .33 .57 Within g roups 37.26 132 .28 Total 37.35 133 Human R esources .66 1 .66 4.11 .05* Within g roups 21.30 132 .16 Total 21. 97 133 Political .38 1 .38 1.70 .19 Within g roups 29.74 133 .22 Total 30.12 134 Symbolic .03 1 .03 .14 .71 Within g roups 31.79 134 .24 Total 31.83 135 p = < ,.05 *Note: Significance level was rounded to t wo (2) decimals. Actual significance was p = 045. Table 4 9. Single and m ultiple f rame p references N = 145 Frame Preference n Frames None 60 Single f rames 50 9(ST) 33(HR) 3(PO) 6(SY) Multiple 21 Two -f rames 11 7(ST+HR) 1(ST+PO) 3(HR+ SY) Three-f rames 6 1(ST+HR+SY) 2 (ST+HR+PO) 1(ST+PO+SY) 2(HR+PO+SY) Four -f rames 4 M issing 14 Is there a difference in single or multi-frame preferences between elected and appointed school district superintendents?
79 Research hypothesis: There are no differences in single or multi-frame preferences between elected and appointed school district superintendents in Alabama, Florida, and Mississippi. This hypothesis is accepted as no statistically significant differences were found in single o r multi -frame preferences between elected and appointed school district superintendents in Alabama, Florida, and Mississippi. A statistically significant difference was found in no preference between elected and appointed superintendents in these three st ates. The data below support the acceptance of this hypothesis. Further statistical study by method of selection (Table 4 10) showed that 47 % of appointed superintendents and 32 % of elected superintendents indicated no frame preference. Twenty -s ix (29 %) of the appointed superintendents preferred a single frame, while 43% of the elected superintendents preferred a single frame. The most preferred frame was Human Resource with 1 7 % of appointed and 30% of elected superintendents indicating this preference. The use of two frames was preferred by 10 % of the appointed superintendents and 4% of elected superintendents. Preference for three frames was indicated by 2% of the appointed superintendents and 8 % of the elected superintendents. None of the appointed superintendents indicated a preference for the use of all 4 frames; 7% of elected superintendents preferred the use of all of the frames. In summary, 12 % of the appointed superintendents and 18 % of the elected superintendents responding indicated a prefer ence for the use of multiple frames. A Chi -square test was used to determine whether or not there was a difference between the use of no preference, single preference, and multi -frame preference by appointed and elected superintendents. The test for no fr ame preference showed a
80 statistically significant difference of .037. The results of the Chi -square procedure are reported in Table 4-11. Table 4 10. Single and m ultiple f rame p references by m ethod ( N=145) Frame Preference N % Appointed % Elected % N one 60 41 42 47 18 32 Single 50 35 26 29 24 43 Structural 9 6 6 7 3 4 Human Resource 3 2 22 15 17 17 30 Political 3 2 1 1 2 4 Symbolic 6 4 4 5 2 4 Multiple Frames 21 15 11 1 2 10 18 ST+HR 7 5 6 7 1 2 ST+PO 1 1 0 0 1 2 HR+SY 3 2 3 3 0 0 ST+HR+SY 1 1 1 1 0 0 ST+HR+PO 2 1 0 0 2 4 ST+PO +SY 1 1 1 1 0 0 HR+P O+SY 2 1 0 0 2 4 All Frames 4 3 0 0 4 7 Missing 14 10 10 10 4 7 Table 4 11. Chi -square (x 2) of n o p reference, s ingle and m ulti -f rame u se by a ppointed and e lected s uperintendents (N = 145, Missing = 14) Fra me Appointed Elected Total x2 (Pearson 2-sided) None 42 18 60 .037* Single 26 24 50 .127 Multiple 11 10 21 .418 p < .05 Multiple Regression Collinearity is always a concern when conducting multiple regressions. Prior to performing the m ultipl e regression procedures, diagnostic measures were conducted to determine the relationships between the eight explanatory variables (age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, y ears as superintendent in current district, and size of
81 district) The greatest concerns of collinearity were between : (a) A ge and Years as Superintendent ( -.297) ; (b) Years in Education (-.438) and Gender ( -.438) ; (c) Years as Superintendent in Current D istrict and Total Y ears as Superintendent ( -.712). Some degree of col linearity also existed between Gender and Highest A ca demic Degree (.329) Total Years as Superintendent, and Years as Superintendent in Current D istrict (.745) Due to the number of res pondents and the fact that participants were primarily white males, some collinearity had to be tolerated; however, i n order to reduce the collinearity, the demographic characteristics were broken into smaller variables utilizing the responses available on the survey. The independent variables selected for use in the regressions were: Age (1, 2, 3, 4), Ethnicity (1, 2, 3, 4), Gender, Highest Academic Degree (1, 2, 3), Years as Superintendent (1, 2, 3, 4), and District Size (1, 2, 3, 4). Years in Education and Years in Current District were not selected for use in the regressions due to collinearity between these variables and Age and Years as Superintendent. The dependent variables for the regressions were the four frames (Structural, Human Resource, Poli tical, and Symbolic) and the selection variable was Method of Selection (Appointed, Elected). The .05 level of s ignificance was used for each model Eight regression models were tested to investigate the influence of these explanatory variables on each of the leadership orientation frames (dependent variable). Is there a difference in leadership style preferences of elected and appointed school district superintendents when considering selected demographic variables (i.e., age, gender, ethnicity, highest a cademic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district)? Demographic variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district) have no statistically significant effect and/or predictive value on the self -reported
82 leadership frame preference s of elected and appointed school district superintendents. This hypothesis is rejected as significant effects and/or predictive values on the self -reported leadership frame preferences of elected and appointed school district superintendents were found. The data below supports the rejection of this hypothesis. The Structural Frame preference by A ppointed superintendents was the first model f ollowed by Structural Frame by E lected. The other models were similar using each frame separately twice, once by ea ch method of selection. The resulting multiple linear regression output was reported using only three output components (Model Summary, ANOVA, and Coefficients) that were used to interpret the results. The first regression model consisted of the dependent variable, Structural Frame, the independent or predictor variables, and the selection variable (m ethod of selectionappointed). Results showed that the R2 of .266 was statistically significant, F ( 13,67) = 1. 87 p = 050 (Table 411). This model indicated that together the demographic variables accounted for 26.6% of the variance in scores on the Structural Frame reported by the appointed superintendents. Also included in Table 4 11, the second regression model included Structural Frame, the independent variables, and selection variable, (method of selectionelected). Results of this model indicated that the R2 of 287 was not statistically significant, F ( 11,38 ) = 1. 389, p = 218. Tables 4 -12 and 4-13 show the regression coefficients of each of the independent variables included in the two Structural models. In the first Structural model (Appointed), the independent variables G ender ( Beta = .295, p = .02 0 ), Ethnicity 1 (Beta = .327, p = 008 ), and DistSize 4 (Beta = -.274, p = .023) were all statisticall y significant DistSize 4 contributes negatively to the variance. In the second Structural
83 model (Elected) Eth1 (Beta = .314, p =.041) was found to be statistically significant Eth1 African Americans was found to be statistically significant in both S tructural models. Table 4 11. Summary of Multiple Regression An alyses for Variables Related to Superintendents Use of the Structural Frame Model R2 SS df MS F p 1. Structural ( appointed ) 13.67 Regression 5.377 13 414 1. 87 050* Residual 14.821 67 2 21 Total 20.198 80 2. Structural (elected ) .287 Regression 4. 779 11 .434 1. 389 218 Residual 11.887 38 .313 Total 16.666 49 p = <.05 Table 4 12. Summary of m ultiple r egression c oefficients for m odel 1 Structural ( a ppointed) Variables B SE Beta t p (Constant) 2.929 326 12.399 .000 Gender .3 36 .14 0 .2 95 2. 390 .020* Age 2 -.055 229 -.0 29 -. 241 -.810 Age3 -.062 .118 -.060 -.527 .600 Eth 1 .476 .1 73 327 2.749 .0 08* Eth3 .630 .357 .196 1.763 -.083 D egree1 -.135 173 -.089 -.778 439 Degree2 -.107 .163 -.086 -.660 .512 YRSup1 .266 188 .168 1.418 161 YRSup3 .0 01 .1 27 001 008 .994 Y RSup4 .2 70 .207 .1 52 1.302 197 DistSize1 -. 102 490 -. 022 -.207 836 DistSize3 -.064 .126 -.062 -.503 .616 DistSize4 -.436 .187 -.274 -2.327 .023* p = <.05 The third regression model consisted of the dependent variable, Human Resource Frame, the independent variables, and th e selection variable (method of selectionappointed). Results showed that the R2 of 1 40 was not statistically significant, F ( 14 66) = .767, p = 699 (Table 4 -14). This model indicated that together the demographic
84 variables accounted for only 14 % of the variance in scores on the Human Resource Frame reported by the appointed superintendents. Also included in Table 4-14, the fourth regression model consisted of the dependent variable, Human Resource Fram e, the independent variables, and the selection variable (method of selectionelected). Table 4 13. Summary of m ultiple r egression coefficients for m odel 2 Structural ( e lected) Variables B SE Beta t p (Constant) 3.366 .520 6.472 000 Gender -. 110 .2 43 -. 073 -.454 653 Age 4 .3 98 204 309 1.951 .0 58 Eth 1 .605 .2 86 .3 14 2. 113 .0 41 Eth3 -.544 .580 -. 1 32 -.937 .355 Degree2 .0 98 224 0 70 .439 633 Degree3 -.081 .228 -.058 -.354 .725 YRSup1 087 527 .0 36 164 870 YRSup2 -.2 11 183 -. 177 -1 155 255 YRSup4 -.055 .369 -.022 -.148 .883 DistSize2 .080 .193 .068 .417 .679 Dis tSize4 -.133 .450 -.055 -. 295 769 p = <.05 Table 4 14. Summary of m ultiple r e gression a nalyses for v ariables r elated to s uperintendents u se of the Human Resource Frame Model R2 SS df MS F p 3. Human Resource (appointed) .140 Regression 2.471 14 177 767 699 Residual 15. 183 66 230 Total 17. 65 4 80 4. Human Resource (elected) .323 Regression 4 032 11 367 1 649 124 Residual 8.448 38 222 Total 12.480 49 p = <.05 Results of this regression model showed that the R2 of 323 was not statistically significant, F ( 11 38) = 1.649, p = 124 Tables 4 -15 and 4-16 show the regression coefficients of each of the independent variables included in the two Human Resource models. None of the independent variables was found to be statistically significant in the first Hum an Resource model;
85 however, in the second Human Resource model Degree 3 (Beta = -.403, p = .016). Degree3 (doctorate) contributes a negative 40.3% to the variance shown in the model. DistSize 2 (3002,999 FTE) is not quite considered significant ( p = .05 5), and also makes a negative contribution. Table 4 15. Summary of m ultiple r egression c oefficients for Model 1 Human Resource Frame ( a ppointed) Variables B SE Beta t p (Constant) 273 241 1.132 262 Gender .0 25 .1 43 .024 176 860 Age 2 1 22 234 .0 69 524 602 Age3 -.076 .121 -.079 -.628 .532 Eth 1 310 .1 85 21 9 1. 677 .0 98 Eth3 -.213 .366 -.071 -.581 .563 Eth4 -.203 .521 -.048 -.390 .698 Degree1 .107 .177 .0 7 5 605 547 Degree2 -.063 .170 -.051 -.369 .713 YRSup1 295 1 92 199 1 536 129 YRSup3 -.0 3 8 132 -. 0 39 -. 289 773 YRSup4 .076 .202 .049 .377 .708 Dist Size1 -. 286 500 -. 068 -. 571 570 Dist Size3 -.0 83 129 -. 0 86 -. 6 44 5 22 DistSi ze4 -.006 .191 -.004 -.033 .974 p = <.05 The fifth regression model consisted o f the dependent variable (Political Frame) the independent variables, and th e selection variable (method of selectionappointed). Results showed that the R2 of 316 wa s statistically significant, F (14, 67) = 2.208 p = 016 (Table 417 ). This model indicated that together the demographic variables accounted for 31.6 % of the variance in scores on the Political Frame reported by the appointed superintendents. Also inclu ded in Table 4 -17 the sixth regression model consisted of the dependent variable, Political Frame, the independent variables, and the selection variable ( method of selectionelected). Results of th e second Political
86 regression model showed that the R2 of 215 was not statistically significant, F (11 38 ) = .946, p = 509. T able 4 16. Summary of m ultiple r egression c oefficients for Model 2 Human Resource Frame ( e lected ) Variables B SE Beta t p (Constant) 9 65 438 2.201 .0 34 Gender -. 100 2 05 -.0 77 -. 487 629 Age 4 -.0 60 .0 72 -.0 54 -. 348 730 Eth1 388 241 2 33 1. 609 116 Eth3 .294 .489 .082 .601 .552 Degree2 -.1 47 189 -. 122 -. 7 78 441 Degree3 -.487 .192 -.403 -2.532 .016* YRSup1 295 445 140 663 511 YRS up 2 -.0 98 154 -. 0 95 -. 636 529 YRSup4 .036 .311 .017 .116 .908 DistSize2 -. 321 .1 62 -. 312 -1 976 055 Dist Size4 -. 448 380 -. 213 -1.179 246 p = <.05 Table 4 1 7 Summary of m ultiple r egression a nalyses fo r v ariables r elated to s uperintendents u se of the Political Frame Model R2 SS df MS F p 5 Political ( appointed) .316 Regression .616 14 044 2.208 016* Residual 1.335 67 020 Total 1.951 81 6 Political ( elect ed) .215 Regression 1.720 11 156 .946 509 Residual 6.280 38 165 Total 8.000 49 p = <.05 Tables 4 -18 and 4-19 show the regression coefficients of each of the independent variables included in the two Political Frame models. In the first Political Frame model the independent variable Eth3 (Beta = -.486, p = .000 ). The variable Eth3 (Hispanic) contributes negatively to the variance in this model. I n the second Political Frame model, Age4 (55 and over) was found to be statistically significant (Beta = .168, p = .024).
87 Table 4 18 Summary of m ultiple r egression c oefficients for Model 1 Political Frame ( a ppointed) Variables B SE Beta t p (Constant) -. 032 071 -.458 649 Gender .0 32 042 .0 91 764 44 8 Age 2 -. 057 069 -.0 97 -. 830 410 Age3 -.0 54 036 -. 171 -1.524 132 Eth1 -. 023 052 -. 051 -. 442 660 Eth3 .486 107 486 4.533 000* Eth4 .055 153 .0 39 .3 56 723 Degree1 -. 024 052 -.0 52 -.468 641 Degre e2 .0 20 .1 49 .052 411 682 YRSup1 016 056 032 276 783 YRSup3 .0 29 038 0 91 769 444 YRSup4 -.023 060 -.04 4 -.3 85 .70 1 Dist Size1 -. 062 148 -. 0 44 -. 418 677 Dist Size3 .0 35 038 109 .912 365 Di stSize4 -.0 62 056 -. 132 -1.101 275 p = <.05 Table 4 19. Summary of m ultiple r egression c oefficients for Model 2 Political Frame ( e l ected) Variables B SE Beta t p (Constant) 245 378 .647 521 Gender -. 1 19 176 -. 114 -. 672 506 Age 4 347 148 390 2.343 024* Eth1 224 2 08 168 1. 078 288 Eth3 -. 355 .4 22 -. 124 -. 841 405 Degree2 -. 053 1 63 -. 055 -. 328 744 Degree3 069 .1 66 071 415 680 YRSup1 011 383 006 028 978 YR Sup2 -.0 38 1 33 -. 0 46 -. 284 778 YRSup4 -. 388 268 -. 230 -1.445 157 DistSize2 -. 131 .1 40 -. 159 -. 9 38 354 Dist Size4 -. 180 3 27 -. 107 -. 551 585 p = <.05 The seventh regression model consisted o f the dependent variable (Symbolic Frame ) the independent variables, and th e selection variable (method of selectionappointed). Results showed that the R2 of 393 was statistically significant, F (14, 68) = 3.138 p = 001 (Table 4-20 ). This model indicated that together the demographic variables accounted for 39.3% of the varia nce in scores on the Symbolic Frame
88 reported by the appointed superintenden ts. Also included in Table 4 20, the eighth regression model consisted of the dependent variable, Symbolic Frame, the independent variable s, and the selection variable ( method of selectionelected). Results of th e second Symbolic regression model showed that the R2 of 366 was not statistically significant, F (11 38 ) = 1.999, p = 056. Tables 4 -21 and 4-22 show the regression coefficients of each of the independent variables included in the two Symbolic Frame models. In the first Symbolic Frame model the independent variable Eth3 (Beta = 507, p = .000) and DistSize1 (Beta = .361, p = .000) were both found to be statistically significant. These variables represent Hispanic and District Size of less than 300 FTEs. In the second Symbolic Frame model, DistSize 2 (Beta = -.424, p = .008), representing districts from 300-2,999 FTEs, was found to be statistically significant. This variable c ontributed negatively to the variance in this model. Table 4 20 Summary of m ultiple r egression a nalyses for v ariables r elated to s uperintendents u se of the Symbolic Frame Model R2 SS df MS F p 7 Symbolic ( appointed) .393 Regression 3.150 14 225 3.138 0 01 Residual 4.675 68 0 72 Total 8.024 82 8 Symbolic ( elected) .366 Regression 2.206 11 201 1.999 056 Residual 3.814 38 100 Total 6.020 49 p = <.05 Table 4 21 Summary of m ultiple r egression c oefficients for Model 1 Symbolic Frame ( a ppointed) Variables B SE Beta t p
89 (Constant) .204 134 1.516 134 Gender -.048 080 -.067 -.602 549 Age 2 -. 095 130 -.0 79 -. 731 467 Age3 -.037 .067 -.058 -.559 .578 Eth1 .067 098 .073 .678 500 Eth3 1.027 .203 .507 5.052 .000* Eth4 .112 .291 .039 .387 .700 Degree1 .063 .099 .066 .643 522 Degree2 .088 .093 .112 .954 .343 YRSup1 -.037 107 -. 037 -.346 731 YRSup3 -.138 072 -.213 -1.908 061 YRSup4 -.129 .113 -.123 -1.147 .255 Dist Size1 1.030 279 .361 3.695 000* Dist Size3 .0 25 072 038 .344 732 DistSize4 -.045 .106 -.047 -.421 .675 p = <.05 Table 4 22 Summa ry of m ultiple r egression c oefficients for Model 2 Symbolic Frame ( e l ected) Variables B SE Beta t p (Constant) -.248 295 -.843 405 Gender .133 1 38 .148 .970 338 Age 4 163 1 15 211 1.414 166 Eth1 310 162 2 68 1. 912 063 Eth3 -. 182 329 -. 073 -. 553 584 Degree2 .111 1 27 .133 .877 386 Degree3 242 .1 29 289 1.875 068 YRSup1 175 299 .120 585 562 YRSup2 .155 1 0 3 .217 1.501 142 YRSup4 -. 244 .2 09 -. 167 -1.168 250 DistSize2 -. 303 .1 09 -. 424 -2 778 008* Dist Size4 -. 277 255 -. 1 90 -1.087 284 p = <.05 Chapter 4 presented data analysis results and procedures. Descriptive data of the demographic characteristics of the 145 participating superintenden ts (89 appointed, 56 elected) indicated that most of the superintendents were Caucasian, male, and 55 years of age or older. Most had been in the field of education for 21+ years and 16 of them had been superintendents for 16 or more years, although just two (2) had been superintendents in their current district for 16 or more years. District sizes (FTE) ranged
90 from 170 to 128,000 with a median district size of 3,327. Approximately 90% of all participating superintendents worked in districts ranging from 300 to 24,999 student FTEs. Sixty (60 ) of the participating superintendents indicated no particular leadership orientation frame preference, while multiple frames were used by 22 (15%) of the superintendents. The Human Resource Frame was indicated a s th e most preferred frame by 32 (23%) respondents. A statistically significant difference of p = .045 was found in the analysis of variance for the Human Resource Frame between appointed and elected superintendents with the elected superintendents indicat ing a higher preference for this frame. Multiple Regression procedures were performed to determine whether or not the demographic characteristics had any significant effect on the leadership orientation frame choices of appointed and elected. The regression models further indicated any statistically significant influence of specific demographic characteristic s on the leadership frame choices of the participating superintendents. The Structural Frame was found to be statistically significant for appointed su perintendents with an R2 of .266 indicting that the demographic variables accounted for 26.6% of the variance in the model. Gender and African American ethnicity were found to be statistically significant as was District Size 4. District Size contributed negatively to the variance. Neither of the Human Resource Frame models was found to be significant; although, a Doctora l degree was a significant contributor to the variance in the Human Resource -Elected model. Regressions for the Political Frame indicated a significant influence on the variance in the model for Appointed superintendents of 21.5%. Hispanic Ethnicity and
91 Age (55 and over) respectively were statistically significant variable s in the two models. Models for the Symbolic Frame proved to be statistically significant for Appointed superintendents ( p = .001) The independent variables Hispanic ethnicity and District Siz e 2 (300-2999) respectively were found to be statistically significant in the Symbolic Frame models. District Size contribut ed negatively to the model. Ch apter 5 discusses the findings and presents conclusions and recommendations for further study.
92 CHAPTER 5 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Summary Chapter 5 summarizes the study, discusses the findings, draws conclus ions, and presents recommendations for further investigation. This study investigated the cognitive leadership frame preferences as defined by the Bolman and Deal (1984 ) four frame typology, of school district superintendents in the states of Alabama, Fl orida, and Mississippi. These four leadership orientation frames (Structural, Human Resource, Political and Symbolic) were the primary focus of this study of appointed and elected school district superintendents and thei r method of selection. Also examin ed were the effects of demographic factors and leader characteristics on superintendent cognitive frame use as well as the similarities and differences between appointed and elected school district superintendent leadership frame preferences. Four researc h questions framed and structured this study: What are the self -reported leadership orientation frame preferences of the elected and appointed school district superintendents in Alabama, Florida, and Mississippi? Is there a difference between self -reported leadership style preferences of elected and appointed school district superintendents? Is there a difference in single or multi -frame preferences between elected and appointed school district superintendents? Is there a difference in leadership style pref erences of elected and appointed school district superintendents when considering demographic and contextual variables (i.e., age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district)? Chapter 1 established the purpose, boundaries, and significance of the study. Chapter 2 presented a review of the literature on the superintendency, appointed and
93 elected superintend ents, frames, and frame studies relevant to the current investigation. Chapter 3 presented an overview of the design and methodology of the study including information on the population of the study, the Leadership Orientation (Self) Survey, and the stati stical procedures used to analyze the quantitative data. Chapter 4 reported t he results of the data that were gathered and analyzed. The total population of school district superintendents in the states of Alabama, Florida, and Mississippi were surveyed using the LO(S)S. Surveys were sent to 131 Alabama superintendents (39 elected, 92 appointed); 67 Florida superintendents (44 elected, 23 appointed); and 150 Mississippi superintendents (65 elected, 85 appointed). A total of 348 surveys designed to collec t data on frame use and demographic information were distributed by United Stated Postal Service as well as electronically and resulted in a return of 145 surveys (41.7%). The participation rate by state was: Alabama (38.3%), Florida (44.7%), and Mississ ippi (42.1%). The participation rate included 89 appointed superintendents (44.5%) and 56 elected superintendents (37.8%). Discussion This section of Chapter 5 addresses the findings from analysis of descriptive data relative to the participants demograp hic characteristics and use of cognitive frames, as well as their use of single and multiple frames. This is followed by a discussion of the findings from analysis of variance and multiple regression analyses. Demographics Demographic data were obtained t hrough participant responses to the Demographic Addendum portion included by the researcher with the LO(S)S. Data indicated that most of the superintendents were 55 or older (63.4%), male (77.2%), and Caucasian (82.8%). This corresponds closely with data reported by the American
94 Association of School Administrators (AASA) which estimated the mean age of superintendents to be 54.6 years and the estimated median age to be 55 years (Glass & Franceschini, 2007). The AASA study reported on the gr aying of the superintendency because conditions such as health and finances tend to keep individuals in the superintendency longer than previously. Glass (2000) reported that most superintendents typically retire at age 57 or 58 after having spent 17 years in the su perintendency. He stated that t he mean age of superintendents may be expected to increase as individuals enter the superintendency later in their careers; however, the number of years spent in the superintendency may decline as many school districts encourage retirement at age 60 (Glass & Franceschini, 2007). Gender is another characteristic which is changing the face of the superintendency. The percentage of women in the superintendency has increased to 21.7%. Greater numbers of females are becoming sup erintendents as the number of female principals and central office administrators increase (Glass & Franceschini, 2007). Data reported in the current study indicated that 22.8% of the participating superintendents were female. Glass and Franceschini (2007) also reported that although the percentage of females is ris ing the number of minority superintendents is only 6%. Responses to the current study indicated that 17% of the participating superintendents were minorities, and 84% of these were African A merican. It is not surprising that the number of minority superintendents in these three states is higher than average as the US Census Bureau reports that each of these states has more than the US average 12.8% population of African Americans. Table 5 1 shows that Florida has nearly 16% African American population, Alabama has more than double the US
95 average of African Americans, while Mississippis African American population is almost triple that of the US average. Table 5 1. African American population percentage (N = 2008 population) State N % African Americans Al abama 4,661,900 26.4 Florida 18,328,340 15.9 Mississippi 2,938,618 37.2 United States 304,059,724 12.8 Data related to highest earned academic degree indicated th at nearly half (49.7%) of the superintendents in the study had earned a doctorate; another 19.3% held a specialist degree, while the remaining 31% held masters degrees. (Table 52) No s uperintendents reported less than a masters degree. Further statist ical study revealed that of the 33 female superintendents responding, 22 (67%) held doctorates with an additional four (12%) h olding specialist degrees. Forty eight percent (48%) of the minority superintendents held doctorates, and 32% had earned special ist degrees. To summarize this data, 78% of the female superintendents and 80% of minority superintendents had earned advanced degrees beyond the masters level. All (100%) of the African American females had advanced degrees ; 6 (75%) of the 8 female par ticipants held doctoral degrees and the other two (25%) had earned specialist degrees. In the AASA mid -decade study (2007) 51% of the pa rticipating superintendents had earned doctoral degrees. Fifty eight p ercent (58%) of the female superintendents held doctoral degrees. There was no reported breakdown as to the number or percentage of minority superintendents having advanced degrees. Perhaps t he reason for women and minorities having earned advanced degrees may be related to the fact that the superintendency is dominated by white males (Bjork & Rodgers, 1999; Glass,
96 Table 5 2. Participating s uperintendents by g ender, e thnicity and a cademic d egree (N = 1 45) Ethnicity Highest Degree Female Male Total African American Masters 0 3 3 Sp ecialist 2 5 7 Doctorate 6 5 11 Total 8 13 21 Caucasian Masters 7 34 41 Specialist 2 18 20 Doctorate 16 43 59 Total 25 95 120 Hispanic Masters 1 1 Specialist 1 1 Doctorate 1 1 Total 3 3 Other Doctorate 1 1 Total 1 1 Bjork, & Brunner, 2000). Glass, Bjork, & Brunner (2000) reported that 86.8% of AASA superintendents were male, and 94.9% were Caucasian Seven years later Glass and Franceschini ( 2007) reported that 78.3% of the responding AASA superintendents were male, and 93.8 % were Caucasian. It takes women and minorities more years in education to achieve the office of superintendent partly due to the lack of a viable network already in the superintendency. These individuals take more time approaching the superintendency, jumping through the educational hoops (i.e., assistant principal, principal, central office administrator). Earning an advanced degree could increase the credibility of female and minority superintendent candidates, for either appointment or election. Alabama, Florida, and Mississippi are the only US states that continue to have both appointed and elected superintendents. Georgi a, South Carolina, and Tennessee
97 previously relinquished their elected s uperintendent option in favor of an appointed only system of superintendents (Schuh & Herrington, 1990). At the time of this study only 148 elected superintendents remain ed in the Uni ted States (Alabama, Florida, and Mississippi). Superintendents who chose to participate in this stud y represented 89 (44.5 % ) of the appointed superintendents and 56 (37.8%) of the elected superintendents in these states Table 5 3. Appointed and e lecte d s uperintendents by s tate State Appointed Elected Total Alabama 92 39 131 Florida 23 44 67 Mississippi 85 65 150 Total 200 148 348 Superintendent F ram e U se This study investigated the use of the Bolman and Deal ( 1997) four leadership frames. Use of t hese frames is based on the leaders experience (Bolman & Deal, 1991a) and must be adjustable in order to match the needs of the situation if leaders are to be successful. Situations viewed through the wrong lens may be handled inappropriately. The Structural Frame is focused on the rational. It emphasizes rules, policies, and procedures and overlooks the humanistic side of any organization. The Human Resource Frame is more focused on the needs of the employees and aligning them with the needs of the organization. The emphasis is on skill development and relationships. The Political Frame sees organizations as competitive arenas where one must compete for scarce resources. Competition, negotiation, and allianc es a re seen as necessary for s urvival of the group.
98 The Symbolic Frame focuses on the organization al culture, using symbols, ceremonies, and stories to create meaning within the organization. In this study the researcher examined the self -reported frame pre ferences of school superintendents. The pattern of frame use that was revealed was similar to the findings of other leadership studies involving school level principals (Cote, 1999; Hodge, 2003; Rivers, 1996; Suzuki, 1994). In each of these instances it was discovered that the most frequently used frame was Human Resource, followed by Structural, Symbolic, and Political. Durocher (1995) also found the Human Resource Frame as the most frequently used, but indicated that the Political and Symbolic Frames w ere used more frequently than the Structural. Studies of college presidents (Bensimon, 1 987 ; Welch 2002) showed that leaders at this level preferred the use of the Structural and Human Resource frames. Echols -Tobe ( 1999 ) as well as Cantu (1997 ) in a stu dy of college/university deans found that the Human Resource Frame was the most frequently used with the Symbolic Frame being used the least. In other studies of superintendent frame usage (Faverty, 1997; Harlow, 1994; Strickland, 1992; Ward, 2006), the H uman Resource Frame was used most frequently. The least used frame s were Political and Symbolic with Structural Frame commonly reported as the second preference. Primary use of the Human Resource Frame may be related to the perception that schools are si milar to extended families. Superintendents may be working to shape school districts into organizations that are able to meet the needs of the community (students, employees, parents). E ven though such action is not a component of the Political Frame, it may be in some instances, part of a political agenda.
99 The findings of Bolman and Deal (1991a) showed that frame orientations are related to success as manager and leader. Management and Leadership are not viewed the same; therefore, the related frames are also different. The Human Resource and Structural Frame w ere found to be good predictor s of management effectiveness while leadership effectiveness was related to the Symbolic and Political Frames. In the current study the majority of superintendents participating reported their management and leadership effectiveness as above average. This effectiveness rating is not supported by their choice of frame usage. Use of the Political frame appears to be a better predictor of management and leadership effectiveness than the use of the Human Resource F rame; however, in this study Human Resource was used most frequently and Political Frame was least used. Bolman and Deal (1992a) reported that the ability to use multiple frames has a positive effect on t he success of managers and leaders. They hypothesized that leadership is based on context with different situations requiring different thinking patterns. The results suggested that having the ability to employ multiple frames is critical to success as l eader and manager. In this study 41% of the participating superintendents indicated no frame preference. Chi -square procedure indicated a statistically significant difference between appointed and elected superintendents ( p = .037) with appointed superi ntendents (47%) more likely to indicate no preference than elected superintendents (18%) Thirty -five percent (35%) of the superintendents indicated a preference for a single frame, and only 15% indicated a preference for
100 multiple frames. Studies available in the literature were mixed as to multiple frame use results Multiple R egression A nalysis Eight regression models were tested to investigate the relationship of demographic variables (age, gender, ethnicity, highest academic degree earned, experience in education, length of experience as a superintendent, length of tenure as superintendent in current district, and size of district) with each of the four frames while controlling for method of selection. Structural, Political, and Symbolic frame use by appointed superintendents were all found to have statistically significant R2 and F test results. This suggest s that, when taken together, the demographic variables significantly influenced appointed superintendents frame use. With regard to the demograp hic characteristics, years in education, years as superintendent, years as superintendent in current district were combined into one variable (years as superintendent). This characteristic was the only one that did not prove statistically significant in any of the models. One possible reason for this is that it could have been too closely related to age. Age was divided into categories relative to available responses on the Leadership Orientation (Self) Survey Age (55 and over) was determined to be st at istically significant in the R egression model for Political Frame-Elected. Perhaps as elected superintendents age, they become more politically minded. Gender was found to be significant only in the Regression model for Structural Appointed. A statistica lly significant R2 of .266 was found in this model (p = .05). Gender was determined to contribute 29.5% to the variance in this model. This is contrary to the bulk of the literature which shows little or no differences in leadership
101 orientation frame pre ferences in regards to gender (Bolman & Deal, 1991a, 1991b; Burks, 1992; Clisbee, 1994; DeFrank -Cole, 2003; Guidry, 2007; Hodge, 2003; Johnson, 1995; Martinez, 1996; Miro, 1993; Rivers, 1996; Thompson, 2000; Ward, 2006). Age, ethnicity and education have n ot been found to be significant factors influencing leadership orientation frame use (Durocher, 1995; Kelly, 1997; Martinez, 1996; Salley et al, 1979; Ward, 2006). Since the superintendency is a white, male dominated profession, study populations may not have yielded enough racial diversity for ethnicity to have been a factor. Ethnicity 1 (African American) was found to be statistically significant in both appointed and elected regression models for the Structural Frame, contributing over 30% to the varianc e in each model. E thnicity 3 (Hispanic) was also found to be significant in the Political Frame-Appointed and Symbolic Frame-Appointed, contributing almost 49% to the variance in the Political model and 50% in the Symbolic model. The reason for the influence of these ethnicities may be a factor of their culture. African American culture has been traditionally matriarchal based on a strict enforcement of rules (McAdoo, 1997; Johnson & Staples, 2005) The contribution to the Political Frame model could be similar to the reason that so many of the minority and female superintendents in this study have attained higher academic degrees. As minorities their opportunities have traditionally been fewer than their white counterparts. Adopting a political orient ation, competing for scarce resources, may be a way of securing their positions. Attainment of a doctoral degree proved to be a statistically significant factor in the Regression model for Human Resource Elected. As there are fewer and fewer elected
102 super intendents, it is possible that achieving an advanced degree could increase ones appeal to the electorate, giving an individual more status and credibility in a race for the office of superintendent Size of District proved to be statistically significant in both of the Symbolic Frame regressions, with DistSize1 (less than 300 FTEs) proving significant in the appointed model (Beta = .361); and DistSize2 (300 2999 FTEs) showing significance in the elected model (Beta = -.424). The superintendency in a small school district would lend itself more easily to a Symbolic leadership frame as individuals living in a small community are more likely to share the same values. Building on these values in the school system could prove an easier task than attempting this in a larger district. The large negative effect of school district size in the second Symbolic Regression should be interpreted with caution because the test was not statistically significant It would be interesting to ascertain the cause of the negative aspect of this variable. It is possib le that factors other than those included in the regression models (e.g., bureaucratic constraints, community expectations, and cultural differences) may account for the differences in frame use among superint endents. Conclusions Several conclusions may be made from this study First, for the most part, the demographic characteristics included in this study mirror ed the demographics in the larger populations in the states of Alabama, Florida, and Mississippi The pattern of frame use by the superintendents both appointed and elected is similar (Human Resource, Structural, Symbolic, and Political ). The only statistically significant difference between appointed and elected superintendents regarding frame use was that elected superintendent s tended to use the Human Resource Frame significantly
103 more than appointed superintendents. This is consistent with prior findings regarding frame use of college level, school, and school district leaders. T he Bolman and Deal (1997) four -frame typology appears to be useful in providing information explaining the cognitive frame use of school district superintendents. It may also be concluded that the school district superintendents in Alabama, Florida, and Mississippi are effective managers. Their preferred use of the Human Resource and Structural Frames, predictors of managerial effectiveness (Bolman & Deal, 1991a) provides data for that conclusion. Based on current data, it cannot be concluded that these same superint endents are effective leaders based on their leadership orientation frame preferences. Several of the demographic characteristics (method of selection, gender, age, ethnicity, highest academic degree, and district size) proved to be statistically significant r el ated to superintendent frame use; however, there appear to be other factors not included in this study which also influence the superintendents frame preferences. Implications of the Findings The quantitative findings may be generalized to the population of superintendents in Alabama, Florida, and Mississippi. These findings have implications for the professional development of school district superintendents. The results of this study may serve as a basis for superintendents to reflect on their pr actice as school district leaders and the importance of developing their understanding and use of the four frames regardless of their manner of selection to the superintendency. As baby boomers age and enter retirement the need for individuals to assume the responsibilities as school district superintendent becomes increasingly urgent F urther study of the use of the four frames is needed to prepare school district leaders with the
104 skills to develop and use multiple frames, lenses through which effective leadership can be achieved. Results of the study reinforce the need for further development in superintendents awareness and use of the Political and Symbolic Frames. Given the current economic climate and budget cuts and shortfalls, training in the use of the Political Frame could enhance the abilities of superintendents to acquire scare resources through the skills of negotiation and compromise. U se of the Symbolic Frame could enhance renewal of schools as a symbol of American culture, a safe haven fo r children, and a trusted resource for parents and community. Recommendations for Further Study This study was exploratory and revealed information that has not been available from a single source, especially regarding the similarities and differences bet ween appointed and elected superintendents. This was a study of superintendents self reported responses regarding frame use and leadership and management effectiveness. Further studies using other Bolman and Deal surveys that ascertain information about superintendents leadership styles and effectiveness from their peers and colleagues might provide more insight into the way that these leaders frame and solve problems Such stud ies could reveal more comprehensive data about the actual functioning of su perintendents as leaders and managers Qualitative studies that include interviews with superintendents and other school district employees might shed light on other characteristics, personal and otherwise, that affect how leaders function and their effect iveness. Such data could be used in developing and implementing superintendent education and training programs.
105 Conducting this study shined light on the need for a comprehensive, centralized professional database of information regarding the performance of superintendents. The AASA currently rel ies on surveys of participating members and lack extensive performance data regarding superintendents across the nation and in other parts of the world. Current efforts to create effective schools and school syst ems would benefit from centralization of data related to the effective performance of school district superintendents.
106 APPENDIX A REQUEST AND PERMISSI ON TO USE SURVEY INSTRUMENT
107 From: LANDRY,CHERI [ mailto:firstname.lastname@example.org] Sent: Wednesday, November 15, 2006 11:22 AM To: email@example.com Subject: Permission to use LO(S)S Dear Dr. Bolman, I am a University of Florida Educational Administration and Policy doctoral candidate. In order to complete my program of study, I am conducting a research investigation, the purpose of which is to examine the leadership orientation preferences (as developed by Bolman & Deal) of elected and appointed school district superintendents in Alabama, Florida, and Mississippi. Demographic characteristics of these superintendents, including age, gender, race, education, method of selection, years of experience in education, years of experience as a superintendent, size of district (number of students), and length of tenure in district will also be analyzed. I would like to use the Leadership Orientation (Self) Survey developed by you and Dr. Deal. The purpose of this communication is to secure your permission to use this instrument in collecting data for my study. This is an auspicious opportunity for me, in that my study will encompass all 148 elected superintendents remaining in the country. The results of this study could have implications for school districts throughout the U nited States that are considering changing from one type of superintendent selection process to another. It will further provide information regarding the influence of specific demographic characteristics on the leadership orientation preferences of s chool district leaders. I am more than happy to provide you and Dr. Deal with a copy of the results of this study, including any papers or publications that are based in whole or in part on the survey. I will also provide you, upon request, with a copy of my data file. I will be writing to the superintendent associations in these three states to apprise them of the research study and to secure their endorsement. If you have questions, comments, or suggestions please feel free to contact me by email, telephone, or letter correspondence (email is best for me). I would very much appreciate your permission to use this instrument, as I would like to begin the survey as soon as possible. If you like, you can mail and/or fax a letter of permission should you grant it. With deep appreciation, Cheri Landry 11921 NW 8th Road Gainesville, FL 32606 (352)3313618 (home and fax)
109 APPENDIX B SURVEY LETTER, ENDOR SEMENT, AND FOLLOW -UP
110 August 30, 2007 Dear Superintendent S everal weeks ago I sent a request for you to participate in a research study of the leadership styles of elected and appointed school district superintendents. This study has been approved by the Institutional Review Board at the University of Florida. I ncluded in the mailing were a letter of endorsement from Paul Houston of the American Association of School Administrators and a copy of the Leadership Orientations (Self) Survey created by Drs. Bolman and Deal. This survey was sent to all superintendents in Alabama, Florida, and Mississippi, as these are the only remaining states to have elected and appointed superintendents. I have received a fair response, but would like to make this study even stronger by having more participation. The survey is brie f (about 15 minutes) and is on line at https://www.education.ufl.edu:8443/survey/entry.jsp?id=1180833211284. The survey code is the first letter of your state and your di strict number. The password is super. You should be able to click on this link to connect to the survey; otherwise please copy and paste the link in your browser window. Please reconsider your decision not to participate and complete the survey on -line, or if you still have it, on the paper copy. If you have any questions, please do not hesitate to email me at firstname.lastname@example.org or my committee chair James Doud at email@example.com. Your input is extremely important and your participation would be greatly appreciated. Sincerely -Cheri Landry PhD Candidate Dept. of Administration and Policy College of Education University of Florida
111 APPENDIX C LEADERSHIP ORIENTATI ON (SELF) SURVEY LEADERSHIP ORIENTATIONS (SELF) 1990, Lee G. Bolman and Terrence E. Deal, all rights reserved This questionnaire asks you to describe your leadership and management style. I. Behaviors You are asked to indicate how often each of the items below is true of you. Please use the following scale in answering each item. 1 2 3 4 5 Never Occasionally Sometimes Often Always So, you would answer '1' for an item that is never true of you, '2' for one that is occasionally true, '3' for one that is sometimes true of you, and so on. Be discriminating! Your results will be more helpful if you think about each item and distinguish the things tha t you really do all the time from the things that you do seldom or never. 1. _____ Think very clearly and logically. 2. _____ Show high levels of support and concern for others. 3. _____ Have exceptional ability to mobilize people and resources to get things done. 4. _____ Inspire others to do their best. 5. _____ Strongly emphasize careful planning and clear time lines. 6. _____ Build trust through open and collaborative relationships. 7. _____ Am a very skillful and shrewd negotiator. 8. _____ Am highly c harismatic. 9. _____ Approach problems through logical analysis and careful thinking
112 10. _____ Show high sensitivity and concern for others' needs and feelings. 11. _____ Am unusually persuasive and influential. 12. _____ Am able to be an inspiration to others. 13. _____ Develop and implement clear, logical policies and procedures. 14. _____ Foster high levels of participation and involvement in decisions. 15. _____ Anticipate and deal adroitly with organizational conflict. 16. _____ Am highly imaginative and creative. 17. _____ Approach problems with facts and logic. 18. _____ Am consistently helpful and responsive to others. 19. _____ Am very effective in getting support from people with influence and power. 20. _____ Communicate a strong and challenging sense of vision and mission. 21. _____ Set specific, measurable goals and hold people accountable for results. 22. _____ Listen well and am unusually receptive to other people's ideas and input. 23. _____ Am politically very sensitive and skillful. 24. _____ See beyond current realities to generate exciting new opportunities. 25. _____ Have extraordinary attention to detail. 26. _____ Give personal recognition for work well done. 27. _____ Develop alliances to build a strong base of support. 28. _____ Gener ate loyalty and enthusiasm. 29. _____ Strongly believe in clear structure and a chain of command. 30. _____ Am a highly participative manager. 31. _____ Succeed in the face of conflict and opposition. 32. _____ Serve as an influential model of organizational aspirations and values.
113 II. Leadership Style This section asks you to describe your leadership style. For each item, give the number "4" to the phrase that best describes you, "3" to the item that is next best, and on down to "1" for the item that is least like you. 1. My strongest skills are: _____ a. Analytic skills _____ b. Interpersonal skills _____ c. Political skills _____ d. Ability to excite and motivate 2. The best way to describe me is: _____ a. Technical expert _____ b. Good listener _____ c. Skilled negotiator _____ d. Inspirational leader 3. What has helped me the most to be successful is my ability to: _____ a. Make good decisions _____ b. Coach and develop people _____ c. Build strong alliances and a power base _____ d. Energize and inspire others 4. What people are most likely to notice about me is my: _____ a. Attention to detail _____ b. Concern for people
114 _____ c. Ability to succeed, in the face of conflict and opposition _____ d. Charisma. 5. My most important leadership trait is: _____ a. Clear, logical thinking _____ b. Caring and support for others _____ c. Toughness and aggressiveness _____ d. Imagination and creativity 6. I am best described as: _____ a. An analyst _____ b. A humanist _____ c. A politician _____ d. A visionary III. Overall rating Compared to other individuals that you have known with comparable levels of experience and responsibility, how would you rate yourself on: 1. Overall effectiveness as a manager 1 2 3 4 5 Bottom 20% Middle 20% Top 20% 2. Overall effectiveness as a leader 1 2 3 4 5 Bottom 20% Middle 20% Top 20%
115 IV. Demographic Addendum 1. Gender: _____Female _____Male 2. Age: _____2534 _____3544 _____4554 _____55and over 3. Ethnicity: _____African American _____Caucasian _____Hispanic _____O ther 4. Highest Academic Degree Earned: _____Masters _____Specialist _____Doctorate 5. Method of Selection to Superintendency: _____Appointed _____Elected 6. Total Number of Complete Years as a Superintendent: _____Less than 1 _____15 _____615 _____16+ 7. Total Number of Complete Years as Superintendent in Current District: _____Less than 1 _____15 _____615 _____16+ 8. Total Number of Complete Years in Education: _____15 _____615 _____1620 _____21+ 9. Number of Full Time Equivalent Students Enrolled in District: __________Students
116 APPENDIX D IRB FORMS
119 LIST OF REFERENCES American Association of School Administrators (1962) On selecting a superintendent of schools. Washington, DC: AASA. Barnett, M. (1982). Perceptions of the leadership style effectiveness of superintendents in Mississippi. Paper presented at the Annual Meeting of the MidSouth Educational Research Association. New Orleans, LA. Bensimon, E. (1987). The meaning of good presidential leadership: A frame anal ysis. Paper presented at the National Meeting of the Association for the Study of Higher Education, Baltimore, MD. Bensimon, E., Neumann, A., & Birnbaum, R. (1998). Making sense of administrative leadership: The L word in higher education. ASHE ERIC Higher Education Report No 1, Washington, DC: School of Education and Human Development, The George Washington University. Bjork, L., & Rodgers, L. (1999). Opportunity in crisis: Women and people of color in the superintendency. Advancing Women in Leadership Journal. Retrieved on March12, 2001 from http://advancingwomen.com/awl/awl_wordpress/just -do it women -superintendents -speak -to aspiringwomen2/ Bolman, L., & Bloch, H. (1999). Versatile leadership: A comparative analysis of reframing in Venezuelan managers. Paper presented at World Wide Bilingual Conference, Madrid. Bolman, L., & Deal, T. (1984). Modern approaches to understanding and managing organizations San Francisco: Jossey -Bass. Bolman, L., & Deal, T. (1990). Leadership Orientation (Self) Sur vey. Brookline, MA: Leadership Frameworks. Bolman, L., & Deal, T. (1991a). Images of Leadership. Nashville, TN: National Center for Educational Leadership. Bolman, L., & Deal, T. (1991b). Leadership and management effectiveness: A multi frame, multi -sec tor analysis. Human Resource Management 30 509 -534. Bolman, L., & Deal, T. (1991c). Reframing organizations: Artistry, choice and leadership. San Francisco: Jossey -Bass. Bolman, L., & Deal, T. (1992a). Leading and managing: Effects of context, culture, and gender. Educational Administration Quarterly, 28,314329. Bolman, L., & Deal, T. (1992b). What makes a team work? Organizational Dynamics, 21, 34 -44.
120 Bolman, L., & Deal, T. (1994a ). Leading with soul: An uncommon journey of the spirit San Francisc o: J ossey -Bass. Bolman, L., & Deal, T. (1994b ). Looking for leadership: Another search partys report. Educational Administrative Quarterly, 30,77 96. Bolman, L., & Deal, T. (1997). Reframing organizations: Artistry, choice and leadership (2nd Ed. ) San Francisco: Jossey -Bass. Bolman, L., & Deal, T. (200 4 ). Reliability of leadership orientations Retrieved August 12, 2004 from http://www.leebolman.com/orientations.htm Bolman, L., & Deal, T. (2008). Refr aming Organizations: Artistry, c hoice, and l ead ership. (4th Ed.) San Francisco: Jossey -Bass. Brand, C. (1993). An investigation of the appointed versus elected superintendents in the state of Georgia. (Doctoral dissertation, University of TennesseeKnoxville, 1993). Dissertation Abstracts Internatio nal, 55 (03)0428 -A. Brandt, R. (1993). New roles, new relationships: On restructuring roles and relations: A conversation with Phil Schlecty. Educational Leadership 51 (2). Bridges, E. (1982). Research on the school administrator: The state of the art, 19 671980. Educational Administration Quarterly, 18 ,12 -33. Burks, T. (1992). The use of organizational frames in leadership development. (Doctoral dissertation, Peabody College for Teachers of Vanderbilt University, 1992). Dissertation Abstracts Internat ional, 53 (05)1413A. Cantu D. (1997) The leadership frames of academic deans randomly selected and nominated as exceptionally effective public colleges and universities. (Doctoral dissertation, Arkansas State University) AAT9724134. Carlson, R. (1996). R eframing and reform: Perspectives on organization, leadership and school change. New York: Longman. Carter, G., & Cunningham, W. (1997). The American school superintendent: Leading in an age of pressure. San Francisco: Jossey -Bass. Chapman, C. (1997). Becoming a superintendent: Challenges of school district leadership. Upper Saddle River, NJ: Prentice Hall. Childress, G. (1994). An exploratory study of leadership orientation frames of United States -based Japanese and American leaders in the automotive manufacturing industry. (Doctoral dissertation, Peabody College for Teachers of Vanderbilt University, 1994). Dissertation Abstracts International, 56(04)1428A.
121 Clisbee, M. (2004). Leadership style: Do male and female superintendents lead different ly? (Doctoral dissertation, University of Massachusetts -Lowell, 2004). Dissertation Abstracts International, 65 (02)359. (UMI No. 3122967). Cote, AnnaMarie Bastoni (1999) Leadership orientation frames of Florida elementary principals in relationship to s chool context and principal characteristics. Ed.D. dissertation, University of Florida, United States -Florida. Retrieved February 1, 2009, from Dissertations & Theses @ University of Florida FCLA database. (Publication No. AAT 9956561). Creasey, E. (2 002). An exploration of educational change and a leadership framework for change. Unpublished doctoral dissertation, University of Calgary. Cuban, L. (1998) Unchanging public opinion. The American School Board Journal l85 (5) p. 4243. Davis L. (1996) A case study of the relationship of race, gender and social class on the leadership styles of three African-American female urban elementary school principals. (Doctoral dissertation, Temple University 1996) Publication No. AAT 9623754. Deal, T., & Pete rson, K. (1994). The leadership paradox: Balancing logic and artistry in schools San Francisco: J o ssey Bass. DeF rank -Cole, L. (2003). An exploration of the differences in female and male self perceptions of presidential leadership styles at colleges and universities in West Virginia. (Doctoral dissertation, University of Pittsburgh, 2003) AAT3117728. Doyle, M., & Smith, M. (2001). Classical leadership. The Encyclopedia of Informal Education. Retrieved on March 12, 2001 from http://www.infed.org/leadership/traditional_leadership.htm Durocher, E. (1995). Leadership orientations of school administrators: A survey of nationally recognized school leaders. (Doctoral dissertation, Columbia University Teachers College, 1995). Dissertation Abstracts International, 57(02)525A. Echols -Tobe, D. (2000). The development of cognitive leadership frames among African American female college presidents. (Doctoral dissertation, Columbia University Teachers College, 2000). Dissertation Abstracts International, 60(07)2300. Edmunds, M (2008) The female superintendency: Reframing our understanding of women's leadership styles and behaviors. Ed.D. dissertation, Seton Hall University, United States -New Jersey. Retrieved January 31, 2009, from Dissertations & Theses: Full Text database. (Publication No. AAT 3319297).
122 Englert, M (2008) Leadership orientations of rural community college presidents serving appointed or elected independent gov erning boards: A four -frame analysis. Ph.D. dissertation, Colorado State University, United States -Colorado. Retrieved January 31, 2009, from Dissertations & Theses: Full Text database. (Publication No. AAT 3321274). Faverty, P. (1997). A causal -compar ative study of the effects of the ACSA Small School District Superintendents Academy of the leadership orientations of California superintendents. (Doctoral dissertation, University of LaVerne). Dissertation Abstracts International, 58 (11)4136A. Flak, T (1998). An inquiry into the leadership orientations of selected school superintendents Unpublished doctoral dissertation, University of Maryland. Gardner, J. (2000). The nature of leadership. The Jossey -Bass Reader on Educational Leadership. San Francisco: Jossey Bass. Georgia Governors Education Review Commission, Atlanta. (1994). Priority for a quality basic education: A first report to Governor Joe Frank Harris and the Georgia General Assembly. (ERIC Document Reproduction Service No. ED253624 ). Gilson, J. (1994). A leadership study: Public higher education administrators in Missouri compared with the Bolman-Deal database on multi -frame, multi -sector leadership. Dissertation Abstracts International, 55(05)1199A. Glass, T., Bjork, L., & Brunn er, C. (2000) The Study of the American School Superintendency 2000: A look at the superintendent of education in the new millennium. Arlington: American Association of School Administrators. Glass, T. & Franceschini, L. (2007). The State of the America n School Superintendency: A mid-decade study Lanham, Md.: Rowman and Littlefield. Guidry, T (2007) Women deans' perceptions of their leadership styles: A study based on Bolman and Deal's four frame theory. Ed.D. dissertation, East Carolina University United States -North Carolina. Retrieved January 31, 2009, from Dissertations & Theses: Full Text database (Publication No. AAT 3255414). Harlow, J. (1994). Educational leadership: A frame analysis. Dissertation Abstracts International, 55(8)2227A. (UMI No. 9427416). Hersey, P. & Blanchard, K. (1977). Management of organizational behavior. Englewood, NJ: Prentice Hall. Hodge, E. (2003) School leadership in Belize: The interrelationships of context, cognitive frames, and leader characteristics. Ph.D. dissertation, University of Florida, United States -Florida. Retrieved February 1, 2009, from Dissertations & Theses @ University of Florida FCLA database (Publication No. AAT 3105615).
123 Holloway, J. (2001). Evaluating educators: Research link Se tting standards for the school superintendent. Educational Leadership, (58) 5. Houston, P. (2001). Superintendents for the 21st century: Its not just a job, its a calling. Kappan Professional Journal (82) 428433. Hoyle, J. (1999). The Triumphant Sup erintendent. The School Administrator 8(59). Hoyle, J., Bjork, L., Collier, V., & Glass, T. (2005). The Superintendent as CEO. Thousand Oaks, CA: Corwin Press. Ireh, M., & Bailey, J. (1999). A study of superintendents change leadership styles using t he situational leadership model. American Secondary Education 27(4)22-32. Johnson, A. (1995). Frames and effective leadership: A study of the use of organizational frames by elementary principals. (Doctoral dissertation, University of LaVerne, 1995). D issertation Abstracts International, 56(7)2507A. Johnson, L., & Staples, R. (2005). Black families at the crossroads: Challenges and prospects San Francisco, CA: John Wiley % Sons, Inc. Jordan, C. (1999). Presidential leadership in three Southern wom ens colleges, 19961998: A frame analysis. Dissertation Abstracts International, 60(4)1041. y Kelly J. (1997). Leadership orientations of executives in business and industry and administrators in higher education. PhD. dissertation, University of Missou ri Columbia. Dissertations & Theses: Full Text database. (Publication No. AAT 9841309). Konnert, M., & Augenstein, J. (1995). The school superintendency: Leading education into the 21st century. Lancast er, PA: Technomic. Martinez, R. (1996). Self and admi nistrative designee perceptions of leadership orientations of elementary principals who have and have not completed the California School Leadership Academy. Dissertation Abstracts International, 57(8)3343A. McAdoo, H. (1997). Black families Thousand O aks: Sage Publications. McCartney -Infelise, R. (1999). A study of the leadership orientations of MAIS school directors and a comparison of their leadership orientations with those of small school district superintendents in California. Dissertation Abstr acts International, 60(05)1411. McGriff, T., Bishop, H., & Rice, M. (1997). Work related behaviors of elected and appointed local superintendents of education in the state of Alabama. Paper presented at the Annual Meeting of the MidSouth Educational Research Association (Memphis, TN).
124 Meade, R. (1992). The California School Leadership Academy: Its effect on the leadership orientations of California elementary school principals. Dissertation Abstracts International, 53(11)3761A. Miller Whitehead, M. (200 0 ). High performing Alabama school systems: What do the best have in common? (Eric Document Reproduction Service No. ED447209). Miro, A. (1993). A comparative study of leadership orientation frames between public high school principals serving in school s that applied and did not apply for California Senate Bill 1274 restructuring grants. Dissertation Abstracts International, 55(5)1158A. Morgan, G. (19 97). Images of organization (2nd Ed ). Newbury Park, CA: Sage. Moss, W. (2002). Superintendent leadershi p orientations and its relationship to school board satisfaction. Dissertation Abstracts International, 63 (12)4174. Norton, M., Webb, L., Dlugosh, L., & Sybouts, W. (1996). The school superintendency: New responsibilities, new leadership. Boston, MA: Al lyn and Bacon. Owens, R. (1991). Organizational behavior in education (4th Ed). Englewood Cliffs, NJ: Prentice-Hall. Parry, A., & Horton, M. (1998). Finding, not looking for leadership: A case study using the Bolman and Deal model. Journal of Leadership S tudies 5(4)122 132. Petersen, G., & Barnett, B. (2003). The superintendent as instructional leader: History, evolution and future of the role. Paper presented at the Annual Conference of the American Educational Research Association, Chicago IL. Redman M (1991). A comparative study of the leadership orientation frames of administrators in private Japanese and American institutions of higher education (Private institutions). (Doctoral dissertation, University of La Verne). Dissertation Abstracts Internat ional 57(07)2785A. Rivers, P. (1996). A frame analysis of principals leadership orientations. (Doctoral dissertation, University of Central Florida, 1996). Dissertation Abstracts International 57(7)2785A. Salley, C., McPherson, R., & Bachr, M. (1979). What principals do: A preliminary occupational analysis. In D. A. Erickson, & T. L. Reller (Eds.), The principal in metropolitan schools (pp.22 -39). Berkley, CA: McCutchan Publishing Corp. Schuh, J., & Herrington, C. (1990). Electing versus appointing d istrict superintendents of schools: An analysis of the issue in six states: Tennessee, Mississippi, Alabama, Georgia, Florida, & South Carolina. Research Park, NC: Southeastern Educational Improvement Lab.
125 Senge, P. (1990). The fifth discipline: The art a nd practice of the learning organization. New York: Currency Doubleday. Senge, P. (2000). Give me a lever long enoughand singlehanded I can move the world. The Jossey -Bass Reader on educational leadership. San Francisco: Jossey -Bass. Sharp, W., & Walte r, J. (2004). The school superintendent: The profession and the person (2nd Ed). Lanham, MD: Scarecrow Education. Snider, J. (2006). The superintendent as scapegoat. Education Week. Retrieved on January 11, 2006 from http://www.edweek.org/ew/articles/2006/01/11/18snider.h25.html?r ... Southern Regional Educational Board. (2000). Focus on school district superintendents. SREB Legislative Action. Retrieved on December 13, 2003 from http://www.sreb.org/scripts/focus/focus1.asp?code=12 Sumner, T. (1986). A comparison of principals percept ions of leadership behavior of elected and appointed school superintend ents in the state of Mississippi Dissertation Abstracts International, 47 (8)2840A. Strickland, J. (1992). Leadership perspective of Tennessee school leaders. Dissertation Abstracts International, 53(4)1018A. Suzuki, Y. (1994). A comparative study of the leadership orientation frames of California, Asian and other public school principals. Dissertation Abstracts International, 55(7)1780A. Tedesco, P. (2004). Leadership style comparison: Iowa community college presidents and superintendents and the ideal leadership style perceived by the board presidents of these organizations. (Doctoral dissertation, Colorado State University, 2004) (UMI No. 3143865). Thompson, M. (2000). Gender, leadership orientation, and effectiveness: Testing the theoretical models of Bolman & Deal and Quinn. Sex Roles: A Journal of Research. Waller, R., & Richardson, M. (1997). Georgia longterm superintendents: Perception and attitudes. Retrieved on March 13,2001 from www.edweek.org/ew/articles/2000/11/08/10glass.h20.htm Ward, D (2006) Dominant leadership frames of reference for active Indiana public school superintendents. Ph.D. dissertation, Indiana State University, United States -Indiana. Retrieved January 31, 2009, from Dissertations & Theses: Full Text database. (Public ation No. AAT 3251420).
126 Welch, C. (2002). Leadership frames of female presidents of American research universities. Unpublished doctoral dissertation, University of North Texas. Willower, D., & Fraser, H. (1980). School s uperintendents and their work. Admi nistrators Notebook 28 1 6 27. Wood, R. (20 01). Oklahoma student performance: No dramatic changes in the 1990s. Open their eyes that they may see. Retrieved on March 17, 2001 from http://www.ocpathink.org/research-ideas/education/?module=news&id=428 Zehr M. (2001). Indiana urged to switch to appointed state school chief. Education Week, 20(22)25.
127 BIOGRAPHICAL SKETCH Cheri was born in Jacksonville, Florida where she graduated from Landon High School. She then attended Florida State University where she earned a Bachelor of Science in English Education. Ten years later she graduated with a Master of Education from the University of North Florida in Jacksonville in 1977; and completed requirements for certification in Administration and Supervision in 1978. Cheri taught Language Arts (English, speech, reading, drama) as well as classes for the educationally mentally handicapped in various school systems in North and Central Florida until 1980 when she was promoted to assistant principal and then vice p rincipal in Duval County. Later she moved to Hamilton County where she served as high school principal and principal of the juvenile justice school located in Jasper, Florida. She also taught as an adjunct for North Florida Community College in Madison Florida. In 2001 after retiring from the public school system, Cheri moved to Gainesville, Florida to pursue her doctorate in educational leadership. She worked as an adjunct lecturer at the University of Florida and also served as an instructor in the Ed ucator Preparation Institute at Santa Fe College. Cheri hopes to pursue a career in teaching at the university level or serving as a central office administrator for public schools. She would also like to become involved in volunteer work that benefits ch ildren.