HULTIVARIATE ANALYSIS OF VARIANCE OF THE EFFECT OF SELECTED FACTORS ON THE EFFECTIVENESS 'OF LEADERSHIP STYLES OP TEACHING STAFF IN SOUTH AFRICAN UNIVERSITIES By $.54 Jeffrey Mkhudlwana'n e \.+“.glu'~ J ,. w (J4 A DISSERTATION Submitted to Miehi an State University in partial ful illment of the requirements for the degree of DOCTOR or entLosopm a”? 3 §M&%lm& 4 z 1 turn Department of Administration and Higher Education 1983 _ uflljnnmmmmml 1293107986071 )ViESI_J RETURNING MATERIALS: Place in book drop to L13RAR155 remove this checkout from __ your record. FINES W11] be charged if book is returned after the date stamped be10w. é/m/S’z .651, $1502 411ch/Zacte yflJ’UAA/ ’ I’ll/(rflbb (ILLG‘La/ xl‘rt-M d [1’ % g'u’wb“ 11/314} (,2 [EA/CK Ag’g Ala, %%’11/ t CCUVX /{,¢(L{,:[Z;v;?/b¢€77 fl4§ //L&// 4066/ ’ Qy’w’l’é ”'7 a fl/Z-a j flaw; / 7149061,- /(7{£’LL7,A' ”/hu4/auw/17 /7/5;L W 305,]; 67/ {2/47, 356 [(6 ABSTRACT A MULTIVARIATE ANALYSIS OF VARIANCE OF THE EFFECT OF SELECTED FACTORS ON THE EFFECTIVENESS OF LEADERSHIP STYLES OF TEACHING STAFF IN SOUTH AFRICAN UNIVERSITIES BY Jeffrey Mkhudlwana Beka Hadebe Basically, this study is an attempt to determine whether, and to what extent, population grouping (Blacks, Coloureds, Indians, and Whites), Faculty (Arts, Business, Education, Law, and Science), and academic position (junior and senior) have a statistically significant effect on the effectiveness of leadership styles of the teaching staff in South African universities as evaluated by students. This study was conducted on a random sample of 200 senior and junior lecturers in the Faculties of Arts, Business, Education, Law, and Science. The participants were selected proportionally from four types (Black, Coloured, Indian, and White) of universities in the Republic of South Africa. The collection of data was carried out by adminis-. tering the'Michigan Four-Factor Theory Questionnaire (Taylor and Bowers, 1972) on leadership styles to five students taught by each lecturer concerned. This questionnaire con- sists of twenty items, grouped into four subscales: support Jeffrey Mkhudlwana Beka Hadebe of members, interaction facilitation, goal emphasis, and work facilitation, which are functions of effective leader- ship styles. This tool measures the effectiveness of leader- ship style. A.high score on this instrument is associated with effective leadership style, a low one, with ineffective leader behavior. To test the null hypotheses in this study, Finn's (1978) Multivariance - Uhivariate and Multivariate Analysis of Variance and Covariance: A FORTRAN IV program.- was pre- pared and employed to analyze the data in a 4 §_5 g 2 factorial arrangement (design). Major Findings 1. Leadership styles of the teaching staff in White uni- versities are more effective than those of lecturers and senior lecturers in non-white universities. 2. Among nonawhites alone, leadership styles of the teaching \ staff in the Indian university are the most effective while those of the faculty in Black campuses are the least effective. The leadership styles of the faculty in the Coloured University are somewhat middling though slightly above average. 3. The teaching staff from the Faculties of Law and Arts have the most effective leadership styles whereas those from the Science Faculty display the least effective leader behavior. 4. Academic position has no statistically significant effect Jeffrey Mkhudlwana Beka Hadebe on the effectiveness of leadership styles in South African universities. 5. Leadership styles of faculty in all population groups are 'most effective in the support subscale and least effective in the interaction facilitation subtest. 6. In goal emphasis (subtest 3) the effect of population groups on the effectiveness of leadership styles of teaching staff interacts with the influence of Faculty, and vice versa. Future studies in this area may attempt to determine and explore possible factors associated with each of these findings . A.MULTIVARIATE ANALYSIS OF VARIANCE OF THE EFFECT OF SELECTED FACTORS ON THE EFFECTIVENESS OF LEADERSHIP STYLES OF TEACHING STAFF IN SOUTH AFRICAN UNIVERSITIES By Jeffrey Mkhudlwana Beka Hadebe A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Administration and Higher Education 1983 Copyright by Jeffrey Mkhudlwana Beka Hadebe 1983 ACKNOWLEDGEMENTS Great appreciation goes to Dr. Eldon R. Nonnamaker, the chairperson of my guidance committee, for assistance and support he gave me on this dissertation as well as throughout my career on this program. I also wish to acknow- ledge the other committee members: Drs. William H. Schmidt, the chairperson of the Educational and Counselling Department, for his statistical advice and for his assistance in helping wdth the data analyses, Larry W. Foster, a Professor of Business Management, for his expertise in the field of leader- ship, and Richard L. Featherstone, a Professor of Administra- tion and Higher Education, for his extensive knowledge in the field of college and university administration. Special thanks are due to Mr. Isaac Kubheka, the Principal of Vukusakhe High School, and his staff for their varied and substantial contributions to the crystallization and development of my interest in the area of leadership and administration. I am also very grateful to Messrs. John D. Coley and Jeremia D.L. Kumalo for their invaluable support during the middle phase of my educational career. Lastly, I acknowledge all other individuals and institutions whose contributions made this study a success. ii‘ DEDICATION Noma isezizweni, phesheya kwezilwandle;, lesisithukuthuku indodana yakho isethula kuwe baba, Levy Gandaganda: Mathumbentwala awabonwa Abonwa zingapheli, Potolozi kwabamhlophe abelungu. Lapho ulele khona esibomvini. Nawe mama, Lydia Heli, ntombi yese-Mbo e-Magogo, eyayincelisa -ngebele onObisi ldkukhuthala, ikubopha ngasibopho sinye. Bafowethu nodadewethu mukelani nakhu 6kuncane kwamandla avela kowakwenu. iii TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES . . . . . . . . . . . . . I. II. INTRODUCTION . . . . . Statement of the Problem . Need for the Study . Bmportance of the Study Basic Assumptions Limitations of the Study . Definition of Terms Organization of the Study REVIEW OF SELECTED LITERATURE Definition of Leadership . Theories of Leadership . Trait Theories . Physical Characteristics . Social Backgromd Intelligence and Ability . Other Abilities Personality . Task—Related Characteristics . Social Characteristics . Behavioral Approach iv' vii ix 11 12 13 14 15 15 l7 18 18 18 l9 19 20 20 21 21 III. IV. TABLE OF CONTENTS (Continued) Situational Theories . Focus on the Leader--Fiedler's Leader Match Theory . Focus on Task/Situation Vroom's Contingency Theory . Types of Management Decision Styles Focus on the Subordinates--Hersey and Blanchard's Situational Theory . Focus on Leader Behavior--Argyris's Model I and Model II . Summary DESIGN OF THE STUDY Hypothesis . Population . Sampling Procedures Instrumentation Collection of Data . Processing of Data . Analysis of the Data . Summary . . . . . . . DATA ANALYSIS AND INTERPRETATION . Introduction . Preliminary Findings . Relationship Among the Four Dependent Variables . . Level of Significance Analysis of Variance of Total Scores . General Inspection of Population Grouping Faculty Differential Effects . . . . 23 23 26 26 27 33 37 39 39 41 41 42 49 50 50 55 56 56 56 56 57 61 63 TABLE OF CONTENTS (Continued) Multiple Comparisons . . . . . . . . . . . . . . 65 Faculty . . . . . . . . . . . . . . . . . . 65 Population Groups . . . . . . . . . . . . . 67 Multivariate Analysis of Variance . . . . . 69 Profiles of Leadership Behavior . . . . . . 78 Summary . . . . . . . . . . . . . . . . . . 85 SUMMARY, CONCLUSION, AND RECOMMENDATIONS . . . . 86 Summary . . . . . . . . . . . . . . . . . . . . 87 Conclusion . . . . . . . . . . . . . . . . . . . 91 APPENDIX A . . . . . . . . . . . . . . . . . . . 95 APPENDIX B . . . . . . . . . . . . . . . . . . . 96 APPENDIX C . . . . . . . . . . . . . . . . . . . 98 REFERENCES . . . . . . . . . . . . . . . . . . . llO vi .10 .11 LIST OF TABLES Sample Distribution of Lecturers and Senior Lecturers Acdording to Population Grouping and Faculty . Indices, Their Definitions, and Constituent Items of the Michigan Four- Factor Theory Questionnaire . . . . . . . . . . . . . Indices and Their Internal consistency of the Michigan Four-Factor Theory Questionnaire . Discriminant Validity of the Indices of the Michigan Four-Factor Theory Questionnaire . Codebook Correlation Matrix of Four Dependent Variables . . . . . . . . . . Principal Components of the Correlation Matrix . . . . . . . . Variation Accounted for by the Eigenvalues Three-Way ANOVA of Total Scores . Means and SD of Various Faculties in Their Respective Population Groups . . . . . . Planned Orthogonal Comparisons Between Population Group Rating Means . . Multivariate Analysis of Variance . . Univariate Analysis of Variance . . . . Subtest 3 (Goal Emphasis) Means of Various Faculties in Their Respective Population Groups . . . . . . Descending Rank Order of Faculty Rating Means in Each Population Group . . . . . . . . . Univariate Analysis of Variance . . . . . . 44 45 46 48 51 58 59 60 62 64 68 7O 72 73 75 76 4.12 4.13 4.14 5. 1 LIST OF TABLES (Continued) Univariate Analysis of Variance . Faculties and Their Respective Rating Means in Each SUbscale . . . . . . . . . . . . . Population Groups and Their Rating Means in Each subscale . . . . . . . . . . Results of Hypothesis Tests . . . . . viii 79 81 83 88 LIST OF FIGURES Decision Process Flow Chart . Basic Leadership Behavior Style . . . . . What a Leader Behavior Should Do in Terms of Degree of subordinate Maturity . . . . . Graphic Representation of Various Faculty Means in Their Respective Population Groups . Graphic Representation of Subtest 3 (Goal Emphasis) Two-way (population by Faculty) Interaction . . . . . A Profile of Faculty Degree of Emphasis on Four Key Dimensions of Leadership Behavior A Profile of Each Population Group Degree of Emphasis on Four Key Dimensions of Leadership Behavior ix' 28 30 32 66 74 82 84 CHAPTER I INTRODUCTION The researcher's interest in leadership and higher education was initiated and further enhanced by a number of different public speeches delivered both in and outside the Republic of South Africa (R.S.A.). In some, the concept of leadership was expressed directly and, in others, indirectly. These were some of them. In his address at Umphumulo Theological School, Hadebe (1977) said, "Our leaders will dream dreams, their followers will see visions." This is consistent with Gardner's (1963) belief that a leader should be a person of visions. However, Rupert (1980) cautioned that a constructive leader ought to be: . . . agpractical dreamer and a realistic optimist. He should, therefore, Have the necessary skill or organizing which really implies that he must also be able to dele- gate efficiently, for he can only manage and lead as far as he can delegate, and again, the basis for all delegating of authority is trust. He must trust others to be trusted. Trust, like confidence, is contagious. But he must know'whom to trust and therefore selection also forms a basic part of efficient long-range organizing. According to King (1963), a leader with dreams and visions is the one who goes to the mountain top to see what 1 lies on the other side before taking followers along. Rupert (1980) also suggested that an effective leader ought to have a cause--something worthwhile--a Leitmotif-- a spirit of service--a sense of mission: A further requirement for a constructive leader is that he should have something worthwhile to strive for--a noble cause to which he is dedicated (and, one could say, the greater the cause the greater the man).* Delivering his speech at the University of Sheffield on June 25, 1946, John Masefield said: There are few earthly things more beautiful : than a University. It is a place where those who hate ignorance may strive to know: where those who perceive truth may strive to make others see: where seekers and learners alike, banded together in the search for knowledge, will honor thought in all its finer ways, will ‘welcome thinkers in distress or exile, will uphold ever the dignity of thought and learning and will exact standards in these things. It is only in the climate of responsible and construc- tive leadership that a university can be such a place. To express the importance of leadership in the future of a university, Gardner (1963) remarked: Now I believe--and will never cease to believe--that men can shape their institu— tions to suit their purposes--provided that they are clear as to what those purposes are: and provided that they are not too gravely afflicted with the diseases of which institutions die—-among them com- placency, myopia, and unwillingness to choose, and an unwillingness on the part of individuals to lend themselves to any worthy common purpose. * For Rupert's speech, see Appendix C. Borrowing some few lines from General Eisenhower, Rupert (1980) added: No leader can win men if he is guilty of arrogance, particularly if it is accompanied by ignorance. A leader must seek to under- stand, not merely to control as a martinet. If he is truly to understand his men he must have the quality of humility. This virtue is the main theme of the motto of the University of Alberta in Canada, which says, "Quaecumque vera, meaning "Whatsoever is true". This motto was adopted from.a verse in the Bible which reads: Finally, bretheran, whatsoever is true, (quaecumque vera), whatever is honorable, whatever is just, whatever is pure, what- ever is lovely, whatever is gracious, if there is any excellence, if there is any- thing worthy of praise, think about these things (haec cogitate), (Phillipians 4:8). This means that the authors of this motto saw the University working to achieve ”whatsoever things are true". "We are called upon to give leadership in discovering who we are and which direction we are to follow." Seeing himself and his staff also called upon to give effective leadership to the University of Durban- Westville, Greyling (1980) said, "Proper communication between University authorities and students is one immediate challenge facing us." Referring to leadership qualities in higher education, Krog (1981) commented: "It is, however, essential that a leader acts in a responsible manner at all times and not to exploit his position. . . ." In his address on leadership responsibility, the Rt. Reverend Bruce Evans (1976), Bishop of Port Elizabeth, remarked: The beautiful word of musical poetry, shalom,* cannot be used of South Africa. There is no harmony, no caring community and little awareness of God. There are many reasons for this, as we all know. One is that of 'leadership responsibility. . . By leadership I mean that quality of responsibility that initiates training, direction and action which encourages another towards the achievement of the same level of competency and ability as the leader himself (p. 6). The Rev. Evans also expanded on the need for leader- ship training in South Africa. This cry for leadership training was also the main theme of Mphahlele's (1981) article entitled "What is the Black Management Forum?" On May 17, 1976, just before the students' riots in South Africa, Bishop D. Tutu published an article sounding a possible danger of ineffective leadership. Hadebe (1977) also dwelt, at length, upon the need for constructive leadership and training thereof in South African schools. Commenting on Black Education, Kambule (1978), the former Principal of Orlando High School, said: "To ignore the facts now facing South Africa (i.e., need for effective leadership) is merely courting trouble." (p. 16). Perhaps it was this need for formally trained leaders in the South African higher educational system, for leader- ship programs, and for any literature on leadership that made Rupert (1980) focus his lengthy speech of characteris- tics of a leader. He emphasized that: fA HebreW'biblical term meaning peace among people of various communities, among and within the individuals. U kan sulke leiers wees. Ons hoef nie te wag om 'n krisis in Suid-Afrika vir die nodige leiers om na vore te kom nie. Laat ons nou die leiers kweek wat ons land op elke terrein kan dien en kan help bou aan 'n gelukkige en meelewende gemeenskap.* These were some of the speeches that initiated the researcher's interest in the area of leadership, which even- tually made him to decide to investigate whether certain selected factors have any influence on the effectiveness of the leadership style of the teaching staff in South African universities. Statement of the Problem Basically, this study is an attempt to determine whether, and to what extent, population grouping, faculty, and academic position have any statistical effect on the effectiveness of leadership styles of lecturers and senior lecturers in South African universities as evaluated by students. This was determined by a survey of a selected sample of university students from four population groups (Blacks, Coloureds, Indians, and Whites) in the Republic of South Africa. Need for the Study There are various reasons why the researcher decided to investigate whether, and to what extent, population grouping, * For English translation, see Appendix A. Faculty, and academic position have any statistically signi- ficant effect on the effectiveness of leadership styles of lecturers and senior lecturers in South African universities. 1. Among non-white population groups (Blacks, Coloureds, and Indians) there is a growing concern that the currently prevailing leadership styles of the teaching staff in South African universities is more effective for white students than for them. This concern has been expressed in a number of articles and remarks. These are some of theme Commenting on education in Black universities, Turflux (April 1982)* said: : Education is essentially for the minority of exploiters and not education for the people. . . . - Education encourages the desire to be like those who are determined to keep the majority of people down. - It (education) helps people to adapt to a world designed and created by others. ° It (education) is conducted in a way that authority is accepted without ever questioning it. We strive to satisfy the authority. . - It is there to entrench the exploitative system and white culture models. : The present methods used in our education must be rejected in their entirety. . . The present teacher - student relationship must be chan ed, both teacher and student must be simu taneously teachers and students. - Students must learn how to learn to solve prdblems.(p. ll) *A magazine published by the Students Representative Council of the University of Turfloop, R.S.A. The National Union of South African Students (NUSAS) (1982)* cited the leaders of the Soweto Students Representa- tive Council as having said: We shall reject the whole system of Bantu Education whose aim is to reduce us, mentally and physically, into "hewers of wood and drawers of water." (p. 8). Reporting on the student school boycott in September 1980, the Radio of South Africa, the External Service of the South African Broadcasting Corporation, directed to North America and Canada said: Student unrest (in South Africa) is linked to a protest against low standards of and inferior education for non-whites. Turflux (1982) added that: "Education (in Black universities) is book-oriented without emphasis on practical experience," (p. 11). This remark agreed with a comment in the NUSAS (1982) that: The quality of education provided in these campuses (universities for Black students) is often inferior, with great emphasis placed on single-textbook courses and rote learning. Libraries are generally poorly stocked. (p. 24) Turflux continued to say that: "Education for Blacks has had effect of creating a Black middle class that is far removed from the true aspirations of the people," (p. 11). That leadership styles of the teaching staff in Black uni- versities tends to create a Black middle class is reflected in a remark on the inequalities of the school system as published by the NUSAS (1982). It cited Gerrit Viljoen, the *A.magazine published by the National Union of South African Students (NUSAS). Minister of Education to have said: If it had not been for the differentiated university system only a very small elite of blacks would have had training opportuni- ties, however wide open the doors of the white universities were. (p. 32) Hence, there is strong need for a thoroughly researched work across population groups to investi- gate the validity of this concern. 2. There is a need for a study to test the validity of the prevailing belief that members of certain faculties have a higher probability of having more effective leadership styles than those from other faculties. 3. Some South African academicians desire to know 'whether academic position has any statistically significant effect on the effectiveness of leader- ship styles of university teaching staff. 4. Some South African higher educational institutions are in the process of introducing courses on leader- ship. One of the problems facing them is that, up to the present moment, neither papers nor longer works have been published in the Republic of South Africa. Most studies have been conducted in the United States although some research has also been done in certain European countries. These institu- tions have expressed a strong need for any literature on leadership conducted in/or based on South Africa. 5. In South Africa, there is a strong need, in general, for studies conducted across color lines or population groups. Importance of the Study In the Republic of South Africa, this study is of significant value in many ways. These are some of them" 1. It is one of the pioneer studies on leadership, and the first study on leadership effectiveness, in South Africa. Up to the present moment, most studies on leadership have been conducted outside South Africa, especially in the United States. It is the first study on leadership ever conducted across color lines or population groups in the Republic of South Africa. By comparing these groups, this work might be said to encourage and facilitate constructive and meaningful interaction among them. The findings of this study may either support or invalidate the concern the non-white population groups have about the effectiveness of the currently prevailing leadership style of the teaching staff in South African universities. If the results are consistent with this concern, perhaps, certain courses and/or in-service training programs on leadership may be introduced to help the teaching staff in non-white universities learn more effective skills so that they can better meet the needs of their subordinates (students). The present work will also check the validity of 10 the prevailing belief that members of certain faculties have a higher probability of having more effective leadership style than those of others. It may also demonstrate whether academic position has any statistically significant effect on the effectiveness on leadership style of the South African university teaching staff. This study may also be a source of reference for courses on leadership and in-service training pro- grams in and outside South African universities. By focusing on leadership in general and on ef- fectiveness of leaderShip of the Faculty in partic- ular, this study, according to MCBeath and Andrews (1960), has implications for the introduction of courses and in-service training programs on leader- ship in and outside universities. By focusing on effectiveness of the university teaching staff, this study shows an indirect bearing on student performance. For studies (e.g., Morall, 1974: Cullers, Hughers and McGreal, 1973) reported a correlation between leadership style of a teacher and student performance. According to McBeath and Andrews (1960), a study that deals with leadership style of the university teaching staff has implications: (1) for educators who are interested in helping teachers to improve (in their ‘work). . . ., 10. ll. 12. ll (2) for selection of candidates for training teachers, (3) for undergraduate courses for teachers in classroom administration (that) might snythesize the teacher's knowledge of subject matter, his understanding of the situation with regard to pupils (psychology) and with regard to the community and groups (sociologY). and the behavior needed to be an effective leader, and (4) for courses on the programs of teacher training colleges (pp. l7-18). If this study finds that selected factors have significant effect on the effectiveness of leader- ship styles of the university teaching staff, further studies may be generated to determine why this is so. According to Greenfield and Andrews (1961), any study on teacher leadership behavior has implications: (1) for the selection of teachers for academic and administrative positions, and (2) for the evaluation of their work (p. 101). According to Evan (1959) and Morall (1974), in- service training on leadership may improve leader— ship skills of the teaching staff or faculty. This, on the other hand, may make lecturers more effective as leaders both in their respective classrooms and in the communities they serve. Basic Assumptions This study was based on the following assumptions: University teachers are leaders (Ivancevich, 12 Szilagyi, and Wallace, 1977: Tannenbaum, Weschler, and Massarick, 1961) . The effectiveness of the leadership style of Faculty can be best judged by its students who, in the teaching process, are followers (Burn, 1978: Cartwright, 1959). The instrument used in this study is based on this assumption. Though the Michigan Four-Factor Theory Questionnaire was constructed and validated in the United States, it can be applied meaningfully in a South African educational setting. Limitations of the Study The Michigan Four-Factor Theory Questionnaire is subject to leniency (Edwards, 1970: Scott, 1968). Leniency is defined as the tendency of people to attribute socially desirable traits to people they like and attribute socially undesirable traits to people they dislike (Edwards and Abbott, 1973: Vroom, 1964). The fact that the Michigan Four-Factor Theory Questionnaire was not validated in the South African population might interfere with the results. Because of the limited nature of the scope of this study, owing to financial and time constraints, its findings should be viewed with caution. The researcher hoped that participants would be 12 Szilagyi, and'Wallace, 1977; Tannenbaum, Weschler, and Massarick, 1961). The effectiveness of the leadership style of Faculty can be best judged by its students who, in the teaching process, are followers (Burn, 1978: Cartwright, 1959). The instrument used in this study is baSed on this assumption. Though the Michigan Four-Factor Theory Questionnaire was constructed and validated in the United States, it can be applied meaningfully in a South African educational setting. Limitations of the Study The Michigan Four-Factor Theory Questionnaire is subject to leniency (Edwards, 1970: Scott, 1968). Leniency is defined as the tendency of people to attribute socially desirable traits to people they like and attribute socially undesirable traits to people they dislike (Edwards and Abbott, 1973: Vroom, 1964). The fact that the Michigan Four-Factor Theory Questionnaire was not validated in the South African population might interfere with the results. Because of the limited nature of the scope of this study, owing to financial and time constraints, its findings should be viewed with caution. The researcher hoped that participants would be 13 sincere in their responses. Commenting on this weakness, Chuenyane (1981) said: ". . . another limitation (in test and inventory administration) is the assumption that the respondents will be candid in expressing their views. . ." (p. 18). This observation was also made by Borg and Call (1971). Definition of Terms Following are some of the concepts or terms and their definitions that will be used during this study: Population Grouping refers to the manner in which the South African population is, at the moment, officially cate- gorized as follows: Blacks are people whose origin is indigenous to Africa, Coloureds are ”mixed race" people in South Africa, Indians are people who originated from India, and Whites are people who are identified as members of the Caucasian race. Faculty is to the South African higher educational system as college or school is to American universities. Faculty of Arts in South Africa is roughly an equiva- lent of the College of Arts and Letters in American univer- sities. The same is true as of the Faculties of Business, Education, Law and Science. Work Grogp refers to any group of individuals who cooperate to achieve a certain goal(s). Non-Whites refers to Blacks, Indians, and Coloureds 14 in the Republic of South Africa (R.S.A.) Organization of the Study Chapter I includes the introduction, statement of the problem, need for the study, importance of the study, basic assumptions, limitations of the study, and definition of terms. In Chapter III, a review of the related literature is presented. Chapter III discusses hypotheses and details the procedures employed in gathering and analyzing the data. Data analysis and interpretation is the focus of Chapter IV while Chapter V presents summary, conclusions, and recommen- dations for future research. CHAPTER II REVIEW OF SELECTED LITERATURE Extensive research has been conducted on leadership and on the determinants of its effectiveness. This review is a microcosm of the macrocosmic studies already done in this field. Many are still underway. Definition of Leadership The term leadership has been defined in many ways. One of these includes that of French and Synder, as cited by Cartwright (1959), who believe that: ". . . leadership is the potential social influence of one part of the group over another." (p. 118) One weakness of this definition is that it does not include the aim of the part of the group that has social influence upon the other. On the other hand, Burn (1978) defines leadership as: . leaders inducing followers to act for certain goals that represent the values and the motivations--the wants and needs, the aspirations and expectations--of both leaders and followers. (p. 19) According to Turker (1981), this definition of leadership has serious shortcomings. One, it does not see 15 16 dictators like Adolf Hitler as leaders. Two, it disregards the leader-follower relationship which may locate itself any- where between dictatorial forms of leadership, on the one hand, and participatory or democratic leadership, on the other. According to Ivancevich, Szilagyi, and Wallace (1977) . leadership is the relationship between two or more people in which one attempts to influence the other towards the accomplishment of some goal or goals. (p. 273) In more specific terms, Tannenbaum, Weschler, and Massarik (1961) define: Leadership as interpersonal influences, exercised in situation and directed, through the communication process, toward the attainment of a specific goal or goals. Leadership always involves attempt on the part of a leader (influencer) to affect (influence) the behavior of a follower (influencee) or followers in situation. (p. 24) Implicit in these definitions is the idea that the effectiveness of a leader partially depends upon the nature of power base available to an individual that she/he decides to use at a given moment (Cartwright, 1953; Janda, 1960): upon the degree to which she/he is accepted by the sub- ordinates (Ivancevich, et a1., 1977: Patchen, 1974): as well as upon the presence of some goal desirable to followers and to which they cannot only be actively influenced by the leader but also cooperated toward its achievement (Tead, 1935: Stogdill, 1950). Power may be defined as the ability to make people carry out instructions (Etzioni, 1964). Various forms of power a leader may employ include the fol- lowing: legitimate, reward, coercive, expert, and referent power (French and Raven, 1959). Ivancevich et al., (1977) 17 define these different forms of power: Legitimate power is derived from an individual's position in the structure or hierarchy of the organization. The organization usually sanctions this form of power by recognition from higher management or by the use of titles, such as manager, director, or supervisor. Reward power is based on the ability to control and administer rewards to others (e.g., money, promotions, praise) for compliance with the leader's direct- ives or requests. Coercive power is derived from the ability to control and administer punishment to others (e. g. , reprimand, termination) for noncompliance ‘with the leader's requests and directives. Expert power is based upon a special ability, skill, expertise, or knowledge base exhibited by an individual. . . Referent power is based on attractiveness or appeal of one person to another. A leader may be admired because of certain characteristics or traits that inspire or attract followers (e.g., charisma). Referent power may also be based on a person's reference or connection with another powerful individual. (p. 271) These are some of the common definitions of leaderev ship, and there are many others. In the next section major theories of leadership will be discussed. Theories of Leadership Theories of leadership have shown a developmental process which consists of three major approaches: trait, behavioral, and situational theories. 18 Trait Theories During the 1940's and 1950's, theories of leadership focused on the traits and determinants of leadership. In view of the research done in the 1970's and earlier, Stogdill (1974) identified six broad classes of leadership traits. , They are described below. Physical Characteristics. Some studies (e.g., Hornaday, and Bunker, 1970: Pinkle, and Friedlander, 1967: Powell and Nelson, 1969: Wollowick and McNamara, 1969: Harrell, 1969) suggested that leaders tend to be character- ized by a great amount of energy and stamina to maintain a high rate of physical activity. Others (e.g., Bass, 1964: Johnson and Peterson, 1968: Kumar, 1966: Stanton and Poors, 1967) reported that people who become leaders tend to show signs of accomplishment which are above average early in life. However, Standard and Poors (1967) Observed that in American corporate structure even good managers take time to reach the tap because promotion tends to depend upon administrative knowledge and demonstration of success which is the function of experience and age. Social Background. Most researchers (e.g., Miller and Dirksen, 1965: O'Donovan, 1962: Sanders, 1968: Stanley, Mann and Doig, 1967) found leaders to come from higher socio- economic backgrounds while a few (e.g., Baldwin, 1932: and Goodenough, 1930) reported no difference. Matthews (1954) 19 and Porter (1965) found that fathers of most leaders were in the professional, proprietor or official occupations. However, Newcomer (1965) and Scientific American (1965) found that leaders came from both lower and middle classes, and not mainly from the latter. Acdording to Powell (1969), in some systems the rise of the leader to the top is influ- enced by ethnic background and religious affiliation. Harrell, Burnham, and Lee (1963) and Lewis (1960) reported that almost all leaders in organizations are better educated. That is, they hold, at least, a college degree. Intelligence and Ability. A number of studies (e.g., Bray and Grant, 1966: George and Abraham, 1966: Kiessling and Kalish, 1961: Roadman, 1964) reported positive correlation between leadership, intelligence, and ability. However, according to Ghiselli (1963b), there is a curvilinear rela- tionship between leadsrship and managerial success. That is, people with very low and very high intelligence tend not to be good leaders. Most successful managers are usually of \/// average intelligence. Other Abilities. Effective leaders are usually characterized by superior judgement and/or decisiveness (George and Abraham, 1966; Miles, 1964: Stryker, 1960): knowledge (Harris, 1964: Robert and Dunnette, 1967: Thornton, 1968): and fluency of speech (Hardesty and Jones, 1968: Harrell, Burnham, Hunt and Lee, 1964; Harville, 1969: Kay and Meyer, 1962: MacNaughton, 1963). Bass (1981) remarks -. 20 that : Persons with high abilities may suffer from extreme self-preoccupation. A large discrep- ancy in capability between leader and led may 'make communication difficult. The leader's ideas may be too far in advance of the followers so they are rejected. (p. 79) Personality. Most leaders are characterized by ability to adjust to new situations (Fitzsimmons and Marcuse, 1961; Ghiselli, 1964: Krishnan, 1965: Mann, 1960): by agres- siveness or assertiveness (Bentz, 1964: Miner, 1968): by independence (Blake and Mouton, 1961: Gordon, 1966), by Objectivity (Albrecht, Glaser, and Marks, 1964: Krishnan, 1965): by resourcefulness (Hemphill, Griffiths, and Frederiksen, 1962): by enthusiasm (Leman and Soloman, 1952); by tolerance of stress (Grant and Bray, 1969: Lange and Jacdbs, 1960): by alertness (Coates and Pellegrin, 1957); by originality (Argyris, 1953: Gordon, 1963) by personal integrity (Gruenfeld, 1966: Stryker, 1960): and by self- confidence (Baker, Ware, and Osborn, 1966: Brown, 1964). Task-Related Characteristics. Studies indicate that A leaders are characterized by a high need for achievement \} (Andrews, 1967: Cummings and Scott, 1965): by high sense of responsibility (Borgatta, Ford, and Bohrnstedt, 1968: Gordon and Medland, 1965): by a high degree of task orientation (Dunteman and Bass, 1963: Harrell and Lee, 1964): by depend- ability in pursuit of objectives (Medow and Zander, 1965): by enterprise and initiative (Harrell, 1966; Helfrich and Schwirian, 1968): and by perservence in overcoming obstacles 21 (Bass, 1981). In short, leaders are usually characterized by strong motivation, drive, and persistence. Social Characteristics. Harrell (1966): Harrell, Burnham, Hunt and Lee (1964): Harrell and Lee (1964): Hornaday and Bunker (1970): Miller and Diksen (1965): Moment and Zaleznik (1963): and Rose (1962) reported that leaders tend to cooperate easily with a wide range of people. Link (1929) found that leaders take part in a number of activities.f Sorokin (1927): and Thresher (1927) also associated leadgfggjfl with higher mobility. According to Bentz (1964): Brown (1964» Carp, Vitola and McLanathan (1963): and DuBrin (1965) most leaders exhibit interpersonal skills which, according to Hemphill, Griffiths, and Frederiksen (1962): and Tornopol (1958) include tactfulness. Leaders with these social char- acteristics can easily get followers (Ghiselli, 1959: Grant and Bray, 1969: Harrell, 1966) and gain more popularity (Harrell and Lee, 1964). Although the researchers of trait theories of leader- ship have successfully identified certain important char- acteristics of leaders, they never attempted to explain the determinants of effective leadership (Ivancevich, Szilagyi, and Wallace, 1977). Behavioral Approach The failure of the trait approach to account for the effectiveness of a leader made social scientists in the 22 1950's and 1960's focus their attention on what the leader does and how she/he does it or, in other words, leadership style. Two major works on leadership style were conducted at Ohio State University and at the University of Michigan. Both these studies investigated leadership style(s) that increased work group performance and satisfaction. The Ohio State studies developed two separate ques- tionnaires. One measured leadership style as perceived by the leader himself/herself--the Leadership Opinion Question- naire (Fleishman, 1957). The other measured leadership style as perceived by the subordinates of the leader-~the Leader Behavior Description Questionnaire (Hemphill and Coons, 1957). Using these two instruments, Ohio State studies identified two independent leadership dimensions: imitating structure (i.e., task-oriented leadership style): and consideration (i.e., employee-oriented leadership style). Korman (1966) remarks that the main weakness of the Ohio State studies is that measurement of leadership style by the leader does not agree with that done by the sub- ordinates. However, despite this criticism, the Ohio State studies provided a firm base for the knowledge of leadership and for current and future research on leadership style(s). University of Michigan studies, conducted almost concurrently with the Ohio State research work, identified two distinct leadership styles: jdb-centered (i.e., task- centered) leadership style and employee centered (i.e., employee oriented) leadership style (Morse and Reimer, 1956: and Likert, 1967). The main weakness of both the University 23 of Michigan and the Ohio State studies was that they ignored situational factors that influence effectiveness of leader- ship style (Ivancevich et al., 1977). Situational Theories The failure of both trait and behavioral theories to include situational factors in their attempts to explain leadership, leader's behavior (style) and leadership ef- fectiveness led to the emergence of situational approaches in the late 1960's. The exponents of situational theory maintain that effectiveness of a leader is the function of the combination of leadership style(s) and situational // factors in the leadership environment. According to Ivancevich et al., (1977), the latter: . . includes the characteristics of the leader and the subordinates, the nature of the task and the structure of the group. (p. 274) Focus on the Leader--Fiedler's Leader Match Theory. A contingency leadership model, developed by Fiedler and his associates, was one of the first situational models of leader- ship ever developed. Conducting research on leaders of task- oriented groups, Fiedler (1967, 1971) emerged with a leader-match theory which initially concentrated on least preferred co~workers (LPC), a tool he had formulated to measure the leader's orientation. This instrument requires the leader to think of all people he/she has ever worked 'with and then think of the persons with whom he/she worked 1.- ‘54“ 24 least well. On an 18 bipolar dimension, e.g., pleasant- unpleasant, trustworthy-untrustworthy, etc., the leader is asked to rate that particular individual in question. Fiedler (1971) associated low LPC (i.e., mainly negative ratings) with task-oriented leadership style while high LPC (i.e., mainly positive ratings) was associated with relationship-oriented leadership style. However, "actual leader behavior was not explicitly studied or correlated with LPC scores in Fiedler's research." (p. 118) Because of his failure to empirically demonstrate the first postulation of his theory (i.e., there is a rela- tionship between absolute LPC score and team performance) Fiedler revised his VIEW'Of leadership. He formulated a contingency theory which suggested a relationship among the leader, the subordinates and the nature of the task. This theory maintains that if the leader is highly accepted or highly rejected by the group, low LPC (i.e., highly task- centered) leaders work best, whereas if the leader is moder— ately accepted by the group, high LPC (relationship-oriented) leaders work best. Perrow (1979) remarked that with this revision: . . . he (Fiedler) demonstrated that the 'climate' of the group had a substantial impact upon the effectiveness of leader- ship style. (p. 106) According to Ivancevich et al., (1977), Fiedler's leader-match theory has a number of strong points. One, it referred to three aspects which are part and parcel of leadership: the leader, the subordinates, and the task. 25 This was the first leadership theory which included these situational factors; This was its main strength. Two, it explicitly pointed out that the main problem of a leader is how to harmonize his/her personality with what the given situation requires, especially because a leader may be effective in one area and not in the other. Fiedler's theory also shows a number of weaknesses. One, the theory both ignores the characteristics of subor- dinates and does not say anything about the technological competence of the leader and subordinates (Graen, Alvares, Orris, and Martella, 1970). Two, the concept of least preferred coaworker (LPC) itself is not well understood (Cronbach and Meehl, 1955: Graen, Orris, and Alveres, 1971: Hosking and Schriesheim, 1978: and Stinson and Tracy, 1974). In his earlier research, Fiedler used LPC as a measure of the individual's personality. Later, he employed the very same concept to mean an individual's social needs. Conse— quently, it is no longer clear whether LPC reflects person- ality or motivational needs in respect to a group. Three, the model ignores the relationship between the leader and the situational variables (Lars Larson and Rowland, 1974; and McMahon, 1972). Bass (1981) comments that: Despite a vast array of publications on reliability, validity, and meaning of LPG and situational favorableness or control and despite many supportive tests of the model, the validity of the model is still contested. (p. 592) Though Fiedler's leader-match theory is one of the oldest theories and research programs on leaders of 26 task-oriented groups, these limitations have succeeded in making it not only the most controversial theory in the field but also in denying it acceptance and recognition as a general strategy to both understanding and solving leadership problems. However, Schein (1980) argues that these shortcomings are not all that serious. Focus on Task/Situation Vroom's Contingengy Theory. While Fiedler's theory concentrated on measuring leader's LPC, Vroom's approach focused on task and subordinates. Vroom (1975) recently introduced his contingency theory by citing from Stogdill (1948): The pattern of personal characteristics of the leader must bear some relevant relation- ship to the characteristics, activities, and goals of the followers. . . . It becomes clear that an adequate analysis of leadership involves not only a study of leaders, but also of situations. (p. 64-65) According to Schein (1980), this contingency theory consists of five "diagnostic criteria that would help leaders decide in which situations to use which kind of behavior," (117, p. 118). These criteria as presented by Vroom and Yetton (1973) are shown below. Types of Management Decision Sgyles AI. You solve the problem or make the decision yourself, using information available to you at that time. All. You Obtain the necessary information from your subordinate(s), then decide on the solution to the problem.yourself. You.may or may not tell your subordinates what the problem is in getting the information from 27 them” The role played by your subordinates in making the decision is clearly one of providing the necessary information to you. CI. You share the problem.with relevant sub- ordinates individually, getting their ideas and suggestions without bringing them together as a group. Then you make the decision, which may or may not reflect your subordinates' influence. CII. You share the problem with your subordinates as a group, collectively obtaining their ideas and suggestions. Then you make the decision, which may or may not reflect your subordinates' influence. GII. You share the prOblem with your subordinates as a group. Together you generate and evaluate alternatives and attempt to reach agreement (consensus) on a solution. Your role is much like that of chairperson. You do not try to influence the group to adopt "your" solution and you are willing to accept and implement any solution which has the support of the entire group. (p. 118) According to Schein (1980), this model suggests: . . . a specific way of analyzing problems by means of (seven) ordered questions which the leader can ask himself or herself and a set of decision rules put in the form of a decision tree which leads to the most desir- able option to be employed (See Figure 2.1.). (p. 118) A leader may analyze a problem by answering "yes" or "no" to some of these seven questions to arrive at possible decision alternatives. Focus on the Subordinates--Hersey and Blanchard's Situational Theory. unlike Fiedler, who asserts that leaders I should loOk for situations that suit their leadership styles,( and Vroom, who suggests that leaders ought to modify their ‘ leader behavior to meet the nature of the task, Hersey and 28 HHo Ham "me HHO "NH Hem .HHo "Ha HHo .Ho .HH< HHO flea and .Hao .Hu .HH4 mp Hem .HHo .Ho .HH< "a up be Hen .Hu .HH< "m Ham .HHO .Hu .Hfla um I .5393 use mason a o m c. weaowusaom wouuowouo aw maoxwa mouuofiouocsm macaw nowamaou waofinoua manu mofl>Hom a“ ponamuuu on ou mHoom Hudoauunwnmwuo wan muons mononwouonom smouonwvuonsm he venomous on oasos_uw umsu namuuoo handsommou aw OH .wHusz an cowmaooo ecu oxma ou mums H ~coHumuemaoaaaw O>Hu00umo ou Hmowuwuo moumnqvuonom an dogmaooo mo ooouuoooom woousuonuum aoanouo any «coamfioou euaamsv one: a came on Own“ uuowuammsm o>m£ H «nuances can» Hoeowumu once on On haoxaa ma soaunaom moo eaaucsosm.ucuaoufiswue zuwamsv m omega Dunno.3on;mmooonm.noemaoonnna.~ ounwwm .HU HHU HHU .H< .H< HHU .H< mH on MH mH mH on imH HNMQ’M‘DN .m .9 .U .m .¢ 29 Blanchard (1977) postulate that: The more managers adapt their style of leader behavior to meet the particular situation and the needs of their followers, the more effective they will be in reaching personal and organizational goals. (p. 101) This leadership theory has two dimensions: consider- ation or employee-oriented leadership style, and initiating structure of task-oriented leadership style. As shown in Figure 2.2, these two dimensions, forming an angle of 90 degree with each other, encompass a space with four cells, each with its own leadership style. Hersey and Blanchard gave these four basic leadership styles shorthand labels of: telling (high-task and low-relationship ‘- behavior), selling (high-task and high-relationship behavior), participating (low-task and high-relationship behavior), delegating (low-task and low-relationship behavior). (p. 124) According to this theory, effective leadership style is a leader's behavior that agrees with the situational demands on a larger environment. On this point, Hersey and l Blanchard generally agree with other contingency theorists. However, according to Schein (1980), the former introduces a certain dimension of the environment--"maturity of the subordinates, or their readiness to tackle the task facing the group." (p. 124) Hersey and Blanchard (1977) defined maturity as: 30 Figure 2.2--Basic Leadership Behavior Sty1e* High Participating .Selling Low task High task High relationship High relationship Consideration Delegating Telling Low task High task Ehw Low relationship Low relationship Low <1 .5 High Initiating,Structure *Taken from Schein, E.N. Or anizational Ps cholo . Prentice-Hall, Inc., Englewood Cliffs, N.J., 3rd Ed., 1980, p. 119. 31 . the capacity to set high but attainable goals (achievement motivation), willingness and ability to take responsibility, and education and/or experience of an individual or a group. (p. 161) It is essential to note that, in this context, the term."maturity" is defined with reference to the task facing the group and that this definition focuses on characteristics expected of the subordinates if leadership is to be effective. One of these characteristics is motivation or psychological maturity, i.e., a sense of self-confidence and self-respect at work that makes an individual not only responsible in his/her work but also aims at possible maximum productivity. The other characteristic is competence or job maturity, i.e., the individual's ability and technical competence to do the job (Schein, 1980). With reference to Figure 2.3, Schein (1980) outlines the practical application of this theory as follows: 1. If the maturity is low, the leader should be high in task and low in relationship behavior to help the group to achieve some success and begin to learn. If the level of maturity of followers in- creases, the leader should begin to reduce task behavior and increase relationship behavior to help the group grow in its own competence. As the level of maturity continues to increase, the leader should begin to reduce both task and relationship behavior, because the group is developing self- confidence and the ability to work on its own. As the group achieves maturity, the leader can continue to reduce task and relationship behavior, and essentially delegate task to the group and expect them to be accomplished. (p. 125) 32 Figure 2.3--What a Leader Behavior Should be in Terms of Degree of subordinate Maturity.* Low Task ‘ High Task Low Relationship Low Relationship Low Task High Task High Relationship High Relationship Mature ¢ 1» Immature *Schein, op. cit., p. 125. 33 Hersey and Blanchard's theory suggests that leader- ship effectiveness depends not only upon the leader's style and situational factors but also upon the level of the group's development. It should be noted that as group maturity increases counter-clockwise (i.e., from right to left) in a normal curve pattern, the leader continues to reduce task and relationship behavior in the same direction. Described differently, there is a negative correlation between leader task-relationship behavior and subordinate maturity. Focus on Leader Behavior--Argyris's Model I and Model II. Argyris's leadership theory is based on the assumption that leadership effectiveness in an organization is the function of the level of psychological maturity of both the manager and the subordinates. However, he maintains that the crystallization and development of this maturity is the responsibility of the leader. In this context, the expression "psychological maturity" refers to the degree to which the leader's overt verbal behavior is congruent with what really goes on in himself or herself. That is, in a psychologically mature leader, overt verbal behavior agrees \h ——..-..~..-..-A...-_.. A _ ,. _- _,.‘ 'with the inner feelings. Argyris (1957, 1964) found that leaders of most organizations are the first in the system to exhibit psy- chological immaturity which they unconsciously pass on to their employees. For example, a manager may say to his/her employees, "This is going to be your job. I trust that you .‘will do it well. I won't interfere with you." That this 34 manager trusts the subordinate may just be lip service. Inside him/her, the manager may have no trust in the employee. This type of communication where two conflicting messages are sent out simultaneously is called "double bind" (Satir, 1964). Argyris argues that an employee perceives such double level messages sent to him/her by his/her leader. If the manager continues to communicate with his/her employees in this manner, the latter eventually learns to communicate to the former in a double-bind fashion, too. This type of communi- cation where both the speaker and the addressee communicate in double binds is called "calibrated communication cycle" (Bandler, Grinder, and Satir, 1976). Argyris observed that leadership styles of most managers are based on two ”theories". He labeled the first one as "espoused theory", which consists of the manager's overt verbal behavior, e.g., his goals, assumptions, values, etc. This may range from participative to autocratic leader- ship. He terms the other "a theory-in-use". This consists of implicit assumptions, what really goes on in the leader, and that which actually governs his/her overt verbal behavior (Argyris and Schon, 1974; Argyris, 1976). According to Schein (1980), a theory-in-use or Model I, as Argyris calls it, consists of four factors or governing variables". 1. that one must achieve one's goals as one sees them; 2. that one must win rather than lose; 3. that one must minimize eliciting negative feelings in relationships; and 35 4. that one must be rational and minimize any emotionality. (p. 127) According to Argyris, these "governing variables" lead to behavior that is controlling of others, that increases one's security, and that makes an individual less confrontive of others. If the leader is confronted by others, she/he becomes defensive, thereby decreasing chances of learning more effective behavior (Schein, 1980). Further- more, a less confrontive behavior does not allow experience to test the validity of these assumptions. A leader who reflects Model I behavior not only has psychological prob- lems himself/herself but also infests his/her subordinates ‘with it. At the end, the whole organization gets psychologi- cally sick. Leadership then becomes less effective. Argyris's leadership theory maintains that psychologi- cal maturity is a prerequisite for a more effective leadership style. The leader must not communicate with his/her subordinates in double binds if she/he does not want them to do the same toward him/her. Schein (1980) observes that Argyris's theory suggests that it would be better for leaders to adopt the Model II approach which states: 1. that action should be based on valid information; 2. that action be based on free and informed choice; and 3. that action should be based on internal commitment to choice and constant monitoring of efforts to implement one's choices. Here the theory builds on the basis of 'learning how to learn'. (p. 128) 36 Argyris's theory shows a number of strengths. One, it specifically describes what a leaderwganmdo to learn_ __.____.—-“_—.. participative leadership style. Vroom, and Hersey and Blanchard, on the other hand, seem to suggest that a leader can learn participative style by simply being told what to do. This is not true because a change in leadership style means behavior change, an experience which involves feelings and change in self-concept. This is not at all easy to achieve. Two, it is true that most managers behave on the basis of Model I. Argyris's theory also shows a number of weaknesses. One, it is not true that most managers were found to be the first individuals in the system to teach their employees double-message communication style because Argyris's studies were not longitudinal. To reach this conclusion, he should have observed the manager's communication style and compared it with that of the employees at the point when the manager first occupied his/her managerial position. Such observa- tions should have been repeated over a certain period of time. If the employees' communication style changed with that of the manager, Argyris and his co-worker could have concluded, at least to some extent, that employees learn double-bind communication from their employers. However, even after such Observations, this conclusion; has some invalidity since there are so many other factors that can make employees learn this type of communication. Two, very few leaders can afford to learn and adopt Argyris's Model II of leadership because that would mean they should have 37 had extensive training in psychology and in group dynamics. This is not always possible. Schein (1980) remarks that Argyris's Model II assumptions should be taken with caution because they had not been supported by extensive research. However, this does not mean that this model is not essential and useful. He also notices that Argyris's theory does not accommodate the possibility that Model I may be quite effec- tive in some situations. Argyris's leadership style theory suggests that training in groupémgyggmiggggaymaggment”leadershipwgffective- ness. The import of Model II takes this discussion to the domain of leadership and small groups, e.g., encounter groups; sensitivity groups, etc. (Bradford, Giff, and Benne, 1964; Schein and Bennis, 1965). However, this area will be excluded because this study focuses on leadership styles of the university teaching staff or faculty. /fl,,~w“ Summary Extensive research conducted on leadership, as demon- strated by this brief review of selected literature, indi- cates hOW'EUCh interest has been shown by social scientists in this field. This may also be an adequate justification of the degree to which effective leadership is a sine gua 223 of success in any system, organization or institution. It is essential to note that the general focus of these pm} studies has not just been on leadership, but chiefly on the 1 question: What is it that determines the effectiveness of 38 a leader? The quest for the solution to this question began roughly in the dawn of the 20th Century (e.g., Terman, 1904; Webb, 1915; Kohs and Irle, 1920; Caldwell, 1920; Bowne, 1926; Sheldon, 1927). Today, it is still in progress (e.g., Schein, 1980; Bass, 1981). The coming years may generate even more literature on leadership. At first, trait theory was suggested. This was pOpu- W I lar before the 1970's, especially during the l940's-1950's. g :1m{n Owing to its fatal weaknesses, it died a natural death. (”U-J Leader approach then came to the scene. Its shortcomings I gave birth to situational theory which, step-by-step, devel- 5 1 oped to include emphasis on the leader (Fiedler, 1967, 1971);} task/situation (Vroom, 1975); subordinates (Hersey and : Blanchard, 1977); and the leader behavior (Argyris, 1957, 1964). This chapter reviewed selected literature related to major theories of leadership. The next chapter will discuss methodology of this study, research instrument, hypotheses, and data analysis. CHAPTER III DESIGN OF THE STUDY The purpose of this study was to determine whether, and to what extent, population grouping, Faculty, and academic position have a statistically significant effect on the effectiveness of leadership style of lecturers and senior lecturers in South.African universities as evaluated by students. This chapter will describe hypotheses, the p0pulations of the study, sampling procedures, instrumenta- tion, collection of data, processing of data, and analysis of the data. Hypothesis In order to examine whether selected variables have any statistically significant effect on the effectiveness of leadership style of lecturers and senior lecturers in South African universities, the following null hypotheses were tested in the study. Thesewnpll hypotheses were_stated in the non-dirsstional £91791.beaaussegprevious Stuslie,§-.havs.. been conducted in this area to provide any clear expectations. 39 40 Null Hypothesis 1 Population grouping (Black, Coloureds, Indians, & Whites) has no statistically significant effect on the effective- ness of leadership styles of lecturers and senior lecturers in South African universities. “B = “c = “I = “w where: ”B = mean for Blacks "C = mean for Coloureds “I = mean for Indians “W = mean for Whites Null Hypothesis 2 Faculty (Arts, Business, Education, Law, & Science) has no statistically significant effect on the effectiveness of leadership style of lecturers and senior lecturers in South African universities. “FA = "FB = “FE = “FL = “Fs where: “FA mean for Arts Faculty “FB = mean for Business Faculty "FE = mean for Education Faculty “FL = mean for Law Faculty "FS = mean for Science Faculty Null Hypothesis 3 Academic position has no statistically significant effect on the effectiveness of leadership style of lecturers and senior lecturers in South African universities. 41 mean for lecturers 'U H II “p2 = mean for senior lecturers Null Hypothesis 4 There will be no statistically significant twoaway inter- actions between levels of independent factors. M“ h.- Null Hypothesis 5 There will be no statistically significant three-way interaction among levels of independent factors. Population In the Republic of South Africa, higher educational institutions are classified into four categories, on the basis of population groups: Black, Coloured, Indian, and White. This study;was conducted on lecturers and senior lecturers of selected institutions from four types of univer- sities. The university teaching staff was selected as the population for this investigation because teachers are leaders (Tannenbaum, Weschler, & Massarik, 1961; Smith, 1960; Gibson, 1974; Belasco, 1973; Evans, 1979; Morall, 1974). SamplinggProcedures By stratified random method, the researcher pro- portionally selected certain universities from higher educa- tional system of the four population groups: Blacks, 42 Coloureds, Indians, and Whites. From this sample of univer- sities, the researcher then randomly selected 200 lecturers and senior lecturers from five Faculties: Arts, Business, Education, Law, and Science. From each population group, random selection of participants was done as follows: a) 50 teaching staff members from Black universities as follows: i) 10 participants in the Faculty of Arts randomly selected as follows: 5 lecturers, and 5 senior lecturers. ii) Exactly the same procedure was followed in each of the remaining Faculties (Business, Education, Law, and Science) to obtain a sub-sample of 50 participants from Black universities. b) The process employed to generate participants from Black universities, as described under (a) above, was followed for lecturers and senior lecturers of each of the remaining population groups (Coloureds, Indians, and Whites) to get the sample, N = 200 (See Table 3.1.). Instrumentation The Michigan Four Factor Theory Questwonnaire (See Appendix I.) was developed by Taylorwand Bowers (1972) at the Institute for Social Research at the University of 43 OON on On on on mHOHOH Os OH OH .OH OH Hmuou now ON m m m. m nonpuoom uoHaom mocoHom ON m m m m HohpuooH OO OH OH OH. .OH. Hmuou ppm ON m m m. m nonpuoou HOHaom. _BMH ON m m m m HounuooH OO OH OH OH. ..... OH. .HMuou now. ON m m n. m HounuooH HoHnom GOHumonvm ON m m n m HounuooH Oq OH OH OH OH Hnuou ppm ON m m m m nonpuooq HoHaom mmoGHmpm ON m m m m Housuooq OO OH OH OH. OH, kuou ppm ON m ..m m, m amusuumq uoaamm mun< ON m m m m MounuooH Hmuoa moanz momanH muounOHou mxomHm mmnomu. onHOHDmom . onHHmom UHSNQ¢U< MHHDU .01). Thus pl = "p2 Null hypothesis 2 was that Faculty has no statisti- cally significant effect on the effectiveness of leadership style in South African universities. This omnibus null hypothesis was rejected, (F = 4.34 p<1.01). Therefore “EA # "FB ¢ “FE ¢ “FL + “F5 Null hypothesis 1 was that population grouping has no statistically significant effect upon the effectiveness 62 Table 4.4-~Three-Way ANOVA of Total Scores ........ p less Effect df MS F than Population 3/160 144.31 36.13 .0001 Faculty 4/160 17.35 4.34 .0024 Academic Postition 1/160 .58 .15 .7041 Population x Faculty 12/160 6.94 1.74 .0633 Faculty x Position 4/160 9.78 2.45 .0485 Population x Position 3/160 3.82 .96 .416 Population x Faculty x Position 12/160 6.32 1.58 .102 63 of leadership styles of teaching staff in South African universities. As can be seen in Table 4.4, this omnibus null hypothesis was rejected, (F = 36.13 p (.01). Thus “3*uc*UI¥I-‘w This suggested that the effect of population grouping dif- fered from one group to the other. General Inspection of Population Grouping FacultyiDifferential'Effects Because population grouping and Faculty had statisti- cally significant influence on the effectiveness of leadership styles of the teaching staff in South African universities (See Table 4.4.), a tw0dway (population by Faculty) Table (See Table 4.5.) was prepared to explore these differential effects. Inspection of this Table revealed that the effect of Faculty might have been statistically significant pri- 'marily due to a low rating mean for the Science Faculty teaching staff. The means of the remaining Faculties looked almost equal, with no statistically significant differences. As can be seen in Table 4.5, the effect of population grouping might have been statistically significant primarily due to a markedly low rating mean for the faculty in Black campuses. The means of the remaining population groups looked almost equal. Apparently, there were no statistically significant differences among them. The foregoing general observations or findings were highlighted in Figure 4.1, a graphic 64 Table 4.5--Means and SD of Various Faculties in their Respective Population Groups. B I c w Total 2 53 63 66 60 61 an; SD 7.0 8.4 10.3 6.4 8.1 i 46 66 58 6o 58 Bus. SD 6.0 7.0 9.9 . 9.1 8.0 i 43 67 57 65 58 13.9... SD 6.5 18.2 8.9 5.7 9.8 i so 63 61 69 61 £1; SD 4.3 11.5 12.2 9.1 9.3 i 35 62 59 55 53 §£; SD 8.8 14.2 11.3 9.6 11.0 i 45 64 6o 62 Total SD 6.5 11.9 10.5 8.0 Key: Blacks (B) Arts (Ar.) Indians (I) Business (Bus.) Coloureds (C) Education (Ed.) ‘Whites (W) Law (La.) Science (Sc.) 65 Figure 4.1--Graphic Representation of Various MEAN FACULTY RATINGS 50 Faculty Means in Their Respective Population Groups. l I l l s I c w POPULATION GROUPS 66 representation of Table 4.5. Multiple Comparisons The validity of the preceding general inspection of differential effects of population grouping, and Faculty was examined, using multiple comparison techniques. Also this further step would help to determine if any additional infor- mation might be gained from this alternative approach. Faculty Statistically significant differences among Faculty means were examined, using a post—hoc measure, called Tukey A, with .01 level of significance. The critical value was found to 4.1 for all comparisons. Contrasts were done by first arranging the Faculty means in an ascending order as shown below. Faculty Means in an AscendinggOrder Science 53 Business 58 Education 58 Arts 61 Law 61 Comparisons were then arranged in the following way: 67 58 (Business) minus 53 (Science) 5:>4.l; Ho: reject 61 (Arts) 'minus 58 (Education) 3<<4.1; Ho: stands 61 (Law) minus 53 (Science) 8>4.1: Ho: reject Thus the results of multiple comparisons indicated a) that Arts and Law Faculties had the highest means while Science Faculty had the lowest, and b) that these two extreme rating means were statisti- cally different from each other. These findings corroborated the findings of general inspection performed in the previous section. The above multiple comparisons also provided addi- tional information, which showed that the rating mean: a) for Business Faculty is statisticallygggegger than that for the ScienceEéeulty, and b) for Arts Faculty is statistically greater than that for the Education Faculty. Population Groups Planned orthogonal comparisons were conducted to determine which means were statistically different from each other. This approach to comparisons was adopted because there were a priori hypotheses as to the differences to be found between the four means. Contrasts prepared and the results of computations are presented in Table 4.6. As can be seen in Table 4.6, the teaching staff in White universities had the higher mean ratings compared with 68 Table 4.6--Planned Orthogonal Comparisons Between Population Group Rating Means Contrasts Results Whites vs. non-whites (Blacks, Coloureds, and Indians) Blacks vs. Indians and Coloureds Coloureds vs. Indians Whites were higher than non-whites. Indians and Coloureds were higher than Blacks. Indians were higher than Coloureds. 69 those of their colleagues in non-white campuses. The faculty in Indian and Coloured universities had higher mean ratings than that of the teaching staff in Black universities. Lecturers and senior lecturers in an Indian University had higher mean ratings than that of teachers in a Coloured university. 4 §_5 §_2 Multivariate Analysis of Variance In the previous section, ANOVA was used to analyze the data because the correlation matrix and the principal components analysis indicated that all four subscales (dependent variables) could be combined into a single com- posite measure without much loss in terms of variation accounted for. To investigate if any additional information could be obtained, a multivariate analysis of variance (MANOVA) was also used to analyze the data, utilizing each subscale as a separate dependent variable. This analysis was carried out using the Finn program available through the College of Education at Michigan State University. The results of the analysis are presented in Table 4.7. Null hypothesis 5 was that there is no statistically significant threedway interaction involving the three independent factors (population groups, Faculty, and aca- demic position). The data presented in Table 4.7 indicate that the findings of this study failed to reject this hypothesis, (F = 1.3899 p>>.01). Null hypothesis 4 was that there are no statistically 70 Table 4.7--Mu1tivariate Analysis of Variance p less Source df Multi-F than Population 12/416 9.48 .0001 Faculty 16/480 3.64 .0001 Academic Position 4/157 .65 .6278 Population by Faculty 48/607 1.67 .0039 Faculty . by Position 16/480 1.32 .1798 Population by Position 12/416 1.09 .3686 Population by Faculty by Academic Position 48/607 1.39 .0460 71 significant twoaway interactions between the factors. The data presented in Table 4.7 indicate that this hypothesis was rejected since one two-way (population by Faculty) interaction was statistically significant (F = 1.6714 p<:.01L Other tdeway interactions were not statistically significant. The univariate analysis of variance in Table 4.8 displays that the population by Faculty twoeway multivariate interaction was statistically significant due primarily to subtest 3, namely, Goal Emphasis (F = 2.5692 p<<.01). Table 4.9 shows subtest 3 means of various Faculties in relation to their respective population groups. Figure 4.2 is a graphic representation of Table 4.9, which illus- trates a twoeway (population by Faculty) interaction in subtest 3. The interaction is reflected by the crossing of lines. Arts Faculty rating means were consistently the highest among non-whites (Blacks, Indians, and Coloureds) whereas it was the lowest for Whites. Law Faculty rating means were average for Blacks and Indians whereas these means ‘were the highest among Coloureds and Whites. Education and Business Faculty rating means seem variable in all four population groups: 'middle for Blacks and Whites, highest for Indians, and relatively lowest for Coloureds. While Science rating means were the lowest among Blacks, Coloureds, and Whites, it was the highest among Indians. These variations are further highlighted in Table 4.10 which presents a descending rank order of Faculty rating means by various population groups. Null hypothesis 3 stated that their academic Table 4.8--Univariate Analysis 72 of Variance Variable Hypothesis Mean SQ Univariate F P less than 1 Subtest l .5081 1.1727 .3070 (Support) 2 Subtest 2 .6625 1.7249 .0660 (Interaction Facilitation) 3 Subtest 3 .7699 2.5692 .0039 (Goal Emphasis) 4 Subtest 4 .3154 1.1804 .3014 (Work Facilitation) 73 Table 4.9--Subtest 3 (Goal Emphasis) Means of Various Faculties in their Respective Population Groups. Arts Business Education Law .Science Blacks 13.8 11.3 11.3 13.7 8.4 Indians 15.9 16.4 16.3 15.7 15.0 Coloureds 15.1 13.5 12.7 15.2 14.8 \. Whites 13.2 14.7 13.9 17.7 12.9 74 Figure 4.2--Graphic Illustration of Subtest 3 (Goal Emphasis) Two-Way (population by Faculty) Interaction.* ”E A 18— Law 17.. 16b 15- ° - . Business / /’ o \ \ / 1 4 / '\ \/ /. Education 13 - / \,../'/ Arts/Science ,/ 12 -— / 1 1 ‘IO- MEAN FACULTY RATINGS l l I I B I C W POPULATION GROUPS *Numerical values rounded 75 Table 4.10--Descending Rank Order of Faculty Rating Means in Each Population Group. Descending Rank': Population Groups Order of Means :i Blacks Indians Coloureds Whites 1 Arts (Education (Arts/Law Law (Business (Science 2 Law Arts Business Business 3 (Education Law' Education (Business 4 Science Science Arts/Science 76 position has no statistically significant effect on effec- tiveness of leadership style in South African universities. Like in ANOVA, in Table 4.7, it was shown that this h,pothesis was not rejected, (§_= .6500 p >.01). Thus, p1 = p2 Null hypothesis 2 stated that Faculty has no statis- tically significant effect upon the effectivenesscufleadership styles in South African universities. As can be seen in Table 4.7, like in ANOVA (See Table 4.4.), this hypothesis was rejected multivariate (§== 3.6353 p*<.01). Therefore, “EA # "FB # “FE # “FL # “Fs Given the significant two-way (population y Faculty) interaction involving subtest three (Goal emphasis) the univariate analysis of variance in Table 4.11 can be examined only for: a) Subscale 1, univariate F = 3.9993 (p<<.01). (Support) The univariate F ratio's for subtests 2 (Interaction Facilitation) and 4 (Goal Emphasis) were not statistically significant at .01 level. Multiple (Tukey A) comparisons were conducted to determine the mean differences that were statistically dif- ferent from each other. Using df=160 and .01 level of significance, the critical value for all comparisons was 1.9. The results of the computations were similar to those obtained under ANOVA. That is, Arts and Law Faculties had the highest rating means while Science, the lowest. These 77 Table 4.1l--Univariate Analysis of Variance ........ Variable Hypothesis Mean SQ Univariate F .P less than Subtest 1 1.7327 3.9993 .0041 (Support) Subtest 2 1.2203 3.1773 .0153 (Interaction Facilitation) Subtest 3 1.8058 8.0262 .0002 (Coal Emphasis subtest 4 .8068 3.0195 .0196 (Work Facilitation) 78 extreme values had statistically significant differences. Hypothesis 1 stated that population grouping has no statistically significant effect on the effectiveness of leadership styles of the teaching staff in South African universities. The data in Table 4.7 show that this hypoth- esis was rejected (multivariate F = 9.4764 p<<.01). This means that “B 3* “c a! “I a4 "w However, the univariate analysis of variance in Table 4.12 indicates that the effect of population grouping was statis- tically significant primarily due to three indices (again ignoring subtest three due to its involvement in the twoaway interaction). These were: a) Subtest 1 univariate F = 27.8791 (p<<.01) (Support) b) subtest 2 univariate F = 27.8265 (p<<.01) (Interaction Facilitation) c) subtest 4 univariate F = 32.1528 (p<<.01) (Work Facilitation) Planned orthogonal comparisons were conducted to determine the mean differences that were statistically dif- ferent from each other. In all selected three subtests, the results were similar to those obtained in ANOVA (See Table 4.6.). Profiles of Leadership Behavior Profiles of leadership behavior in all four dependent variables were prepared and explored. This further spelled 79 Table 4.12--Univariate Analysis of Variance ...... ......... --------- Variable Hypothesis Mean SQ Univariate F P less than Subtest 1 12.1219 27.9791 .0001 (Support) subtest 2 10.6874 27.8265 .0001 (Interaction Facilitation) Subtest 3 5.9883 19.9841 .0001 (Coal Emphasis) Subtest 4 8.5915 32.1528 .0001 (Work Facilitation) 80 out the influence of each statistically significant main effect on the main dependent factor, namely, leadership style. Table 4.13 shows Faculties and their respective means in each subtest. This information was then graphically illustrated in Figure 4.3. Science Faculty rating means were consistently the lowest in all subtests whereas Arts Faculty rating means were consistently the highest. Law Faculty rating means were variable: high in subscales l and 3, and average in subscales 2 and 4. Education and Business Faculty rating means were somewhat in the middle. It is essential to note that in each Faculty, subscale 2 rating mean was always the lowest. Table 4.14 presents population group rating means in each subtest. The profile of this data is given in Figure 4.3. According to this Figure, Black university teaching staff received the lowest rating means in all subscales. In subtest 2, they were even markedly very low (10). On the other hand, Whites and Indians rating means were con- sistently the highest. Their rating means for subscale l were even markedly very high (17). Coloureds rating means were always in the middle though slightly closer to Whites and Indians. 81 Table 4.13--Faculties and Their Respective Rating Means in each Subscale. Faculties Subscales 1 2 3 4 Support Interaction Goal Work Facilitation Emphasis Facilitation Arts 16.4 14.8 14.0 14.8 Business 14.8 13.5 14.0 15.3 Education 15.3 14.5 13.6 14.7 Law' 16.0 14.0 15.6 15.2 Science 13.8 12.5 12.7 13.5 82 mm popcpou oHoS mosHm> HmoHHQEDZ¥ Em: cozooaom Irina—am. mmoEmsm IIIAd... 26.. .III .69 oocflom ES 22 Eczoezoou 6:25. ON 0% m. 3. 0.. or . "V Boomnzm d d . EmacoEm Rog "m Enomnam Eozozzuon. 203093.... "N 0.33.6 mu \uo-o—o—.J/ Atoooom. "w ofiomnam _ P p on o. m P 3 or or m .234 .2528. do . «HOH>m£om aHnmuoomoH mo mGOHmcoBHn hex pooh co mHmwneem mo acumen auHSUmm mo mHHmoum HNUHHMEZyn l I I I20. 3050.00 "cs. oozes. ”.m. 9.665 mu ON .nH or mp up or up pp 0 . H H o H d u H 1.: ’. x x , .. . l. , . _ _ _ _ _ . _ \ \ \ \ .I — h p n - - p n ma“ .oNn .np .ww m.. «e. n.. NF. Pp .uw m .55 e o m sH0H>m£om oHSmhoomoH mo mconcmEHn mom pooh co m .cozoezoou 6:25. "v 0.3.3.5 .emofism Koo. "n o.noon:m .coaoezoou 203093.... "N 0.3.6.95 .uoaaao. 3 0.33.5 mHmmneEm mo common esouw ooHumHsaom Loom mo mHHmon