For Reference

NOT TO BE TAKEN FROM THIS ROOM

©x mm

University of Alberta Printing Department

]

Digitized by the Internet Archive in 2018 with funding from University of Alberta Libraries

https://archive.org/details/dimensionsofconfOOmikl

X'Wfc i$, V2>- ^

THE UNIVERSITY OP ALBERTA

DIMENSIONS OP CONFLICTING EXPECTATIONS AND THE LEADER BEHAVIOR OP PRINCIPALS

by

ERWIN MIKLOS

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOB THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF EDUCATIONAL ADMINISTRATION

EDMONTON, ALBERTA MAY, 1965

ABSTRACT

The study reported in this thesis explored the utility of Guttman scale analysis procedures for quantifying and for analyzing the structure of the expectations which teachers hold for the "behavior of principals. The results of the scale analysis were used for testing hypothesized relationships among the degree of consensus on expec¬ tations within a school, the degree of teacher- principal agreement on expectations for the role of the principal, and the teachers' de¬ scriptions of the leader "behavior of principals. Leader ambivalence was defined as the intensity with which principals held expectations for their role, and the relationship of this variable to leader behavior descriptions was also explored. Data for the study were obtained from fifty-six principals and 765 teachers in fifty-six non-urban centralized schools .

Items were constructed for each of authority, personal, status, and means-ends categories and the responses were tested for unidi¬ mensionality using scale analysis techniques. None of these sets of items satisfied the criteria for the existence of a scale or even a quasi scale. However, approximately scalable subareas were identified and were used for quantifying the expectations held by teachers and principals for the role of principals. From these were derived indices of Agreement and Consensus while scale analysis of questionnaire responses was used for deriving an index of principal Ambivalence.

No significant linear or curvilinear relationships were found to

iii

. '

«

.

« v ' ' '• ;

'

-

exist between Ambivalence and teachers’ descriptions of the behavior of principals on the Leader Behavior Description Questionnaire. However, when schools were classed as being either high or low in Consensus, it was found that Ambivalence correlated positively with Structure in low Consensus Schools while the same correlation was negative in high Con¬ sensus schools. The difference between the two correlation coefficients was statistically significant and was interpreted as suggesting that Ambivalence is related to what might be considered desirable leader behavior only if it is appropriate to the situation.

The intensity with which principals held their self-expectations was not related to either Consensus or Agreement, but these variables were related to style of leader behavior as defined by Initiating Structure and Consideration scores. Leader behavior characterized by scores high in both Structure and Consideration was found to be associated with significantly higher Agreement scores than was any other style of leader behavior. The same trend was evident when leader behavior was related to Consensus but the results were not statistically signifi¬ cant. For principals who were high in Structure, both Agreement and Consensus correlated positively with the average length of time that principals had been in the present school with the present teachers while the same relationship was negative for principals classed as low in Structure. In the case of Agreement the difference between the two correlations was highly significant.

iv

.

-

.

-

'

. '

.

.

ACKNOWLEDGEMENTS

The writer wishes to express his thanks to the supervisor of the thesis, Dr. J.H.M. Andrews, for his valuable guidance at various stages of the study. Thanks are extended also to the other committee members, Dr. L.W. Downey and Dr. R.L. James, for their assistance during the investigation and during the writer's program of studies in general.

Appreciation is expressed to the many teachers and principals whose cooperation made possible the collection of the necessary data.

The courtesy with which the writer was received by the principals at a busy time of year is evidence of their concern for research in education the interest of teachers in such studies w as indicated by the excellent response in the form of completed questionnaires.

Mention must also be made of the fellow students, other friends, and family without whose assistance, encouragement, and consideration study and research would have been much more arduous.

Finally, financial assistance in the form of an Alberta Teachers* Association Fellowship in Education and a Province of Alberta Graduate Scholarship which enabled the writer to pursue graduate study is grate¬ fully acknowledged.

v

. » . -

.

■>

'

TABLE OF CONTENTS

CHAPTER PAGE

I. INTRODUCTION . 1

Purposes of the Study ... .

Need for the Study . 4

Limitations of the Study . 8

Organization of the Thesis . 10

II. THEORETICAL BASES OF THE STUDY . 13

A Theoretical Model . 13

Institutional and Individual Conflict . 17

Role Conflict . 17

Dimensions of Conflicting Expectations .... 19

Value Contradictions . 20

Relation to Role Conflict . 21

Leader .Ambivalence . 23

Attitude Intensity . 24

Generalized Intensity . 26

Ambivalence . 27

Leader Behavior . 28

Research Hypotheses . 30

Definitions of Terms . 30

Scale Analysis . 31

Hypotheses . 32

III. RELATED RESEARCH . 36

Research on Role Conflict . 36

vi

l' . .

- ; / ;•

.

CHAPTER

PAGE

Non -Education Role Conflict Studies . 36

Research in Education . 42

Intensity Function and .Ambivalence . . . 45

Attitude Intensity . 45

Ambivalence . 46

Some Studies of Leader Behavior . 48

IV. SCALE ANALYSIS: THEORY AND METHODOLOGY . 55

The Theory of Scale Analysis . 55

The Universe of Attributes . 55

Example of a Dichotomous Scale . 58

Definition of a Scale . 59

Measurement of Error . 60

Quasi Scales . 62

The Methodology of Scale Analysis . 63

The Scalogram Board . 63

Initial Arrangement . 64

Ranking of Respondents . 65

Combining Categories . 66

Final Arrangement . 67

Using one Board . 67

Testing for Scalability . 67

Scoring the Questionnaires . 68

Discussion . 68

The Utility of Scale Analysis . 69

vii

;• ••

( *

f i . •. to i li

- * . •- .

. . . .

.....

t

*

••••* o

... v.' . . I

CHAPTER

PAGE

Defining a Problem . 69

Implications of Scores . 70

Prediction . 70

Some Problems in Scale Analysis . 71

Sampling of People . . 71

Sampling of Items . 72

Test-Retest Reliability . 74

Summary . 74

V. SCALE ANALYSIS: CRITICISMS, DEVELOPMENTS , AND

APPLICATIONS . 76

Some Criticisms of Scale Analysis . . . 76

Universe of Attributes . 78

Item Construction and Arrangement . 79

Interviewer Effect . 82

Coefficient of Reproducibility . 82

Some Developments . 84

Applications of Scale Analysis . 86

Checking Existing Scales . 87

Data, from Interviews . 88

Church Orthodoxy . 89

Urban Structures . 89

Student Attitudes . 90

Voting Behavior . 91

Delinquent Behavior .

Concept Development . 93

viii

:

CHAPTER

PAGE

Aggressive Fantasy . 93

Object Scales . 94

Scale Analysis in the Study of Role Conflict ... 94

Conclusion . 97

VI. COLLECTION OF DATA . 101

Teachers' and Principals’ Questionnaires . 101

Selection of Items ..... . 101

Pretests, Validity, and Revisions . 102

LBDQ, . 105

Principal Ambivalence . 105

Selecting the Sample and Obtaining Data . 105

The Sample . 105

Visiting Schools . 106

Response . 107

Characteristics of Respondents . 109

VII. SCALOGRAM ANALYSIS OF DATA . 114

General Procedures . 114

Approximately Scalable Areas . 116

Status Dimension . 117

Authority Dimension . 119

Personal Dimension . 123

Means -Ends Dimension . 126

Scoring . 127

Scale Analysis of Principals’ Responses . 129

ix

.. ' J'

!

CHAPTER PAGE

Ambivalence Scores . 131

VIII. TREATMENT OP DATA . 134

Leader Behavior Scores . 134

Ambivalence, Consensus, and Agreement . 137

Ambivalence Scores . 137

Consensus . 138

Teacher -Principal Agreement . 139

Statistical Techniques . . . 142

Differences Between Means . 143

Measures of Relationship . 145

Correlation Ratio . 145

IX. TESTING THE HYPOTHESES . 148

Ambivalence and Leader Behavior . 148

Testing Hypothesis One . 148

Discussion . 151

Relationships of Structure and Consideration to Agreement and Consensus . 155

Testing Hypothesis Two . 153

Discussion . 157

Testing Hypothesis Three . 157

Discussion . 161

Relationship of Ambivalence to Consensus and

Agreement . 161

Testing Hypotheses Pour and Five . 161

x

. V-.'H

M

. . •-•. i

........on.... i... . . v

........... •■'£'■• /.

............

. . . i

fjOl

CHAPTER

PAGE

Discussion . 164

Relationship of Style of Leader Behavior to

Consensus . I65

Hypotheses . . . 166

Discussion and Further Analysis . 168

Relationship of Style of Leader Behavior to

Agreement . 172

Hypotheses . 172

Discussion . 17 6

Style of Leader Behavior, Tenure and Consensus. 176

Testing Hypothesis Six . 177

Discussion . 179

Style of Leader Behavior, Tenure and Agreement. 181

Testing Hypothesis Seven . 181

Discussion . 184

Summary . « . I84

Ambivalence, Structure, Consideration . . . 185

Ambivalence, Agreement, Consensus . 185

Structure, Consideration, Agreement .... 186

Structure, Consideration, Consensus .... 186

Discussion . 186

X. SUMMARY AND CONCLUSIONS . 188

Summary of the Study . 188

Conclusions and Implications . 194

Conclusions . 195

xi

l

O J'

r s.

-L .

•» O V.

1< •. . ••

1 1

i r v .

o :

Q

CHAPTER

PAGE

Implications for Research . 19 6

Implications for Administrators . 199

BIBLIOGRAPHY APPENDIX A APPENDIX B APPENDIX C APPENDIX D APPENDIX E

EXPECTATIONS FOR PRINCIPAL LEADER BEHAVIOR LEADER BEHAVIOR DESCRIPTION QUESTIONNAIRE

PRINCIPAL'S QUESTIONNAIRE .

QUESTIONNAIRE RETURNS BY SCHOOLS . . . NORMALIZED DATA BY SCHOOLS .

203

212

222

225

231

233

xii

. . . . .

. .■

r».«

.

...... '

......

LIST OF TABLES

TABLE PAGE

I. Number of Questionnaires Distributed and Completed

by Size of School . 108

II. Selected Characteristics of Respondents . Ill

III. Distribution of Principals by Experience and Teachers

by Tenure . 112

IV. Frequency Distribution of Teachers* Scale Scores .. 129

V. Coefficients of Reproducibility for Teachers' and

Principals' Responses to Seven Sets of Items .... 130

VI. Frequency Distribution of Principals' Scale

Scores (n=60) . 131

VII. Coefficients of Reproducibility for Principals'

Content and Ambivalence Scales . 132

VIII. Frequency Distribution of Principals' Ambivalence

Scores (N=60) . 133

IX. Frequency Distribution of Raw and T -Scores for Consideration and Initiating Structure LBDQ

Scores (N=56) . 135

X. Frequency Distribution of Raw and T-Scores for

Principals' Ambivalence Scores . 138

XI. Frequency Distribution of Raw and Normalized

Consensus Scores . 140

XII. Frequency Distribution of Raw and Normalized

Teacher- Principal Agreement Scores . 141

xiii

TABLE

PAGE

XIII. Mean Structure and Consideration Scores, Variances, and Significance of Differences for Principals

of High and Low .Ambivalence . 149

XIV. Linear and Nonlinear Correlation Coefficients for the Relationship Between Ambivalence and Two

Dimensions of Leader Behavior (N=56) . 151

XV. P Ratios for Significance of Linear and Curvi¬ linear Relationships Between Ambivalence and Two

Dimensions of Leader Behavior . . . 152

XVI. Mean Agreement Scores, Variances, and Significance of Differences for Principals of High and Low

Structure and Consideration . . 154

XVII. Linear and Nonlinear Correlation Coefficients for the Relationship Between Agreement and Two

Dimensions of Leader Behavior (N=56) . 155

XVIII. F Ratios For Significance of Linear and Curvilinear Relationships Between Two Dimensions of Leader

Behavior and Agreement .

XIX. Mean Consensus Scores, Variances, and Significance of Differences for Principals of High and Low

Structure and Consideration . 158

XX. Linear and Nonlinear Correlation Coefficients for the Relationship Between Consensus and Two

Dimensions of Leader Behavior (N=56) . 180

XXV

0

-O' s . : t

v )

TABLE

PAGE

XXI. F-Ratios For Significance of Linear and Curvi¬ linear Relationships Between Two Dimensions of

Leader Behavior and Consensus . . 160

XXII. Mean Consensus and Agreement Scores, Variances, and Significance of Differences for Principals

of High and Low Ambivalence . . . 162

XXIII. Linear and Nonlinear Correlation Coefficients for the Relationship Between Ambivalence and

Teacher Consensus and Agreement . 163

XXIV. F-Ratios for Significance of Linear and Curvi¬ linear Relationships Between Ambivalence and

Teacher Consensus and Agreement . I64

XXV. Mean Consensus Scores by Leadership Style of

Principal (N=56) . I67

XXVI. F-Ratios Obtained Through Analysis of Variance of Consensus Scores Grouped by Principals'

Style of Leader Behavior . 168

XXVII. Intercorrelations of Three Variables and

Significance of Differences For Schools of High

and Low Structure Principals . 170

XXVIII. Intercorrelations of Three Variables and

Significance of Differences for High and Low

Consensus Schools . 171

xv

'

TABLE

PAGE

XXIX. Mean Agreement Scores by Leadership Style of

Principal (N=56) . 175

XXX. F Ratios Obtained through Analysis of Variance of Agreement Scores Grouped by Principals'

Style of Leader Behavior . 174

XXXI. Analysis of Variance of Agreement Scores for

Structure and Consideration . 174

XXXII. Significance of Differences Between Mean Agreement Scores for S-+C + Group and Agreement Means of

Three Other Groups . 175

XXXIII. Mean Consensus Scores by Style of Leader

Behavior and Tenure . 178

XXXIV. F Ratios Obtained Through Analysis of Variances of Consensus Scores Grouped by Principals'

Style of Leader Behavior and Tenure . 179

XXXV. Linear Correlation of Tenure with Consensus for both High and Low Styles of Leader Behavior

With Significance of Differences . 180

XXXVI. Mean Agreement Scores by Style of Leader Behavior

and Tenure . 182

XXXVII. F Ratios Obtained Through Analysis of Variance of Agreement Scores Grouped by Principals' Style of Leader Behavior and Tenure . 185

xvi

r

o

. .

I

TABLE

PAGE

XXXVIII.

XXXIX.

Linear Correlation of Tenure and Agreement for Both High and Low Styles of Leader Behavior

with Significance of Differences . 183

Correlation Matrix for Five Variables . . . 185

xvii

.

LIST OF FIGURES

FIGURE PAGE

1 The Social Systems Model . 14

2 Response Patterns for Three Dichotomous Items .... 57

3 Scalogram for Three Dichotomous Items . 58

4 Scalograms for Status Items . 120

5 Scalograms for Authority Items . 122

6 Scalograms for Personal Items . 125

7 Scalogram for Mean-Ends Items . 128

8 Distribution of Principals in Four Quadrants

Derived from LBDQ Dimensions (U=56) . 138

xviii

r '■

CHAPTER I

INTRODUCTION

Daring the past decade there has "been a significant change in the methodology and the content of investigations carried out hy students of educational admini strati on. Whenever there is such a change there probably is a. danger that new concepts and methods will be introduced rapidly, will be dealt with superficially in research, and will then be passed over in favor of still newer approaches. An additional danger in the study of administration is that topics which are dealt with at a theoretical level may never be developed to the point where their full implications for the practice of administration become known. It was concerns such as these which led to the develop¬ ment of this study. Neither the concepts nor the methods are particu¬ larly novel; they have been used before. The unique contribution of this study lies in the attempt to combine methods and concepts in such a way as to make more explicit the implications of administrative theory for administrative practice in a small but significant area.

The methodology employed in the study may also suggest fruitful approaches for extending our knowledge in a problem area where there is need for further developments.

The sections in this chapter outline the purposes of the study, indicate the need for extending knowledge in this aspect of administ¬ ration, and also give recognition to some of the limitations of the study. The concluding section contains a guide to the organisation of the thesis.

: - c7 7 J c 7

- 7' . r : 7 -jJ ' 7->- 01 '

-- :7 ■' 7 i 77 : ' 7

0 0 . . C i : 7 .7, _ . ;

- -- - xoxi < 77 . ' .. oi :

. ■' 777

j i j' - _r«.

. V. . ' c 77 , 7 j 7.'; . J-.,' - •,

' : 7 : 7 77 : 7 77 7 7 ' - ,,

J.lXJ - x '

^ ‘-I ' -W U

i

7

; .

> o. . : -.to "

7 .

,r ■.

O' i

7 7 7 t

- o

- . . ; 7 ;

7 r/7c 7J ci

7

--- o 7 7 77 0 '

i 8 7 7

i . X ; . ; ; t ;

2

I. RJRFOSES OF THE STUDY

Much of the recent research in educational administration has focused on the social nature of the tasks of an administrator. Such research has been aided hy various theoretical formulations and models (8, 10) which serve to order some of the earlier studies and to stimulate new enquiry. One of the concepts or variables to which these draw attention is the expectations which different individuals in the administrator's life space hold for his behavior. In discussing the importance of these various expectations Campbell (3) states:

An understanding of these expectations, often conflicting in nature, may appear most frustrating. Only by such under¬ standing, however, can the administrator anticipate the reception of specified behavior on his part, Such anticipation seems necessary if the area of acceptance is to be extended and the area of disagreement minimized. Moreover, such understandings are necessary if a program of modifying expectations is to be started, (3, p. 264)

Before an administrator can understand, before he can modify, in fact before he can act in any way with respect to the expectations of others, he must have adequate knowledge about the nature of these expectations.

So far research has given some general clues as to what these expectations may be, but this same research has also shown that there are wide variations in expectations from situation to situation. One of the purposes of this study was to explore a means, scale analysis, for assessing the expectations of people who are of importance to an administrator in such a way as to give him greater understandings upon which to base his actions. It was hoped that the results of the study

would not only give administrators information about the structure of

» ' ' ' ' i " ' j :

' '' ' : '

" o ' ' .

'■ '• :

. ' ' T :

1

,

' ' ' '

: :

« " - ' '

'■ ■' ' '• - ' .

- > ' ,

r ’• r . i

i . .V •J JFj'l o -U

5

expectations which are held, for their behavior, but that these would also suggest how expectations might be studied comparably in a variety of situations.

A second purpose of the study was to explore the possibility of determining from self-report instruments variations in the abilities of administrators for dealing with the frustrations which result from the fact that different people hold different expectations for the behavior of an administrator. If this were a normal state of affairs in the reafity of administration, and research seems to indicate that it is, then it would be of obvious value during the selection process to know which administraf ors are better able to deaf with this situa¬ tion. This study subjected the concept of ambivalence to more intensive testing than was done in previous research and also studied the relationship of this variable to the way in which teachers described the leader behavior of the administrator .

A third purpose of the study was to utilize the results of the scale analysis for testing a number of hypotheses concerning relation¬ ships among the leader behavior of an administrator, the extent to which there is agreement between the administrator and his subordinates on role-expectations, and the degree of consensus within the group.

In this particular study the administrator who was under investi¬ gation was the principal of a school, and the expectations studied

were those of the teachers in his school.

.. . - i; •' . ::

:■ . -x. dx - c . n I

J 2 *' .. d , ;; hr: oor.

. 1> ... .:. >.'• . * c, . :: J: , .5".

o .ex d v . r ■■ x . .

: V. , id . e . ;

. X . . ' > ' u" ' X X L ■. ' .. . -LVo . 1

x V o' c x ; xd ' \ xx.' .x x

x . . ■. .

- xx x . •. . ; x ex . x x

x . o' . d c- ' 1 x ' . xx

"• ox

o . c >0 G.r v .f xd j' . \r

> L

J .. ' J x 1 . ■. : . < 1 x

i o X ' xx J! ;

?• " i r Xf. < ' ri.'r . i J:

J . . . . o ■■ ..

' > : ... . .& .

' . '

- d 5 x \ 'x . i r

. ? ,.i: . o x J ' .u o lc!

II. NEED FOR THE STUDY

4

Reference to the need for studies such as this has already "been made hut should he elaborated. If principals need to know what expectations are held for their behavior, and if the expectations change from situation to situation and from time to time, then there is a need for methods which will facilitate comparable research under these conditions. Such research may be carried out using an item by item approach as has been done in past studies, but this has all the dis¬ advantages of opinion polling which uses only unrelated items. There is a need for studies which will explore the structure of the expectations and show directions in which further progress might be made.

The importance of having adequate knowledge about the expectations which are held for a role incumbent has been illustrated by studies such as that carried out by Femeau (7) who investigated the expectations for the consultant role as perceived by administrators and by the consultants themselves and then related the degree of similarity between the two to the extent which consultant services were considered to be effective. The study revealed that when there was general agreement between administrators and consultants on the consultant role, the services were rated as more favorable than when there was a large measure of disagreement between the expectations of the administrator and the consultant. Femeau concluded that consultants and administrators must perceive each other as function¬ ing according to their expectations if the consultation is to be effective.

This problem may not be overcome as easily as it might appear for the study

.

. rf

. ' ' '

' ; ' : ' r

' ' ' i s':

.... . .

'■ •:

-v u C - i : ut

.• '•...! v )r.a 2"

, . : V : ' ' j

rf ;i\ t .'

... '• / '/•; r.

. '• v ' ' ^ '

. ; ' - ' '

5

revealed, the presence of regional differences in the expectations which were held for the role of the consultant. This suggests that a consultant cannot assume that he has acquired a thorough knowledge of the role as a result of his experiences in one situation; he may well have to adjust his ideas about the role when he moves to a different position.

The importance of the expectations which subordinates and super- ordinates hold for each other is also illustrated in the study reported by Moyer (13). Using only a small sample- which places serious limitations on the interpretation of his findings- Moyer investigated the relationship between teachers' attitudes toward leadership and teacher satisfaction. He suggests as a tentative conclusion that teachers do have a mental picture of an ideal leader and that the closer this approaches the type of leader¬ ship which they perceive to exist, the higher will be their satisfaction.

He states further that:

... the principal or superintendent, to be the leader, must be aware of the attitudes of the teachers, their individual, sub¬ group, and collective differences and similarities. Equipped with this knowledge of the nature of his group, he could increase the effectiveness of the group and his function by seeking to unify and harmonize the prevailing differences among them and, at the same time, attempt to bring his own leadership attitudes into a compatible relationship with his group. His position of leadership will probably be even stronger if his attitudes are beyond the aspirational tendencies of the group. (l3> p.3)

Moyer also suggests that in order to become keenly aware of his

subordinates' attitudes toward leadership, a school administrator should

use some system for determining what the expectations are as a first step

toward, unifying these and modifying his own behavior and the expectations

of subordinates.

It ; t'.:'

.

. '

.

'

» •. .

'

....

'• ' ' '

. >') •'

, : '

6

Chase (4,5) has reported on a number of studies from which a

general conclusion may be drawn to the effect that when teachers'

expectations with regard to leadership are fulfilled their morale

rises but when they are not fulfilled their morale drops. These

same studies also indicate that the problem is complicated by the

fact that the expectations vary from community to community. Yet,

if he is to fulfill expectations . . .

The administrator needs accurate information regarding the teachers' expectations as to what functions should be carried out by the school, who should carry out particular functions, and how the person or persons should perform these functions.

(5, P- 1)

Chase suggests that the expectations can be determined through such procedures as day to day contacts, the setting up of grievance committees and such devices as questionnaires on which subordinates indicate their expectations for the behavior of the administrator.

It would seem, though, that before this approach can be truly useful, the administrator needs to know what questions to ask and how to ask them. Both of these are problems for empirical research before they can be utilized by practitioners.

Two other investigations carried out independently by Valenti and Smith (summarized in 6) support the observation that the process oi determining expectations is complicated by variations from situation to situation. Valenti developed a scale to measure attitudes toward seventeen personnel problem areas; his investigation revealed that different teachers take different approaches to their work and that there is really nothing like a "typical" teacher. Nor was he able to predict teachers' attitudes

' ,'V A.

' ' x

. V- ^

J >. . .

< - j >

;i ' o

•' i ..V

z

X

. . : . :

V. . . * . . .

. .. f

. :

. . .• : :■ .. •>.!’ 1

. !'■ : ' . r. ;

c, ; . >. ' . ' '

. c

•:

'

1 c

. .

^ •- ;

/ .

" : -■ , . r ■. ' •: ,• ^

. ;

- - - ioxv i

o . ' x;.J

r.-: ; ■'

». . . i.

•• j.\. X

. . .

' ' . r

.

v - y ...

r . , IX

: f c . , . ; r

' :u a : ;' r.

:.y jr

7

from other characteristics of the teachers. As a result of this information one tends to agree with the conclusion . . .

The school situation in which a teacher or administrator works appears to shape his attitudes far more than his personal characteristics. Wide differences were found in the responses among school systems and among schools in the same systems.

(6, p. 2)

The Smith study supported the findings and the conclusion stated above.

When he investigated the attitudes of teachers in a seven-school system, Smith found that the composite attitudes of the staff in one school could not be inferred with accuracy from the knowledge of the total system score. This emphasizes the need for developing methods and instruments for assess¬ ing these various attitudes and expectations.

The above studies and others such as those by Bidwell (l),

Campbell (2), and Hal pin (ll) tend to support the need for further research into the expectations which are held for the incumbents of administrative positions. Furthermore, training programs are in need of predictors which will aid in the selection of those individuals who may be better able to cope with this reality of administrative life; it is hoped that the study of leader ambivalence might give some insights into this problem.

There has also been much conjecture but comparatively little evidence on the extent to which educational administrators are able to influence the expectations which 3taff members hold for their behavior. The first step toward utilizing knowledge in this area for the preparation of adminis¬ trators would be to determine the extent to which practicing administrators differ in ability to do this. For a field of study which is oriented

.

' >J

'

'

. > ' ’O '

: •' 1

.

.... , :

'

. [»;■• (j

' ' 1

8

toward practice to such an extent as is administration, it is not enough to theorize about influencing expectations. The theory must suggest hypotheses and the hypotheses must then be tested so that the implications of theory for practice can be determined.

III. LIMITATIONS OF THE STUDY

One of the problems with which this research project was forced to cope involved the application of methods which had not been used in this particular type of investigation but which had been employed success¬ fully in related social sciences. It was felt that if the same methods appeared to be useful in the study of administrative phenomena, then what was initially a limitation might become a major contribution.

This research was also plagued by another problem which is common to all the social sciences in that it sought to find significant relation¬ ships between a few variables from among the many which no doubt were present in the situations. One can assume that all other things were equad, but in the complex field of human relations one can rarely place much confidence in such an assumption. It was recognized that if the hypothesized relationships did exist at all, it was likely to be at a low level of significance . However, even such slight relationships may serve to indicate fruitful directions for more intensive and more care¬ fully controlled investigation.

When hypothesized relationships are found to exist there is also a great temptation to attribute cause and effect without due caution. This

i' ; '

„C;

i i ! . ).

.

i :

) ' .xi 1 up

. •: : c . '■ : . .u vo - V.

i; - i h ©i. - >' vI

. - V. ' - . 0

r f,r

? >

: : ■;

: i i >

; .....

•.

i -

: . . \ .

j: ;.

T

' ) i. . _ J ' - ... \

j

: :: bst

.' v. . . ' . t :

. J ' C ' U . ,

'

. 1 ,,

: c .

c . : e : ; .

j;

v r

.V' . v L" 1 .

: ' '

: r.

9

is particularly true in the case of an exploratory study which does not have an extensive grounding in theory. In this study there has been a deliberate attempt to avoid such a pitfall, but there is always the danger that not enough caution has been observed.

The exploratory nature of the study did not make possible a design as rigorous as might be desirable. This has its attendant problems but some possible advantages as well. It may be that what is a limitation in one respect is an advantage in another. Merton has indicated that empirical research may serve