• TABLE OF CONTENTS
HIDE
 Title Page
 Table of Contents
 Abstract
 Introduction
 Hypotheses
 Methodology
 Data analysis
 Findings
 Conclusions
 Bibliography
 Supplemental bibliography
 Appendices
 Biographical sketch














Group Title: exploratory field study involving the effects of supervisory succession on comparative work units /
Title: An exploratory field study involving the effects of supervisory succession on comparative work units /
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097454/00001
 Material Information
Title: An exploratory field study involving the effects of supervisory succession on comparative work units /
Physical Description: vii, 121 leaves : ill. ; 28 cm.
Language: English
Creator: Taylor, Gordon Arthur
Publication Date: 1980
Copyright Date: 1980
 Subjects
Subject: Supervisors   ( lcsh )
Supervision of employees   ( lcsh )
Management thesis Ph. D   ( lcsh )
Dissertations, Academic -- Management -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 89-95.
Additional Physical Form: Also available on World Wide Web
Statement of Responsibility: by Gordon A. Taylor, Jr.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097454
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000098498
oclc - 06766302
notis - AAL3945

Downloads

This item has the following downloads:

exploratoryfield00taylrich ( PDF )


Table of Contents
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
    Abstract
        Page v
        Page vi
        Page vii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
    Hypotheses
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
    Methodology
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
    Data analysis
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
    Findings
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
    Conclusions
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
    Bibliography
        Page 89
        Page 90
    Supplemental bibliography
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
    Appendices
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
    Biographical sketch
        Page 122
        Page 123
        Page 124
        Page 125
Full Text













AN EXPLORATORY FIELD STUDY
INVOLVING THE EFFECTS OF
SUPERVISORY SUCCESSION ON COMPARATIVE
WORK UNITS






BY


GORDON A.


TAYLOR, JR.


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1980
















Copyright 1980

by

Gordon A. Taylor, Jr.
















TABLE OF CONTENTS


CHAPTER PAGE

ABSTRACT ................................................. iii

ONE INTRODUCTION................................................ 1

The Succession Territory ................................... 1
Succession as a Significant Investigative Area........... 3
Supervisory Succession ........... .................... .... 6
Purpose of the Study .................... .................. 8

TWO HYPOTHESES ................ ................................ 10

Predecessory/Successor Leadership in Terms of Congruency
and Successor Acceptability ............................ 10
Hypothesis One ......................................... 11
Hypothesis Two ......................................... 11
Postsuccession Frustration and Successor Acceptability... 12
Hypothesis Three. ....................................... 13
Level of Subordinate Satisfaction and Successor
Acceptability. .......................................... 13
Hypothesis Four......................................... 14

THREE METHODOLOGY ............................................. 15

The Sample................................................ 15
The Questionnaires. ....................................... 17
Description of Previous Supervisor...................... 17
Description of Current Supervisor....................... 18
Management Questionnaire ............................... 19
Questionnaire Scoring..................................... 19
Control Variables ........................................ 20
Task Rigidity........................................... 20
Supervisor Impact ...................................... 22
No Successor Mandate for Change from the Formal
Organization. ......................................... 24
Questionnaire Administration.............................. 25

FOUR DATA ANALYSIS.. .......................................... 28

Results of Factor Analysis............................... 28
Hypotheses One and Two .................................... 30
Hypotheses Three and Four................................ 34









CHAPTER PAGE

FIVE FINDINGS.......... ....................................... 36

Hypothesis One--Groups.................................... 36
Hypothesis Two--Groups.................................... 40
Hypothesis Three.......................................... 41
Hypothesis Four .......................................... 46
Analysis of Individual Respondents........................ 50
Acceptance and Leadership Style of the Successor....... 52
Subordinate Liking of Successor and Predecessor........ 65
Other Findings .............. ............................. 69

SIX CONCLUSIONS .......... .................................... 82

BIBLIOGRAPHY. ............................................ 89

SUPPLEMENTAL BIBLIOGRAPHY................................. 91

APPENDICES

A SUPERVISORY SUCCESSION QUESTIONNAIRE ..................... 96

B SUPERVISORY SUCCESSION QUESTIONNAIRE .....................101

C MANAGEMENT QUESTIONNAIRE ................................. 108

D SUPERVISORY SUCCESSION EPISODE INFORMATION FILE..........110

E SUPERVISORY SUCCESSION JOB CLASSIFICATION FILE...........115

F ORIGINAL CLASSIFICATION OF DELETED QUESTIONS ............. 116

G FACTOR ANALYSIS NUMBER ONE ...............................118

H FACTOR ANALYSIS NUMBER TWO ............................... 121
















Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


AN EXPLORATORY FIELD STUDY
INVOLVING THE EFFECTS OF
SUPERVISORY SUCCESSION ON COMPARATIVE
WORK UNITS

By

Gordon A. Taylor, Jr.

June 1980

Chairman: William M. Fox
Major Department: Management


Succession involves the placement of an individual into a leader-

ship position which his predecessor has vacated due to promotion,

demotion, lateral transfer, hiring, movement to a job with another

company, retirement, death; or any other reason resulting in his

departure from the position. The successor is expected to effectively

handle the requirements of the position formerly held by his prede-

cessor and perform in a manner which will satisfy those who have

selected him for his new job. In this study, supervisory succession

applies to succession situations at the lowest level in the organiza-

tion where the supervisor is responsible for the work efforts of non-

supervisory personnel. From a time perspective, succession concerns

itself with the interplay between what the individual leader does and

various environmental factors during the initial period after assum-

ing his new position rather than what he does after a year on the job.









The primary purpose of this study was to examine the impact of

supervisory succession of specific work groups and the people who make

up these groups. Such information could be beneficial to organizations

if certain factors were usually present that were associated with high

levels of subordinate acceptance of new supervisors. Of the four

hypotheses tested in the study, two dealt with the level of behavioral

congruency between the successor and his predecessor. It was thought

that a new supervisor would want to be thought of as behaving in a

fashion similar to that of a well-liked predecessor and behaving in a

manner dissimilar to that of an unpopular predecessor.

A third hypothesis stated that there would be a negative correla-

tion between subordinate frustration and the acceptability of the

successor to the work group and a final hypothesis stated that there

would be a positive correlation between subordinate job satisfaction

and the acceptability of the successor to the work group.

One hundred and eleven episodes (559 total respondents) were

examined with each episode characterized by task rigidity, strong

supervisor impact on the work group and little supervisor mandate to

initiate change in the work group. Subordinates were asked to

describe the predecessor immediately after his departure and the

successor after he had been on the job approximately six weeks. This

information was gathered from questionnaires that provided data on

predecessor/successor behavior, subordinate frustration and satisfac-

tion levels, subordinate liking of predecessor, subordinate acceptance

of successor, etc.

On the basis of the data gathered there was no difference between

the acceptance levels for congruent versus not congruent successors

vi









who followed well-liked supervisors. Hypothesis two which states that

successors who behave in a manner highly congruent with that of an

unpopular predecessor will have lower acceptance scores than those who

behave in a manner incongruent with an unpopular predecessor could not

be tested due to an inadequate sample size. The results of the study

did indicate both a negative correlation between subordinate frustra-

tion and the acceptability of the successor to the work group and a

positive correlation between subordinate satisfaction and the accept-

ability of the successor to the work group. An important result of

the analysis after using stepwise regression was the emergence of

specific leader behaviors which help explain over 60% of the varia-

tion in acceptance scores.

One of the most important results of this investigation would be

the pinpointing of those types of behavior displayed by new super-

visors that seem to explain why they are accepted by those they super-

vise. Also, possiblereasons why the issue of predecessor/successor

behavior congruency had no significant impact on the results of this

study might be explored.















CHAPTER I
INTRODUCTION



The Succession Territory


Succession involves the placement of an individual into a leader-

ship position which his predecessor has vacated due to promotion,

demotion, lateral transfer, hiring, movement to a job with another

company, retirement, death, or any other reason resulting in his depar-

ture from the position. The successor is expected to effectively

handle the requirements of the position formerly held by his prede-

cessor and perform in a manner which will satisfy those who have

selected him for this new job.

In searching through the available literature, I find that little

has been done from a field research perspective in the entire suc-

cession area. The preceding is not to be construed as a declaration

that we are lacking in data on leadership as such a statement would be

patently absurd. However, succession is essentially the study of

leadership in its formative or early stages as contrasted with much of

the current leadership literature which is postsuccession in scope.

In other words, succession concerns itself with the interplay between

what the individual leader does and various environmental factors

during the initial period after assuming his new position rather than

what he does after a year on the job. The study of succession, then,

is concerned with what happens immediately after a position is vacated





2


and subsequently filled. It is a process that takes place at every

level in the organizational hierarchy from the Chief Executive Officer

(CEO) to those functioning at the foreman or supervisory level.

As noted previously succession is primarily concerned with what

transpires as soon as a successor takes over from his predecessor.

This study did not involve a longitudinal analysis of the succession

phenomenon due to practical limitations but this is certainly not to

be interpreted as a denial of theconceptual relevance of long range

surveys. Likert and others have investigated the affects of time on

leadership styles and concluded that "many leader behavioral changes

will not impact significantly upon subordinate perceptions, inter-

personal behavior, or job performance for anywhere from six months to

two years" (6, p. 7). Consequently, the short-run duration of a

succession study-does not facilitate the use of productivity measures

though this type of criterion is certainly feasible in longitudinal

studies (17). In addition, in a study of brief duration, problems

develop in attributing increases or decreases in productivity to the

right source.

It is important to note that the term, succession, as it is used

here has a broader meaning than that usually employed by those involved

in manpower management or personnel policy decisions. Succession, from

the viewpoint employed here, does not refer to personnel strategies

which are concerned with overall programming of a manager's movement

through various corporate levels.








Succession as a Significant Investigative Area


A given succession episode can have profound effects on an

organization. The transfer of power and responsibility from one

individual to another that takes place when succession occurs can be

disruptive to the organization in general, to a particular department

or departments, or to specific individuals involved contingent on the

combination of relevant variables that surrounds the succession

incident as well as the organization level at which the succession

episode transpires.

The presence of the successor might cause changes in the status

quo regarding acceptable behavior (both job and nonjob related), the

existence of informal groups as well as their structure, the organiza-

tion's decision-making process, the balance of power existing prior to

the successor's entrance, the roles played by various organization

members and a myriad of other consequences. In turn these factors

tend to influence the successor's behavior in his new role. Thus, in

analyzing succession one desires to determine, at least in part. how

the successor affects the everyday operations of the area he now heads

and how he is in turn affected by them.

The subject of succession in industry is by no means new as can

be evidenced by the classic studies of the succession phenomenon con-

ducted by Gouldner (9) and Guest (13). In his book,Patterns of

Industrial Bureaucracy (9), Alvin W. Gouldner outlines the events

which transpire in a gypsum plant when there is an unexpected change in

plant management and an "outsider" is brought in to take charge of the









plant. The subsequent succession crises faced by both the workers and

the new plant manager, Peele, is a classic case study relating to the

issue of succession and its effects on organizational behavior.

Robert H. Guest in Organizational Change: The Effect of

Successful Leadership (13) discusses another succession situation

which invloves an automobile assembly plant. The contrasting behavior

between the actions of Gouldner's plant manager, Peele, and Guest's *

plant manager, Cooley, vividly depicts the difficulties which must be

faced in a succession situation. The different views of their predica-

ments by Peele and Cooley indicate the importance of not only the

successor's impression of his position but also the significance of

the environment into which he is thrust. A given successor with a set

approach will evoke different reactions in different organizations.

The reaction of the workers at Gouldner's gypsum plant to Peele's suc-

cession technique of "punishment-centered" administration might have

been different at Guest's Plant Y. By the same token, there is no

guarantee thatCooley's "representative" administrative approach would

be successful in all instances. Also, there is evidence that the

abilities of the predecessor and the pattern of relationships asso-

ciated with him can significantly moderate the impact on the succes-

sor's behavior (9, p. 17). In addition, the leadership milieu at the

top of the organization, the type of supervision that the supervisor

receives from his boss and the system of rewards and penalties utilized

in the organization can have a dramatic impact on a supervisor

(successor) (7, p. 221).

There is little question that succession is a subject of con-

siderable value to those of us who take an interest in the interpersonal





5


dynamics of organized group activities be they in business, education,

the military, hospitals, or other such goal oriented groups. As

Oscar Grusky has so cogently stated:

. all organizations must cope with succession.
One reason for this is so obvious that it is easily
overlooked--man is a mortal creature. Thus, all
organizations must at some time be confronted with
the necessity of making substitutions at all levels
of the hierarchy if the system is to continue
functioning. (11, p. 106)

Grusky correctly states that succession is worthy of our time if for no

other reason than it is an inevitable aspect of organized group

behavior.

Edwin P. Hollander, in identifying areas in leadership that

warrant further research, mentions succession when he states, "Indeed,

succession is an area of leadership which urgently demands more study"

(14, p. 30). His concern involves the use of leader "credits" to

provide support for the successor. The more "credits" an individual

has the easier it will be for him to become established in the work

group. Fiedler and Chemers in a vein similar to Hollander's also

comment on the relative lack of research that has been conducted in the

area of leadership succession (4. p. 115).

Bernard Levenson looks upon succession as a process that affects

the entire organization.

Succession to office at the top, of course is some-
times accompanied by explosive changes in organiza-
tional structure and policy. But the viability of
an organization requires that vacancies be filled
on all levels, not merely on the top level. (17, p. 363)

As mentioned earlier, succession is an area of inquiry which

has applicability to all sorts of organized group endeavors (1, 2. 8,

12, 21). As much of the preceding material indicates, a somewhat









perplexing aspect of the succession issue as it concerns business

revolves around the fact that it has been largely the province of

sociologists, political scientists, anthropologists and those inter-

ested in public educational administration--not those in business and

organizational behavior. This is particularly strange when one con-

siders the preoccupation many business scholars seem to have in the

areas of interpersonal relations and group dynamics. Succession

certainly is an intriguing and important aspect of corporate life which

would lead one to suspect that it should be of interest to those

interested in business research.


Supervisory Succession


The specific subject area to be investigated in this study

involves supervisory succession. As discussed here, supervisory suc-

cession applies to succession situations at the lowest level in the

organization where the supervisor is responsible for the work efforts

of non-supervisory personnel. Examples would be a foreman on a pro-

duction line or an individual responsible for the work of clerks such

as an office supervisor. This is to be differentiated from executive

or managerial succession which is involved at levels above the first

line supervisor. Paramount to the distinction between the terms

"executive" or "managerial" and "supervisory" employed here is the

fact that supervisors head a nonmanagement family--the workers

(3, p. 119).










A supervisor's job differs from that of other
managers because the group he supervises is
different. This situation requires him to
interact in an authority relationship with two
groups: (1) workers, who are his subordinates,
and (2) managers, who are his superiors. Any
manager who is not a supervisor interacts
primarily with two levels of only one group,
namely, managers, who are both his superiors
and his subordinates. '(3, p. 119)

The reason this investigation will concentrate at the supervisory

level is that, by analyzing comparable work groups, associations can be

found concerning the relationship of succession on such groups.

Keying the analysis at the supervisory level is necessary in order to

facilitate the generation of an adequate sample size. One of the

critical problems of undertaking a study of succession at levels

above that of first line supervisors is that of generating a satis-

factory sample size in a given time frame. As one moves to the levels

of managerial or executive succession, the possibility of getting

similar work groups in satisfactory numbers within a specified time

limit becomes increasingly remote.

As previously mentioned, by specifying that one examine similar

work groups the possibility of using chief executive officers, division

vice presidents or plant managers is precluded out of hand. A major

barrier to deriving information at the Chief Executive Officer (CEO)

level is that each CEO position is uniquely different from all other

such positions including the environments surrounding and influencing

each organization. Studies of executive succession are interesting

and worthwhile but there is some doubt regarding the statistical

reliability of the information generated between CEO positions. The

classic studies of Gouldner and Guest make fascinating reading but









their value is anecdotal in nature whereas v /th supervisory succession

one can come up with solid research that cr I be used to compare one

group and supervisor with similar groups aAd supervisors. The short-

comings outlined herein for executive succession apply to a lesser

degree for managerial succession and th' s supervisory succession

becomes the primary vehicle for the replication of data on the suc-

cession phenomenon.


Purpose of t\l& Study


There are numerous compelling reasons which justify the need for

scholarly research in the entir- succession area. Briefly stated,

they are:

1. Succession talas place in all viable organizations.

2. Succession c n cause organizational instability.

3. Succession -an be a major organizational change
strategy.

4. Very litf-ie research exists which has as its
primary 'ocus, the effects of succession on
organi? ,tions and the individuals who make up
those organizations.

These reasons certa- ily apply to the whole succession spectrum, be they

at the executive, anagerial or supervisory level. The need for

further research .nto the mechanics and effects of succession is

obvious and longer overdue. The primary focus of this study is to

explore number four above which involves the impact of supervisory

succession or specific work groups and the people who make up these

groups. In analyzing the impact of succession on work groups an

attempt will be made to determine if it is possible to develop a









contingency framework for studying succession situations such that

when a new supervisor is brought in, his chances of success in terms

of being accepted by the group can be better determined in advance.

Since succession can cause a certain amount of discomfort for

those affected, a possible by-product of the study would be to pre-

scribe specific behavior patterns to the successor such that the

trauma experienced by organization members can be kept at a low level

in situations where such a state is desired by management. If the

transition between predecessor and successor can be made smoothly with

the new supervisor being readily accepted by his subordinates, the

desirability of such a condition in terms of reduction of subordinate

anxiety and uncertainty is apparent.

The fact that this study concentrates on supervisory succession

as it affects work groups and the people who make up these groups is

not meant to slight other reasons for studying the succession phe-

nomenon but rather to suggest the breadth of the succession area and

indicate the need to begin work analyzing the myriad aspects of

succession.















CHAPTER II
HYPOTHESES


The primary purpose of this study is to examine the impact of

supervisory succession on specific work groups and the people who make

up these groups. Four hypotheses were analyzed regarding different

aspects of succession. The experimental nature of this analysis will

no doubt lead to other hypotheses which can be investigated in sub-

sequent research.


Predecessor/Successor Leadership in Terms of
Congruency and Successor Acceptability


Two hypotheses deal with the issue of behavioral congruency be-

tween successor and predecessor as viewed from a leadership perspec-

tive. Hard data regarding these hypotheses at the supervisory level

are lacking and consequently the justification of these statements as

viable hypotheses is based upon reasonable speculation on the part of

the researcher.

It is anticipated that congruency of behavior between predecessor

and successor and subordinate like or dislike of the predecessor will

have an effect on the results of the succession episode.

Subordinates can be friendly or unfriendly toward
a successor, and what they think of him depends
very much upon what they thought of his predecessor.
The successor's fate at first depends less upon
what he does or who he is and more upon how he is
seen to compare with "Old So-and-so," his predecessor.
(10, p. 26)










Specifically, one would expect that if the successor's behavior is

highly congruent with that of a well-liked predecessor, then the suc-

cessor will be more readily accepted by his subordinates than if his

behavior is highly incongruent with that of a well-liked predecessor.

The predecessor was used as a standard against which the new super-

visor could be judged. This standard may be high or low depending on

the favorability with which the subordinates viewed the predecessor.

Furthermore, the similarities between the predecessor and successor as

perceived by the subordinates are a key point here. Intuitively, one

would expect the new supervisor to fare batter if he behaves in a

consistent manner with a highly regarded rather than unpopular prede-

cessor. A new supervisor, therefore, will wish to be thought of as

behaving in a fashion similar to that of a well-liked predecessor and

behaving in a manner dissimilar to that of an unpopular predecessor.


Hypothesis One


If the successor's behavior is highly congruent with that of a

well-liked predecessor, then the successor will be more readily

accepted by his subordinates than if his behavior is highly incongruent

with that of a well-liked predecessor.


Hypothesis Two


If the successor's behavior is highly congruent with that of an

unpopular predecessor then the successor will not be as readily accepted

by his subordinates than if his behavior is highly incongruent with

that of an unpopular predecessor.









Postsuccession Frustration and Successor Acceptability


It is felt that if subordinates feel frustration after the suc-

cession episode has taken place, it is quite possible that this dis-

comfort will manifest itself, at least in part, by a low level of

subordinate acceptance of the new supervisor. The tension or frustra-

tion which subordinates might experience is viewed here as resulting

from those factors that relate to the worker's overall job experience

in terms of such things as having adequate authority to carry out one's

responsibilities, the clarity of his assignment, work load reasonable-

ness and availability of information necessary to complete one's job

satisfactorily.

The succession episode itself may or may not be partially respon-

sible for discomfort felt by subordinates due to inadequate support in

the aforementioned areas. The key issue revolves around whether or

not subordinates feel frustrated and not necessarily who is account-

able for the existing frustration. It is possible that frustration,

if there is any, existed prior to when the new supervisor took over

and thus the successor could not be held accountable for this state

of affairs. Inadequate information and resources necessary to do the

subordinate's job properly in many cases may not be directly under the

control of a first line supervisor. Consequently, nonsuccessor induced

frustration will continue to burden a work group, making it difficult

for a new supervisor to gain acceptance from his subordinates. Though

the successor may not be able to resolve the existing frustration he

may find that he is hindered by it as he attempts to gain work group

acceptance.










Hypothesis Three


There is negative correlation between subordinate frustration and

the acceptability of the successor to the work group.



High



Group
Subordinate
Frustration




Low _
Low High

Acceptability of successor



Level of Subordinate Satisfaction and Successor Acceptability


If, after the transfer of power from predecessor to successor has

been completed,the work group is relatively content regarding such

issues as pay, work relations, appreciation by other workers and the

nature of specific job assignments then it can be assumed that the

new supervisor will be more readily accepted than if such is not the

case. The reasoning behind this supposition is that if workers are

basically satisfied with their overall work situation after the

exchange of responsibility from predecessor to successor then there

is little need for the employee to reject the new supervisory incum-

bent. However, if one perceives his pay to be inadequate for the

work effort expended, if coworkers are distant and uncooperative on

the job, if he feels generally unappreciated and if one finds his work

assignment to be less than interesting, it is quite possible that he









might attempt to vent his dissatisfaction on the new supervisor since

he is the worker's closest contact with the management team of the

organization.

It is important to add that whereas Hypothesis Three is specifi-

cally concerned with worker frustration due to inadequate time, infor-

mation and resources with which to carry on one's job assignment,

Hypothesis Four states that even if such resources are adequate/

inadequate the worker can still be satisfied/dissatisfied with his

overall job situation relative to what he does, whom he does it with

and to what extent he is remunerated for his efforts. To reiterate,

it is quite possible that an individual truly enjoys what he does on

the job, whom he does it with, and how much he gets paid while at the

same time feels frustrated in that he perceives he could do even

better were he given more time, more resources and a clearer delinea-

tion of his specific work assignment.


Hypothesis Four


There is a positive correlation between subordinate job satisfac-

tion and the acceptability of the successor to the work group.



High


Subordinate
Satisfaction
Level

Low
Low High

Acceptability of
Successor
















CHAPTER III
METHODOLOGY



The Sample


Various institutions were selected contingent on organizational

receptiveness to the goals of the research effort and apparent organi-

zation compatibility with the level of task structure necessary in the

work groups to be studied. Specifically, access to organizations was

the primary factor in determining if a given firm would be included in

the initial screening process. Access was gained in three major ways:

1. researcher association with members of management
of prospective firms,

2. contacts of researcher's colleagues with members of
management of prospective firms, and

3. cold calls by researcher on area organizations that
might prove to be suitable.

Once an organization had been contacted and found to be poten-

tially receptive to the project it was necessary to determine if they

expected any turnover in the months ahead among their first-line super-

visors. If they did, then they were still a possible participant in

the study. If not, they were eliminated from further consideration.

After this step, the researcher usually observed the kinds of work

being done to determine in a broad sense the degree of task structure

present. Any organization that was characterized by nonrepetitive task

assignments at the operative level was eliminated out of hand. This










further reduced the pool of potential prospects since the constraint

of high task structure was a critical control variable in the imple-

mentation of the study.

Finally, if it was established that the work done in a given area

was repetitive and that there was to be a change in supervisors then

an informed member of management was asked if the new supervisor was

expected to initiate change in his area. If so, then the organization

was not considered for inclusion in the study. If not, the organiza-

tion was considered as a possible participant.

The size of the organizations involved varied from a low of

around thirty employees to corporate giants numbering in the thousands.

Large organizations, due to the number of employees, generally had

more turnover and consequently more succession episodes. Of the

thirty-five organizations utilized in the study, nineteen were

directly involved in manufacturing yet these nineteen organizations

accounted for eighty-three of the 111 episodes included in the study.

Initial screening concerning possible turnover, level of task

structure and apparent lack of mandate for change resulted in 115

episodes which appeared to be usable. However, after analyzing the

groups in terms of the control variables employed the net sample con-

tained 111 separate work groups. The data are characterized by the

following:

1. 111 episodes with a range from 2-15 subordinates with
a total of 559 subordinates involved.

2. Average group size is 5.036 subordinates.

3. 35 organizations used in the study--number of episodes
per organization ranged from 1-18.









4. Number of subordinate occupational classifications was
11 with 90 of the 111 being either production workers
or clerks.


The Questionnaires


Copies of the questionnaires employed in the study can be found

in the Appendices. A brief explanation of the make-up of the three

questionnaires employed follows.


Description of Previous Supervisor


Questions 1-44 in Appendix A deal with the leadership behavior

of the predecessor on the job as viewed by the subordinates. Questions

were utilized which were intended to measure decisiveness, goal

emphasis, support, consultative-participative decision making, work

facilitation, management of contingencies and rule enforcement on the

part of the previous supervisor. Many of the questions employed in

this part of the questionnaire were taken from the Modified Leader

Behavior Description Questionnaire developed by Fox (5, p. 4) with

assistance from Stogdill and Coons (20). The questions dealing with

management of contingencies were taken from Reitz (18, p. 230) and

modified to maintain continuity with the form utilized in the

questionnaire.

Questions 45 and 46 deal with the acceptability of the previous

supervisor to the work group. Question 47 concerns the extent to

which subordinates respected their previous supervisor. Questions

48-52 were designed to generate an indication of how much subordinates

liked their previous supervisor. The questions used to generate a

measure of liking were inspired by terms used by Scott (19) though they










have been placed in a Likert-scale format to conform to the structure

employed throughout this study.


Description of Current Supervisor


Questions 1-52 in Appendix B in the subordinate description of the

current supervisor are identical to those used in the subordinate

description of the previous supervisor except that for the successor

all statements are in the present tense. Additional questions beyond

those used to describe the successor's behavior are included in that

instrument and they begin with five questions (55, 57, 60, 63, 66)

concerned with subordinate postsuccession frustration which deal with

the subordinates' discomfort due to inadequate time, information and

resources with which to carry on one's job assignment. These questions

have been adapted from items used by Indik, Seashore and Slesinger in

an article on employee psychological strain (16, pp. 28-29). Ten

questions (53, 54, 56, 58, 59, 61, 62, 64, 65, 67) also adapted from

Scott's semantic differential (19) were initially used to derive a

measure of subordinate job satisfaction regarding such issues as pay

and working conditions.

Question 68 deals with whether or not subordinates felt there was

a need for major change in their department. Questions 69-74 provide

a descriptive measure of task rigidity as viewed from the subordi-

nates' perspective (15, pp. 44-45). Finally, question 75 asks why the

previous supervisor departed and question 76 asks the subordinate

where his new supervisor previously worked.









Management Questionnaire


A total of fourteen questions were asked of an individual in the

management hierarchy who has considerable familiarity with both the

type of work the subordinates in question are performing and the level

of power held by the supervisor to whom these subordinates report.

Questions 1-5 and 14 deal with task rigidity and are the same questions

posed to subordinates in the current supervisor questionnaire.

Question 13 concerns whether or not management expected the new super-

visor to initiate major changes in the work group. Questions 6-12 deal

with the ability of the new supervisor to exact compliance from his

subordinates. As such, an attempt was made to determine how much

genuine clout the new supervisor has over his subordinates (4,

pp. 68-69).

The aforementioned provides a brief summary of questionnaire

content for the three instruments used.


Questionnaire Scoring (Items 1-67)


To determine relevant factors all questions were scored A=1,

B=2, C=3, D=4, E=5. One analysis was made for questions 1-52 and

included subordinate responses concerning both predecessors and suc-

cessors. The results of this analysis and resulting factor loadings

on leadership dimensions, acceptance and liking can be found in

Appendix G. A second analysis was made involving just the responses

of subordinates as they described their current job situations

(Questions 53-67) in terms of levels of frustration and satisfaction

perceived on the job. The resulting factors and their loadings can be









found in Appendix H. It is important to note that some questions have

negative loadings due to the manner in which they were worded (3, 7,

14, 15, 19, 23, 25, 28, 31, 42, 49, 50, 52, 55, 57, 58, 60, 62, 63,

66).


Control Variables


Questions dealing with task rigidity, successor supervisory impact

and successor mandate for change (control variables) were designed to

ensure that the researcher was working with comparable work groups.

It was assumed that the aforementioned variables would change little

from succession group to succession group. Some variation along these

dimensions was certainly possible but the control variables need to be

roughly equivalent in the succession groups studied in order to match

the work units involved.


Task Rigidity


Task rigidity involves the extent to which a job activity is well-

defined, clear-cut, and unambiguous. As the level of task rigidity

increases, the degree of creativity and ingenuity potential on the

part of subordinates lessens. Task rigidity is significant as a con-

trol variable since it is important that workers involved in the study

are limited in the amount of originality they can bring to their job.

By making sure jobs performed by the various respondents were similar

in the degree of structure present, similarity of the various groups

being studied was increased.

Task rigidity was determined by administering essentially the

same job related questions to both subordinates performing the tasks










and management personnel familiar with the job being analyzed. Spe-

cifically, questions 69 and 70 on the subordinate questionnaire

(Appendix B) and questions 1 and 2 on the management questionnaire

(Appendix C) were used to derive a measure of task rigidity. The

Appendices can be consulted for specific questions related to this and

other dimensions employed in the study. Response opportunities ranged

from A (which translates numerically to 1) or a high degree of

structure to E (which translates numerically to 5) which indicated a

low level of structure. Averaging questions 1 and 2 on the management

questionnaire resulted in the following distribution:


Average Frequency

1.00 32
1.50 41
2.00 35
2.50 7

TOTAL 115


All 115 averages fall within the high task rigidity portion of the

response profile. None of the averages fell in the area characterized

by low task rigidity (4-5) and consequently the managerial appraisal

of the degree of structure present is high in all cases. Even the

seven episodes with an average of 2.5 are countered by subordinate

averages that also fall in the high task rigidity range:


Episode Number Management Average Subordinate Average

25 2.50 1.87
27 2.50 1.83
57 2.50 1.33
63 2.50 2.75
93 2.50 1.87
105 2.50 1.5
111 2.50 2.0










Averaging the responses of Questions 69 and 70 for the subor-

dinates resulted in the following distribution:


Average

1.00-1.50
1.50-2.00
2.00-2.50
2.50-3.00


Frequency

24
53
30
8


Again, no averages appear in the range (4-5)where subordinates per-

ceived that there was a low level of task rigidity present in their

jobs. The eight episodes that do fall within the 2.50-3.00 range are

countered by management averages that are no higher than 2.5.


Episode Number

63
76
77
79
81
88
90
99


Subordinate Average

2.75
2.67
3.00
2.67
3.00
2.58
2.72
3.00


Management Average

2.5
1.0
1.5
1.0
1.0
2.0
2.0
2.0


None of the original 115 episodes needed to be lost based on the issue

of task rigidity since there were no episodes in which both management

and subordinate averages to the questions involved fell in the low

task rigidity range.


Supervisor Impact


Supervisor impact concerns the ability of the new leader to in-

fluence the work group in terms of such issues as subordinate perfor-

mance evaluation and supervisor direction of workers in the fulfill-

ment of their job requirements. In addition, the support given by the










formal organization to the new supervisor to help him exact subordinate

compliance to department objectives is involved. It is felt that

individuals who have a strong supervisor impact will find it easier to

get cooperation from subordinates than those having weak supervisor

impact. However, it should be pointed out that the degree of super-

visor impact alone will not assure subordinate adherence to leader

directives since his acceptance by group members as well as other

factors could very well serve to mitigate or counterbalance his

influence on the group.

To derive a measure of supervisor impact, questions 6-12 in the

management questionnaire (Appendix C) were averaged for each episode

with the lower the score the higher the degree of supervisory impact

indicated. The following distribution resulted:


Average Frequency

1.29 2
1.43 2
1.57 14
1.71 14
1.86 16
2.00 27
2.14 19
2.29 11
2.43 7
2.57 3

TOTAL 115

All averages are below three (none in low impact range) for all

115 episodes so it is assumed that for the purposes of this study suc-

cessors find themselves in situations where they have a substantial

degree of supervisor impact on their respective work groups.










No Successor Mandate for Change From the Formal Organization


For the purposes of this study, only succession episodes in which

the successor had little or no mandate to initiate changes in the work

were analyzed. The reason for the imposition of this limitation on

the study group is quite simple. :In examining situations at the

supervisory level, the majority of cases will frequently involve

transfers of power where no great changes in the status quo are sought

by management. The formal organization structure stresses the need

for continuity in such cases such that the work group can continue to

function in as undisturbed a fashion as possible during the transition

period when the predecessor departs and the successor arrives on the

scene.

There are no doubt situations when management feels that change

is needed at the operative level and the best way to stimulate that

change is to bring in a new supervisor and expressly inform him of the

fact that part of his job is to shake-up the status quo and get his

work group moving in a more productive fashion than was the case under

the predecessor. Yet, from the standpoint of matching groups, such

situations were deemed to be the exception rather than the rule. The

greatest number of succession episodes at the supervisory level are of

little or no mandate for change variety and consequently only such

groups were included in the study.

A member of the management team familiar with a given episode

was asked if the new supervisor was expected to act as a major change

agent in the work group. This information was found in question 13

(Appendix C) which was reverse scored such that A=5 and E=l. This was









done in order to make it easier to conceptualize the results in that a

high score indicates that management feels a major change is in order

whereas a low score indicates that little, if any, change is deemed

necessary by management. The original 115 responses from management

on the issue of mandate for change (Question 13) resulted in the

following frequency distribution:


Response Frequency

1 71
2 40
3 3
4 0
5 1

TOTAL 115


Using three as a midpoint it was decided to drop the four episodes

with responses of three or higher since in these situations there was

no clear mandate for the maintenance of the status quo. Therefore, the

sample ultimately used in the study consisted of 111 separate succession

episodes all characterized by task rigidity, strong supervisor impact,

and little, if any, supervisor mandate to initiate change in the work

group.


Questionnaire Administration


As can well be imagined, the topic of supervisory succession is a

difficult one to investigate due to the very real problem of generating

an adequate sample size over a specific time frame. Since succession

is a continuing process that takes place over time in a less than pre-

dictable manner, the effort involved to generate a sample size of 111

episodes was considerable. However, as mentioned earlier, the case









study approach simply was not acceptable considering the goals of this

study. For this same reason, in-depth interviews were also inappro-

priate. This left the researcher with the questionnaire format as the

only viable tool with which to gather data. In this regard, a mail

questionnaire was also not feasible due to the complexities involved

regarding both timing and administration of the survey instrument.

It was necessary to administer the questionnaire over a five-month

period since over 100 supervisory succession episodes, each charac-

terized by, a high degree of task rigidity, strong supervisor impact,

and little, if any, mandate for change from the formal organization,

were included in the study.

The questionnaire concerned with the previous supervisor

(Appendix A) was given as soon as possible after that person departed

to insure that subordinate reaction was obtained while the image of

the predecessor they held was still fresh in their minds.

The questionnaire pertaining to the current supervisor was

administered six weeks after that individual had taken over. The

decision to wait six weeks between subordinate description of the

previous supervisor and subordinate description of the current super-

visor was based on the results of a pilot study. The purpose of the

trial run was to determine at what point in time subordinates felt

most comfortable in describing the successor. The key was to deter-

mine the minimum waiting period between predecessor departure and

successor takeover in terms of insuring objectivity on the part of the

subordinates in describing their new supervisor. In addition, a time

gap in questionnaire administration was critical in order to minimize

the impact of the predecessor's behavior on subordinate description of










the current supervisor. Questions regarding successor behavior were

given to three test groups at three-week intervals (3 weeks, 6 weeks,

and 9 weeks) after the new supervisor had taken over. After the third

and final administration of the successor questionnaire (9th week),

subordinates were asked collectively at which administration of the

questionnaire they felt they could be most objective in describing the

successor's behavior.

It was felt that giving the questionnaire immediately after the

successor takes over would be unwise since subordinates would not yet

have had sufficient time to be able to describe the successor's

behavior. In addition, there might be evidence of a "honeymoon"

period between successor and subordinates immediately after the trans-

fer of power and time would be needed for this to wear off, if, in fact,

it exists at all. On the other hand, if too much time passes between

successor takeover and subordinate description, the study runs the

risk of being an examination of leadership rather than succession.

The results of the pilot study indicated that each group overwhelmingly

supported a six-week time span between the subordinate descriptions of

the previous and current supervisor. Each group stated that they felt

that three weeks was not enough time to get to know the successor but

by the same token it didn't take nine weeks to feel comfortable

describing his behavior either.

A questionnaire was also administered to the successor's super-

visor and this instrument can be found in Appendix C. Timing was not a

critical issue in the administration of this instrument and it could

therefore be completed at the discretion of the successor's supervisor.
















CHAPTER IV
DATA ANALYSIS



Results of Factor Analysis


The 44 questions on behavior'which were originally thought to

indicate seven aspects of leadership were factor analyzed with sub-

ordinate responses for both the predecessor and successor. After

factor analysis some questions were dropped and some dimensions were

modified based on the results of the factor analysis. A listing of

specific questions dropped from the final analysis can be found in

Appendix F. The three leadership dimensions which resulted from the

factor analysis and their respective loadings in descending order can

be found in Appendix G. The original dimension to which each question

had been assigned prior to factor analysis is also indicated. To be

included in a dimension, questions had to have a factor loading of .5

or higher and the loading had to explain at least twice as much vari-

ance as any loading for the other two factors. Dimension One was

called Supportive Recognition due to the overall orientation of most

of the questions which fell there. Dimension Two is called Decisive-

ness since all four of the questions which fell there were in the

original list of questions which it was initially thought would give

a determination of that dimension. Dimension Three was called Rule









Enforcement since all three questions initially thought to measure

that dimension did so in addition to the other two which also have a

Rule Enforcement orientation.

Factor analysis produces three types of factors. A general

factor has all variables (questions) with sizable loadings. A group

factor has a set (or group) of variables that load together. A

specific factor has only one variable with a significant loading. The

three leadership dimensions in this study are group factors.

Acceptance was determined by averaging responses to questions 45

and 46 as they both loaded together and were the same questions it was

felt initially would generate a measure of acceptance. In order to get

a measure of acceptance, questions 45 and 46 were reverse scored (A=5,

B=4, C=3, D=2, E=l) such that a high score is indicative of a high

degree of successor acceptability to subordinates. Subordinate liking

of the predecessor was determined by averaging questions 48-52 since

these five questions had significantly high loadings in one factor and

were the questions it was initially thought would generate a measure

of liking. Questions 48 and 51 were reverse scored such that a high

liking score indicates that predecessors were liked by their

subordinates.

Frustration was determined by averaging questions 55, 57, 60 and

66 since these four questions had significantly high loadings in one

factor and are four of the five questions it was felt would initially

measure frustration. These questions were reverse scored so that a

high score would indicate a high level of frustration for subordinates.

Satisfaction was determined by averaging questions 53, 59, 61, 64 and

65 since there were loadings for questions 53 and 65 in one factor,









questions 59 and 61 in a second factor, and question 64 in a third

factor. The reason for using all five questions is that it would seem

implausible to rely on just two of the original ten questions

designed to measure subordinate satisfaction. All three factors seem

to embrace different elements under the umbrella of satisfaction. To

get a satisfaction score questions 53, 59, 61, 64 and 65 were reverse

scored so that a high score indicated a high level of subordinate

satisfaction.


Hypotheses One and Two


Hypotheses One and Two involve a measure of congruency between

the predecessor and successor within each episode. Three leadership

dimensions were used to generate a congruency measure. The final

make-up of these dimensions was determined after the forty-four

leadership questions in the 111 usable episodes were factor analyzed.

Both common space and principle component were used to factor

analyze questions 1-44 in order to determine which questions had

salient loadings in a particular factor. As mentioned earlier, ques-

tions were used in a dimension only if they met two criteria. First,

the loading had to explain at least twice as much variance as any

loading for the other two factors and,secondly, had to be high enough

(.50 or better) to merit inclusion in a dimension. As mentioned

earlier, negative loadings exist for some questions due to the way they

were worded (3, 7, 14, 15, 19, 23, 25, 28, 31, 42).

There are no statistical tests to determine if a given dimension

is more important than the others in determining congruency. Although

Factor (Dimension) One has more questions with large loadings than Two










or Three, one cannot argue that it is a "better" measure of con-

gruency. Since there is no way to determine the "best" dimension,

there is no way to determine if a combination of two dimensions is

more significant than another combination. In other words, con-

gruency on Dimensions One and Two must be interpreted as being as

important as congruency on Dimensions Two and Three or One and Three.

A number of statistical procedures were considered to establish

a measure of congruency on the three dimensions. Several of these

methods had to be eliminated since they were not relevant to the data

being used. A test of significance between mean responses on prede-

cessor and successor dimensions cannot be used. This technique would

involve the assumption that the employee responses used (the sample)

were randomly selected from the population of employees reporting to a

given predecessor and successor. The questionnaire was administered

to all employees in a given succession episode rather than to some

randomly selected sample.

Discriminant analysis is a method of determining linear combina-

tions of the original predictor variables that show large differences

in group means. The original variables are derived from two or more

known groups. The analysis, then, is usually used to determine the

ability of these variables to discriminate among the groups and to

predict the group membership of new observations. In this study, the

groups are not known. Rather, it is the group classification which is

being determined.

Cluster analysis is a term used to describe a collection of

techniques used to group multidimensional entities according to various

criteria of their degrees of homogeneity and heterogeneity. Many of









these techniques involve a minimum variance or minimum distance

function. Several attempts were made using various alternative

methods with the data from the original 111 episodes. It became

apparent that the results were based on a comparison of one episode

to another (between episodes) rather than the degree of congruency

of the predecessor and successor (within an episode).

The problem of congruency is in determining whether the prede-

cessor and successor were similar in leadership style. This is not

the same as determining if the predecessor was in some way "better"

than the successor or vice versa. For example, suppose the following

averages were calculated for the significant questions:


Dimensions

One Two Three

Pred. Succ. Pred. Succ. Pred. Succ.

Episode A 2.8 1.9 2.2 2.4 2.2 3.7

Episode B 3.1 3.0 2.4 3.4 1.2 2.3


Assume the questions are coded such that "1" is "always" (good) and "5"

is "never" (poor). In episode A the predecessor could be considered

"worse" on Dimension One and "better" on Dimension Three. In this

case the differences between predecessor and successor averages would

be .9 (One), -.2 (Two) and -1.5 (Three). Using +.5 as the criterion,

the predecessor/successor combination is not congruent on Dimensions

One and Three even though one difference is positive and one negative.

In episode B the differences are .1 (One), -1.0 (Two) and -1.1

(Three). In this case they are not congruent on Two and Three and are

both negative. In episode A the predecessor was "better" once and










"worse" twice. If the problem is simply one of determining simi-

larity, then it can be argued that they are not congruent in both

episodes. Hypotheses One and Two are initially concerned with the

issue of congruency and not whether one supervisor was better or worse

than any other.

The technique ultimately used was to calculate the average of

the subordinate responses to the questions comprising each dimension

for the predecessor and for the successor. Within each episode the

difference between the predecessor and successor means was determined

for each dimension. If this difference was less than +.5, the prede-

cessor and successor were similar on the dimension. Finally, if a

successor/predecessor combination was similar on two or three

dimensions, they were said to be congruent. Likewise, if they were

dissimilar on two or three dimensions, they were defined as not

congruent.

This technique may appear to lack a degree of statistical sophis-

tication. However, it overcomes the problem of comparing predecessors

and successors in one episode with those in another episode which is

not the purpose of this study. It is a way, however, of comparing a

predecessor to a successor within a work group on the three established

dimensions. The difference of +.5 was not arbitrarily selected. For

each of the three dimensions, the average of the responses for the

successor was subtracted from the average of the responses for the

predecessor. For Dimension One, for instance these differences ranged

from +2.56 to-2.27. As mentioned above, the purpose was not to

determine if the predecessor was better (worse) than the successor but

rather if their leadership styles were congruent (not congruent). For










each of the three dimensions, the differences between averages were

arranged in an ordered array from most positive to most negative. In

each array, the differences appeared to be greater than +.5, between

+.5 and -.5, or less than -.5. Based on this information, it was

subjectively decided that differences of less than +.5 would be

considered similar and differences of more than +.5 would be

dissimilar.


Hypotheses Three and Four


For Hypothesis Three, acceptance and frustration were determined

as outlined previously. To analyze the data, the responses for the

frustration questions and acceptance questions were averaged for each

individual. These averages were correlated over the 559 subjects to

determine the relationship between acceptance and frustration. If the

correlation coefficient was significant and negative, Hypothesis Three

would be substantiated. The average of the responses in each episode

were also correlated to determine the group acceptance/frustration

relationship. Hypothesis Three could also be phrased as the degree of

subordinate acceptance being dependent on the level of frustration.

Frustration and acceptance for each individual were divided into high,

medium and low categories. Chi-square analysis was employed to test

this dependency relationship. The same procedure was applied to test

the dependency within episodes.

For Hypothesis Four. acceptance and satisfaction were again deter-

mined as outlined earlier. To analyze the data, the responses for the

satisfaction questions and the acceptance questions were averaged for

each individual. These averages were correlated over the 559 subjects









to determine the relationship between acceptance and satisfaction. If

the correlation was significant and positive, Hypothesis Four would be

substantiated. The averages of the responses in each episode were also

correlated to determine the group acceptance/satisfaction relationship.

Hypothesis Four could also be phrased as the degree of subordinate

acceptance of the new supervisor being dependent on the level of

satisfaction. Satisfaction and acceptance for each individual were

divided into high, medium and low categories. Chi-square was employed

to test this dependency relationship. This procedure was also applied

to each episode to test the dependency relationship.
















CHAPTER V
FINDINGS



Hypothesis One-Groups


Hypothesis One states that if the successor's behavior is highly

congruent with that of a well-liked predecessor, then the successor

will be more readily accepted by his subordinates than if his behavior

is highly incongruent with that of a well-liked predecessor. The

manner in which congruency, liking and acceptance were determined has

been explained earlier in this paper. However, the congruent/

noncongruent dichotomy has not been presented. Group acceptance and

liking scores have also not yet been presented. This information can

be found in Tables One, Two and Three.


Table One

EPISODE CONGRUENCY

Three Dimensions Two Dimensions Subtotal

Congruent 25 30 55


Noncongruent 22 34 56


TOTAL 111










Table Two

SUCCESSOR ACCEPTABILITY TO SUBORDINATES

Acceptance Scores Number of Episodes

1.00 1.49 1
1.50 1.99 0
2.00 2.49 3
2.50 2.99 6
3.00 3.49 10
3.50 3.99 20
4.00 4.49 36
4.50 5.00 35

TOTAL 111


Table Three

SUBORDINATE LIKING OF PREDECESSOR

Liking Scores Number of Episodes

1.00 1.49 1
1.50 1.99 1
2.00 2.49 3
2.50 2.99 5
3.00 3.49 10
3.50 3.99 21
4.00 4.49 47
4.50 5.00 23

TOTAL 111



The congruent/noncongruent breakdown resulted in a highly

fortuitous 55/56 split which provided numerous episodes in both the

congruent and noncongruent groups. As Tables Two and Three indicate,

the results of the group acceptance and liking scores were not as

evenly divided with the majority of scores being found in the high

acceptance and high liking range.

Using the data in Tables Two and Three. liking and acceptance

were broken down so that scores less than three are representative of a










low level of subordinate liking of the predecessor and low subordinate

acceptance of the successor and scores greater than three represent a

high subordinate acceptance of the successor. The resulting classifi-

cation for congruent episodes is found in Table Four.


Table Four

WITHIN EPISODE CONGRUENCE

Liking of Predecessor


0
O CO
Low


U High
< o


Low High

0 2


0 53


0 55 55


Again using data from Tables Two and Three, liking and acceptance

were broken down so that scores less than three indicate low liking

and acceptance, a score of three on acceptance indicates a medium

level of acceptance, a score greater than three indicates high liking

and high acceptance. The resulting classification for noncongruent

episodes is found in Table Five.


Table Five

WITHIN EPISODE NONCONGRUENCE

Liking of Predecessor
Low High

4 Low 7 1 8
0

o Medium 2 0 2

< High 1 45 46
< o High 1 45 46


46 56










In order to analyze Hypothesis One it was necessary to combine

relevant portions of Tables Four and Five so that groups with liked

predecessors containing both congruent and not congruent successors

could be viewed together. This configuration can be seen in Table

Six.


Table Six

LIKED PREDECESSORS WITH CONGRUENT/NOT CONGRUENT
SUCCESSORS AND THEIR ACCEPTANCE LEVELS

Congruent Not Congruent TOTAL

High Acceptance 53 45 98

Low Acceptance 2 1 3

55 46 111


On the basis of the data presented above there is clearly no

difference between the acceptance levels of congruent versus not

congruent successors who followed well-liked supervisors. All 55

successors in the congruent classification followed predecessors who

were well-liked by their subordinates. Only two of these 55 suc-

cessors had a low group acceptance level. Of the 46 not congruent

successors who followed supervisors that had been well-liked only

one had a low acceptance score from his subordinates. Group

acceptance scores were very high for both congruent and not congruent

successors and thus Hypothesis One is clearly not supported. It is

worth noting that acceptance levels were determined using relative

rather than absolute acceptance scores and thus there is the possi-

bility that groups with a very high absolute acceptance score

(4.00-5.00)might have fallen more consistently into the high congruency










category or vice versa. The same would hold true for groups that had

a lower positive acceptance score (3.00-3.99) in terms of a possible

clustering of these groups in the congruent or not congruent cells.

However, in the final analysis it still appears that predecessor/

successor behavior congruency does not play a role in determining the

acceptance level of successors in episodes where the predecessor was

well-liked by his subordinates.


Hypothesis Two--Groups


Hypothesis Two states that if the successor's behavior is highly

congruent with that of an unpopular predecessor then the successor

will not be as readily accepted by his subordinates than if his

behavior is highly incongruent with that of an unpopular predecessor.

Table Seven provides a breakdown of episodes containing unpopular

predecessors for both congruent and noncongruent groups and their

acceptance levels.


Table Seven

UNPOPULAR PREDECESSORS WITH CONGRUENT/NOT CONGRUENT
SUCCESSORS AND THEIR ACCEPTANCE LEVELS

Congruent Not Congruent TOTAL

High Acceptance 0 1 1

Medium Acceptance 0 2 2

Low Acceptance 0 7 7

0 10 10









As can be seen from Table Seven, there were no cases in the con-

gruent, low liking classification. Consequently, Hypothesis Two can-

not be tested due to an inadequate sample size and thus it cannot be

substantiated in this study.


Hypothesis Three


Hypothesis Three states that there is a negative correlation

between subordinate frustration and the acceptability of the successor

to the work group. This proposition was tested by analyzing available

data for each individual queried in the study regardless of group

affiliation and by analyzing the information by episode. Testing was

done by using the correlation coefficient as well as by chi-square

analysis.

Acceptance involves the extent to which subordinates concur that

the new supervisor is doing what needs to be done and that they accept

his direction. The results of averaging the two questions dealing with

acceptance for each of the 559 respondents can be found in Table Eight.


Table Eight

SUCCESSOR ACCEPTANCE TO SUBORDINATES

Acceptance Scores Number of Respondents

1.00 12
1.50 12
2.00 13
2.50 23
3.00 60
3.50 55
4.00 136
4.50 108
5.00 139

TOTAL 558a

aOne person did not answer both questions relating to acceptance
and was excluded from the analysis.









Frustration is concerned with subordinate discomfort on the job

for reasons outlined earlier. Averaging the questions dealing with

frustration results in the information provided in Table Nine.


Table Nine

SUBORDINATE FRUSTRATION

Frustration Score Number of Respondents

1.00 44
1.25 47
1.33 1
1.50 71
1.67 2
1.75 65
2.00 63
2.25 45
2.33 2
2.50 47
2.67 1
2.75 42
3.00 47
3.25 27
3.33 2
3.50 19
3.67 1
3.75 9
4.00 6
4.25 4
4.50 6
4.75 3
5.00 4

TOTAL 558a
a One person did not answer the questions relating to frustration
and was excluded from the analysis.


After deriving individual respondent scores for acceptance and

frustration, these scores were correlated to determine if there was a

significant relationship between acceptance and frustration. The

correlation coefficient between acceptance and frustration for the

557 usable respondents had a value of -.50321 which is significant at









the .0001 level. Partial correlation was also used to test the

relationship between subordinate acceptance of the successor and the

frustration felt by subordinates. A first-order partial correlation

coefficient measures the association between acceptance and frustra-

tion by holding the effect of satisfaction constant. The partial

correlation coefficient between acceptance and frustration for the

557 usable respondents controlling for satisfaction had a value of

-.3926 which is significant at the .001 level. In other words, after

controlling for satisfaction there is still a significant negative

relationship between subordinate acceptance of the successor and the

level of frustration felt by subordinates. Hypothesis Three is

therefore substantiated using this procedure on individual responses

regardless of group membership.

Chi-square analysis using a 3 by 3 contingency table was also

employed to test the dependency relationship between acceptance and

frustration for individual respondents. Using the data from Tables

Eight and Nine acceptance and frustration scores were broken down such

that scores less than 3 indicate low frustration and low acceptance,

a score of 3 represents a medium level of frustration and acceptance,

and scores greater than 3 indicate high frustration and high

acceptance. The resulting classification of the subordinates is

shown in Table Ten.





















Low


Medium


High


TOTAL


Table Ten

LEVEL OF SUBORDINATE FRUSTRATION
AND SUBORDINATE ACCEPTANCE OF SUCCESSOR

Subordinate Frustration

Low Medium High

23 8 29


41 12 7


365 27 45


60


60


437


557


The value of X2 is 81.37 which is significant at the .00005 level.

This procedure also validates Hypothesis Three for individual

respondents.

Hypothesis Three was also tested using the same procedure as

outlined previously but with each of the succession episodes rather

than by individuals. Acceptance scores by episode can be found in

Table Two. Group averages for frustration by episode can be found in

Table Eleven.


Frustration Score

1.00 1.49
1.50 1.99
2.00 2.49
2.50 2.99
3.00 3.49
3.50 3.99
4.00 4.49
4.50 5.00


Table Eleven

GROUP FRUSTRATION SCORES

Number of Episodes

9
30
39
26
6
0
0
1

TOTAL 111


0
o m
u en
c 4-J U
0. Z
(D1 W
U
<: 0










The correlation coefficient between acceptance and frustration for the

111 groups was -.49173 which is significant at the .00001 level. The

partial correlation coefficient between acceptance and frustration for

the 111 groups controlling for satisfaction had a value of -.3250

which is significant at the .001 level. There is still a significant

negative relationship between acceptance and frustration after control-

ling for satisfaction. Hypothesis Three is, therefore, substantiated

using this procedure on each episode.

Chi-square analysis using a 2 by 2 contingency table was also used

to test the dependency relationship between acceptance and frustration

for each episode. Using the data from Tables Two and Eleven, accep-

tance and frustration scores were broken down such that scores less

than 3 indicate low acceptance and low frustration, and scores greater

than 3 indicate high acceptance and high frustration. The medium

level of acceptance where A=3 was omitted since there were only two

such cases and the "rule of thumb" is that each cell should have at

least five observations for X2 to be meaningful. This classification

is shown in Table Twelve.


Table Twelve

LEVEL OF GROUP FRUSTRATION
AND SUBORDINATE ACCEPTANCE OF SUCCESSOR

Group Frustration
a0o
u U) Low High


Low

0
High


TOTAL


7 3


95 4


102 7









Note that two cells have less than five observations so the value of

X2 may not be valid for groups. However, Hypothesis Three is sub-

stantiated using the correlation coefficient and chi-square for

individual respondents and the correlation coefficient for groups.


Hypothesis Four


Hypothesis Four states that there is a positive correlation

between subordinate satisfaction and the acceptability of the suc-

cessor to the work group. This proposition was also tested by

analyzing available data for each individual questioned in the study

regardless of group affiliation and by analyzing the information by

episode. Testing was again done by using the correlation coefficient

as well as by chi-square analysis.

The acceptance results outlined in Hypothesis Three apply

here as well and can be found in Table Eight. Satisfaction concerns

worker contentment with his overall job situation in terms of such

factors as pay, coworker rapport and desirability of specific work

assignments. Averaging the questions dealing with satisfaction

results in the information provided in Table Thirteen.

After deriving individual respondent scores for acceptance and

satisfaction, these scores were correlated to determine if there was

a significant positive relationship between acceptance and satisfac-

tion. The correlation coefficient between acceptance and satisfac-

tion for the 558 usable respondents had a value of .4159 which is

significant at the .0001 level. Partial correlation was again used to

test the association between subordinate acceptance of the successor

and the level of satisfaction felt by subordinates. The partial










correlation coefficient between acceptance and satisfaction for the

558 usable respondents controlling for frustration was .2479 which is

significant at the .001 level. Hypothesis Four is therefore sub-

stantiated using this procedure on individual responses regardless of

group membership.


Table Thirteen

SUBORDINATE SATISFACTION

Satisfaction Score Number of Respondents

1.40 3
1.60 3
1.75 1
1.80 2
2.00 9
2.20 16
2.40 12
2.60 31
2.67 1
2.75 1
2.80 26
3.00 49
3.20 50
3.25 3
3.40 55
3.50 8
3.60 58
3.75 4
3.30 47
4.00 63
4.20 51
4.40 34
4.50 1
4.60 15
4.75 1
4.80 10
5.00 5

TOTAL 559









Chi-square analysis using a 3 by 3 contingency table was also

employed to test the dependency relationship between acceptance and

satisfaction for individual respondents. Using the data from Tables

Eight and Thirteen, acceptance and satisfaction scores were broken

down such that scores less than 3 indicate low satisfaction and low

acceptance, a score of 3 represents a medium level of satisfaction

and acceptance and scores greater than 3 indicate high satisfaction

and high acceptance. This classification is shown in Table

Fourteen.


Table Fourteen

LEVEL OF SUBORDINATE SATISFACTION
AND SUBORDINATE ACCEPTANCE OF SUCCESSOR

Subordinate Satisfaction

Low Medium High


Low


Med ium


High


33 7 20


15 11 34


57 31 350


TOTAL 105 49 404 558


The value of X2 is 77.16 which is significant at the .00005 level.

This procedure also validates Hypothesis Four for individual

respondents.

Hypothesis Four was also tested using the same procedure as just

outlined but with each of the succession episodes rather than by









individuals. Acceptance scores by episode can be found in Table Two

whereas group averages for satisfaction are located in Table Fifteen.


Table Fifteen

GROUP SATISFACTION SCORES

Satisfaction Scores Number of Episodes

1.00 1.49 0
1.50 1.99 0
2.00 2.49 4
2.50 2.99 8
3.00 3.49 41
3.50 3.99 35
4.00 4.49 21
4.50 5.00 2

TOTAL 111


The correlation coefficient between acceptance and frustration for the

111 groups was .47093 which is significant at the .00001 level. The

partial correlation coefficient between acceptance and satisfaction

for the 111 groups controlling for frustration had a value of .2864

which is significant at the .001 level. There remains a significant

positive relationship between acceptance and satisfaction after

controlling for frustration. Hypothesis Four is, therefore, sub-

stantiated using this procedure on each episode.

Chi-square analysis using a 2 by 2 contingency table was also used

to test the dependency relationship between acceptance and satisfaction

for each episode. Using the data from Tables Two and Fifteen, accep-

tance and satisfaction scores were broken down such that scores less

than 3 indicate high acceptance and high satisfaction. There were

only two cases where acceptance =3 and two cases where satisfaction =3.

Since this would have resulted in cells with extremely small frequencies,










these four cases were eliminated from the analysis. The classifica-

tion of the 107 usable episodes can be found in Table Sixteen.


Table Sixteen

LEVEL OF GROUP SATISFACTION AND
SUBORDINATE ACCEPTANCE OF SUCCESSOR

Group Satisfaction

Low High


V 0
U )

S( Low

U U0
High


2 7


8 90


TOTAL 10 97 107


Since one cell has an extremely low frequency X2 is not meaningful

when analyzing responses by episodes. However, Hypothesis Four is

substantiated using the correlation coefficient and chi-square for

individual respondents and the correlation coefficient for groups.


Analysis of Individual Respondents


In addition to analyzing the data pertaining to Hypotheses One and

Two using groups, an attempt was also made to look at this information

using individual responses regardless of subordinate episode member-

ship. Further, it would have been ideal to be able to compare

individual subordinate responses concerning the predecessor to the same

subordinate's responses concerning the successor. However, this was

not possible since there was no way of matching up individual

predecessor/successor combinations when the questionnaires were

administered. This was due to the highly sensitive area addressed by









the questionnaire as well as the view by management that their very

endorsement of the project could trigger-off volatile responses in

certain work groups. Further, in most cases management iself was

extraordinarily apprehensive about the ultimate intent of the study

and great pains had to be taken to placate them on this issue. In

addition, both management and the subordinates involved had to be

continuously assured that subordinates participating in the study

would have their anonymity guaranteed. Participants were concerned

that it would be possible to compare the responses of subordinate A on

the predecessor to his responses on the successor. For this reason,

each questionnaire administered explicitly stated that the respondent

was not to reveal his name anywhere on the instrument during either

response period. There obviously was a complicated set of factors at

work that made research of this sort highly suspect. Management was

apprehensive that their endorsement of the study would be interpreted

by subordinates as an attempt by management to interfere with the work

groups involved for whatever reasons they might have. Subordinates

were sometimes afraid that those above them in the organizational

hierarchy would use the information provided to single out potentially

troublesome employees. It is for these reasons that the predecessor/

successor match-up for an individual respondent was not feasible which

eliminates the possibility of deriving an individual congruency/

noncongruency measure for each of the 559 respondents. In spite of

the aforementioned limitation, it was possible to utilize information

on all 559 respondents regardless of work group membership.

Several approaches were utilized with the 559 respondents to

determine the relationship between successor acceptability to the work









group with leadership style of the successor and subordinate liking

of both the predecessor and successor. The key issue is still sub-

ordinate acceptance of the successor and whether or not there is some

definite factor that helps explain subordinate acceptance or rejection

of the new supervisor.


Acceptance and Leadership Style of the Successor


In order to look at this relationship simple correlation co-

efficients were generated between acceptance and the Leadership

Behavior Questions (1-44).


Table Seventeen

SIMPLE CORRELATION COEFFICIENTS BETWEEN SUBORDINATE
ACCEPTANCE OF THE SUCCESSOR AND LEADERSHIP BEHAVIOR QUESTIONS (1-44)


Simple
Correlation

.4400
.5995
-.2195
.6165
.5708
.1478
-.3873
.5465
.4086
.5937
.4482
.4451
.6080
-.2570
-.4034
.5509
.6083
.4756
-.5186
.1802
.6260
.2269


Significance

.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001


Question

23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44


Simple
Correlation

-.5630
.6836
-.4906
.6074
.5856
-.4339
.5771
.1652
-.3992
.6523
.6521
.7442
.5584
.4718
.6962
.6003
.6024
.6130
.6518
-.4214
.0109
.6724


Significance

.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.399
.001


Question

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22










The data above indicate that there is an association between the

leadership behavior questions and subordinate acceptance of the suc-

cessor. With the exception of question 43, the simple coefficients

are all significant at the .001 level. It is interesting to note that

the Rule Enforcemnt Questions (6, 30, 43) had comparably lower

coefficients than most of the other leadership behavior questions.

This might indicate less of an association between the particular

leadership dimension and subordinate acceptance of the successor.

Partial correlation coefficients were derived between the accep-

tance questions and leadership behavior questions controlling for sub-

ordinate liking of the predecessor. These results can be found in

Table Eighteen.


Table Eighteen

PARTIAL CORRELATION COEFFICIENTS BETWEEN SUBORDINATE ACCEPTANCE
OF THE SUCCESSOR AND LEADERSHIP BEHAVIOR QUESTIONS (1-44) CONTROLLING
FOR SUBORDINATE LIKING OF THE PREDECESSOR (48-52)


Partial
Question Coefficient


.4344
.5984
-.2163
.6135
.5710
.1546
-.3921
.5440
.4108
.5903
.4460
.4465
.6155
-.2521
-.4190
.5579
.6088
.4820
-.5240
.1787
.6233
.2324


Significance

.001
.001
.001
.001
.001
.014
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.005
.001
.001


Question

23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44


Partial
Coefficient

-.5656
.6837
-.4943
.6150
.5816
-.4328
.5777
.1636
-.4033
.6529
.6502
.7439
.5680
.4712
.6968
.5946
.6030
.6143
.6501
-.4216
.0098
.6704


Significance

.001
.001
.001
.001
.001
.001
.001
.010
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.001
.445
.001









After controlling for subordinate liking of the predecessor, a

relationship continues to exist between the acceptance questions and

leadership style questions. As with simple correlation, questions 6,

30 and 43 (Rule Enforcement) again stand out due to their low partial

coefficients and higher levels of significance. It is possible that

these questions may not be particularly important to the issue of

subordinate acceptance of the successor. Possibly, subordinates accept

the fact that rules and regulations need to be enforced if an effective

work group is to be maintained.

Another point of major interest is that the partial correlation

coefficients generally remain quite high (absolute sizes do not change

much) when comparing them to the simple correlations derived prior to

controlling for subordinate liking of the predecessor. It appears

that liking of the predecessor may not be that significant in determin-

ing whether or not a new supervisor is accepted by individual subordi-

nates in relation to the leadership style questions. However, such is

not the case when generating partial correlation coefficients between

the acceptance questions and leadership behavior questions controlling

for subordinate liking of the successor. These results can be found

in Table Nineteen.

By controlling for subordinate liking of the successor it is

possible to see if that enters into the relationship between accep-

tance and the leadership style questions. After controlling for sub-

ordinate liking of the successor there is still a significant relation-

ship between subordinate acceptance of the successor and most of the

leadership behavior questions. However, in some cases the partial










correlation coefficients drop dramatically and the significance levels

also rise markedly on some questions. This can be contrasted to the

results that come from controlling for subordinate liking of the

predecessor. Whereas liking of the predecessor does not appear to

have much of an impact on acceptance of the successor in relation to

the leadership style questions, liking of the successor definitely

appears to be an important factor. In looking over the leadership

behavior questions where coefficients and significance levels fall

appreciably when correlated to acceptance there does not appear to be

any particular pattern of questions which seem to lose their signifi-

cance. Yet, the results indicate an association between subordinate

liking of the successor and whether or not a successor will be accepted

by his subordinates in relation to the leadership style questions.

It is necessary to explore some potential reasons that various

questions are no longer significant when controlling for subordinate

liking of the successor while looking at successor acceptance as a

function of the leadership behavior questions. As mentioned earlier,

several questions that have high coefficients and low levels of

significance when using simple correlation have lower partial coeffi-

cients and higher levels of significance when using partial correla-

tion. Using a significance level of .05, questions 14, 16 and 42 are

no longer helpful in explaining the relationship between the leader-

ship style questions and subordinate acceptance of the successor.

Conceptually then, for a group of equally liked successors, no rela-

tionship was found between needling behavior and acceptance

(Question 14),between treating members as equals and acceptance










(Question 16), and between feeling closer to management and acceptance

(Question 42). Whether or not the successor engages in the behavior

outlined above appears to be inconsequential as long as the group likes

him.


Table Nineteen


PARTIAL CORRELATION COEFFICIENTS BETWEEN SUBORDINATE ACCEPTANCE OF THE
SUCCESSOR AND LEADERSHIP BEHAVIOR QUESTIONS (1-44) CONTROLLING FOR SUB-
ORDINATE LIKING OF THE SUCCESSOR (48-52)


Partial
Coefficient

.2241
.3908
-.1235
.2294
.2378
.1608
-.1541
.2128
.1758
.2309
.2202
.1688
.2987
-.0214
-.1773
.0871
.3588
.3291
-.2940
.1609
.2891
.2035


Partial
Significance Question Coefficient Significance


.001
.001
.040
.001
.001
.011
.014
.001
.006
.001
.001
.008
.001
.381
.006
.108
.001
.001
.001
.011
.001
.002


-.1860
.3629
-.3335
.2405
.3587
-.1193
.1970
.1798
-.2421
.2521
.3579
.5189
.2137
.3439
.4066
.2261
.1760
.2936
.3481
-.0854
.1172
.3199


.004
.001
.001
.001
.001
.045
.002
.005
.001
.001
.001
.001
.001
.001
.001
.001
.006
.001
.001
.113
.048
.001


A further look at the issue of leadership behavior and acceptance

was taken using a stepwise regression on an individual basis using

each person's averaged acceptance score as the dependent variable and


Question

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22








all the leadership behavior items (1-44) as the predictor variables.

A 5% level of significance for predictors to enter and stay in the

stepwise equation was utilized.

It is important to point out the distinct possibility of bias

resulting from the fact that the same respondents (subordinates) who

described the behaviors of the predecessor/successor are also scoring

the successor on acceptance. It was not possible to get a rating of

the successor's performance by an individual or group other than the

subordinates and thus this group was used to generate both acceptance

scores and leadership behavior measures.

After only six steps, the level of significance for the entering

variable would have been greater than 5%. The results after six

steps are shown in Table Twenty.

Three interesting observations merit attention at this point

concerning the results displayed above. The first is that there were

only six variables that were significant when related to average

acceptance scores. It was thought that a higher number of questions

would both enter and stay in the equation but such was not the case.

Secondly, two of the variables that do appear in the equation (34 and

5) were not included in the three leadership behavior dimensions that

were used only to determine congruent/noncongruent, predecessor/

successor match-ups. Thirdly, and most importantly, the six behaviors

that did enter and stay in the equation had a multiple correlation

coefficient of .85 and explain 72% of the variance in the successor

acceptance score. Pursuing this further, partial correlation co-

efficients for each of these six leader behavior questions and sub-

ordinate acceptance of the successor controlling for the other five











Entering
Variable
(Question)

34

44

13

25

2

39


Table Twenty

STEPWISE REGRESSION-LEADER BEHAVIOR AND SUCCESSOR ACCEPTANCE

Multiple Partial Correlation
Regression Level of Correlation Coefficient Controlling
Coefficient Significance Coefficient for other Five Variables

.27999 .000 .74422 .4079

.15012 .002 .79937 .2188

.17687 .000 .82202 .2720

-.10974 .002 .83365 -,2165

.12156 .005 .84145 .1947

.11575 .011 .84688 .1772


Level of
Significance

.001

.001

.001

.001

.001

.001


Step
Number

1

2

3

4

5

6










leader behavior questions were generated. These results can also be

found in Table Twenty. All partial correlation coefficients are

significant at the .001 level indicating that questions 34, 44, 13,

25, 2 and 39 are very helpful in explaining the association between

leader behavior and subordinate acceptance of the new supervisor.

The data presented in Tables Seventeen through Nineteen indicate

a possible connection between acceptance of a successor and that

individual's leadership behavior. Yet, the results provided in Table

Twenty more clearly and precisely indicate which questions concerning

leader behavior on the job are most closely associated with sub-

ordinate acceptance of the new supervisor.

With the exception of question 25, the other five questions all

deal with the supervisor's concern with group and individual welfare

in terms of being responsive to their needs while at the same time

doing what needs to be done to ensure group viability in the future.

As a group, these questions indicate a supervisor who knows what needs

to be done but also attempts to solicit and cultivate group support of

his actions by behaving in an empathetic fashion towards his sub-

ordinates.

Question 25 concerns the need for the supervisor to stand firm

and not back down when confronted by certain pressures which are not

specified in the question. The negative sign on the regression co-

efficient for 25 indicates that the supervisor who does not back down

will be more likely to be accepted by the work group than his counter-

part who does acquiesce and give in when he should stand his ground.

At this point, an attempt was made to cross-validate the regres-

sion model by randomly dividing the 559 respondents into groups of 280










and 279 respectively. This was done by randomly generating 280

numbers ranging from 1 to 559. These individuals were put in one

group and the remaining 279 in a second group. The stepwise regres-

sion program with the 44 leadership behavior items and acceptance was

run for the group of 280 subordinates and these results appear in

Table Twenty-One. The variables that remained after utilizing the 5%

level of significance for predictors to enter and remain (13, 24, 26,

34) were then put in a multiple regression with the other 279 subjects

to see if these same questions would satisfy the 5% criterion to enter

and stay and determine the resulting correlation coefficients. These

results are also found in Table Twenty-One. As the results in Table

Twenty-One indicate, questions 13, 24 and 34 show up in both runs and

consequently there is a definite degree of cross-validation between

the two groups. These results indicate that these are not chance

models. The questions that remain in both models indicate a super-

visor who not only handles difficult or emergency situations effec-

tively but also is concerned with the welfare of group members. It

seems that successor acceptability might be associated with an

individual who can handle a stressful situation in an adroit manner

while simultaneously showing concern and high regard for those he

supervises.

It is important to note that unlike question 13, questions 34 and

24 are not included in any of the three leadership dimensions. In

particular, question 34 which is the single most important item in the

stepwise regression fails to turn up at all, even in a single item

factor. Item 34 is also worth noting due to its emphasis on super-

visor response to emergency situations. In light of the short









Table Twenty-One


STEPWISE REGRESSION-LEADER BEHAVIOR AND SUCCESSOR ACCEPTANCE/SPLIT SAMPLE



N-559 N=280 N=279
Entering Entering Entering Multiple ,Multiple
Step Variable Regression Level of Variable Regression Level of Correlation Variable Regression Level of Correlation
Number (Question) Coefficient Significance (Question) Coefficient Significance Coefficient (Question) Coefficient Significance Coefficient

1 34 .27999 .000 34 .34478 .000 .78109 34 .36605 .000 .69943

2 44 .15012 .002 13 .22001 .001 .82869 13 .22909 .000 .76074

3 13 .17687 .000 26 .14063 .010 .85189 24 .20110 .002 .78048

4 25 -.10974 .002 24 .15140 .017 .86071

5 2 .12156 .005

-'










duration between predecessor departure and subordinate description of

the successor's behavior, it is entirely possible that no truly

legitimate emergencies arose that were under the supervisor's juris-

diction in the work place. In this instance, it is conceivable that

subordinates are actually reacting to how the successor behaves when

he is under a little pressure, be it from his supervisors, peers, sub-

ordinates or some undetermined source. In other words, it is possible

that subordinates are not really identifying a specific emergency

situation, but rather describing how the new supervisory incumbent

keeps his cool when things heat up a little on the job.

Additional runs were made using all 76 questionnaire items

(except 47), again retaining acceptance (45 and 46) as the dependent

variable. Question 47 was not used in deriving an acceptance score but

its similarity to 45 and 46 necessitated dropping it from the analysis

when using all questionnaire items (1-76). As with the leadership

behavior items, after a stepwise regression was performed on the full

model consisting of all 76 items (except 45, 46 and 47) the two split

samples of 280 and 279 respondents respectively were tested in the

same fashion as had been done earlier to see if the same questions

entered and stayed in both groups. These results can be found in

Table Twenty-Two.

In the run using all 559 respondents question 51 fits into the

mold that emerged when using only the leadership questions as 51

involves whether or not the supervisor is thoughtful. The same holds

true with question 48 as it concerns whether or not the supervisor is

fair in his treatment of his workers. Question 55 has a negative

regression coefficient since it is worded in such a fashion that a










high score indicates that subordinates who feel they have adequate

authority with which to carry out the responsibilities assigned to

them will tend to be more receptive of their supervisors than sub-

ordinates who feel they lack adequate authority in fulfilling their

job requirements.

An attempt to cross-validate the regression model using all 76

items was done in the same manner as outlined earlier when using the

44 leadership behavior items. As Table Twenty-Two indicates, ques-

tions 2, 13, 34 and 51 show up in both runs and consequently there is

considerable cross validation between the two groups. These results

echo the data found in Table Twenty-One which suggest that accepted

supervisors are those possessing empathetic skills as well as being

capable of making decisions when they need to be made. Thus Tables

Twenty through Twenty-Two indicate specific characteristics that may

be significant predictors of successor acceptance by subordinates.








Table Twenty-Two


STEPWISE REGRESSION-ALL ITEMS AND SUCCESSOR ACCEPTANCE*


FULL AND SPLIT SAMPLE



N=280 N=279
Multiple Entering Multiple Entering Multiple
Step Variable Regression Level of Correlation Variable Regression Level of Correlation Variable Regression Level Correlatioo
Number (Question) Coefficient Significance Coefficient (Question) Coefficient Significance Coefficient (Question) Coefficient Coefficient Coefficient

1 34 .25716 .000 .74422 34 .26884 .000 .78109 34 .31569 .000 .69943

2 51 .14648 .004 .81362 51 .12719 .030 .83126 51 .29239 .000 .79253

3 13 .16024 .000 .83314 13 .15673 .003 .85387 13 .15829 .003 .80927

4 25 -.11215 .001 .84468 24 .14438 .003 .86641 02 .14331 .012 .81927

5 48 .14976 .001 .85319 49 -.16178 .001 .87643

6 55 -.08808 .009 ,85859 02 .09963 .046 .88024

7 02 .10043 .017 .86291


*except 45. 46, 47










Subordinate Liking of Successor and Predecessor


Another issue to be explored concerning successor acceptability

to his subordinates is the possibility that acceptance is in some way

a function of subordinate liking of the successor or subordinate

liking of the predecessor. Evidence gathered in analyzing Hypotheses

One and Two for groups (episodes) lends credibility to this view,

particularly when congruency between predecessor/successor leadership

behavior appears not to be the major factor it was expected to be.

Table Twenty-Three provides the simple correlation coefficients

between the acceptance questions and subordinate liking of the

predecessor. Where appropriate, questions 49, 50 and 52 were reverse

scored since they were worded in a negative fashion such that initially

a high score (5) was good and a low score (1) was bad.


Table Twenty-Three

SIMPLE CORRELATION COEFFICIENTS BETWEEN SUBORDINATE
SUCCESSOR AND INDIVIDUAL SUBORDINATE LIKING
THE PREDECESSOR (48-52)

Simple
Question Correlation

48 -.0305
49 -.0628
50 -.0674
51 -.0728
52 -.0912


ACCEPTANCE OF THE
ITEMS FOR



Significance

.236
.069
.056
.043
.016


Although only two of the coefficients are significant at the .05

level, these results indicate an inverse relationship between sub-

ordinate acceptance of the new supervisor and subordinate liking of

the previous supervisor. Proceeding further, the simple correlation

coefficient between acceptance and the combined liking questions for










the predecessor is -.0737 which is significant at the .041 level. It

appears that when the predecessor was well-liked subordinates were

less likely to accept the successor in his new role. Table Twenty-

Four provides the simple correlation coefficient between successor

acceptability to subordinates and.subordinate liking of the successor.


Table Twenty-Four

SIMPLE CORRELATION COEFFICIENTS BETWEEN SUBORDINATE ACCEPTANCE OF THE
SUCCESSOR AND INDIVIDUAL SUBORDINATE LIKING ITEMS FOR
THE SUCCESSOR (48-52)

Simple
Question Correlation Significance

48 .6857 .001
49 .5740 .001
50 .5789 .001
51 .6928 .001
52 .5632 .001


It is quite clear that there is a very high correlation between the

liking of the successor by his subordinates and his acceptability as

a supervisor to those same individuals. In addition, the correlation

.coefficient between acceptance and the combined liking questions for

the successor is .7345 which is significant at the .001 level. There

was initial concern during the development of the proposal used as a

foundation for this study that subordinates might like a supervisor and

still not accept him as their legitimate leader due to a lack of faith

in his judgment, ability or support of the work group. The data in

Table Twenty-Four indicate that this is not the case with the

individuals who were involved in this study.

It was felt that more information was needed to better explore

the relationship between liking and successor acceptability so partial









correlation coefficients between the acceptance questions and sub-

ordinate liking of the predecessor controlling for subordinate liking

of the successor were generated. These results are found in Table

Twenty-Five.


Table Twenty-Five

PARTIAL CORRELATION COEFFICIENTS BETWEEN SUBORDINATE ACCEPTANCE
OF THE SUCCESSOR AND INDIVIDUAL SUBORDINATE LIKING ITEMS FOR THE
PREDECESSOR (48-52) CONTROLLING FOR SUBORDINATE LIKING OF THE
SUCCESSOR (48-52)

Partial
Question Correlation Significance

48 -.0441 .151
49 -.0444 .149
50 -.0256 .274
51 -.0607 .078
52 -.0706 .049


Controlling for subordinate liking of the successor has some

effect on the results. Evidence indicates that there is little

connection between successor acceptability and subordinate liking of

the predecessor when controlling for subordinate liking of the suc-

cessor. The relationship that does exist is again negative but the

levels of significance are so high that it is not possible to speculate

with any support that liked predecessors will be followed by suc-

cessors who will not be accepted by the work group. Furthermore,

using a single acceptance and liking score results in a partial cor-

relation coefficient of -.0553 which is significant at the .097 level.

Contrast the data in Table Twenty-Five to that found in Table Twenty-

Six where one finds partial correlation coefficients between accep-

tance and subordinate liking of the successor controlling for sub-

ordinate liking of the predecessor.









Table Twenty-Six

PARTIAL CORRELATION COEFFICIENTS BETWEEN SUBORDINATE ACCEPTANCE
OF THE SUCCESSOR AND INDIVIDUAL SUBORDINATE LIKING ITEMS FOR THE
SUCCESSOR (48-52) CONTROLLING FOR SUBORDINATE LIKING OF THE
PREDECESSOR (48-52)

Partial
Question Coefficient Significance

48 .6856 .001
49 .5779 .001
50 .5800 .001
51 .6922 .001
52 .5628 .001


Table Twenty-Six amplifies the notion that liking of the prede-

cessor has little impact on successor acceptability to his subordinates.

Evidence indicates that subordinate liking of the new supervisor

covaries with the matter of whether or not that individual will be

accepted by his subordinates. In addition, it is important to note

that the acceptance and liking items used loaded independently and

significantly on two different factors. Using single acceptance and

liking scores reiterates this point since the partial correlation co-

efficient is .7344 which is significant at the .001 level. Certainly

liking of one's superior is not the only criteria for accepting that

individual but it does seem to be of some importance. It is worth

mentioning that all the groups in this study were characterized by

situations where subordinate tasks were structured, supervisors had

considerable power and there was no expectation or mandate for

immediate change in the work group. Had these control variables not

been present, the results may naturally have been quite different.

The preceding information supports the proposition that to some

degree acceptance of the new supervisor is associated with that










individual's leadership behavior. In addition, acceptance of the

successor also appears to be associated with subordinate liking of the

successor when analyzing the results using individual respondents.


Other Findings


In addition to analyzing the four major hypotheses already dis-

cussed, other aspects of supervisory succession were investigated.

In analyzing four of these other issues tangible results using chi-

square were difficult to come by due to the number of empty cells

existing after group classifications were made. Even collapsing

certain parts of tables would have done little to facilitate the

analysis due to the clustering of group responses in specific classi-

fications. However, it is worth discussing this additional material

as it sheds further light on our understanding of what takes place

after an organization experiences a supervisory succession episode.

It was thought that there might be an association between the

fate of the predecessor and the extent to which the new supervisor was

accepted by the work group. The issue concerned whether or not there

might be a connection between why the previous supervisor left and how

his successor was received by the subordinates. Would a fired or

demoted predecessor impact on the acceptability level of his replace-

ment? Evidence provided in Table Twenty-Seven indicates little in

this regard except that in 81% of the episodes in question the prede-

cessor either transferred or left the organization altogether.










Table Twenty-Seven

SUCCESSOR ACCEPTABILITY TO SUBORDINATES AND FATE OF PREDECESSOR

Fate of Predecessor

Another
Transferred Fired Demoted Firm Died Retired TOTAL

1.00-1.49 1 1

1.50-1.99

o 2.00-2.49 1 2 3

u 2.50-2.99 3 1 2 6

3.00-3.49 8 2 10
0
w 3.50-3.99 17 2 1 20

4.00-4.49 26 1 1 6 2 36

4.50-5.00 22 4 2 1 6 35

TOTAL 78 6 5 12 1 9 111

Column Mean 3.9940 4.2472 4.0221 3.4945 not 4.4198
applicable


Very few departed supervisors fell into the fired, demoted, or retired

categories. A series of tests for the difference in average accep-

tance scores for a new supervisor contingent on the fate of the prede-

cessor were performed. There was no significant difference in the mean

acceptance scores between any of the groups except when the episodes

where the predecessor went to another firm were compared to those where

the predecessor had retired.

It was felt that there might be a connection between successor

origin and successor acceptability to the subordinates. If the new

supervisor came from within the organization (either within the

facility or another facility in the company) that individual would have










a better chance of getting favorable treatment from the subordinates

than if he came from totally outside the organization. Subordinates

would feel more at ease with someone who came from within the organi-

zation and was familiar with the existing culture and group norms. An

outsider would not be privy to such knowledge and consequently might

behave in a manner inconsistent with subordinate desires and expecta-

tions and therefore be less palatable to the group. Table Twenty-

Eight shows that the nature of the sample provided little help inves-

tigating this possibility since almost 90% of the successors were

insiders leaving too small a group of outside successors upon which to

draw any conclusions. Also, the overall high rate of successor accep-

tance found in the study would make an insider/outsider dichotomy less

meaningful. A series of tests for the difference in acceptance scores

contingent on the origin of successor were performed with no signifi-

cant differences found between the groups.


Table Twenty-Eight

SUCCESSOR ACCEPTABILITY TO SUBORDINATES AND SUCCESSOR ORIGIN

Successor Origin

Another Part Another Facility
of Facility of Organization Outsider TOTAL

1.00-1.49 1 1

1.50-1.99 0
0
M0 2.00-2.49 2 1 3

S2.50-2.99 3 2 1 6

S3.00-3.49 7 2 1 10
0
U 3.50-3.99 18 0 2 20

4 4.00-4.49 28 6 2 36

U 4.50-5.00 25 4 6 35

TOTAL 84 14 13 111

Column Mean 4.0117 3.9601 3.0107










An attempt was made to see if an association existed between the

origin of the successor and the level of satisfaction experienced by

the work group. Would subordinates be more content in their jobs with

someone promoted from within corporate ranks or someone from outside

the existing organization structure? A high level of satisfaction

might be associated with inside succession due to its reaffirmation of

the normal succession process of promoting from within. As Grusky

points out:

Succession can also improve morale in the organiza-
tion and facilitate the proper socialization of
bureaucratic personnel. By systematically making
changes in top positions, organizations allow for
upward mobility, thereby promoting the develop-
ment of motives to succeed and produce for the
system. (11, p. 111)

Though Grusky's comments refer to high level positions their applica-

bility would also seem to be germane in exploring the effects of

internal promotion at the supervisory level.

The presence of an outside successor might result in either high

or low levels of subordinate satisfaction contingent on how the work

group perceives the arrival of the new supervisor. If the predecessor

was disliked by his subordinates and had been closely associated with

the current management team, an outsider might be viewed as a welcome

change of pace as yet relatively uninfluenced by the whims of higher

management. However, an outsider would also have a negative impact on

satisfaction levels due to uncertainties as to what his tenure will

mean to the group in terms of the existing status quo. He could be

perceived as a threat to the current state of affairs.











The results of the study were inconclusive concerning the issue

of inside versus outside succession as a factor associated with the

level of subordinate satisfaction.


Table Twenty-Nine

JOB SATISFACTION OF SUBORDINATES WITH INSIDE AND OUTSIDE SUCCESSORS

Origin of Successor

Inside Organization
Framework Outsider TOTAL

1.00-1.49

1.50-1.99
o
' 2.00-2.49 4 4
UO
m 2.50-2.99 6 2 8
4J
S 3.00-3.49 35 6 41

c 3.50-3.99 33 2 35
r-,
4 4.00-4.49 18 3 21
0
-o

n 4.50-5.00 2 0 2

TOTAL 98 13 111

Column Mean 3.5233 3.4677


As Table Twenty-Nine indicates, both inside and outside successors

generally enjoyed relatively high levels of satisfaction. Only twelve

of the 111 cases had satisfaction scores in the low satisfaction range.

A test for the difference in satisfaction levels of the subordinates

contingent on the origin of the successor was performed with no

significant difference found between insiders and outsiders.

A similar analysis was also used to determine if there might be a

link between subordinate frustration levels and the degree of congruency










between the predecessor and successor. It was initially thought that

the greater the similarities between the predecessor and successor,

the less would be the disruption forced on the work group and conse-

quently low levels of frustration would be present when congruent

predecessor/successor combinations were in evidence. Similarly,

noncongruent match-ups would result in relatively high levels of

group frustration. Table Thirty indicates that such is not the case.


1.00-1.49

1.50-1.99

2.00-2.49

2.50-2.99

3.00-3.49

3.50-3.99

4.00-4.49

4.50-5.00

TOTAL

Column Mean


Table Thirty

SUBORDINATE FRUSTRATION AND PREDECESSOR/SUCCESSOR
BEHAVIOR CONGRUENCY

Predecessor/Successor Behavior Congruency

Congruent Congruent Not Congruent Not Contruent
on Three on Two on Three on Two TOTAL

3 1 1 4 9

7 5 8 10 30

6 12 9 12 39

7 10 3 6 26

2 2 0 2 6


25

2.1212


30

2.2619


22

2.1961


34

2.0905


Regardless of the extent to which predecessor/successor dyads

were congruent or not congruent subordinate frustration levels remained

generally low. A series of tests for the difference in frustration







75


levels of subordinates contingent on the level of predecessor/

successor behavior congruency were performed with no significant

differences found between the four groups.

Three other succession related subjects were investigated using

correlation analysis. As Table Thirty-One indicates there is an

association between subordinate respect for the new supervisor and

the level of acceptance for that individual.


Table Thirty-One

SUCCESSOR ACCEPTABILITY TO SUBORDINATES AND SUBORDINATE RESPECTS
FOR THE SUCCESSOR

Successor Respect

1.00- 1.50- 2.00- 2.50- 3.00- 3.50- 4.00- 4.50-
1.49 1.99 2.49 2.99 3.49 3.99 4.49 5.00 TOTAL

1.00-1.49 1 1

1.50-1.99 0

m 2.00-2.49 1 2 3
U
| 2.50-2.99 3 3 6

o 3.00-3.49 1 2 1 4 1 1 10

r 3.50-3.99 6 5 5 4 20
4-.
4.00-4.49 1 3 19 13 36

4.50-5.00 2 1 10 22 35

TOTAL 1 2 4 4 16 10 35 39 111


The correlation between subordinate respect of the successor and sub-

ordinate acceptance of the successor was .8313 which is significant at

the .001 level. It was initially felt that the level of respect and

level of acceptance would move in the same direction and results bear










out this proposition. As can be seen from Table Thirty-One there is

a heavy concentration of responses in the high respect/high acceptance

area. This parallels findings reported earlier in this document

that indicate an association between subordinate liking of the suc-

cessor and a high acceptance level.

Another issue explored involved the probable association between

the degree to which subordinates liked the predecessor and the amount

of frustration felt by subordinates when working for the new super-

visor. Robert Guest points out that:

.. if the subordinates liked the predecessor,
and liked the way he worked with them and the
organization, they will be more anxious about
a change. (10, p. 24)

If the old supervisor was well-liked by the group and if his suc-

cessor poses a threat to group norms in terms of traditional ways of

carrying out job assignments and providing needed resources and time

commitments then group frustration may manifest itself under the new

supervisor. Part of this liking predecessor/frustration under suc-

cessor dilemma could be caused by the uncertainty felt by subordinates

as they try to adjust from working under someone they knew and liked

to someone that may not embrace the same policies that made the prede-

cessor so desirable. It is this change in priorities and ways of

doing things that can cause frustration for subordinates when the new

supervisor takes over.

Table Thirty-Two indicates that in this particular study there

was not a positive correlation between high subordinate liking of the

predecessor and a high level of subordinate frustration under the

successor.










Table Thirty-Two

SUBORDINATE LIKING OF PREDECESSOR AND SUBORDINATE
FRUSTRATION UNDER THE SUCCESSOR

Subordinate Frustration

1.00- 1.50- 2.00- 2.50- 3.00- 3.50- 4.00- 4.50-
S1.49 1.99 2.49 ,2.99 3.49 3.99 4.49 5.00 TOTAL
01

1.00-1.49 1 1



on
e 1.50-1.99 1 1

t 2.00-2.49 1 1 1 3

r 2.50-2.99 3 1 1 5

3.00-3.49 1 3 4 2 10

c 3.50-3.99 3 4 13 1 21

4.00-4.49 4 20 17 6 47
0

4.50-5.00 5 6 10 1 1 23

TOTAL 9 30 39 26 6 0 0 1 111


The correlation coefficient between subordinate liking of the prede-

cessor and subordinate frustration under the successor was -.5410

which is significant at the .001 level. In this study most groups

were characterized as both liking the predecessor and having a low

level of frustration when the new supervisor took over. This might

be partially explained by the fact that only groups without a strong

mandate for change on the part of the successor were included in the

analysis.

Finally, an analysis was made to determine if the level of accep-

tance for successors would be associated with the level of frustration

experienced by subordinates in groups where the predecessor was well

liked and there was high predecessor/successor behavior congruency.






78


In other words, given the same levels of liking of the predecessor and

predecessor/successor behavior congruence, a lower level of acceptance

would be associated with subordinates who experience higher levels of

frustration. No attempt was made to look at low liking/low congruence

groups since so few episodes fell into that classification.

As Table Thirty-Three demonstrates, when both liking and con-

gruence are high and the frustration level is low, acceptance of the

successor is high which is to be expected. The correlation coefficient

between subordinate frustration and acceptance of the successor is

-.4122 which is significant at the .001 level.


Table Thirty-Three

SUBORDINATE ACCEPTANCE OF SUCCESSOR AND FRUSTRATION OF SUBORDINATES
FOR GROUPS CHARACTERIZED BY HIGH SUBORDINATE LIKING OF PREDECESSOR
AND HIGH PREDECESSOR/SUCCESSOR BEHAVIOR CONGRUENCY

Subordinate Frustration

1.00- 1.50- 2.00- 2.50- 3.00- 3.50- 4.00- 4.50-
1.49 1.99 2.49 2.99 3.49 3.99 4.49 5.00 TOTAL

1.00-1.49

1.50-1.99
o
Q 2.00-2.49
U
S2.50-2.99 1 1 2

o 3.00-3.49 1 1 3 2 7
a)
= 3.50-3.99 2 4 7 13
4J
S4.00-4.49 1 4 9 3 2 19

4.50-5.00 3 5 3 3 14

TOTAL 4 12 18 17 4 0 0 0 55










Again, the importance of this finding is tempered by the fact that

acceptance levels of the successors were generally high throughout the

study.

The findings reported in this section might be affected by the

high acceptance, satisfaction and respect levels found in most groups

as well as the relatively low level of frustration experienced in the

vast majority of the episodes. Such findings may not be typical of

work groups in general but may be partially a function of the control

variables used to select episodes for inclusion in the study. Yet,

the data examined do shed additional light on the succession phe-

nomenon not previously addressed in the discussion of Hypotheses 1-4.

The other issues, all related to the interaction between sub-

ordinate liking of the predecessor, predecessor/successor leadership

behavior and subordinate acceptance of the new supervisor, were also

explored. The first item involved the possibility that acceptance

levels for successors might vary when the level of congruence in the

groups was lowered while subordinate liking of the predecessor was

held constant. In order to accomplish this, a test for the difference

in acceptance scores between high like/congruent groups and high like/

noncongruent groups was performed. The mean acceptance score for the

9 congruent episodes with liking scores of 4.00-5.00 was compared to

the mean acceptance score of the 14 noncongruent groups with similar

liking scores. There was no significant difference in the mean accep-

tance scores between the two groups. Similar results were obtained

when liking was lowered to 4.00-4.49 which included 24 congruent and

23 noncongruent episodes and when liking was further lowered to

3.50-3.99 which included 14 congruent and 7 noncongruent episodes.










It was not possible to compare the results between the congruent and

noncongruent groups when liking was lowered to 3.00-3.49 since only

two groups fell within the noncongruent category. These results

indicate no effect on successor acceptance when liking is held con-

stant and predecessor/successor congruence is lowered.

A second issue along similar lines involved a comparison between

acceptance levels for two groups: congruence with the highest level

of liking and the same level of congruence with a lower level of

liking. A test for the difference in acceptance scores between

congruent/high like groups and congruent/medium-high like groups was

performed. The mean acceptance score for the 33 congruent episodes

with liking scores of 4.00-5.00 was compared to the mean acceptance

scores of the 22 congruent medium-high like episodes (3.00-3.99).

There was a significant T ratio between the groups indicating a

possible connection between liking and level of acceptability within

congruent groups. The congruent groups with a high liking score had

a significantly greater acceptance level than their congruent/medium

high liking counterparts.

Such was not the case when noncongruent groups with the highest

level of liking were compared to noncongruent groups with a lower level

of liking. There was no significant difference in the mean acceptance

scores between the 37 noncongruent/high like groups and the 9

noncongruent/medium-high like groups. It might also be added that

when all episodes are put together (congruent and noncongruent) and

the level of liking is lowered there is no significant difference in

the mean acceptance scores between the two groups.









A third topic involved acceptance levels when both congruence

and liking were lowered. To generate this information, a test for the

difference in acceptance scores between congruent/high like groups and

noncongruent/medium-high like groups was performed. The mean accep-

tance score for the 33 congruent/high like groups was compared to the

mean acceptance score of the 9 noncongruent/medium-high like groups.

There was no significant difference in the mean acceptance scores

between the two groups. In addition, no significant difference

existed when the 22 congruent medium-high like episodes were compared

to the 37 noncongruent/high like episodes.

As can be seen from the three topics just discussed, altering

predecessor/successor congruency levels appears to have little bearing

in determining whether or not a new supervisor is accepted by his

subordinates. Subordinate liking only stood out as being significant

in those situations where congruent groups were being analyzed and

levels of liking were altered. In the aggregate, it appears that

predecessor/successor congruency and subordinate liking of the prede-

cessor have little to do with the acceptance level of the successor

in his work group.
















CHAPTER VI
CONCLUSIONS



The results of the analysis as they relate to Hypotheses One and

Two merit additional attention. Possible explanations for the

hypotheses substantiation or rejection need to be explored. It was felt

that supervisors who behave in a manner similar to that of a well-liked

predecessor would be more readily accepted by the work group than

supervisors who behave in a fashion dissimilar to that of a well-liked

predecessor. While congruent predecessor/successor match-ups in

which the predecessor was well-liked had high levels of successor

acceptability to subordinates, the same is true of episodes where

predecessor/successor behavior was noncongruent. Table Six bears out

these findings. Possible explanations for this state of affairs need

to be examined in regard to why both congruent and noncongruent suc-

cessors enjoyed such high acceptance levels in situations where the

predecessor was well-liked by the work group.

Congruent successors might profit from a carry-over effect from

predecessor to successor such that the subordinates feel at ease with

someone who exemplifies the qualities of the supervisor who has

recently departed. There might be some benefit to the new supervisor

that derives from his perceived familiarity to the group in terms of

his leadership behavior. In situations where the successor is per-

ceieved as noncongruent with the predecessor, acceptance levels remain









high possibly indicating subordinate optimism that everything will

work out for the best and there is no need to be concerned just

because the new supervisor does not behave like the old one. Data do

not substantiate the original expectation that successors who behave in

a fashion incongruent with a well-liked predecessor will be less well

accepted by their subordinates than their congruent counterparts due

to the fact that only one episode met the aforementioned criteria

(Table Five).

Of the 111 episodes analyzed in this study, 98 fell into the high

predecessor liking/high successor acceptance category and only three

groups were found in the high predecessor liking/low acceptance cells

(Tables Four and Five). In most cases, predecessors were well-liked

and successors were accepted regardless of the congruency level

between predecessor and successor. This implies the possibility that

subordinates tend to accept new supervisors during their first few

weeks on the job and the issue is just that simple. The results of

this study do nothing to refute this view since in 99 of the 111

groups studied, new supervisors were found acceptable to their work

groups.

The above finding provides a possible limitation to the study as

it was conducted. If subordinates overwhelmingly accept their new

supervisor during the first six weeks of his tenure then possibly

that is not an adequate time frame for discord or a deterioration of

harmony to manifest itself in such groups. Maybe there is a longer

"honeymoon" effect than initially expected between subordinates and

supervisor and problems will only emerge after a lengthier period of

time has passed. Future research should focus in on this time frame










issue in terms of how long subordinates are willing to accept their

new supervisor with few, if any, reservations. In spite of the

results of the pilot study, it may take more time than initially

thought for subordinates to become disenchanted with their super-

visor for a variety of reasons. However, this limitation was not

germane to the current study since as stated in the Introduction,

succession is primarily concerned with what transpires as soon as a

successor takes over from his predecessor and not what happens six

months or a year later. The stated purpose of this project was to

explore the succession phenomenon and this emphasis should not

detract from the validity of the results derived.

The entire congruency issue shed little light on any possible

connection between leadership behavior and successor acceptability to

the work group. As Table Twenty indicates, the congruency measures

used in this study do not appear to sufficiently reflect those be-

haviors which are crucial in determining whether or not a new super-

visor will be acceptable to the work group. Of the six predictors

found in the stepwise regression, only four were in the three dimen-

sions used to determine predecessor/successor behavior congruency.

Consequently, the technique used to measure congruent/noncongruent

behavior in this study does not provide a meaningful guide concerning

the association between this issue and successor acceptability to the

work group. However, this does not mean that congruency per se is

unrelated to acceptance but rather there appears to be little associa-

tion between the two as measured in this study.










As was discussed in the previous chapter, the results of the data

analysis provide no support for the anticipated outcome of Hypothesis

Two. In this study, of the 111 groups tested there were none in which

the predecessor/successor combination was congruent, the predecessor

was disliked by the group and the successor had a low acceptance from

the work group (Table Seven). There were only seven episodes where the

predecessor/successor match-up was noncongruent, the predecessor was

disliked by the group and the successor had a low level of acceptance

from the work group (Table Seven). These results could be due to the

fact that the groups involved simply did not meet these criteria or to

some factor the study did not anticipate being present. It is con-

ceivable that a successor would not want to behave in a similar manner

to an individual who was disliked by his subordinates and thus no suc-

cessors would mimic a disliked predecessor. This is merely a specula-

tion based on a sample of ten cases that might be investigated since

of the 111 episodes studied only ten had unpopular predecessors and in

none of these did successors behave in a manner congruent with that of

the predecessor.

The results of the analysis demonstrated generally high liking

scores for predecessors throughout the analysis. A possible explanation

for a high liking score might be that once a supervisor has departed,

subordinates tend to overlook his faults and decide that, now that he

is gone, the "old" supervisor wasn't so bad after all and they really

did sort of like him.

There is information in Table Seven that suggests a particular

bias on the part of subordinates towards their supervisor. For









noncongruent groups where the predecessor was not liked, only one of

ten successors was readily accepted by the work group with seven

falling into the not accepted category. This is interesting when one

considers that, of the 111 total episodes studied, only ten even fall

into the low acceptance area and seven of these are bunched together.

It may be that when subordinates hold negative cognitions about the

predecessor these thoughts carry over and affect the new supervisor.

When the previous supervisor was disliked it may be that those feelings

create a negative perception of the position regardless of the incum-

bent. In other words, if the subordinates did not like the predecessor

then they are not about to accept the successor due to their negative

predisposition towards that position. It is possible that there might

be a negative halo effect operating in this regard.

It is also possible that timing is a factor in this type of situa-

tion. Subsequent research might attempt to discern a change in sub-

ordinate willingness to accept the successor of a disliked predecessor

over a lengthier period of time. However, as stated earlier, it is

worth noting that all the low liking episodes were found among the non-

congruent predecessor/successor combinations (Table Seven). These are

considerations that warrant further study with adequate sample sizes

in subsequent research.

It is important to note that one of the control variables may have

had an impact on the study. It would seem quite plausible that the

fact that there were no groups included in which management intended

to significantly alter the work group might have been instrumental in

high liking/high acceptance response ratio on the part of subordinates

concerning their past and current supervisors.









It is possible to conceptualize that if a mandate for change had

been in existence that subordinates might have perceived a non-

congruent successor coming into replace a well-liked predecessor

with a resulting low acceptance of the successor. Subordinates might

have been less likely to accept the new supervisor if they felt

unwanted change was coming. Consequently, the fact that there was

little or no mandate for change in the groups examined may have had a

moderating effect on the association between change and successor

acceptability to the group.

Valuable information did present itself when an attempt was made

to cross-validate each of the randomly generated groups taken from

the population of 559 respondents. The multiple correlation coeffi-

cients were high when using stepwise regression on an individual basis

to determine the association between leader behavior and successor

acceptance to the work group (Table Twenty-One) and again when using

all items in the questionnaire (except questions 45, 46, 47) with

successor acceptance (Table Twenty-Two). These data support the value

of the overall model in terms of specifying those types of supervisory

behavior that lead to acceptance by subordinates.

The replication of questions 13, 24 and 34 in the split samples

in Table Twenty-One and the replication of questions 2, 13, 34 and 51

in Table Twenty-Two indicate considerable cross-validation between the

randomly generated groups of 280 and 279 respondents respectively.

After using the significant leader-behavior variables in one sample as

independent variables in the second sample, questions 13, 24 and 34

explain approximately 61% of the variation in acceptance scores.









Using all of the questionnaire items (except 47) in a similar manner

results in questions 2, 13, 34 and 51 explain 67% of the variation

in acceptance scores. The high multiple correlation figures indicate

that the variables that remained in the model, after the imposition

of the 5% criterion to enter and stay, account for a large percentage

of the variance in the model in terms of those questions which are

most important in explaining the acceptance behavior of subordinates.

One of the most important results of this investigation would be the

pinpointing of those types of behavior displayed by new supervisors

that seem to explain why they are accepted by those they supervise.

In conclusion, it would seem that the primary thrust of future

research efforts should be directed at investigating further the

reasons why specific leader behavior questions had a strong associa-

tion with the acceptance level of the new supervisor. As mentioned

earlier, the seeming importance of question 34 to subordinates in

terms of successor acceptance merits specific attention, especially

since it did not appear in any of the leadership dimensions resulting

from the factor analysis. Possible causes of this result are cer-

tainly worth investigating. Certainly, the data gathered in this

study lead one to new questions and new paths to explore concerning

the overall topic of supervisory succession and its impact on work

groups.















BIBLIOGRAPHY


1. Carlson, Richard 0. Executive Succession and Organization Change.
Chicago, IL: Midwest Administrative Center, University of
Chicago, 1962.

2. School of Superintendents: Careers and Performance.
Columbus, OH: Chas. E. Merrill, 1972.

3. Davis, Keith. Human Behavior at Work. 4th Edition.
Hightown, NJ: McGraw-Hill, 1972.

4. Fiedler, Fred E. and Martin M. Chemers. Leadership and Effective
Management. Glenview, IL: Scott-Forsman, 1974.

5. Fox, William M. An Analysis of Military Leadership in a Realistic
Field Setting. Technical Report 70-6. Organizational
Effectiveness Research Programs, Office of Naval Research
Code 452. Arlington, VA 22217, September, 1975.

6. Identifying and Developing Leadership Aspects of
Effective Management in Team-Oriented Task Groups.
Technical Report 70-4. Organizational Effectiveness
Research Programs, Office of Naval Research Code 452.
Arlington, VA 22217, March, 1974.

7. The Management Process. Homewood, IL: Richard D.
Irwin, 1963.

8. Gamson, William A., and Normal A. Scotch. "Scapegoating in
Baseball," American Journal of Sociology, Vol. 70 (July,
1964), 69-72.

9. Gouldner, Alvin W. Patterns of Industrial Bureaucracy.
Glencoe, IL: The Free Press, 1954.

10. "Taking Over," Trans-Action, Vol. 1, No. 3 (March,
1964), 23-27.

11. Grusky, Oscar. "Administrative Succession in Formal Organiza-
tion," Social Forces, Vol. 39 (December. 1960), 105-115.

12. "Managerial Succession and Organizational Effective-
ness," American Journal of Sociology, Vol. 69 (July, 1963),
21-23.









13. Guest, Robert H. Organizational Change: The Effect of Successful
Leadership. Homewood, IL: Richard D. Irwin, Inc., 1962.

14. Hollander, Edwin P. "Processes of Leadership Emergence,"
Journal of Contemporary Business, Vol. 3, No. 4 (Autum
1974), 19-33.

15. Hunt, James G. and Lars L. Larson, eds. Contingency Approaches to
Leadership. Carbondale, IL: Southern Illinois University
Press, 1974.

16. Indik, Bernard, Stanley E. Seashore, and Jonathan Slesinger.
"Demographic Correlates of Psychological Strain," Journal
of Abnormal and Social Psychology, Vol. 69, No. 1 (1964),
26-38.

17. Levenson, Bernard. "Bureaucratic Succession," Complex
Organizations: A Sociological Reader. Edited by Amital
Etzioni. New York, NY: Holt, Rinehart and Winston, 1961.

18. Reitz, H. Joseph. "Managerial Attitudes and Perceived
Contingencies Between Performance and Organizational
Response," Academy of Management, Atlanta, GA: (August
1971), 227-238.

19. Scott, William E., Jr. "The Development of Semantic Differential
Scales as Measures of 'Morale'," Personnel Psychology,
Vol. 20, No. 2 (1967), 179-198.

20. Stogdill, Ralph M. and Alvin E. Coons, eds. Leader Behavior:
Its Description and Measurement. The Bureau of Business
Research. Columbus, OH: The Ohio State University, 1957.

21. Zald, Mayer N. "Who Shall Rule? A Political Analysis of
Succession in a Larger Welfare Organization," Pacific
Sociological Review, Vol. 8 (1965) 52-60.















SUPPLEMENTAL BIBLIOGRAPHY


"An Apprentice President Makes the Grade," Business Week, No. 2142
(September 19, 1970), 60, 65.

"Autos: New Model Chrysler," Newsweek. Vol. 36, No. 3 (July 21,
1975), 55-56.

Baars, Arthur W. "A Preliminary Analysis of Strategy in Executive
Succession." Ph.D. Dissertation, University of Southern
California, 1973.

Baehr, Melany E. and Richard Renck. "The Definition and Measurement
of Employee Morale," Administrative Science Quarterly, Vol. 3,
No. 2 (September 1958), 157-117.

Beall, Harold Addison. "Administrative Succession: A Study of Role
Acquisition and Maintenance." Ed.D. Dissertation, University of
Oregon, 1962.

Behling, Orlando and Chester Schriesheim. Organizational Behavior-
Theory, Research, and Application. Rockleigh, NJ: Allyn and
Bacon, Inc., 1976.

"Big Steel Shuffles from the Inside," Business Week. No. 2257
(December 2, 1972), 24.

Bonner, R. E. "On Some Clustering Techniques," IBM Journal.
(January, 1964), 2-32.

Brown, Michael Craig. "Bureaucratic Dynamics: Leadership Succession
and Change in the Working Organization." Ph.D. Dissertation,
Cornell University, 1976.

"Can Frank Borman Make Eastern Take-Off?" Business Week. No. 241
(December 22, 1975), 40-44.

Carlson, Richard 0. "Succession and Performance Among School Super-
intendents," Administrative Sciences Quarterly. Vol. 6
(September 1961), 210-226.









Cherrington, David J., H. Joseph Reitz, and William E. Scott, Jr.
"Effects of Contingent and Noncontingent Reward on the Relation-
ship Between Satisfaction and Task Performance," Journal of
Applied Psychology, Vol. 55, No. 6 (December 1971), 531-536.

Cooper, W. W., H. H. Leavitt, and M. W. Shelly, II. eds. New
Perspectives in Organization Research. New York, NY: John Wiley
and Sons, 1964.

Cronbach, Lee J., and Lita Furby. "How Should We Measure 'Change'--
or Should We?" Psychological Bulletin. Vol. 74, No. 1 (July,
1970), 68-80.

DuBrin, Andrew J. Fundamentals of Organizational Behavior--An Applied
Perspective. Elmsford, NY: Pergamon Press, Inc., 1974.

Fiedler, Fred E. A'Theory of Leadership Effectiveness. Highstown, NJ:
McGraw-Hill, 1967.

Garner, Warren Kenneth. "Leader Behavior Following Executive Succession
in Selected California School Districts." Ph.D. Dissertation,
Claremont Graduate School and University Center, 1969.

Gordon, Gerald and Selwin Becker. "Organizational Size and Managerial
Succession: A Reexamination," American Journal of Sociology,
Vol. 70 (1964), 215-223.

Grusky, Oscar. "Corporate Size, Bureaucratization and Managerial
Succession," American Journal of Sociology, Vol. 67 (1961),
263-269.

"Succession with an Ally," Administrative Science
Quarterly, Vol. 14 (June, 1969), 155-170.

Guest, Robert H. "Managerial Succession in Complex Organizations,"
American Journal of Sociology, Vol. 68 (1962-1963), 47-54.

Guertin, Wilson H., and John P. Bailey, Jr. Introduction to Modern
Factor Analysis. Ann Arbor, MI: Edwards Brothers, Inc., 1970.

Hall, Richard H. Organizations--Structure and Process. Englewood
Cliffs, NJ: Prentice-Hall, 1972.

Helmich, Donald L. "Corporate Succession: An Examination," Academy
of Management Journal, Vol. 18, No. 3 (September 1975), 429,441.

Helmich, Donald L. and Warren B. Brown. "Successor Type and Organiza-
tional Change in the Corporate Enterprise," Administrative Science
Quarterly, Vol. 17 (1972), 371-381.

Hemphill, John K. Situational Factors in Leadership. The Bureau of
Education Research. Columbus, OH: The Ohio State University,
1948.









Herbert, Theodore T. Dimensions of Organizational Behavior.
New York, NY: MacMillan, 1976.

Hersey, Paul and Kenneth H. Blanchard. Management of Organization
Behavior, 2nd Edition, Englewood Cliffs, NJ: Prentice-Hall, 1972.

Holland, J. L. Making Vocational Choices: A Theory of Careers.
Englewood Cliffs, NJ: Prentice-Hall, 1973.

Manual for Vocational Preference Inventory, Palo Alto, CA:
Consulting Psychologists Press, 1965.

Huneryager, S. G. and I. L. Heckman. Human Relations in Management.
2nd edition. Cincinnatti, OH: Southwestern Publishing Company,
1967.

Jacobs, T. 0. Leadership and Exchange in Formal Organizations.
Human Resources Research Organization, 1970.

Jay, Antony. Management and Machiavelli. New York, NY: Holt,
Rinehart and Winston, 1968.

Jennings, Eugene Emerson. The Mobile Manager: A Study of the New
Generation of Top Executives. Highstown, NJ: McGraw-Hill, 1967.

Kahn, Robert L. and others. Organizational Stress: Studies in Role
Conflict and Ambiguity. New York, NY: John Wiley and Sons, 1964.

Kerle, Robert H. and Hilton M. Bielek. "The Construction, Validation
and Application of a Subjective Stress Scale," U.S. Army Leader-
ship Human Research Unit, Presidu of Montery, California,
(February, 1958), 1-39.

Knedlik, Stanley M. "The Effect of Administrative Succession Pattern
Upon Education Innovation in Selected Secondary Schools."
Ed.D. Dissertation, New York University, 1967.

Kossen, Stan. The Human Side of Organizations. San Francisco, CA:
Canfield Press, 1975.

Kotin, Joel and Myron R. Sharaf. "Management Succession and Administra-
tive Style," Psychiatry, Vol. 30 (1967), 237-248.

Likert, Renisis. "Organizational Improvement and Human Resources
Accounting," paper presented at American Psychological Association
Meeting, Montreal, August, 1973, 1-10.

New Patterns of Management. Highstown, NJ: McGraw-Hill,
1961.

Likert, Rensis, David Bowers, and Robert M. Norman. "How to Increase
a Firm's Lead Time in Recognizing and Dealing with Problems of
Managing its Human Organization," Michigan Business Review,
Vol. 11, No. 1 (January 1969), 12-17.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs