Faculty job attitude

MISSING IMAGE

Material Information

Title:
Faculty job attitude construction and analysis of a series of models
Physical Description:
xi, 103 leaves : ill. ; 28 cm.
Language:
English
Creator:
Wahl, John Charles, 1948-
Publication Date:

Subjects

Subjects / Keywords:
Teachers -- Job satisfaction   ( lcsh )
Community college teachers -- Attitudes   ( lcsh )
Community colleges -- Faculty -- Attitudes   ( lcsh )
Educational Leadership thesis Ph. D
Dissertations, Academic -- Educational Leadership -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1986.
Bibliography:
Bibliography: leaves 98-101.
Statement of Responsibility:
by John Charles Wahl.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000926085
oclc - 15997809
notis - AEN6745
System ID:
AA00002163:00001

Full Text





















FACULTY


JOB ATTITUDE:


CONSTRUCTION AND


ANALYSIS


OF A SERIES OF MODELS


By

JOHN CHARLES WAHL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL


OF THE UNIVERSITY OF


FLORIDA


PARTIAL FULFILLMENT OF THE REQUIREMENTS


DEGREE OF


DOCTOR OF


PHILOSOPHY


UNIVERSITY


OF FLORIDA


1986





























dedicate


this


study


to the three most


important


people


in my


life:


first


to Beth


Leggett


Cameron,


my wife,


provided


inspiration,


encouragement,


support


above and beyond


the call


duty;


second


to Flora


Flint


Wahl,


my mother,


taught me


to love


read;


third


to Gerald Ayers


Wahl,


my father,


taught me


value


perseverance.
















ACKNOWLEDGEMENTS


Several


organizations,


advisors,


friends


have assisted me


with


the development


of this


study.


They


deserve


greater


reward


than


will


can provide.


show the esteem


hope


that


in which


acknowledgement


hold


their


indebtedness


support.


First,


Next,


offer


want


thanks


to thank


to the participating


committee


colleges and


for their willing


faculty.


support


my endeavor,


and for


their willingness


to investigate new theoretical


descriptions of


job attitude.


am grateful


to Dr. Lloyd Robison


of Auburn


University for


introducing me


to Thom'


elementary


catastrophe


theory.


also want


to thank


Dr. Stephen Olejnik


individually for


assistance


in initial


development


of this


study


and for his


contagious


thank


Department

Services,


enthusiasm for


research and


the Graduate Council


teaching


of the University


of Educational Administration,

the University Council for Educ


of Florida,


the Office of


Extended


nationall Administration,


the American Association


of School Administrators


for their


faith


the assistance and


opportunity they


have made


available.


Special


thanks go


to Fred Blakey,


Ann Burch,


Patricia Campbell,



















TABLE OF CONTENTS



PAGE

ACKNOWLEDGEMENTS..................................................iii

LIST OF TABLES.................................................. vii

LIST OF FIGURES.................................................. iii

ABSTRACT............................................................X

CHAPTERS

I THE PROBLEM AND ITS SIGNIFICANCE...........................1


Nature of the Problem..........
Purpose of the Study...........


Statement of the
Models and Assoc
Model I......
Model II.....
Model III....
Model IV.....
Model V......
Importance of th
Assumptions.....
Candor.......
Intrapersonal
Intrapersonal
Cooperation..


Suitabili
Definition o
Perceptio
Emotional
Job Attit
Overview of
Organization


. ...... ... .. ..... .... .1
.........................7


Problem................................7
iated Hypotheses........................8
. .. .. .. .. . .. . . . .0.8


Study..0
Study..


Consistency..
Stability....


ty of Instrumentat
f Terms...........
in of Working Condi
Involvement in th'
ude...............
Methodology.......
of the Study.....


. . . .. .. ... .8
.. .. .. .. .. 0...0...0 ....... 0 .9
... 0........ 0.0.....00000. ...9
. 0................. 0. .. ..9
............ .. .......... .10
.........................10
. .. .. .. . ... .. .10
. .. .. .. ........ .. .11
.. ... ......... .. .. .. .11
.........................11
ion......................11
...... ..... .............. ll

tions....................12
e Job....................12
.........................13
. ... .. . ... . ..13
.........................14


REVIEW OF LITERATURE...................................... 15











Motivation...........
Emotional involvement
Systems Theory.........
Catastrophe Theory......
Summary...................


............................23
and affective arousal......26
. . . . .. .... .29
............................30
.... .. .... 0. ... ....... .. ... 35


MODEL DEVELOPMENT........................................ 37


Intr6duction.............
Single Continuum Models..
Model I...............
Model II..............
Model III.............
Model IV..............
A Model with Multiple Fra


LI


..............................37
...... ... .... .. ..... ........38
.... .....0.. ..................38
. . .. .. .. .. . .40
.. ................. .. .. .42
.. .. ...... ...... .... .. ...44
mes of Reference: Model V.....47


Summary...................................


....51


METHODOLOGY...............................................53


Design................
Population and Sample.
Instrumentation.......
Survey Procedure......
Statistical Analysis..


Data Analysis.....
Survey Response
Demographic Comrn
Model Analysis....
Model I........
Model II.......
Model III......
Model IV.......
Model V........
Summary...........


. . ............ .. ...53
.... ... ...... ...................54
.... .............................55
........................... .....58
.................................59


S..... ..... ....o.... ........ .....61
s......... ....... ........ ...... .....61
parisons.............................62
.. ............ ....................65
.... ............. ......... ..... ... 66
.. ..................0...............66
. ..... ... . .. .. .. ....67
......o...............................69
.....................................70
.. ......... ...... .....o ...........72


SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS.................78


Summary................................
Conclusions............................
Recommendations........................
Instrumentation.....................
Job Attitude Models................
High job attitude faculty........
Low job attitude faculty.........
Summaryv. . .. ..


............... .78
.... .. ...... .82
................83
................83
................84
................85
................85
as;


RESULTS...................................................60












APPENDIX B. PERMISSION TO REPRODUCE COPYRIGHTED


MATERIAL..


REFERENCES.......................*..******* ***** ****.***. ..** *****... 98

BIOGRAPHICAL SKETCH.........................*****************-****102


. .. .. .. .. .. .. .. .96


















LIST OF TABLES


TABLE


PAGE


Sample and State of


Florida Demographic


Comparisons.


and Employment


Demographics


of the


Respondents..


.... .64


Test


of Model


III,


the Planar


Single


Continuum Model


Test


Model


the Planar


Single Continuum


Interaction Model.......


.... ......... .69


Test


of Model


a Curvilinear


Planar


Model...


.........63


........68


...............72
















LIST OF


FIGURES


FIGURE


PAGE


Aspects of work


Emotional


involvement


and job


A job attitude model


analogous


to the cusp model..........34


Model I--A two-dimensional single continuum linear
model relating job attitudes and perceptions of
working conditions................. .......... .... ...39


Model II--A two-dimensional single continuum
linear model relating job attitude and level of
emotional involvement in the job.......................41


Model III--A three-dimens
planar model relating
of working conditions,
involvement in the job


ional single
job attitude
and level o


Model IV--A three-dimensional single
planar interaction model relating
perception of working conditions,
emotional involvement in the job..

Model V--A three-dimensional multiple
planar interaction model relating
perception of working conditions,
emotional involvement in the job..


continuum
, perception
f emotional
.....................43

continuum
job attitude,
and level of
.....................46

continuum
job attitude,
and level of


The best-fit model


for the data


collected


in this


study...71


Plot of job attitude
by perception of
high emotionally


(level
working
involved


of overall
conditions
responses


job satisfaction)
for moderately
in model V.........73


job attitude..........................20


attitude....................27


The Cusp Model.............................................32


....49










Plot


of job attitude


by perception


low emotionally


(level


of working


involved


of overall
conditions
responses


satisfaction)


in model


..75


Plot


of job attitude


perception


low emotionally


(level


of working


involved


of overall
conditions
responses


job satisfaction)
for moderately


in model


V....


Plot


of predicted


attitude


residuals


showing


heteroscedasticity.....


V........


.......81
















Abstract


of Dissertation


of the University


Presented


of Florida


to the Graduate


in Partial


Fulfillment


School
of the


Requirements


for the


Degree


of Doctor


of Philosophy


FACULTY


ATTITUDE:


ANALYSIS OF


CONSTRUCTION AND


A SERIES OF


MODELS


JOHN


CHARLES WAHL


December,


1986


Chairman:


James


L. Wattenbarger


Major


Department:


Educational


Leadership


purpose


of this


study was


to develop a


parsimonious


explanation


of the relationships


among


community


college


faculty


attitudes,


perceptions


of working


conditions,


levels


of emotional


involvement


that


would


permit


community


college


administrators


anticipate

This was a


faculty


n importa


reactions

nt area f


to changes


in the work


or study because


current


environment.

theory,


supported by re

relationships b

The initial


search,

between


presented c

job attitude


problem was


contradictory

and associate


to determine,


using


explanations

ed variables.


a review


literature,


the relationships


between


variables


that


have been


used


to describe


job attitude


in theory


and research.


Subsequently,


series of models


representing


job attitude


relationships was


developed,


culminating


in a model


structurally


analogous


to the


v









A stratified


cluster


sample


of full-time


faculty


from large,


medium,


and small


public community


colleges


in Florida


was surveyed


using a mailed-response questionnaire


(58.1%


returned


of 607


distributed).


This questionnaire


was


used


to ascertain


faculty


attitudes


(measured as


overall


satisfaction),


perceptions


working


Index


conditions

Smith, Ke


(using


ndall,


a shortened

and Hulin, w


version


ith


of the Job


a modified


Descriptive


11-point


Likert-type


answer


format),


level


of emotional


involvement


in the


(indicated


the intensity


component


of the above


Likert-type


format).


Relationships


conditions,


among


level


faculty


of emotional


job attitude,


involvement


perception


the job


of working


represented


the structural


models


were


significant


using


the community


college


faculty


sample.


A curvilinear


model


was


found


to best


represent


job attitude

results were


relationships

interpreted a


for community


s indicating


college


support


faculty.


for the


use of the


elementary cusp catastrophe

framework for representing


model


as an analogical


the relationships


among


structural


attitude,


perception


of working


conditions,


level


of emotional


involvement.


Further


research,


including


a longitudinal


study with a


larger


sample,


needs


to be


conducted


to verify


the current


findings.
















CHAPTER


THE PROBLEM AND

Nature of


ITS SIGNIFICANCE


the Problem


In order


to direct


growth


and development


a community


college


efficiently,


an administrator


should be


able


to predict


changes


faculty


job attitude


that


will


result


from planned


changes


in the


institutional


environment.


A problem in


college


administration


been


that


relationships


between


the variables


that


comprise


attitude


are not


described


consistently by theory.


Instead,


practicing


administrators


are


confronted with


job attitude


theories


that


make


conflicting predictions


of relationships


between similar


variables

measure o


Further,


f empirical


each


of the conflicting


support.


This


theories


problem appears mo


has received a

st distinctly


when


the job


attitudes


of satisfaction


and dissati


sfaction are


considered.


group of


attitude


theorists


pporting what


is called


the single continuum

relationship between


environment


theory,

level o


and perception


theorists maintain


that


holds


if sati


that


there


sfaction


of working


if changes


a linear


perceived


conditions.


are made


a work


Single


the work


continuum


environment


which


result


enhance


will


the worker


be improved


s perception


job attitude.


of working


Conversely,


conditions,


changes which













level


of job attitude.


This


relationship between


job attitude


perception of working


conditions


theorized


to hold


for the whole


range of


job attitudes


from maximum


sati


faction


to maximum


dissatisfaction.


Change


in level


of attitude


along


this


single


continuum


is thought


to be


contiguous


and therefore,


predictable.


the applied context


educational


admini


station,


the single


continuum


theory


indicates


that


if the administrator


understands


perceptions of


present


working


conditions


eld by


faculty,


knows


the attitudinal


valence


and vector


of any proposed


change


in the work


environment,


the resultant


level


job attitude


can be predicted.


Experience,


supported


records


of high


turnover


of administrative


personnel,


indicates


that


although


this


relationship


is frequently


accurate, s

Another


uch a


description


job attitude


of job


theory,


attitude


called


is simplistic.


the two-factor


or two


continuum theory,


was presented by


Herzberg,


Mausner,


and Snyderman


1959).


Using


a semi


-structured


interview technique


in which


subjects were asked


to identify


describe


highly


satisfying


highly


dissatisfying


critical


incidents


from


their working past,


Herzberg


et al. found


that


different


sets


of factors were associated


with


satisfaction and with dissatisfaction.


set of factors,


which Herzberg


et al. called


hygiene


factors,


was responsible


attitudinal


frame of


reference


called


job dissatisfaction


when


they














predominant,


but resulted


a condition


no satisfaction


when


absent.


was a lack


A lack


of motivator


of hygiene


factors


factors was


satisfying.


not dissatisfying;


Change between


neither


these


frames of


reference appears


to be nongradual


or discontinuous--a


rapid,


unpredictable


shift


from


job satisfaction


to job


dissatisfaction or


vice


versa--brought


on by


a minor


change


working


conditions.


type of


change


is exemplified by


a specific


incident


precipitating


a major


change


in level


of attitude


such


as a


-- reprimand being


reason


for quitting


a job.


Change within


each


these


frames


of reference seems


to be


gradual


or continuous,


with


attitude


improving or


of motivator


Results


hygiene


of tests


declining


factors


of hypotheses


in relation


present


designed


to changes


in the work


to examine


in the levels


environment.


the validity


of the single continuum and


the two continuum


job attitude


theories


have


been


inconsistent


contradi


ctory.


Each


theory provides a


valid description


of the job


attitude with


working


conditions


relationship under


given


situations


using


specified methodology.


Each


theory has


the ability


to predict


certain


types


of attitudinal


results


following changes


in the work


environment,


but neither


type


of theory


accounts


for all observed


outcomes.


two continuum


theory


seems


to be bound


to the critical


incident methodology,


while


the single


continuum


theories


lack


the ability


to describe


nongradual





m


I .













comprehensive description


of job


attitude


that


can be used


to predict


attitudinal


consequences of


changes


in work


environment.


It is


important


that


this


encompassing model


not be situationally


methodologically


dependent.


Several


researchers


(Flay,


19781


Poston


Stewart,


1978;


Woodcock


Davis,


1978;


Zeeman,


1976)


have


suggested


that


the elementary


cusp


catastrophe model,


one of


a series


of models


developed


by Thorn


(1969)


to define


the equilibrium solutions


between


cross-pressured


variables,


would be


useful


in the social


sciences


for describing


attitude-related


phenomena.


Flay


(1978)


stated


that


catastrophe


models


have been


shown


to be


parsimonious,


synthesi


zing many


diverse


and sometimes


seemingly


contradictory


findings


theories


from


social-psychological


literature.


He suggested


that


incongruent


attitude-related phenomena might


be modeled


using


a single


catastrophe mod


that


such a


model


could


provide better


understanding of


the relationships


found


in previous


research.


Flay


cautioned

the social


that a major

sciences was


problem with

that much o


catastrophe


f the modelin


theory modeling

a has been


speculative and not


verified


experimentally.


Catastrophe


theory


been


used


in experimentally


controlled


situations


in the


physical


sciences


and has,


in these


situations,


withstood


tests


that


were


acceptable


even


to critics of


catastrophe


theory


in the social














A problem with


verification


of catastrophe models


in the social


sciences


has been


imprecision


of measurement


and scaling of


social


science constructs,


something which becomes


increasingly


important


models


become more


complex.


Imprecision


of measurement


scaling


is not unique


of the social


to catastrophe


sciences


theory modeling


in general.


is a


difficulty


characteristic


in developing


models of


sufficient


generality to warrant


the effort


of colle


cting


data,


while


retaining


enough


specificity


to have applicability,


also contributed


to the current


lack


of verified


catastrophe models.


approach


sciences t

be to draw


to the


hat minimizes


use of catastrophe modeling


these measurement


an analogy between


the dimensional


in the social


and scaling problems would


structure


appropriate catastrophe model


the relationships


between


variables


within a


social


sciences problem.


structure


presented


by the


analogical


model


could be


used as


a basis


for deducing


relationships


between


problem variables


that


could be


tested


using


available


levels


of measurement


scale.


If the analogy


is supported,


inference of


principles


found


in the idealistic


catastrophe model


can be


used


derive new understanding


of the social


problem.


Although


following


this


process would


not imply


that


structure of


a selected


catastrophe model


would be


analogous


to a


given social


problem


in all details,


the model


could


be used as














process


should


be placed


on gaining new knowledge


about


a social


problem of


interest


rather


than on


the precision


a particular


idealistic model

social model is


for describing


determined


a relationship.


through successive


iterati


true nature of

on of testing


and formulation.


Implementation


of the


process


of analogizing


from


catastrophe


theory models


to a


social


problem such as


job attitude


can be


straightforward.


variables


that


interrelate


to define


attitude


problem are determined from findings


and contexts of


previous


research.


Attention


is directed


to the relationships


between


variables


that


have been


experimentally validated but


apparent

research


conflict

contexts


with other


validated


for similarities


findings in order

differences both


to examine

in definition


of variables


investigated and


in the experimental


designs


used


in the


conflicting studies.


The general


literature


of the problem area must


be investigated

been neglected


to be certain


that


that


could account


important


relationships


for the discrepancies


have not


between


verified

reviewing


theories.


Relationships


the literature may


then be


between


test


variables

d using a


identified by


n appropriate data


set to


ensure


that


the predicted


relationships


do indeed


exist


in the


sample under


study.


Once


the number


of variables


involved has


been


determined,


the appropriate


elementary


catastrophe mod


el may be


selected.


Variables may


then be assigned


to the


axes of


are


the














relationships


revealed


by previous


research


(and


the rules


of thumb


of the practitioners


of the profession)


to determine


if the model


face


validity.


If known


relationships


are


accommodated,


the model


would


then be worthy


of further


study.


Purpose of


the Study


purpose of


this


study was


to develop a


parsimonious


explanation of


the relationships


perceptions of working


involvement

anticipate


conditions


that would permit

faculty reactions


among

, and


community

to changes


the faculty's


levels


job attitudes,


of emotional


college adm

in the work


inistrators


environment.


This


explanation


was developed


examining


the following multi-stage


problem.


Statement


the Problem


initial


stage


in this


problem was


to determine,


from a


review


of literature,


attitude


the variables


in theory


that


research.


have been


Next,


used


to describe


the variables


determined


be plausible


from the


review were


used


to assemble


structural


models


from which


relationships


(and


related


hypotheses)


were


derived


that


could


be tested


in an appropriate


population


using


regression


analysis.


A subsequent


stage was


to develop parsimonious models


representing these


relationships


and variables


in a


structures


analogous


to elementary


catastrophe


theory.


The main


problem was














Models


and Associated Hypotheses


A series of


models was


constructed


to represent


the relationships


between


job attitude


factors


detected


in the review of


literature.


These models


their


associated


hypotheses


are


listed


here


in order


from


the least


to the most


complex.


Model


Model


represented


the linear


single continuum relationship


between


purpose of


of working


job attitude and


this model


conditions


was


perception


of working


to demonstrate


as a predictor


conditions.


the efficacy


attitude


of perception


in the obtained


sample.


This


hypothesis


tha


relationship was

t knowledge of f


tested by


acuity


examining


perceptions


the null

working


conditions would not significantly


increase


the ability


to predict


job attitude compared


to knowing


only mean overall


satisfaction


- 0),


at alpha


= .05.


Model


Model


II represented


the linear


single continuum


relationship


between


job attitude


and level


of emotional


involvement


in the job.


The purpose of


testing this model


was


to determine


the efficacy


the level


of emotional


relationship was


knowledge of


involvement


tested by


the level


examining


of emotional


o predict

the null

involvement


ob attitude.


hypothesis

t in the i


This


that


ob expressed














Model


Model


III represented


the planar


combination


of the two job


attitude


factors


tested


independently


in models


and II.


This


relationship was


tested by


examining


the simultaneous


null


hypotheses


that


knowledge of


perception


of working


conditions


would not


improve


prediction of


job attitude when


level


of emotional


involvement


was


known,


knowledge


of the level


of emotional


involvement


would not


improve


prediction of


job attitude when


perception of


working


conditions


was known


and H


- 0)


at alpha


= .05.


Model


Model


IV represented a


planar


interaction


relationship among


level


of emotional


involvement,


perception


of working


conditions,


term representing the


interaction


effect


between


level


of emotional


involvement


This model,

designed tc


perception


suggested by


determine


of working


the elementary


conditions,


and job attitude.


cusp catastrophe model,


if the relationship among


job attitude,


was


level


of emotional


consistent


tested by


involvement,


at all levels


examining the


perception


of all variables.


significance


of working


This


conditions was


relationship was


the interaction


term


- 0),


at the elevated alpha


level


of .20.


This


alpha


level


selected


to limit


the possibility


of creating


a type


II error


in the


was














conditions, in ad

interaction model


edition


to the factors


described


in model


that


contributed


These


curvilin


to the planar

ear effects


were


indicated


analysis


of the residuals


in model


IV and by


deduction


from the elementary


cusp model.


Each


instance


curvilinearity was


tested by


examining


the null


hypothesis


that


successive addition


to the planar


interaction model


of terms


representing


conditions and


the second


level


third


of emotional


powers of


involvement


perception


would not


of working


contribute


significantly


to prediction


of job


attitude


= b5


= b6


= X7


= 0)


In all instances,


the alpha


level


was


set at .05.


Importance


of the Study


Community


college administrators


need


an encompassing


attitude


appropriate


theory that


research


can


organize


findings


already verified,


professional


situationally


body


knowledge.


Integration


of these


sources


of knowledge would


result


greater u

community


understanding


of job attitude


college administrators


relationships,


to plan and


implemen


enabling

t institutional


development


ways


that


would


improve


faculty


job attitude.


Assumptions


Several


important


assumptions


were made


regarding


the problem


under


investigation.


These


assumptions


are detailed below:


Candor














effort


was made


to increase candor


responses


assuring


anonymity,

of much of


stressing


the standard


the need


for candor,


demographic


data


and forgoing

to present an


the collection


appearance


privacy.


Thus


was


assumed


that


the faculty provided


candid


replies.


Intrapersonal


Consistency


It was assumed


that


use of


a long questionnaire


introduced


possibility of


extraneous


changes


of attitude


during


participation.


short


form of


instrumentation


was


used


to control


this error.


Intrapersonal


Stability


was


responses


assumed,

given by


lacking ev

the faculty


idence t

reflect


o the contrary,


a stable


that


integration


the

of their


perceptions and attitude.

Cooperation


Faculty


in public Florida


community


colleges were willing


able


to identify


and report


their


perceptions


of working


conditions,


level


of overall


job satisfaction,


level


of emotional


involvement


in their


current


jobs.


Suitability


of Instrumentation


instrumentation


that


was


used


to measure


the variables


under


consideration


is suffic


gently


reliable,


stable,


sensitive,


valid


for the


purpose.













Perception


of Working Conditions


A faculty member


forms perceptions


of various


factors


in the


environment


in which work


takes


place


evaluating


the effects


these


factors


have


on any


internal


agenda


that


the faculty member


regarding

overall P


the workplace.


perception


These


of working


perceptions


conditions.


may

This


be summed for

term was


operationally


defined


use


in this


study


as the perception


measured by the


averaged


scores of


responses


to adjectival


stimuli


contained


in sections


, II, III,


IV of the Faculty


Attitude


Survey,


an investigator-designed


short


form


(with a modified


response


format)


of the Job


Descriptive


Index


(Smith,


Kendall,


Hulin,


1969)


The questions


in these


sections


of the


instrument


sample


perceptions of


four


job attitude


factors--work


on the present


job,


supervision,


present


pay,


and people on


the present


job--found


by Smith


scored


et al.


for this


satisfaction

undecided one


to indicate


purpose


three p

point,


points ,

and r


sati


crediting

responses


esponses


sfaction.


responses

indicating


indicating


instrument


was


indicating


no feelings


dissatisfaction zero


points.


Emotional


Involvement


in the Job


The level


of emotional


involvement


in the job


a measure of


emotional


intensity,


the affective


arousal


that


is invested


in the














which


extreme


attitudinal


positions


are located at


opposite


ends


the scale with


a neutral


position


in the middle,


absolute


value


is taken


of the


raw score.


This


raw


score


will


indicate


intensity


or level


of involvement


rather


than


polarity


of the


answer.


In this


study,


the level


of emotional


involvement


the job


was operationally


defined as


the average of


absolute


values


of the


raw score


responses


to adjectival


stimuli


presented


in parts


through


IV of the Faculty


Job Attitude Scale.


Job Attitude


The

job sati


specific

sfaction.


job attitude

The most d


under


irect


investigate

measurement


in this


of express


study was

d job


attitude


is overall


job satisfaction,


assessed


asking


respondent


to state


the current


level


of personal


agreement


with


statement


asserting


that


the job


is satisfying.


Therefore,


attitude


was operationally


defined


in this


study


as the average of


the faculty member's


responses


to question


two in part


VI and


question


one in part


VII of the Faculty


Job Attitude


Scale.


Overview


of Methodology


problem was


examined


reviewing


the literature


theory


of job attitude


to determine


relationships


between


variables


that


contradictory predictions.


These


relationships


were


then


examined


light


the situations


in which


the research


was


undertaken and














questioned


regarding


their


perceptions


levels


of the indicated


variables.


The hypotheses were


tested


using


regression analyses


determine


the explanatory power


of the proposed models.


Organization


of the Study


This


study


is reported


in six chapters.


first


chapter


description and


an overview


of the problem.


Chapter


two is


a report


of the review of


appropriate


literature


includes


a synthesis


variables


that


have been


determined


in previous


research


to be


important


in the study


of job


attitude.


Chapter


three


a report


the development


a series


of structural models


representing


relationships


found


in chapter


two.


fourth


chapter


description


of the methodology used


to investigate


the problem,


including


data.


data


collection,


results of


instrumentation,


the analysis


of the


and treatment


survey


of the


are discussed


chapter


five.


summary


of findings


is provided


in chapter


six,


well


as implications


of the study


for administrative


practice


and for


further


research.
















CHAPTER


REVIEW OF


LITERATURE


Introduction


There


have


been


several


major


facets


of development


in the study


of job attitude.


As a consequence,


there now


are several


theoretical


foundations


that


describe


relationships


between


the factors


that


comprise


this


construct.


One must


develop an


understanding


of the


theoretical


construct more


foundations


thoroughly.


in order


This


to examine


understanding


the job attitude


can


be improved


keeps


in mind


that most


foundational


theories


were


developed by


synthesizing


information drawn


from


samples


of populations.


Although


relationships described


in the theories


have


been


generalized


larger


populations,


strictly


speaking,


the theories


are valid


only


within


the limited


samples


circumstances


in which


they were


formulated.


If restricted


to these


samples


circumstances,


theories


are


reduced


to findings,


gaps


between


these


findings


be spanned with new


ideas.


These


ideas


can be tested


to develop a


more


complete


understanding


of the structure


of job attitude.


purpose


contradiction


of this


between


review


accepted


is to identify


theoretical


areas and


foundations


sources


and to


investigate


relationships


between


factors


that


have


inharmonious,


one


can













Theoretical


Scientific


Foundations


Management


In the 19th


commercial


and early


goods was


20th


transferred


centuries,


the manufacture of


from small-scale,


supportive,


family-centered


units


to more


efficient


industrial


production


units


designed


using


organizational


principles


that


had been


effective


the physical


sciences


(Dreeben,


1968).


The management


of these early


industrial


organizations


also


was subjected


to scientific analysis.


Taylor's


(1911)


principles


of scientific management


development


of time


and motion


study by Gilbreth


(1911)


illustrative of


the scientific management


movement.


The major


focus


this movement


was to


secure


increased


productivity


designing


jobs


to use worker


abilities


in the most


efficient


effective way


possible.


An underlying


assumption


was


that


securing maximum


productivity,


the organization


would become more


prosperous.


greater


pool


of wealth


would be


divided


between


capital


and labor


and all would benefit


(Taylor,


1911).


Educators,


emulating


the philosophy


and practice


of the


industrialists who controlled


their


fiscal


welfare,


developed


school


systems


that


were organized


for efficiency


and managed


following


scientific principles.


Students


became


products,


teachers


became


workers,


and administrators


became


managers,


with


policies


determined














capital


interests,


stated


that


greed motivated


each


party


shortchange


the other.


response


to scientific management


been


collectivization


of labor,


a factor


in industry


education.


Mathewson


1931)


found


that


labor


interests


restricted


output


to a level


where


they


felt


that


they were


receiving


a fair


return


for effort


expended.


Human Relations


Movement


Another


possibility


for improving productivity


the industrial


workplace was


detected


in the studies


conducted


at the Western


Electric Company's


Hawthorne


Plant


between


1924


1927.


These


studies,


designed


to evaluate


the effect


of varying


environmental


factors


in the workplace,


found


that


the effect


of paying


attention


to the workers was


as important


in increasing


productivity


as any


the environmental


changes


that


were


tried


(Mayo,


1933;


Rothlisberger


& Dickson,


1939).


scientific methodology


and subsequent


merit


the Hawthorne

Landsberger (


studies


1958)


have


Locke


been

(1976)


criticized by Carey


suggests


(1967)


a footnote


(p.1298)


concerning another


Mayo


study


that


it would be


interesting to


what


would have


happened


if the workers


had been


talked


to and


listened

spite of


and no changes


the apparent


had been made


methodological


in working


shortcomings,


conditions.


the Hawthorne


studies


initiated a


strong


human


relations


movement.


The main


focus


- a .


see


* *


*


1




















implication


that


worker


attitude


could


influence


productivity


resulted


in a


plethora


of studies


investigating


the relationship of


job attitude and


job satisfaction


to productivity


that


extends


from


that


date


to the present.


Hoppock


(1935)


reviewed


32 early


studies


investigating


job satisfaction.


He concluded


that


although


studies were done


using widely


at different


divergent


oCC


times


upational


and places


groups


different


from equally


people


divergent


backgrounds,


and despite


the fact


that


no two studies


used


same


technique,


all the


studies


showed


the same central


tendency


of the


worker


to be satisfied with


work.


an outgrowth


of this


review,


Hoppock


investigated


the job attitudes of


500 schoolteachers


in 51


urban


and rural


sites


using


four


attitude


scales.


He identified


working


conditions and psychological


aspects


of the work


environment


(such as growth,

factors related


satisfied


status, recogni

to the teachers'


teachers were


tion,


and advancement)


job attitudes.


compared with


When


the 100 least


as the major

the 100 most


satisfied,


Hoppock


found


that


the satisfied


showed


fewer


indications


emotional


maladjustment,


the sati


sfied were more


religious,


the satisfied enjoyed better


human


relationships


with superiors


and associates


, (d)


more of


the satisfied


were


teaching


cities


more


than


10,000,


differences


in salaries


between satisfied and


dissatisfied were not


statistically


significant,


the satisfied


- ~ ~ ( -


q


1


m .


I














their


work


interesting,


dissatisfied more


frequently


reported


fatigue and monotony,


the satisfied


averaged


years


older


than


the dissatisfied.


Hoppock


(1935)


described


job attitude as


follows:


A person may
uncertain.


be satisfied,


He may


job and dissatisfied wi
specific satisfactions


composite


satisfaction


dissatisfied,


be satisfied with


others;


some


he may


and dissatisfactions


with


th


satisfaction may vary from day


e job as
to day,


indifferent,


aspects
combine


of his
such


into a


a whole.


it may


Such
be


rationalized;
mechanics of
physiological
situation wil


This


definition


it is


not identical


satisfaction may


1


chemistry,
probably


illustrates


with


interest.


eventually be


but external s
help determine

the traditional


timul


explained by
i in the job


the result.


concept


attitude--a


line,


a single


continuum from satisfaction


dissatisfaction


with


indifference


and uncertainty


somewhere


the middle


(see


Figure


2.1).


If indifference and


uncertainty


coexist


in the center


region,


perhaps


they


imply


an additional


dimension.


Hoppock

combination


defined


the job attitude,


of psychological,


job satisfaction,


physiological,


as "any


environmental


circumstances


that


causes


a person


to truthfully


satisfied with my


job'"


(1935,


47).


Subsequent


researchers


have developed


two concepts


of job attitude


from Hoppock's


work.


first


attitude


measure


is overall


job satisfaction,


measure


of the rationalized


assessment


that


a respondent


holds


p











20








>I
*
rd

10
4J
02


C)
tD,



0 H









E


\ 4-1


\ O
\k k
\-I 0





\ 4


0 <
N


0O



C *-
\ 0















r1 9 C
O 0









OC

N
S4-*
\o "C
\ U)
o 0*


\~ ir 3 Q












O \ O .m

\ ** *P
\ '0
4.r4



0 0


v~ l3 -
\


\ V-o
\ tR


ly +j a
Cii 0




0; 0 4
cU 'U
~~4-4




C cfl *-
'4- IC
~c ;( to
J, 0














responses


to factors


that


contribute


or detract


from overall


satisfaction.


sum-of-facets


technique measures


the cumulative


satisfaction


perceived with


various


work


environment


conditions.


Smith,


Index


Kendall,

JDI), an


and Hulin


| objective


1969)


have designed


sum-of-facets


the Job Descriptive


instrument,


to measure


perceptions of

respondent's a


facets


attention


of working

on the wor


conditions


environment,


focusing

making


the

the instrument


job-referent,


rather


than


on the intensity


of the respondent's


feelings


about


the work


environment


(which


would


be self-referent).


This


accomplished by


use of


a scoring method


that


measures


vector


or directionality


responses


rather


than


response


intensity.


Scarpello and Campbell


(1983),


discussing


the non-equivalence


between


single


item measures


of overall


satisfaction and


sum-of-facets measures


of satisfaction,


stated


that


researchers


have


assumed


that


the single measures


of overall


satisfaction


unreliable and,


additionally,


that


non-equivalence with sum-of-facets


measures may


be due


to differences


in scoring methods


and/or


use of


linear vs nonlinear


methods


for facet


combination.


Scarpello and


Campbell


stated


that


there


no empirical


evidence


that


the single


item measure of


job satisfaction


is unreliable


that,


in fact,


has been


shown


to be


extremely


stable and


reproducible


e.g.,


Staines














be expected because of


responding to each


the difference


instrument.


in frame


Scarpello and


of reference


Campbell


used when


also cited


research


Aldag


and Brief


(1978)


and Ferrat


(1981)


indicating


that


there


is no strong evidence


to show


that


item scoring


or facet


combination strategies


single measures


contribute


facet measures


to the non-equivalence


of job satisfaction


between


(pp.


580-581).


They


summarized


this


discussion


saying


that


"it may be


more


fruitful


to consider whether


low correlations


between overall


and specific

variables fr


satisfactions


are caused


om the measurement


of job


omission


satisfaction"


of important


581).


Thus,


sum-of-facets


satisfaction


will


an approximation


(and


perhaps


a strong predictor)


of job attitude expressed


as overall


satisfaction.


Productivity.


Other


reviews


of research


on job satisfaction,


employee


attitude,


and productivity


have


found


that


satisfaction


with


one's position does


not necessarily


imply motivation


to produce.


Brayfield and


employee


Crockett


attitudes


(1955)


reviewed


to performance.


empirical

studies


literature


they


relating


reviewed


involving performance


on the job showed minimal


or no relationship


between attitude


and performance.


Studies


involving withdrawal


from


the job


showed some evidence of


relationship between


employee


a significant


attitudes


and absence


complex

es. Herzberg,














attitudes


are


related


to increased productivity.


However,


positive


correlations were


in many


of these


studies


the large


number


of studies


in which


they


found


productivity


and attitude


unrelated


must be


stressed.


Herzberg


et al.


also


found


evidence of


relationships


between attitude


and absenteeism.


Brayfield and Crockett


the Herzberg


et al. reviews


seem


to agree


that


the research


findings


prior


to 1955


did not


show


strong


positive


correlations


between


attitude and productivity


(Lawler


Porter,


1967).


Subsequent


reviews


by Vroom


(1964)


and Locke


(1976)


supported


these findings.


Motivation.


Another


area


of development


of job attitude


has been


the investigation


of job


attitude


as a factor


in work motivation.


theories


(1938)


investigating


and Maslow's


(1954)


this

needs


relationship

theories.


stem


from Murray's


McClelland


(1955)


developed a


theory


of motivation


from Murray's work


that


was


based on


an affective


terminology.


Within


the framework


of McClelland's


theory,


a motive


the anticipation


a change


in affective


state.


Every motive


is acquired


in the conditions

of developing corn


under


or learned


which we are re


mon motives.


A motive


the individual.

ared increase t


involves


Similarities


:he probability


two points on


affective


continuum--a


present


state and an


anticipated


state.


difference


in intensity


and desirability


between


the two states













of the affective association,


and of the


at which


the association


was formed.


Similar


theories


have


been


developed specifically


work motivation.


Vroom's


(1964)


valence/instrumentality/expectancy


(VIE)


theory and Porter


and Lawler


1968


performance


satisfaction


model


are examples.


Vroom defined


the forces


of job attitude


as follows:


valence


referred


to the level


of affective


preference


associated with


outcome;


expectancy was


the momentary belief


concerning the


likelihood


that


a particular


action


will


be followed by


a particular


outcome;


instrumentality


evaluated


the belief


that


the first


level


outcome


is a necessary


and sufficient


condition


for the


attainment


an ultimate outcome;


force


motivation)


was a monotonically


increasing


function


of the product


of valences


expectancies


14-19)


. Vroom noted


that


valences


can be either


attractive


(seeking


a pay


raise)


or negative


(avoiding


being


fired).


Important


in Vroom's


theory was


the recognition


that


the relationship between


force


and valence


and expectancy was


a variable


function.


Vroom did


not specify this


variable


function,


nor did he


speculate about


cause of


the variability.


effect


of this


function


being


variable


was that


the relationship between


valence and


expectancy may vary


similar


situations when


force


is held


constant.


This


relationship


characteristic of


an interaction


between


variables.














one level


was satiated,


the next


higher


level


of need


became


dominant.


This changed


the focus


of the individual


thereby


changed


the potency


of goals


as motivators.


Maslow


developed


theory while working as a


clinical


psychologist.


He found


that


patients


(who could be


characterized


as highly


emotional)


were


stratified within a needs


hierarchy.


Applying Maslow's


hierarchical


concepts


of motivation


to the


workplace,


Herzberg,


Mausner,


and Snyderman


(1959)


interviewed


working


accountants


and engineers


to determine


what


factors


these


workers


associated with


job satisfaction


and with


dissatisfaction.


They


found


that


factors


associated with


satisfaction appeared


to come


from


the work


itself,


while


factors


relating


to dissatisfaction


were


related


to the conditions


under


which


the work was


performed.


factors


relating


to work


itself


were


termed motivators,


while


the factors


relating


to conditions of


work were


termed hygienes.


Herzberg


further


theorized


that


these


motivator


and hygiene


factors


represented


two distinct,


separate


relationships


or continua--motivators were associated with a


continuum from no


sati


sfaction


to satisfaction


and hygienes with a


continuum from no dissatisfaction


to dissatisfaction.


findings predicted by


Herzberg's


two-continuum


theory


have


been


replicated


in postsecondary


educational


environments


by numerous














a description


of the relationships


between


Herzberg's


theory


several


psychological


theories


that


lends


logical


support


to a


hierarchical


job attitude approach.


The most


telling criti


cism of Herzberg's


two-continuum theory has


been


Vroom's


(1964)


contention


that


use of the


interview


technique allowed subjects


to take


personal


credit


for satisfying


experiences while casting


blame


on the work


environment


dissatisfying experiences.


Other


major


criticisms


of the theory


dealt


with


the categorization


technique


used


by Herzberg


to analyze


the interviews and with


the lack


of proof


a two continuum


structure versus


a single continuum


(House


Wigdor,


1967).


summary of


supportive studies


was


presented by Whitsett and Winslow


(1967).


Other proposed


hierarchies


or stratifications


attitude


that are applicable


to the workplace are Rotter's


(1971)


theory of


internal


versus


external


locus


of control,


and Mi


skel's


(1978)


three-factor


job attitude


theory,


and McGregor's


(1960)


theories X


and Y.


Emotional


involvement


affective


arousal.


In separate


developments


related


to the human


relations movement,


several


researchers


have commented on


the effect


that


emotional


involvement


and affective


arousal


have on


attitude.


Oppenheim


(1966)


stated


that


a continuum of


attitude


intensity


can be


plotted as


a U-shaped curve










High









Intensity












Law


Unavailable Attitudinal


Position


Dissatisfaction


Satisfaction


Extremes Extraes
Possible Cantent Possibln


Relationship between


intensity


and attitudinal


content.


Satisfied











Range
of
Affect










Dlssatisfied


Unavailable Attitudinal
Position


IliJh


Value Importance


Relationship between range


of affect


value


importance.


Low


I













involvement


on the attainability


of stable attitudinal


positions.


When


job attitude


is arranged


on this


U-shaped


curve,


only


extreme


attitude


positions of


satisfaction


and dissatisfaction


are attainable


at high


levels


of emotional


intensity or


involvement.


At low


levels


of involvement


a complete


spectrum of


attitudinal


content


available.


Locke


(1976)


also


indicated


this


effect,


stating


that


higher


levels


of emotion,


range


of available behavior


becomes


polarized


(see


figure


2.2b).


Although Oppenheim


s conception


of attitude


appears


as valid as


the traditional


identical


single


states of


continuum,


attitude


it may not


at high


refer


low levels


to qualitatively


of emotional


involvement.


It is entirely


possible


that


traditional


continuum


concepts of


indifferent


satisfaction and


evaluations


dissatisfaction


of the workplace,


are


rational,


objective,


do actually belong


on a


straight


continuum.


Perhaps


some


jobs


objectively more


satisfying


than


others


independent


our feelings


emotions.


how do we


interpret


and dissatisfaction


that


the other,

appear to


emotional


be modeled


states of


satisfaction


the U-shaped


curve?


The answer may


in the


use of the level


of emotional


involvement


as an additional d

working conditions


imension


to the job attitude


relationship,


by perception


to appropriately model


the job


attitude


problem.


The added


dimension


would


extend


from a


level


~














Systems


Theory


Another


facet


of job attitude


appears


when


it is examined


in a


systems orientation.


Early


theories


of this


genre


include


homeostasis and equilibrium


theories


proposed


Freeman


(1948),


Helson


(1948,


1959),


and Ashby


(1952,


1960).


homeostatic


behavior


principle


based


on the premise


that


behavior


attempt


an organism


to preserve


the integrity


of its environment


reacting


to any force


that


disturbs


the homeostatic


state


(Freeman,


1948).


ability


to maintain homeostasis


requires


that


an organism be


able


to react


to a


stimulus,


to gauge


the effect


reaction,


through


feedback


techniques,


to determine


the necessity


of future actions


adaptation-level


to maintain


theory


or regain balance.


incorporated


Helson's


the possibility


(1959)


a shifting


homeostatic


state--an adjustable


frame of


reference


that


takes


into


consideration


the accretion


of experience.


In this


theory,


a system


at one stage of


development


prone


to react


quickly


intensively


to a


disruption and


to change


its homeostatic


state more


extensively than would a


system with more experience


(more


reactions


to stimuli).


reaction


a system to a


given


stimulus,


then,


to be evaluated


in light


of the


system


s previous


experience,


making


system history


an important


factor


in prediction


of future


responses


to stimuli.














change


a point


where


a critical


boundary


threshold)


crossed,


the equilibrium shifts


to a new


level


of stability where


system can


continue


to operate.


A familiar


example of


an ultrastable


system


is the automatic transmission


found


in an automobile.


This


system

held c


is bounded by


constant.


engine


If load


speed and


is increased on


load


if the accelerator


the system


climbing


hill,


for instance),


engine


speed


decreases


in some


proportion


to the


increase


in load,


until


a critical


point


beyond which


the engine will


stall.


When


engine


speed


decreases


to this


critical


threshold level,


the transmission will


shift


gears,


resulting


in a


new relationship


changed proportion)


between


load


and engine


speed


that


will


allow the


system


operate


to continue


in at least


to operate.


two separate


This


system


frames


is multistable


of reference)


as well


ultrastable--if


one of


frames


of reference


is threatened,


system can


shift


to another


level


of organization.


Ashby


hypothesized that


the nervous


system


is organized as an


ultrastable


system of


ultrastable/multistable


subsystems.


He has


supported


this hypothesis with a mathematical


model,


and with working


models


called homeostats,


that


demonstrate multistable,


ultrastable,


systems.

Catastrophe Theory


A further


development


in the representation


of relationships


can













models


that


describe


the equilibrium relationships which occur


when


systems of


variables


operate


in cross-pressured


situations.


Each


the models proposed by Thom is


appropriate


for describing


a system


with


a specific number


of variables.


These models show potential


means


to represent


relationships


between


the variables


of human


behavior


(Flay,


1978).


Of particular


interest


is the cusp model,


which


shows


possible equilibrium states


systems


of three


variables.


cusp


model


(Figure


2.3)


illustrates


a plane or


frame of


reference


describing


the equilibrium


relationships


between


variables


This


two-dimensional


plane buckles


under


the pressure of


a third


variable


to form a


three-dimensional


overlapping model.


At low


levels of


variable


a single


surface


represents


the range of


stable


equilibrium relationships


possible


between


variables


X and Y.


As the


level


of variable


increases,


a point


is reached where variable


makes


a difference


in the


relationship


between X


and Y.


This point


is called


the germ of


cusp.


At levels


variable


greater


than


the germ of


cusp,


relationships


between


variables


X and Y


assume multiple


stable equilibrium solutions


over


a portion


of the


possible


range of


relationships.


In the cusp model,


two stable


equilibrium surfaces


located


ranges


greater


than


the germ of


cusp are


joined


an unstable


equilibrium surface.


stable














High


Variable


Low



Low







Variable


Germ


' Cusp


I
Stabile
Equilibrium
Surface
I

--- -A


table


Equilibrium


Surface


4%


Stable


Equilibrium


Surface


High


High


Variable


Figure


2.3.


illustrating
The model co


references.


The Cusp Model.


An elementary


catastrophe


the equilibrium relationships between


ntains


and an


two stable


unstable


(attracting


(repelling


surface)


surface)


frame


three


theory model
variables.


frames of


of reference













An analogy


can be


drawn


between


the relationships


among


attitude,


perception of working


conditions,


level


of emotional


involvements


and variables


in the cusp catastrophe model


(see figure


level


of emotional


Within


this


involvement


analogical


in the job


job attitude model,


increases,


as the


job attitude


becomes


polarized


into


high levels


of satisfaction and


dissatisfaction.


relationship between


job attitude


perception of working


conditions


becomes


less


distinct


at high levels


of emotional


frame of


involvement,


reference appear


and multiple


possible with


stable


interpretations of


equivalent


perceptions


working


conditions,


as described by


Herzberg


et al.


(1959).


However,


at low levels of


emotional


involvement


in the


job,


the relationship


between


job attitude


perception


of working


conditions appears


strong.


The effect


levels appears


of increasing


to produce


emotional


different


involvement


results


from


depending on


initial

the


initial


level


of job attitude.


If the initial


job attitude


low,


the result


of increasing


attitude


is polarized


job attitude;


the opposite reaction


is apparent


an initially


high


attitude.


At low


levels


of emotional


involvement,


and within


each


of the separate


frames of


reference,


the relationships


between


attitude


and working


conditions


theorized by Vroom


(1964)


appear


hold.















Satisfied







Job Attitude








Dissatisfied



Low







Emotional


nfl


Involvement


High


Perceptions


of Working


High


Conditions


Figure


2.4.


A job


attitude model


analogous


to the cusp model


cusp model,


representing


the relationships


between


attitude,


perceptions


of working


conditions,


and level


of emotional


involvement.


I


-a









35



Summary


Several


separate


theoretical


views


of the relationships


between


variables


that


comprise


job attitude


have


been


developed by


researchers


When


studying


generalizations


the construct


from these


from different


views


are


vantage


restricted


points.


to the samples


from which


they


have


been


derived and


the circumstances


under which


they were created have been


clearly


delineated,


there are


contradictions


between


the theoretical


positions.


However,


this


process creates gaps


between


the theoretical


positions


that are


currently unexplained.


independent


influence


variables


job attitude.


have been


These were


identified


perception


that


of working


appeared


conditions


and level


of emotional


involvement


in the job.


Relationships


between


job attitude


and perception


of working


conditions


have been presented as


single and dual


continuum


structures.


single


continuum


structures were presented by


researchers


stressing


objective methodologies,


while


the dual


continuum structure was


detected


at high


levels


of satisfaction and


dissatisfaction.


Level


of emotional


involvement


job attitude are apparently


related


in a


complex way,


with


increasing


emotional


involvement


causing


a polarization


in job attitude.


At high


levels of


emotional














emotional


involvement,


perceptions


of working


conditions


are more


rational,


with a


wider


range of


attitudinal


content


available.


Recent


developments


in mathematics


have


identified and proven a


series of


relationships between


variables,


collectively


called


elementary


catastrophe


theory,


that may provide


structure


for a model


that


would


improve


understanding of


job attitude.


Models


illustrating


the relationships


among


attitude,


level


of emotional


involvement,


and perception


of working


conditions


have been


introduced


that


are


structurally


analogous


to the elementary


cusp


catastrophe model.


This analogical


job attitude model


has face


validity


in that


it apparently


resolves


confusion


between


contradictory theoretical


foundations.


In order


to determine


the explanatory power


of this


proposed


analogical


job attitude model,


a series


of structural models


representing


theoretical


relationships


between


job attitude,


level


emotional

constructed


involvement,


and perception


Hypotheses were


developed


f working

to test


conditions was

these models.


sample of


the population


of interest


was


selected and questioned and


regression analyses were


performed


on the


responses.


These


procedures


are described


in detail


in the following


chapters.
















CHAPTER


MODEL


DEVELOPMENT


Introduction


In this


chapter,


several


structural


models


were


developed


which


the dependent


independent


variables


variable


job attitude was


identified


in the review of


predicted by the


literature.


primary


independent


conditions and


their


variables

level of


were


faculty perceptions


emotional


involvement


of working


in the job.


These models

interaction,


were elaborated by

and the curvilinear


investigating


effects


the first


suggested


level


the analysis


residuals of

developed re


relationships


the less complex models.


presenting the


between


Specifically,


single continuum linear


attitude


models


were


planar


faculty perceptions


working


conditions,


tested by


examining


the null


hypothesis


, at alpha


.05;


job attitude


level


emotional

hypothesis


involvement


in work,

at alpha


tested

= .05;


examining


the null


job attitude


perception


of working


conditions with


level


of emotional


involvement,


using


a simultaneous


test


of the null


hypotheses


-0,


at alpha


= .05;


job attitude


and perception of


working


conditions with


level


of emotional


involvement,


including


a .I


.1 I


I~

















relationships


found


significant


in the planar


interaction


model

terms

working

polyno


were


also


tested


representing the

g conditions and

mial terms were


for curvilinearity


squared and


level

tested


cubed


of emotional

as additions


individually


values


including


of perception


involvement.

to the planar


These

interaction


model

with


at alpha


= .05.


the predicted


In addition,


scores


for this


the plots

curvilinear


comparing

model we


the residuals


re examined


for evidence


of bimodality


in the


responses


at higher


levels of


emotional


involvement,


which


would


indicate multiple


stable


frames of


reference


results


of the empirical


tests


of the models


developed


in this


chapter


are reported


in Chapter


V and


discussed


Chapter


Single Continuum Models


Model


of the most


highly


researched


relationships


in the study


attitude


perception


has been


of working


the predictio

conditions.


of job


As discussed


satisfaction


in Chapter


Smith


et al.


(1969)


developed


their


instrument--the


Descriptive


Index--to


measure


the perceptions


workers


hold


of diff


erent


facets


the working


conditions


in their


jobs.


The relationship between


attitude and


perception


of working


conditions


is assumed


to be


single


continuum,


with


job attitude


improving


as perception


were


I


























(0
C
a




*4i





O^
*-

'U
O
'0
0
H
0)
'a




\ ~31
0


In'ar
X. 4-'




A O



\ d



0
CIC









O: CI
o d <



\~0 c -




r O
.44

CE





3~4 tr'
C O
0 0







10
3 HC
V 4-4 *v-4

O -












arE:
o (RHu







'00



OO
D i W4-i






S0
0 A 0
\ 4J *4J





-ri

'or
\U U) O)




rCII 0"
\a 'U'a*






\ 0 A
\r O 11 ^F lr
\r I -I HC

o r-1 ^ ,C



'U tU 03
U) 44 CA
4t1 I n1r-













However,


relationship


between


job attitude


and perception


working


conditions


to be


significant


before


perception of


working


conditions


can be


included


in more


complex


models


attempting


predict


job attitude.


appropriate


regression model


to test


significance of


of working


the relationship between


conditions was


+ blX1


job attitude


where


and perception


represented


job attitude,


perception


measured as


of working


overall


conditions,


satisfaction,


measured by


represented


the perception of


working


conditions


scale of


the Faculty


Attitude


Survey.


relationship to be


tested was


that


there would be


no significant


difference

working co


between


nditions


knowing the information

and chance as predictors


contained


in perception


of job attitude,


or that


, at alpha


= .05.


Rejection


of the null


hypothesis


would


indicate


that


perception


of working


conditions


significant


factor


in determining


job attitude


and should be


included


in future models.


Model


Analogous


to model


a linear


model


was


constructed showing


relationship between


job attitude


level


of emotional


involvement


(see


Figure


3.2).


This model


represented


a test


of the


power


of the


level


of emotional


involvement


to explain


variation


in job attitude,


and also was


included here


as a


preliminary


test


before


the variable


was


























































C
*r*
CP
43

O
0

0
r-4 *



UI


_ ill













was tested


in model


appropriate


regression


equation


to test


the relationship modeled


in Figure


was


where


Y represented


job attitude measured


as overall


sati


sfaction,


represented


emotional


involvement.


The null


hypothesis


to be


tested


was that


there would


no significant


difference between


knowing


the information


contained


in level


of emotional


involvement


chance as predictors of


job attitude,


or that


alpha


= .05.


Note


that


this


model


eliminated


the influence of


polarity


value of


but here


of emotion by


the emotional


improved


considering


component.


job attitude was


only


Model


the magnitude,


was


predicted


analogous


a higher


or absolute


to model


level


emotional


involvement


in the job.


Model


objection


to the linear


single


continuum model


relating


attitude


and perception


of working


conditions


presented


in Figure


was that


it failed


to account


simultaneously


for the effect


emotional


involvement


and of


perception


of working


conditions on


attitude.


This


factor


was easily


accommodated


by using


a planar


single


continuum model


including


the additional


dimension,


level


emotional


involvement


(see


Figure


3.3a).


planar


increase


single continuum model


in job attitude with an


(model


increase


III)


described


in perception


of working


















































i

-I


* .94
I 4p 44
hi I
.4I A I

-4 h ,

V) a


0 C
*n ( a *<-1

CLi 3 H
C 0*O
*- -- >
J ,Q C
r-4 *rl (0 0


'0 < O

C r-
-%

*rI0

*r OkH


24 *M*
V a

ii4 OV
am a. -
At MA aA














Model


was


tested


to determine


the independence of


the two


variables when


variables


used


duplicated


in the same model,


the explanation


examining whether


of variance


one of


the other.


this


were


the case,


the less


complex model


would be


preferred.


combined effect


single continuum model


of the independent


was


tested by


factors


examining


the planar


the significance of


the terms


representing


perception


of working


conditions


and level


emotional


involvement


the job


in the regression model


+ blX1


where


= job attitude


= perception


of working


conditions


= emotional


involvement.


This


was a simultaneous


test


of H0


, and H0


at alpha


- .05.


Failure


to reject


either


of the null


hypotheses


this


model


would


indicate


that


the factors


duplicate


prediction


variance,


and that


the remaining


significant


factor


adequately


predicted


job attitude


itself


Rejection


of both nulls would


indicate


that


the factors


were


independent,


that


both


should be


included


in future models.


Model


planar


single continuum model


(model


III,


Figure


3.3a)


used


to test


the relationships


among


job attitude,


perc


option


was


= bo














contained


in Chapter


II led to the inference


that


this


relationship


did not


always


hold.


For example,


the change


range


of tenable


attitude positions

(Oppenheim, 1966)


job attitude


at varying

appeared to


perception


levels


of emotional


interfere with


of working


involvement

relationship between


conditions.


A planar


single continuum


interaction model


incorporated


features


necessary


to describe


such


a variable


relationship


(see


Figure


3.4,


model


IV).


This


model


was


developed by


examining


the cusp


catastrophe model.


was inferred


that


an interaction


would


occur


between


perception


of working


conditions


job attitude


different

derived f


levels


rom Vroom'


of emotional


(1964)


involvement.


variable


Similar


relationship.


predictions

In addition,


were

the


cusp structure


supported


the prediction


that


the residuals


from


current


planar


interaction model


would be bimodal


to the


inherent


curvilinearity


of the


cusp.


Model


IV was


designed


to test


interaction portion


of this


prediction.


investigation


curvilinearity was withheld


here


and was


investigated


separately


using model


It could be


working


effect


predicted


conditions,


on job attitude,


involvement


was limited


from model


increasing


while


IV that


emotional


the effect


at higher


levels


with


involvement


of increasing

of perceptions


- a


low perception


considerable


emotional

of working
Ui *


i


i


- _1 II



















,:

P
S)
>

^1
0


rl
0f
-r1


U\

4d 0m
HO H-
(0 rn M
0 0
rH.CH


CM 9 '-4
HW- -M >
O 0 G <0 >

Nrl r-t >
PI W CO
u
Cy > oa r-i
C Et O m
0 r- 0 -r 5
CU 0 Q)g
i-i> N-Hr4-4
4H *dC 4
-d 'U
SO W 0 >,

S-C 4r > -*

0 -P 9- (
3 240 -Jd


E: rlt cr r
O 4 44 4W
0 > 0 (0
0 0) G )

u 1-1 r i 0
't Qa a +
*C GON &
EG C ** 0
*Hd ,U Q


0 44 *.4

0 4 r-I 0 0
0 0) ( *r
0 U r 0 (


a C O
*r C OU


*a 0) toa'














significance


of the interaction


in model


IV was


tested by


examining


the interaction


term


in the regression model


+ blX1


+b2X


+ b3X1X2


where


= job


attitude


= perception


of working


conditions


- emotional


and X1X


involvement


- the interaction between X1


This


was a test


significance


level


of HO:


was


= 0,


set high


at alpha


in order


.20.


to limit


the possibility


creating


a type


II error


in the test


an interaction.


Failure


reject


the null


hypothesis


indicated


that


two variables


have


synergistic effect.


If detected,


this


interaction


effect


would


demonstrate


that


relationships


between


variables were


not constant


throughout


the model.


A Model


with Multiple


Frames


of Reference:


Model


Research by Herzberg


et al.


(1959)


has indicated


that


there are


multiple


work.


stable


frames


structure


reference


of the cusp mod


from which

el appears


workers


view their


to accommodate


this


phenomenon.


following methods were


used


to investigate


possibility


of multiple


stable


relationships.


First,


model


IV was


extended by


including polynomial


factors


to test


for curvilinearity


in model


Second,


the plot


of the residuals


and predicted


scores


- b0













A job attitude model


elementary


with


a structure


cusp catastrophe model,


to be capable of


discuss


representing multiple


pla


analogous to

ed in Chapter

nes or frames


Thom's


appears


of reference


in a


single model.


resultant model


(Figure


3.5)


a more


succinct


explanation


of the relationships


among


job attitude,


perception


of working


conditions,


level


of emotional


involvement


than


the separate


single-


and dual-continuum


job attitude


explanations.


identified


This model


in model


incorporates


expands


the interaction


the development


effects


including


polynomial


factors


variables.


attitude


while


to represent


upper


responses,


the lower


surface


corresponds


surface,


curvilinear


of model


relationships


V, representing


to Herzberg'


representing


satisfied


attitude


between


high


responses,


responses,


corresponds


to Herzberg's dissatisfied


responses.


The bimodal


effect


noted


by Herzberg


under


high


levels


of emotional


involvement


modeled,


but diminishes


a single continuum


as the level


emotional


third


involvement


decreases.


set of possibilities


that


The middle


are


surface


in equilibrium bu


represents

t are


unstable--these would be


the "undecided"


participants,


using


Hoppock's


(1935)


terminology.


region


of the model


characterized


levels


of emotional


involvement


lower


than


that


of the germ of


cusp would


represent


the indifferent


participants.


















O4
-rnO

*n 0
3


f 0
-4 v -4
0 04 U6
-4< 9. -4
U 4) *1
cn a a
* 4. '
C12 'C U


i V

%M &
04 (
-a a a
)4J 'a
4) '.
IC w


/,


4-f


I

S1
\ 1-


4- r 0)

00)






U 04d
,- 4 ) .
0uy )4
3W-H 0
00B-)
HC -4J


4=
0 a
- C
Z 0
*u j
0 4

oo
ec
0 0
LJ
U *-
h iC
0.0
&5


in o ^
r4t(t S

a


45
O


-H O U) 0

> r-IHr-
3 C
wrl o r C
*- 0 0

SW0O
cn 0 .


C0 ,4 (



.) C -
. 0 4g
-H O a n


.- A
-4 .94
1U IA


(U r-t Ui 0
0H 4 0 C ,C
WCO0J>XO
S oa Om tr o Q)
410 0 *r 0H

SOc > -


0) 0- -E,C 4

U M X< -^ 0 *H (U
wr0 0) ) C
0 ft C.*0 n3
LA cU) $401

2' r-I H P-
09d O a,

*. Cq 0 U
,lr +j c r"* a -
irl *rM 0 (0
W4-)C4-i'O 0
4 rd WE-- *r
U.0 Q Q
Cn,0 ^^ C 4 r3


" U1














emotional


involvement;


bifurcation,


or existence of


more


than one


stable


job attitude at


certain


combinations


of emotional


involvement


and perception


of working


conditions;


inaccessibility,


or the


existence


of behavioral


states


that


are


unattainable


under


certain


combinations

conditions;


of emotional

hysteresis, o


involvement


and perception


r fluctuation between


opposing


of working


stable


attitude


positions


available


at certain


combinations


of emotional


involvement


and perceptions


of working


conditions;


and catastrophic


change, th

thresholds


e "sudden


jumps"


that


delimit


that


stable


occur

frames


when


the boundaries


of reference


are crossed.


Not all of these characteristics of


the cusp model


were evaluated


in this


study because


of the static


nature


of the data


collected and


used


in the analysis.


For example,


the properties


hysteresis and


catastrophic change can


only


be analyzed


using


dynamic or


longitudinal


data.


The remaining properties,


bimodality,


bifurcation,


inaccessibility,


were


explored by


examining


the relationships


exhibited between


the predicted and


residual


values


created by


polynomial


regression


models


starting with


+ blX1


+ b3X3


+ b4X4


where


= job


attitude


= perception


of working


conditions














Polynomial


factors were


exploratory process


in order


included


in the regression model


to increase explanatory


power


as an


(increase


R2).


procedure used


to direct


development


of the curvilinear


model


was examination of


predicted and


residual


values


of the


previous


best-fit model


to determine


its deficiencies,


with


subsequent inclusion and/or

of the model to the data.


substitution

This process


of terms


to improve


was halted at


an earl


the fit

y stage


due to the heteroscedastic


nature of


the data


set which


interfered


with


the ability


to distinguish


lack


of fit from inherent


error.


Subsequently,


the variable


(emotional


involvement)


categorized


into


low,


moderately


low,


moderately high,


and high


levels.


This


procedure permitted


graphing


a progression


curves


which


facilitated


interpretation


of the analyses.


Summary


A series


of regression


models


were


proposed


to analyze


a static


data


set comprised of


faculty responses


to a


questionnaire


measuring


overall


level


of job


attitude,


emotional


involvement


work,


and perception


the models,


between


of working


various


conditions.


theoretical


fit,


explanations


as described by


of job attitude


and the collected


empirical


data


was


examined.


A further


series


regression models


was used


to analyze


the data


set for the presence


of multiple stable frames


of reference--a property that


was


was














determination


of the appropriateness of


the cusp catastrophe


structure


for modeling


attitude


relationships


could


have


important


implications


for community


college


administrators.


Multiple


response


patterns


to selected


environmental


factors


(such as


response


to salary


increases)


might


be situationally valid


responses


in differing


frames of


reference.


differences


between


single-


and two-factor


explanations


of job attitude may


be encompassed and


the conflict


between


these


descriptions


of job attitude


relationships


may be


resolved.


The cusp model


of job attitude


permits description


of situations


in which


sudden


or unpredicted


changes


in attitudinal


frame of


reference from satisfaction


to dissatisfaction may


occur.


Intervention


strategies


to precipitate or


prevent


sudden


change


become


apparent.















CHAPTER


METHODOLOGY


purpose


of this


study was


to develop a


parsimonious


explanation


of the relationships


among


the faculty's


attitudes,


perceptions


of working


conditions,


and levels


of emotional


involvement


that


would


permit


community


college


administrators


anticipate faculty


reactions


to changes


in the work


environment.


this


end,


a series


of models was


developed


from


relationships


previously

Hypotheses


described


drawn


in job attitude


from these models


literature


were


tested with


theory.

regression


analyses,


using


data


gathered by


surveying


full-time


community


college


faculty


This


chapter


describes


processes


used


to select


the sample and the


instrumentation,


survey


the sample,


and analyze


the data.


Results


of these


analyses


are reported


in Chapter


with


conclusions and


recommendations


presented


Chapter


Design


A stratified


cluster


sample of


faculty members who were


employed


full-time


in public


Florida


community


colleges was


surveyed,


using


self-report,


mailed questionnaire,


to ascertain


their


perceptions


present


employment.


These


data


were


used


to analyze


relationships


among


job attitude,


perception


of working


conditions,


levels


of emotional


involvement


in their


jobs,


in order


to test














Population


and Sample


population


used


for the study


consisted


of faculty members


employed


full-time


in public


Florida


community


colleges.


Community


Colleges


were


stratified


into


large,


medium,


small


institutions


using


fall


semester


1983


enrollment


reports


(Division


of Community


Colleges,


1985)


cluster


samples


consisting


of all full-time


instructional

faculty member


personnel w

s from each


ere


selected


strata.


to include


Using


this


an equal


process,


number


six colleges


were

small


selected as

* one was m


survey


medium,


sites.

and one


Four


of these


was large.


institutions were


Three


of the selected


institutions were either


multi-campus


organizations


or had a main


campus with


one or more


branch


campuses.


selected


institutions


represented


urban,


suburban,


small


town,


rural


communities.


study


controlled


for size


of enrollment


to insure


that


factors


related


would not


community


to the si


bias


college


of community


the results.


faculty


in which


population


a single state


the college was


was


selected


remove


located


from


potential


variation


due to differing


statutory


and regulatory


requirements.


Since


the model


under


investigation


is not dependent


on personal


characteristics


the subjects,


collection


of demographic


information was

to published po


limited


pulation


to factors n

demographics


necessary

for publ


to compare

ic Florida


respondents

community


I














number


of faculty


reported


employed


full-time


as instructional


personnel


by the


Division


of Community Colleges


1985),


a sample


size


of 578

sample


was


selected with


rates of


a maximum difference


5% at a 99% certainty


level.


between


This


population


size was


rounded


to 600 to permit


equal


representation


from each


strata


without


further


sampling within


clusters.


Instrumentation


The instrument


used


assess


faculty


attitude


variables


(the


Faculty


Job Attitude Survey,


see


Appendix A)


was constructed by


modifying


the previously published


instrument,


the Job


Descriptive


Index


(JDI,


copyright


1985),


from Smith,


Kendall,


Hulin


(1969).


The JDI


was originally


designed


to measure worker


satisfaction


with


various


facets of working


conditions


sampling


five distinct


factorially


determined


job attitude areas--work


on the present


job,


supervision,


present


pay,


people on


present


job,


and opportunities


for promotion--and a


sixth


set of questions


that


measure


general


satisfaction.


forms


of the JDI


were


used


in the present


study.


A short


form of


the JDI


was


created


reducing


the number


of questions


each of


the six


topic areas.


Questions


used


in the shortened


version


were selected


to be common


to both


the 1969


the 1985


editions


the JDI.


A portion


of the questionnaires


used


in this


study


(101














response


rate


to the self-administered,


mailed


response


questionnaire,


as proved


to be the case.


short


form had a


return


rate


of 60.7%


(307


of 506 forms


distributed were


returned)


while


original


form was


returned


44.6%


of the time


of 101 distributed


were


returned),


an overall


return


rate


of 58.1%.


When


original


shortened


versions of


the JDI


were


compared


explanatory power


(using


the regression


formula


from model


Chapter


III),


the original


form was


able


to explain


36% of the


variation


attitude,


while


the short


form explained


28%.


Thus,


while


the short


form had a higher


return


rate,


this advantage was


canceled


its lower


explanatory power.


Subsequently,


the long


form


responses were coded as


short


forms


and were


used with


the actual


short


form


responses


in the larger


study.


Other


modifications


to the JDI


consisted of


changing


response

Likert s


format


cale


from


ranging


the original


from


(yes


-5 (strongly


/ no),


disagree)


an 11-point


through


feelings)


to +5


(strongly


agree).


This


change


retained


the content


of the original


format,


since


positive answers


could be equated with


the JDI


"yes"


response,


negative answers


with


the JDI


"no"


response,


and zeros with


the JDI


This


change


permitted


capture of


additional


information


concerning


relative magnitude


of emotional


involvement by averaging the


absolute


values of


responses














unanticipated problem with


responses.


Several


of the community


colleges


participating


in the study


did not


use


faculty


rank.


Some


faculty members


at these


institutions


indicated


that


they


had no


prospect


for promotion,


while others


interpreted


term promotion


more


loosely,


and answered


as intended.


This


problem was


alleviated


removing


the JDI


promotions


category


from


the analyses.


three


job attitude


variables measured with


the Faculty


Attitude


Survey were perception


of working


conditions,


level


emotional


involvement,


overall


satisfaction.


These


variables


were analyzed


using


the shortened


format.


responses


that


were


returned

contained


on the long f

in the short


orm were

form.


recorded


These


to include


recorded


only


responses w


the questions

ere included


with


the short


form responses


in all analyses.


Faculty perceptions


of working


conditions


were measured by


averaging

the first

Survey.

JDI scori


the sum of


four


scores


categories


Responses

ng format,


to these


the number


of questions

e questions w


as previously


discu


of non-missing


on the Faculty

ere transposed

ssed. As with


responses


Job Attitude

to the original

the original


JDI,


answers


indicating


satisfaction


received a


score


of three,


answers


received a


score of


one,


and answers


expressing


dissatisfaction


category by


received a


dividing


score of


the sum of


zero.


scores


Answers w

the number


ere


averaged by


of non-mi


ssing













Level

absolute


four


of emotional


values


categories


involvement


of the non-zero,


of the Faculty


was


assessed by


non-missing


Attitude


averaging


responses


Survey


to the first


following


Suchman


emotional


1950).


A response


involvement


while


represented a


five


represented


a high


level

level


emotional


involvement.


Zero was


undefined,


as an undecided


response


was not necessarily an uninvolved


response.


Overall


job satisfaction


was measured


averaging


responses


the statements


"most


of the time,


find


this


job satisfying"


"overall,


find


this


is satisfying,


" with


responses


indicated on


same


11-point


scale


used


on the rest


of the instrument.


Survey Procedure


investigator


contacted


the chief


administrative officer,


chief


academic officer,


or the designated


representative at


each


institution


giving


selected,


assurances of


participants.


to appeal fo

institutional


The officer


contact


r their


cooperation


and personal

d was asked


in the study,


anonymity


to designate


a contact


person and to grant


permission


to contact


faculty members


through


campus mail


system


solicit


their


cooperation


to distribute


survey


instruments).


On campuses with


established procedures


to gain


permission


to distribute materials


to faculty,


these


procedures


were


followed,


and permission


to use


institutional


facilities


was














Responses were transcribed


to data


entry


sheets


from


the returned


questionnaires,


coded


in machine


readable


form,


processed,


printed,


and checked against


the original


questionnaires


for clerical


errors


in preparation


for further


processing


computer.


Statistical Analysis


relationships


between


the variables


were


examined


using


regression analysis


technique


to test


the appropriate


hypotheses with


the Statistical


Analysis


System


(SAS)


program General


Linear


Model


(GLM)


as described


in Chapter


III.


Results


of these analyses


reported


in Chapter V.


are
















CHAPTER V
RESULTS


purpose of


this


study was


to develop a model


representing


relationships


among


the faculty's


attitudes,


perceptions


working


conditions,


levels


of emotional


involvement


that


would


permit


community


college administrators


to plan


and implement


organizational


changes


effectively.


To this


end,


models


representing


theoretically


derived


relationships


between


attitude


variables


were


developed,


community


college


faculty were


surveyed


to obtain a


cross-sectional


data set


illustrating


their


perceptions


expressions


of the job attitude


variables


comprising


the models,


hypotheses were


tested


using


this


data


set to determine


appropriateness of


relationships.


specific models


tested


were


as follows:


Model


I--Job attitude


a linear


function


of faculty perception


of working


conditions.


Model


II--Job attitude


a linear


function


of faculty


level


emotional


involvement


in the job.


Model


III--Job attitude


perception of


working


a planar


function


conditions


of faculty


level


of emotional


involvement


in the job.


Model


IV--Job attitude


a planar


function


of faculty














Model


relationship found


in model


IV can be extended


addition


perception


of exponential


of working


components


conditions


(created


and emotional


raising


involvement


scores


to the second


powers).


This


chapter


reports


the results of


the tests


of these models.


Conclusions and


recommendations


based


on these


results


are


reported


in Chapter VI.


Data Analysis


Survey Responses


Examination


of the data


from


the returned questionnaires


revealed


two problems with


the original


plan


for aggregating


responses


into


the variables


to be


used


in the model


tests.


These problems


were


several


returns


had data missing


and (b)


Section


requesting


information


on promotions on


the current


job,


had been


interpreted by


respondents


in two distinctly


different


ways


that


jeopardized accurate


interpretation


of the


responses.


The problem of


missing


data


was


alleviated


changing


the method


in which


answers


were


scored


from a


simple


sum


(where missing


data


caused


the entire


questionnaire


to be rejected


from analysis)


a sum divided by the


number


non-missing


responses.


The effect


of this


change was


increase


the number


of usable


responses


while


slightly


decreasing


comparability with other


studies


using


the JDI


instrumentation.














second problem was


specific


to the Florida


community


colleges


selected


for inclusion


in the study.


Several


of these


colleges


use


responded


the academic

nts indicated


ranking

that s


common


ince


in higher


the colleges


education.

did not us


Some


e academic


ranking,


there was


no promotion


policy


and consequently


answered


that


the question


did not apply


leaving


it blank


or by


registering


feelings.


Others


interpreted


the question


as anticipated,


that


raises


and favored


assignments


(and


even


the absence of


a clear


promotion


relation


policy)


to this


constituted


structure.


a promotion


Marginalia


structure,


assisted


answered


in identification


the problem and


eliminating


in these


section


interpretations.


This


from analysis.


problem was


main


effects


resolved


of the


latter


change were


that


the error


term


in the models


tested was


inflated across


all models


and that


the variance


explained


in this


study was


lower


than


in similar


studies


using


same


instrumentation.


Demographic


Comparisons


The demographic


information


obtained


from


survey


summarized,


compared with appropriate


system statistics


(Division


Florida


community


of Community Colleges,


1985)


college

where


applicable and presented


in tables


and 5.2.


figures


presented


in these


tables


are


for full-time


instructional


personnel,


















Table
Sample


and State of Florida


Demographic Comparisons.


Demographic Category Response Percent State Figures
for for Frequency in for
Comparison Comparison (Sample) Category Comparison


Employment


Basis


Full-Time
Part-Time
Non-response


94.6


Sex of Respondent


Female


Male


41.7
57.7


42.8
57.2


Non-response

Race of Respondent
Black


White


89.7


86.0


Hispanic
Other
Non-response

Teaching Assignment:


Advanced


Prof.


Vocational
Developmental
Combinations/Other
Non-response


46.7
36.3


College


Size


Employing Respondent


Large
Medium
Small


23.0
33.0
44.0


(38%


(73%


response)
response)
response)


__














Table 5.2
Age and Employment Demographics of


the Respondents.


Demographic Category Response Percent Cumulative
for for Frequency in Percent in
Comparison Comparison (Sample) Category Category


Age of Respondent


nder
5 -
0 -
5-
0 -
5 -
0 -
5 -
0-
5-
0-


-response


Years


at this


Institution
Under 5
5 9
10 14
15 19
20 24
25 & Over


-response


1 2
3 1
5 2
6 2
0
4
2


Years


Employed


as Educator
Under 5
5 9
10 14
15 19
20 24
25 29
30 34
35 & Over


Years


Employed


in Post


10


second
der 5
- 9
- 14


ary Educ
5
5
8


tion













not be determined.


sex of the respondents


appeared


approximate


the statewide


distribution,


as did


the race of


respondents.


comparison.


Respondent


response


teaching


rates


assignments


from the


were


different


listed


strata


for future


of colleges


size


indicated


that


the questionnaires were


returned


from


smaller


institutions


more


frequently


than


from


the larger


institutions,


with


the medium sized


institutions


represented


proportionately.


Table


5.2 reports


age and employment


demographic


information


for the sample.


The median age


bracket


the respondents


was


45 to 49


years


age.


This


median


together


with


the distribution


reported


indicated a mature


faculty


at the


institutions


selected.


This


was confirmed by


reported number


years


employed


as educators.


While


these


factors


indicate


stability


in the faculty at


present,


they


indicate


potential


future


large


faculty


turnovers


due to retirement


and may


help explain


the low


levels


of dissatisfaction


that


were


reported.


Model Analysis


hypotheses proposed


to verify


the regression models


representing


theoretical


positions


in job attitude were


subjected


test


using


the General


Linear


Model


program


(GLM)


from


the SAS


statistical


results


package


of these


(release


tests are


5.08


discussed


SAS Institute,


in this


1984).


section starting with














Model


The model


in which overall


perception of working

models proposed as an


conditions

explanation


job satisfaction


was one of


was


the leas


of job attitude.


predicted by

t complicated


regression


model


examined


to test


the ability


of perception


of working


conditions


to explain


job attitude was


stated


+ blX1


where


= job attitude


and X1


= perception


of working


conditions.


hypothesis


tested


was that


there


would be


no increase


predictability


of job attitude


knowing the worker


's perception


working


This


conditions,


or that


hypothesis was


= 0,


rejected


at F


alpha


= 115.88,


=.05.


.0001


.2646.


This


result


indicated


that


when


considered alone,


perception


of working


conditions was


capable


of explaining


26.5% of


the variation


in overall


job satisfaction


and was


a statistically


significant


predictor


of faculty


job attitude.


Thus,


the single


continuum view of


sample,


job attitude


the variable perception


is appropriate


of working


for use with


conditions


this


should be


included


in future models.


Model


similar


model


was constructed


to examine


the ability


of the


i b0














= b0


+ b2X


where


= job


attitude


= level


of emotional


involvement.


hypothesis


under


test


was


that


there would be no


increase


predictability


of job attitude by


knowing


the worker'


level


emotional


involvement


in work,


or that


: X2


= 0,


alpha


.05.


This


hypothesis was


.1181.


Rejection o


rejected

f this h


at F


= 43.11,


ypothesis


p m .0001,


indicated


that


and

the level


of emotional


involvement


in work,


taken as


an independent


factor,


a statistically


significant


predictor


of job attitude,


capable of


explaining approximately


11.8%


of the variation


in job attitude.


Therefore,

predictor


emotional


involvement


of job attitude,


althou


also was

gh it has


a statistically


less


significant


explanatory power


than perception


of working conditions.


Model


Model


three combined


the factors


found


significant


in models


and II


a single model


to develop a


planar


single continuum model


of job attitude.


hypothesis


to be


tested was


that


there would be


no difference


combined effect


in ability to


of emotional


predict


job attitude


involvement


knowing the


perception of


working


conditions,


or that


= 0,


alpha


.05,


in the model


+ blX1


+ b2X2


and X2


was














Table


5.3 displays


the results of


regression.


As can be


seen,


the hypothesis was


rejected at


an P


value of


73.52,


.0001,


and R2


.3142


independent


for the overall model.


variables were


statistically


n addition,

significant


both

contributors


the model


(p =


.0001),


indicating that


they


contributed


to the


prediction of


job attitude


independent


each


other.


Thus,


both


factors


should be considered


in the predi


action


of job attitude.


Together


the variables


explain


31.4%


of the variation


in job


attitude.


Table


Results of


the Test


of Model


III,


the Planar


Single Continuum Model.


Source DF Sum of Mean F p R
Squares Square

Model 2 399.20 199.60 73.52 .0001 .3142
Error 321 871.54 2.72
Total 323 1270.74

Source DF Type III SS F p

X 1 249.16 91.77 .0001
X1 1 62.92 23.18 .0001
2














Model


This model


expanded


the previous


planar


single


continuum model


including


relationships


an interaction


term to determine whether


in the previous model


were consistent


or not the


throughout


model.


hypothesis


to be


examined was


that


the interaction


term


made


no difference


in prediction


of job attitude,


or HO:


= 0,


alpha


- .20,


in the regression model


+ blX1


+ b2X
2 .


+ b3XlX
3 1


where


blX1


= perception


of working


conditions


2 level
2


b3X1X


of emotional


= the interaction


involvement


effect


This


hypothesis was


rejected with


an F


of 3


.66,


- .0565,


indicating that


there was


a significant


interaction among the


previously


significant


predictor


variables


in the job attitude


model.


Table


shows


that


inclusion of


the interaction


term


altered


the explanation of


variation.


Table


Results


of the Test


of Model


the Planar


Single Continuum


Interaction Model.


Source DF Sum of Mean F P R
Squares Square

Model 3 409.07 136.35 50.64 .0001 .3219
Error 320 861.68 2.69
Total 323 1270.74














Model


A model


was examined


that


compared


the curvilinear


effects


of the


initial


variables.


First-


second-power


factors were combined


a regression model


as independent


variables


to select


regression


model


that


best


fit the data


obtained


in this


study.


Model


represents


that


final model.


As can be


seen


in Figure


5.1 and


table


this


model


is very


similar


to model


the planar


single


continuum model,


the only change being the


inclusion


second-power


factor


representing the


curvilinearity of


the perception


of working


explanation of


conditions.


variance


This additional


in the data


factor


set over


improved


the planar


the

interaction


model.


It should be noted


that


there was


a difference between


numbers


of low and high


job attitude


responses


in the sample


used


test


the model.


It was assumed


that


this


imbalance was


representative of


the true


nature of


the distribution


in the


population


under


examination.


Others


investigating


the level


satisfaction of


faculty


in the Florida


community colleges have


found


similar

One


distributions


result


(Benoit


of the imbalance


Smith,


1980).


in representation between


low and


high


job attitude responses was


program to emphasize


the tendency of


the importance of


the predominant


regression


group








































m C


U &
C


U J
n h


C u


Z I.


Su


CU


Urn


CC


.... 4. -...-. 4 ...... 4 +--- +--- +--. .. ---+ -.. ++* .... + .... +... .... -... 4 ..... .. ..... 4'.-.


< A
















Table


5.5.


Test


of Model


a Curvilinear


Planar


Interaction Model.


Source DF Sum of Mean F p R
Squares Square

Model 4 449.27 112.32 43.62 .0001 .3535
Error 319 821.47 2.58
Total 323 1270.74

Variable Estimated t PR > t Std. Error of
Regression the Estimate
Coefficient

X 4.980 3.77 .0002 1.322
X -0.813 -1.60 .1100 0.507
-1.303 -3.95 .0001 0.330
X *X2 0.694 2.97 .0033 0.234
1 2


deviations


from the predicted model,


suggesting


the distribution of


unexplained


variance,


and yielding


information


that


could be


useful


in developing


further


models.


resulting pattern


in the plots


job attitude

this model w


scores by perception


would be bimodality


of working


in perception


conditions


working


scores


conditions


high

this


levels


of emotional


bimodality


in contrast


involvement.


Figures


to the patterns


5.2 and


found


at low


illustrate


levels of


emotional


involvement


(Figures


5.4 and 5.5),


indicating that


the cusp


model


was appropriate as


a structure


for the job attitude problem.


Summary


Results of


the analyses


indicated


that


perception of working


































.C




a U 'C


a u a u C


S a 0 4 u


C <

* < U Iu


C 'C C 'C


C U



uC


'C C


C C


C


IC


















I0
+0
In
I
I
a










fM
InI
I
I
I
I
I






< <
< a

to






to
I
IN






C Sc






*6
I.





C CI
IO









< < 3N


19 >
/: c

I
I







SII o





I
i C
I
SC















)OO
I
I
I




















10
I
C I
I
4 I
4.4








I
C ~I
I
13



90
I c
I- -
I
I
CI
I
I0
I
0
I
I
o A.
0
#0
*-






a-
6





I




















4 4


Sm


m m


4 c C


U U


C C


C O


C C


C a a

ma a a


SI


< a U


am


< < <*
C a


m 4


I
C
I
+
1

I

I
I

I


I
I

+
1
1
I
I
I


I
I
I
I
I
I
I
I
I
I
4
I
I
I
I
II
I
I
I
I
I
Il



- --- + ---.-.- 4 -------- 4-.-. --- 4-+- ---- -*+ .----*-.-- + -- ---+--4 -- ---- --4-- ----- .- -.- --+. -- -*-- -- -1I


,QrI

n4
'U

O 0

0 a
0 S3














regression model


developed


from model


IV, the planar


interaction model


for the overall


sample,


was


then


examined for


curvilinearity


by testing model


V for the significance of


perception


of working


conditions


level


of emotional


involvement


each


raised


to the second power.


Perception


of working


conditions was


shown


have a


significant


involvement


did not


curvilinear

. Therefore


component,

* the most


but level

complete


of emotional

regression model


examined


included the variables


of perception


of working


conditions,


level


of emotional


squared,


involvement,


the interaction


perception


term,


perception


of working


of working


conditions


conditions


level


of emotional


involvement,


as significant


factors


explaining


job attitude.


Further


interpretation


of the residuals


from the best-fit


regression model


(model


for the overall


sample,


together with


indications of bimodality


and polarization


in the sample,


indicated


that


the cusp catastrophe model


analogical


structure to model


was


appropriate


the relationships


to use


among


as an


job attitude,


perception of


working


conditions,


level


of emotional


involvement.
















CHAPTER


SUMMARY,


CONCLUSIONS,


RECOMMENDATIONS


Summary


purpose of


this


study was


to develop a


model


representing


relationships among


the faculty's


job attitudes,


perceptions


working


conditions,


levels


of emotional


involvement


that


would


permit


community


college


administrators


to plan and


implement


organizational


reviewing


changes effectively.


the theoretical


This


research


problem was


literature


approached by


to determine


primary j

variables


ob attitude

expressed


variables.

in various


Predicted

foundational


relationships

job attitude


between

theories


that


were


in conflict


were


noted,


sources


of the conflict


were


postulated.


These


conflicting


relationships


then


served as


the basis


for constructing


structural


models


(with


attendant


null


hypotheses)


that


could be


tested with


regression analysis


techniques.


Two primary


independent


variables


associated with


job attitude


were


identified by


reviewing


the research


literature.


These


associated


variables were


the faculty members


perception


of working


conditions


and level


of emotional


involvement


in the job.


Several


structural models were


constructed


that


represented


theoretical


relationships


between


these


variables


and job attitude.


These













Model


II--Job attitude


is a


linear


function


of level


of emotional


involvement


in the job;


Model


III--Job attitude


a planar


function


of perception


working


conditions


and level


of emotional


involvement


in the


job;


Model


IV--Job attitude


a planar


function


of perception


working


conditions


and level


of emotional


involvement


in the


job,


and of


the interaction between


these


terms;


Model


V--The


relationship


found


in model


IV can be


improved by


addition


perception


of exponential


of working


components


conditions


created by


and emotional


raising


involvement


scores)


to create


a curvilinear


model.


null


hypotheses


for the first


three


models


stated


that


variables


included


in each


of the successive models would not


statistically


significant


at the


.05 level


of confidence.


interaction


term in model


IV was


tested


in a


similar manner,


but at


an elevated


alpha


level


of .20,


to avoid


creating


a type


error.


Model


V was


tested by


examining


the null


hypothesis


stating


that


exponential


components


would


not be statistically


significant


predictors


of job attitude


(alpha


.05).


A cross-sectional


sample


of Florida


community


college


faculty


members was questioned


using


long


short


forms


of the Faculty













emotional


involvement


in the job,


attitude.


participating faculty were


full-time


instructional


employees


at six


Florida


public


community


colleges.


A total


of 607 questionnaires


were


distributed,


352 (58


.1%)


were


returned.


Of these,


324 were used


in analyses


of the models


that


had been


proposed.


The balance of


returned questionnaires


were not


used


either


because


they were


incomplete,


or because


the respondents were not


members


of the target


population


of full-time


faculty.


Initial


analysis of


the demographic data obtained


from


questionnaire


responses


indicated


that


the respondents were


comparable


to the Florida population


community


college faculty


the factors


of race and sex.


Analyses


of the data


testing


proposed null

of working co


hypotheses


editions


confirmed


and emotional


that


the variables


involvement


were


perception


statistically


significant


predictors of


job attitude


individually,


in combination,


as an interaction.


On this


basis,


the null


hypotheses


for models


I, II,


III,


IV were


rejected at


preset


levels of


alpha.


The null


hypothesis


for model


V stating


that


the inclusion


of the


second-power

significant


term for perception


predictor


of working


of job attitude when appended


conditions would not be a


to model


(alpha


was


rejected while


hypotheses


for all


other


exponential


terms


failed


to be rejected


at alpha


= .05.


Analysis


of the model




















































u C
C U


U

C


44
" o2
U<

C OW
SQC


OC


C <


- 4......... 4 4 -4 + + -+ 4 -


-*4- -


A *-- 4. 4 .**.. + .*- 4. --- + .-.. 41 -- *


I **
















ability


to interpret


the degree


to which


variance


in the model


due to lack


of fit of the model


compared


to error.


Several


factors


could have


contributed


instrumentation


that


to this


was


outcome.


used


Foremost,


study may not


be sensitive


the perceptions of working


conditions


of low


job attitude


faculty to


the same degree


external


that


was


factors not accounted


for high


job attitude


for in this


study


faculty.


influence


Perhaps


the low


job attitude


respondents more


than


the high


job attitude


respondents.


Conclusions


The

Florida


results of

community


the project


college


faculty


indicate


that


are determined


the job attitudes of


in part by the


influences

emotional


of their p

involvement


erceptions

in their


of working


work,


conditions,


and by the


the level


interaction between


these


factors.


These


findings


are


in accordance with


earlier


theoretical


descriptions


of relationships between


job attitude and


the individual


variables.


In addition,


the theoretical


relationships


are extended by


combination


of the separate


relationships


in a


single


encompassing model

catastrophe model.


structurally


This


analogous


to the elementary


cusp structure appears


to be


cusp


appropriate as


a model


to represent


the relationships


among


these


job attitude


variables.


The current


data


for high


job attitude


responses


resemble


upper


plane of


the analogical


cusp


job attitude model.


For this


was













relationships


between


variables


for low


attitude


responses


insufficiently populated


to conclusively


demonstrate


relationships.


There was,


however,


a different


relationship between


the variables


for the low job attitude


responses


for the overall


sample based


on the observed


bimodality


that


occurred at


high


levels


of emotional


involvement.


Among


the low


job attitude


responses,


increased


emotional


involvement


intensified


low job attitude.


When modeled


the cusp structure,


this apparent


contradiction between


low and high


responses


is logically


sound.


Recommendations


Instrumentation


The use of


the shortened


form of


the Job


Descriptive


Index with a


modified answer


format


was


a qualified success.


instrumentation


successfully measured


the faculty perceptions of working conditions


as designed.


However,


the instrument


apparently measures


perceptions of


responses with high


job attitude more accurately than


for low job attitude


responses,


which


creates


a heteroscedastic


condition


in the data set.


The modified answer


format


This

made


limits

possible


interpretation of


collection


the data.


of information


about


faculty's


level


of emotional


involvement


in their


jobs.


These


variables,


perception


of working


conditions


level


emotional


involvement,


were


found


to be


independent


of each other


was













the original


form in


this project


in which


participation was


voluntary


and anonymous.


shortened


form was


as efficient


the long


(original)


form in


capturing


information


from each


respondent


which


largely


offset


the advantage


gained by


the higher


return


rate.


The difficulties


encountered


assessing perceptions


concerning promotion were considered limited


to the specific academic


situation


and not a limitation


on the applic


ability


of the


instrumentation


in general.


The directions


for completing the


questionnaire should be


clarified


if further


use of


instrument


planned.


Therefore,


it is recommended on


the basis


of these


findings


that


in situations


in which


response can


be controlled,


long


form of


the JDI


instrument be used.


In voluntary


response conditions,


shortened


instrument


increases


the representativeness


of the


responses but


decreases


the sensitivity


of each


reply


In all


situations


it is recommended


that


the modified answer


format


be used


to permit


collection of


the information


on emotional


involvement


which


was shown


to be


a significant


predictor


of job attitude.


Job Attitude Models


There was a


among


complex


job attitude,


emotional


involvement.


relationship


perception


This


for community


of working


relationship was


college


conditions,


curvilinear,


faculty


level


and was













High


attitude


faculty.


Increases


in level


of emotional


involvement


and increases


in perceptions


of working


conditions


were


associated with


expressions


of more positive


job attitude by these


faculty


This


confirms


earlier


single continuum theory predictions.


Within


the high


job attitude


plane of


the cusp model,


there


is a


fairly


strong


relationship between


job attitude and perception


working


conditions at


levels


of emotional


involvement near


the germ


of the


cusp.


This


indicated


that


changes


in working


conditions would


be most


influential


(holding


emotional


involvement


constant)


moderately involved individuals.


This


relationship was


less


pronounced f

unaffected by


or both highly


changes


involved


in working


individuals--who are


conditions--and


relatively


for uninvolved


individuals who evidently


relationship between


do not


job attitude


care

and


about

level


working conditions.


of emotional


involvement


was also greatest


perceptions of working


the region near


conditions,


cusp.


the changes


At elevated


emotional


involvement have a


reduced


effect.


two variables


influence


attitude


in unison,


creating


a synergistic effect.


job attitude


faculty.


These


faculty react


to increased


emotional


involvement


intensifying


low job attitude.


Faculty with


low job attitudes


react


to increasing perception of


working


conditions


in different ways


in different


portions of


the plane.


- S *


I
















require an

the frame o


increase in

f reference.


the level


This


of emotional


creates


involvement


to maintain


an antagonistic situation


where


improving perception of working


conditions


intensifies


emotional


involvement,


lowering


job attitude.


Summary.


It is apparent


that


equivalent


actions with


faculty


members

similar


or groups


of faculty possessing


conditions of work


different


or with different


level


perceptions of

s of emotional


involvement


can produce


opposite


results.


For this


reason,


it is


necessary


to know the


job attitude history of
illli... --- .-- ___ _- ,*w --f- -t ^ **-^-" -- '* -


any individual


or group


before


attempting


to predict


the outcome of


change of


working


conditions or


any program to


change


the level


of emotional


involvement.


Increasing


emotional


involvement


will


intensify


and polarizs job


attitudes


in a


group.


Decreasing


the emotional


involvement


will


increase


and,


the effect


in certain


of changes


circumstances,


in perception of

may precipitate


working

sudden c


conditions


change


from a


low job attitude condition


to a high


condition,


or vice versa.


Without making a


qualitative


judgment


about


the matter,


becomes


apparent


that


workers with a


low to moderate


level


emotional


involvement


in work


will


be most


responsive


to changes


perception


of working


conditions,


since


the correlations are


high and


sudden


or catastrophic


change


is not a factor


that will













improvement


can be obtained


in the prediction


of job attitude by


increasing emotional


involvement.


Highly


involved


individuals will


remain


polarized at


either


high


or low


job attitude


levels


throughout


a wide


range


of working


conditions and will


change between


these


polarized extremes without


Recommendations


warning


for Future Research


The initial


recommendation


for future


research


is that


current


study be


replicated and expanded


to ensure


the accuracy


these


findings.


The current


study


should also be extended by


starting a


longitudinal


study


in a


controlled


environment


to detect


the presence of


catastrophic change


in faculty


job attitude with


changes


in perception


of working


environment


and/or


level


emotional


involvement.


At the


same


time,


alternate methods


of measuring


each


variable


should be

the data.


attempted


to alleviate


suggestion would b


the problem of

e to collect i


heteroscedasticity


information concerning


the variables


of interest


with


an adjective


checklist


designed


measure emotional


intensity


as well


as perception


of working


conditions.


This


information


could


be analyzed


with


discriminant


analysis


to select


terms most


highly associated with


high and


attitude


An additional


recommendation


is that


alternative methods of


-~~~~~ --- N -- --


. f


m ad


J 1


I







































APPENDIX A.


INSTRUMENTATION




















COLLEGE


EDUCATION


Jntve ty


GanewvlNE.


of Florlde
Fla 32611


February 4.


Dear Colleague,


I need


your


assistance.


relationships between


some jobs


a developing


certain


to be satisfTing and


dissatistyinq. A


better


attitude


other


understanding


g a mode

factors


very similar j


of these


I to describe

that may cause

oba to be


relationships may


help improve


faculty


work situations.


The attached questionnaire


has been designed


to assess these


job attitude


factors.


I would


appreciate


your taking the


time to


fill it


out and return


it to


me.


I do not believe


participation


will take


more than ten minutes.


Participation


is voluntary,


responses will be treated


confidentially.


Thank


your time,


John C. Wahl, Graduate


Department


of Educational


Student


Leaderbsip


1986


--I


_ _


UnC