The relationships between selected variables and full-time instructional salaries

MISSING IMAGE

Material Information

Title:
The relationships between selected variables and full-time instructional salaries in community colleges in the State of Florida
Physical Description:
vii, 97 leaves : ill., map ;
Language:
English
Creator:
Diesen, Jeanne K., 1935-
Publication Date:

Subjects

Subjects / Keywords:
Community college teachers -- Appointments, promotions, salaries, etc -- Florida   ( lcsh )
Community colleges -- Administration -- Florida   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1982.
Bibliography:
Includes bibliographical references (leaves 93-96).
Statement of Responsibility:
by Jeanne K. Diesen.
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 - 000334703
notis - ABW4346
oclc - 09483288
Classification:
lcc - LB2334.3.F6 D54
System ID:
AA00002182:00001

Full Text











THE RELATIONSHIPS BETWEEN SELECTED VARIABLE
FULL-TIME INSTRUCTIONAL SALARIES IN COMMUNITY
IN THE STATE OF FLORIDA


JEANNE


A DIS
THE


AND
COLLEGES


DIESEN


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
















ACKNOWLEDGMENTS


This writer wishes to extend her appreciation to

Dr. James Hale, Chairman of her Supervisory Committee,

for his counsel, assistance, and constant support through-


out this study.


Thanks are also extended to other members


of her committee, Dr. Kern Alexander and Dr. Robert Jester.

A debt of gratitude is owed Dr. Charles Polk,

President, and Dr. Paul Thompson, Provost Open College,


of Daytona Beach Community College,


for their support and


guidance throughout this study.


Special


thanks go to


members of the staff and faculty at Daytona Beach Community

College; especially the South Volusia Center, all of whom


cooperated with their


love, support, and understanding.


Special


thanks go to Dr. Charles Dzuiban,


University of


Central Florida,


for his help with the statistical program.


The researcher expresses deepest gratitude to friends
I


and family for support and understanding,


children, Bill and Michele,

Mamie and William Kitt. TI


especially her


and her mother and father,


ie author thanks Janet Jones


for her able assistance in the typing of the manuscript.


















TABLE OF CONTENTS


Page


ACKNOWLEDGMENTS

LIST OF TABLES


* *8* S S S to S S S
S.* .....*.... S S ...... St *...S....... S


ABSTRACT


CHAPTER


INTRODUCTION


The Problem ................
Limitations ................
Delimitations ..............
Justification for the Study
Assumptions ........ ........
Definitions of Terms .......
Procedures .................
Variables of Interest ....
Statistical Procedures ...
Data Sources .............


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


Organization of the Research Study


REVIEW OF RELATED LITERATURE


Growth in Higher Education .......
Community Colleges ...............
Community College Financing Plans


Florida
Formula
Teache?
Faculty
Adjunct
Summary


Community College Funding


*a Sd *.. ***. .......
Salary Studies .......
Salary Studies .......
Faculty ..............


S* S
*5S* S St
* S S S 555
* S S S S S S
* S S S S S S


ANALYSES


OF DATA AND FINDINGS


Discussion









CHAPTER


Page


Conclusions ......
Recommendations ...... ... ...

APPENDIX A SURVEY INSTRUMENT ..............


APPENDIX B MEANS AND STANDARD DEVIATIONS
OF THE VARIABLES ............


REFERENCE LIST ..... . .

BIOGRAPHICAL SKETCH .........................
















LIST


OF TABLES


TABLE


Page


FLORIDA
COUNTIES


COMMUNITY
SERVED .


COLLEGES


AND
..*.. ... ....


IDENTIFICATION


OF VARIABLES


PEARSON


CORRELATION


COEFFI


CLIENTS


PARTIAL


REGRESSION


CORRELATION


ANALYSIS


COEFFICIENTS


OF THE


RELATIONSHIP


OF THE


INDEPENDENT


DEPENDENT
VARIABLES


VARIABLE
.. ** *


REGRESSION


ANALYSIS


RELATIONSHIP


OF THE


INDEPENDENT


DEPENDENT
VARIABLES


VARIABLE
. .* .* .t.. .


REGRESSION


ANALYSIS


OF THE


RELATIONSHIP


OF THE


INDEPENDENT


DEPENDENT
VARIABLES


VARIABLE
. ........


STEPWISE
DEPENDENT
SUMMARY T

ANALYSIS


MULTIPLE
VARIABLE


REGRESSION
MAS


ABLE


OF VARIANCE


(ANOVA)


BETWEEN


SUBPOPULATIONS
















Abstract


of Dissertation


Presented


he University
Requirements


THE RELATIONSHIPS BETWEEN


Florida


in Partial


the Graduate Council


Fulfillment


Degree of Doctor of


SELECTED


Philosophy


VARIABLES


FULL-TIME


INSTRUCTIONAL SALARIES


IN COMMUNITY


COLLEGES


IN THE STATE OF FLORIDA


By


Jeanne


August


Diesen


1982


Chairman:


James A.


Hale


Major


Department


Educational


Administration


Supervision


This


study


sought


to determine


certain


socio-


economic


characteristics were


related


instructional


salaries


community


colleges


state of


Florida.


Three


levels of


salary


were


used


the dependent


variables.


related


determine


literature


relevant


research


independent


were


variables.


reviewed


five


variables


selected were enrollment,


total


population










The sample of


institutions included the twenty-eight


community college districts in the state of Florida.


The percentage of


classes taught by adjunct faculty


contributed little to variance in faculty salaries.


Strongest


variables were Full-Time Equivalent student


enrollment, and Effective Buying Income per capital; the


latter served

districts. T


as a wealth measure of


he fact


the community college


that institutions used adjunct faculty


to teach approximately


25 percent of their classes led


to the hypothesis that full-time faculty salaries would


be systematically related to use of adjunct faculty.


the relationships were very weak.


However,


It is suggested that


other research designs and other variables be utilized to

explore these relationships.

The five independent variables explained only 30


percent of


the variance


in average salaries of persons with


doctoral


degrees but 51 percent and 41 percent of the


variances in the two master's degree salary


levels.


suggested that


the cost-based,


lull state funding method


of community college support


in Florida may have reduced


the expected salary ranges among institutions.


Therefore,


is recommended that total compensation (salary plus


benefits) of instructional faculty be used in other studies















CHAPTER 1
INTRODUCTION


Educational institutions in America experienced


unusual


enrollment growth in the period between World


II and the early


197O's.


Every sector of education


grew;


elementary, secondary, higher education, public


and private.


Few other institutions have experienced


the dynamic growth exhibited by American higher education


during those years.


To support this unprecedented growth,


an accompanying seven-fold increase was made in higher


education budgets.


State government appropriations for


public higher education increased 375 percent during the


ten-year period of 1960-1970 (Quindry and Masten,


1976).


A large part of this phenomenal growth in public

higher education can be attributed to a new institution,


the community college.


This different institution,


which had evolved from the junior college, got its

direction and emphasis from President Harry Truman's

Commission on Higher Education.


Whatever form the community college takes,


purpose i


educational service to the entire


community, and this purpose requires of


it a










attempt
needs of


to meet the total
its community.


post-high school
(President's Commission


on Higher Education,


cited in Fields,


p.63).


Public community college enrollments increased 930


percent between 1960 and 1979.


After


1975 approximately


50 percent of


all first-time college students enrolled


in a community college,


including increased numbers of


older part-time students who entered college to get the


education they had missed.


(Breneman and Nelson,


1981)


1980 postsecondary education began facing grave


financial


difficulties.


Inflation was double digit.


Federal funding was decreasing,


thus shifting more of


the burden to the states.


Many states'


revenues were


not keeping pace with inflation.


More agencies began


competing for available dollars.


While it


is generally


agreed that all


levels of


government have a responsibility


to support higher education,


that support is dependent


upon expenditure priorities and available revenue at


any given time.


(Quindry and Masten,


1976)


Since public community colleges employed more than

87,000 full-time and 115,400 part-time faculty members


during 1979 (Breneman and Nelson,


1981),


it was apparent


that adjustments would be necessary if funding continued

to diminish.
m ^ -- 1 1 j -. __ i- 1._ 1 ii 4 iA *r







3

literature which addresses funding changes for community


colleges.


Further, Breneman and Nelson make the point


that of the ten states with largest enrollments in two-

year institutions, only Washington and Florida do not


depend to some degree on local financing.


that


This means


in most states, community colleges face the same


basic monetary issues


as elementary and secondary schools,


a condition which makes their funding subject to local

property tax limitation efforts.


Florida's community college system


is cost-based.


The funding formula


based on actual costs


instruc-


tion per Full Time Equivalent (FTE) student.


statutes further place a


Florida


limit on the number of FTH students


to be funded for each institution.


This


limit is based on


enrollment projections,


which create a number of problems


relative to faculty employment and assignments.


enrollments exceed projections, an adjustment has to be

made either to teach larger classes or to hire part-time


instructors at lower salaries.


This use of part-time


instructors can provide an important management option.


Rather than cutting or increasing permanent staff,


employer can simply reduce the number of part-time


personnel


to accommodate fluctuating enrollments.


If~1 finnn iinr oorolhnfrintt,..









be made?


Since salaries comprise the major part of any


institution's budget it is important to know what factors

determine salaries paid by the institution.


The Problem


The purpose of this study was to determine


relationships exist,


what


if any, between selected socio-


economic variables and salaries paid to full-time

community college instructional personnel in the twenty-

eight community college districts in the state of Florida.


Further,


the study evaluated the importance of other


selected factors as they may relate to instructional

salaries, specifically the effect that utilization of

part-time instructors may have on full-time professional

salaries.

The specific questions of this study were:

1. ls there a relationship between the widespread

use of adjunct faculty and salaries paid to

full-time instructional personnel?


Is there a relationship between the size of the

population of the community college district and

salaries paid to full-time instructional personnel?

Is there a relationship between the size of the
-^lV^4 no+ 4* .,+ 4- r1 ^\V nei/ lnn4/ i- "- an n\ Ai A- 4 1 *4*nlInn t^









Effective Buying Income per capital of the

community college district?


Is there a relationship between the presence

of a faculty bargaining unit and full-time

instructional salaries of the community college

district?

Which independent variables account for the

largest percentage of variance in the mean daily

salary for individuals holding doctoral degrees?

Which independent variables account for the


largest percentage of


variance in the mean daily


salary for individuals with master


degree plus


30 credit hours?

Which independent variables account for the

largest percentage of variance in the mean daily

salary for individuals at the master's degree

level?


Limitations


This study is confined to the twenty-eight community


college districts in the state of Florida.


The method of


funding for these institutions is a cost-based formula.


Therefore,


this study cannot be generalized beyond these


ttwun f vt-p1 i arhlr1 i s t ri ts -









Delimitations


In salary studies reviewed fringe benefits were

sometimes included, especially if unionization was the


basis for study.


Estimates of fringe benefits were


not included in this study because part-time instructors

do not receive the same fringe benefits as full-time

instructional personnel and it appeared that salaries


were a more accurate basis of comparison.


Quality is


sometimes an important factor when comparing public and

private institutions and two-year and four-year insti-


tutions.


Because the community colleges in Florida have


a common funding formula, and common governance, no

attempt was made to attach a degree of quality to the

institutions in this study.

Many studies have dealt with differences in salaries


paid to males and females.


What may appear as sexual


discrimination must be researched to determine if the

discrimination issue is valid and not related to male


longevity in the position and/or higher


levels of education.


It was beyond the scope of this study to deal with this

issue.


Collective bargaining was considered


variable.


as a uummy


This may allow conclusions as to whether or










This study


looked at data of


twenty-eight community


college districts.


No attempt was made to compare


salary data within the institutions.


Therefore,


this


study should not be compared with studies dealing with

factors inside a single institution.


Justification for the Study


Many economic changes are taking place in institu-


tions of higher


learning in general and in institutions


that offer the first two years of postsecondary instruc-


tion in particular.


This study should lead to a better


understanding of the factors which have the greatest


impact on instructional personnel costs.


This study may


also add to our understanding of the practical use of


large numbers of


adjunct


faculty and the implications


thereof.


Finally,


the study may suggest substance and


methodology


for a broad-based research effort that utilizes


a national sample of


community colleges.


Assumptions


Population figures were taken from the Florida


County Comparisons


1981


The source for these


data


was


the 1980 Census of Population,


i n .- 1 .. __ -, A t ------------ -_ A L L -


r\ r.









made to contact individual institutions for

verification of these data.


Effective Buying Income by County


was


secured from


Sales and Marketing Management Magazine.


These


data were used as the basis to aggregate from county


to districts.


of data


It was necessary to assume reliability


as given.


Definition of Terms


Adjunct Faculty.


Part-time instructional personnel


as opposed to contracted full-time instructional

personnel.


Doctoral Level Personnel.


Instructional personnel with


an earned PhD


, EdD, or DBA.


Dummy Variable.


Method for including qualitative


variable in regression model.


Effective Buying Ilcome.


Personal income less personal


tax and non-tax payments, i.e. disposable personal income.

Full-Time Equivalent Enrollment. The total annual


student semester hours registration divided by thirty.


Master's Level Personnel.


Instructional personnel


with an earned master's degree.


Master's Level plus 30 Personnel.


Instructional personnel


.~.44-k ~ ~ .~.- --a a- I.. rr~nr~c' 4.nt.rr*. A. ah










Unionization.


Whether or not institution has a


collective bargaining agreement covering instructional

personnel with a unioa or unions.


Procedures


Variables of Interest.

In order to compile information regarding adjunct

faculty employed by the twenty-eight community college


districts in Florida,


a data collection instrument was


devised and sent to each community college.


classes taught b


was one variable of


The percentage


adjunct faculty in each institution


interest to study.


From a review of literature and research on instruc-


tional salaries the following variables were chosen:


FTE enrollment of institutions studied,


community college districts,


total


population of the


per capital Effective Buying


Income of the community college district, and unionization.


Along with percentage of classes taught by adjunct


faculty,


these variables were chosen for study because they seem

to be related to an institution's ability and perhaps

its propensity to spend for instructional salaries.

The dependent variables chosen were mean daily rate

of salary for instructors with doctoral degrees, mean










mean daily rate of salary for instructors with master's


degrees.


These variables were chosen because they


represent typical salary factors utilized by community


colleges.


The five independent variables were used in


regression analysis with each training level for all

twenty-eight community colleges in Florida.


Statistical Procedures.


This study


sought to determine factors systematically


associated with various salary levels in community colleges


in Florida.


After variables related to instructional


salaries were identified from a review of research and


related literature,


these variables were subjected to


multiple regression anal


ysis.


The three dependent variables


were identified for use in the multiple regression equation

as follows:

Y1 mean daily rate of salary for full-time instruc-

tional personnel with doctorate degree.

Y2 mean daily rate of salary for full-time instruc-

tional personnel with a Master's degree plus 30

additional graduate hours.

Y3 mean daily rate of salary for full-time instruc-

tional personnel with a Master's degree.


I, f..-J


41,0ro -inrlanantil~ont lr^ 0 vnlf ltnrU vnarin T hlpg iit









total population of the community college

district.

per capital Effective Buying Income of the

community college district.


percentage of


classes taught by part-time faculty.


unionization (dummy variable)


- No


Multiple regression is a general statistical


tech-


unique through which relationships between a dependent or


criterion variable and a s

variables can be analyzed.


;et of predictor or independent

In this study multiple


regression was used


a descriptive tool.


The most


important uses of the technique


are to


as a descriptive tool


ind the best linear prediction equation and


evaluate


prediction accuracy, and to control for


other confounding factors in order to evaluate the


contribution of


a specific variable or set of variables


(Nie et al.,


1975, p.


321).


Multiple regression was the method of analysis used

in all of the'salary studies reviewed; however, some of

the reviewed studies failed to look at partial correlation


coefficients.


Partial correlation provides the researcher


with a single measure of association describing the

relationship between two variables while adjusting for







12

Partial correlation analyses were an important procedure

for inclusion in this study.


Data Sources:

Data relating to full-time instructional salaries

and FTE generation in community colleges in Florida were

obtained from the Division of Community Colleges, Depart-


ment of Education, State of Florida.


taken from the Florida County Comparisons,


Population data were


1981.


source for this data was the 1980 Census of Population,


Department of Commerce, Bureau of the Census.


Effective Buying Income data were generated by using the


July,


1981, Sales and Marketing Management Magazine


information


the data base for aggregation.


pertaining to adjunct faculty was secured by


Information


use


of a survey


instrument sent to each community college in Florida.

Organization of the Research Study
III I n in l. i il a li l+ ii. II II II ii ii


The research study contains the following four

chapters:


Chapter


An introduction to the study and an overview


which includes the statement of the problem,

delimitations and limitations, justification

for the study, assumptions, definitions, and









Chapter


III: A presentation and analyses of the study


results.


Chapter


Summary,


conclusions, and recommendations.















CHAPTER


REVIEW OF RELATED


LITERATURE


To provide


salaries


a basis


community


for examination


colleges


instructional


the state of Florida,


an understanding of


community


college


funding


is essential.


This


review of


literature


briefly


addresses higher education


growth


in general


and an


in-depth


discussion of


community


college


funding.


Formulas employed


the United States are


examined along with a more


thorough


review of


the Florida


funding


formula


community


colleges.


Teacher


faculty


salary


studies were


reviewed


purposes of


deterrm:i.ing and selec :ing of


variable es


study.


C'jr~


studies dealing with


effects


union-


ization


anE


included


only for


purposes of


verification


relevant


variables


salary


studies.


important


section


this


literature


review deals


with


institutional


uses of


adjunct


faculty


the effect


their use may


have on


full-time


faculty


salaries.


This


review of


literature and related


research


provided


for the


variables


selected


for use


study.








15

found that there were consistent increases in the level of


financing until 1968 but a decline thereafter.


reasons


A number


existed for the strong support afforded higher


education during those early years.


The nation suffered


an intellectual humiliation following the Russian Sputnik


achievement.


A rush began to support educational efforts


to keep this country at least even with Russian technology.


Veterans of World War

of strong support. T


II were another factor in this era


he government guaranteed veterans


an education, and with thousands taking advantage of the

educational opportunities enrollments burgeoned and


accompanying financial support increased.


changes were taking place,


While these


the national economy was grow-


ing and federal funds were plentiful in the 1960-1968

period but dissipated rapidly, according to Quindry and


Masten (1976).


One reason for higher education growth


during the early part of the 1950's was the move to make

education available to citizens who were heretofore denied


for economic and/or


logistic reasons.


This move brought


about expansion most notable in the public sector.


the middle 1960


public higher education enrollment


was exploding


across


the landscape (Jenny,


1975,


29).


about this time the community college began growing at







16
represented an increase of one million students from the


previous peak year of 1949 (Fields,


1962,


p.4).


In 1960,


315 public two-year colleges existed with


an enrollment of 392,000 students or


higher education enrollment.


11 percent of total


1979, 926 public two-year


colleges enrolled 4,057,000 students, 35 percent of the

total higher education enrollment and more than the cornm-


bined higher education enrollment for


1959 (Breneman and


Nelson,


1981).


The increase in enrollment in public community


colleges from 1960 to 1979 was 930 percent.


Since 1975


about one half of all first-time enrollees attend community

colleges.
The same reasons that higher education in general

expanded during this time were responsible for the community


college growth.


Additionally,


the community colleges


pioneered the "open door" policy which offered greater


accessibility (Breneman and Nelson,


1981).


The timing was


ideal since the national emphasis was shifting to one of


providing equal opportunity for everyone.


Community


colleges could. respond quickly to community needs.

adults, mostly part-time students, were also find


Older


ng their


places in community colleges, further expanding enrollments.


The community college,


because of


its community base for


enrollment, grew in both traditional and non-traditional










Community Colleges


Junior


century


colleges were


according to Medsker


a product


(1960).


early


As more


twentieth


young


people


attended


and graduated


from high


school


early


1900's,


increased


interest


higher


education


created


a strain


the universities.


Pressure


from


universities caused


public


schools


to begin


offering


"post


graduate"'


courses


to meet


growing


need.


This


push


to extend high


school


education


beyond


twelfth


grade,


while stopping short


four-year


college,


was

year


important


college


strand


(Breneman


in the development

and Nelson, 1981).


Two-year


insti-


tutions were encouraged by


both high


schools and


univer-


sites.


1920 the


idea of


junior


college was


deeply


rooted

types


in educational


junior


programs and


establishment


colleges was taking place


throughout


various

the


nation.


1930,


there were


these


year


insti-


tutions with


1940


a combined enrollment


number


enrollment


over


institutions had grown


149,854.


numbers continued


55.000.


to 456 with


to grow.


Four

over


hundred

242,000


eighty-three


1950


two-year


(Biennial


institutions


Survey


enrolled


Education in the


United States,


1958,


p.18)







18
Commission on Higher Education called for a new emphasis


in postsecondary education.


This commission


the tone


for the "new" community college.


While these two-year


institutions had been predominately private in governance,


this trend was reversed by


1952 when public junior colleges


outnumbered private ones (Gleazer,


1959).


'The gradual


emergence of


the concept


of a community-serving, institution


marks the development of


the community college movement


out of


junior college efforts" (Fields,


1962, p. 2).


The returning military with unprecedented veterans'


benefits and a new social


emphasis on equal opportunity


for education caused great enrollment increases in insti-


tutions of higher education during the 1950's.


Acceptance


of the lifelong-learning concept brought adults into these


institutions,


while great increases in the number of high


school graduate


was another major growth factor.


Accord-


ing to Fields (1962)


, four-year universities were more


research oriented and tended to be uninvolved with the


local community.


was


The pervading feeling before World War


that it was broadening for youngsters to get away


from home for college.


However,


the expense greatly


limited the number of persons able to attend distant


universities.


No financial assistance was available in


the form of government grants and loans.


In order to







19

at such a tremendous rate that state governance boards


became alarmed.


Concern was generated by the rapid growth


because it appeared that state legislators were promising


a junior college in each of their districts (Conant,


1964).


These concerns were based on the prospects that uncontrolled


community college growth might have on programs ant


at four-year institutions.


funding


No one was quite certain of


the mission of the community college;


therefore,


it was


to be feared if not contained.

Much discussion began about the community college


mission without much consensus.


however


It was generally agreed,


that these institutions should be integral parts


of the particular communities served and that they should


serve all


the needs possible.


The Southern Regional


Education Board Commission on Goals for Higher Education

in the South recommended the development by each state of

a strong system of community colleges and further defined


their scope (Report for Florida Community Colleges,


1980, Part


II,


1979-


This Report stated the following:


TheSe non-residential institutions, generally
located in urban areas, can serve a variety of
functions for which four-year institutions are not


required.


Among these are freshman and sophomore


college courses,


vocational and technical programs,


guidance and counseling services, specific programs
to meet community needs, and adult education.

Tho rnmmiin i tu ornl 1o ia c nnnnmi n fnr hnth









system.
however

1.


Whatever the basis of the organization,
, three things are essential:

They must be integral parts of the state
system of higher education and fully


coordinated with the other part


of the


system.

They must resist pressure to expand into
four-year institutions, concentrating
rather on achieving excellence in their
two-year programs.

Their distinctive function must be


recognized and respected.


They


are


neither mere extensions of the high
school nor decapitated versions of
the four-year college.

By the late 1970's educators began questioning the


future mission of the community college movement.

in the number of college-age students were certain.


prospect of less federal funding in the 1980'


real.


Declines

The


was very


Not only were funds cut for the institution programs


but also funds were cut


for student grants and loan


well.

With limited funding a reality, other institutions

were anxious to tap the student population heretofore the


exclusive clientele of


community colleges.


After all,


community colleges had not proven they could do the job


better;


they simply had expressed a willingness to do the


job that society wanted done (Cross,


1981).


In a:i earlier


era community colleges were not challenged in their effort







21

Competition for both students and funds will be


intense in the 1980's.

students and funds, th


With increased competition for


lose states with planned systems of


two-year institutions could better respond to these changes.

One state with a master plan for community colleges was

Florida.

The first public two-year institution in Florida was


Palm Beach Junior College.

the only public two-year col


From 1933 until 1947 it remained

lege. In that year, St. Peters-


burg Junior College changed from private to public


and in


1948 Pensacola Junior College was established, and Chipola


Junior College changed from private to public.


In 1955 the


legislature established the Community College Council.


Council's report in 1957 recommended a plan for establishing


a public community college system in Florida.


institutions,


These


which would provide postsecondary education,


were mandated to be within commuting distance of 99 percent


of the population.


The report was approved by the State


Board of Education and the legislature authorized creation

of the Division of Community Colleges under the State


Department of Education.


Funds were appropriated for six


new community colleges:


Gulf Coast Community College,


Central


Florida Community College, Daytona Beach Community College,

Manntip .Thninr Cfnllea. Nnrth Finrida .Tuninr Cnllpev. nnd







22

Community colleges in Florida are a part of the State

system of public instruction:


State community coll


eges


shall consist of all public


educational


district
and rule
maintain
undergra
be autho
operate
education
schools.
in arts
awards,
of the c
limited
to the a
and dipl
leading
basic el
general
of the c
community


inst
rds
the
pri
e in
d by


a depart
n school
These
and asso
and dipl
community
to, cour
bove-men
omas; vo
directly


itu
of
St
mar
str
th
men
an
ins
cia
oma
co
ses
tio
cat
to
. a


ementary


or liberal


t
t


ions operated by
trustees under st


ate Board of
y responsibil
auction. A co
e State Board
t designated
d authorized
titutions may
te in science
s. The total
alleges may in
as component
ned degrees,
ional and tec
employment;
nd secondary
arts courses


Educ
ity
mnun
of


community
atutory au
action and
for lower-
ity college
Education


as an area
to operate
grant the
degrees,
program o


clude,
s of p
certif
hnical
compen
educat
sought


college
thority
shall
level
e may
to


-vocational
adult high
associate
certificates,
offerings


but not
programs
icates,
offering
story,
ion; oth
by the


be
leading
awards,
gs
adult
er
citizens


community for personal development; and other
y services. (Section 240.301 Statutes of the


State of Florida)

The Report for Florida Community Colleges 1979-1980,


Part II,


gives


the following descriptions and relationships


of the various organizational entities to which each

community college is related:


The State Board of Education is the


designate
work with
operate i
Statutes,
tion shal
of state,
treasurer
commission
that the
board and
be its se


d t
in
n F


o p
whi
lor
ate
ons
tor
omm
of


governor


hat
of
gen
ione
ucat
shall


and es
commnun
section
the St
the go
eral,


tablish
ity coll
229.012
ate Boar
vernor,
comptrol


r of agricultu


ion. It fur
1 be the cha


the commissioner of educ
retarv and executive off


th
ir
at
ia


state agency
the frame-
eges may
, Florida
d of Educa-
secretary
ler,
re, and the
er states
man of the
ion shall
er. The


I


)

r

l
)
r







23

prescribe minimum standards for community
colleges.
approve planning and construction of


facility
author
fees to
adopt r
approval
adopt r
relatin


of absen


ies.


the matriculation or tuition


charged
relati
common
for co
certif
of all


to students
ng to preparation and
ity college budgets.
immunity college teachers
ication, tenure, leaves
types, including


sabbaticals,


etc.


The State Comm
is comprised o
in accordance
Statutes. The
nine (9) encum
trustees, the


College
lay memb
approved
Education


Student
er. Al
by fou
n, and


unity Coll
f eleven (
with Secti
Coordinat
bent membe
president


Government
1 are appoin
r members of
confirmed by


Coordinating Board


) members


240.
Boa
of 1
the
Asso
ted
the
the


307, Fl
rd cons
ocal bo
Florida
ciation
by the
State
Senate


point
orida
ists
yards
Juni
, and
Gover
Board
in


ed


of
of
or
one
nor,
of


regular session. Each member is charged with
the responsibility for serving the entire state
in terms which vary from one to four years.


With the help of its standing committees--


executive,


finance
and pol
procedu
ance of
mendati
Council


S
he


account
capit
s--the
and po
state


to
f Pr


of trustees,
also serve on


tabili
al out
Coord
licies
commu
State
ents,
commun
ending


ty, programs and quality,
lay, rules, governance,
inating Board develops
pursuant to the govern-
nity colleges for recom-
Board of Education. The
members of local boards
ity college employees
committees to advise and


recommend actions to the Coordinating Board.

Specific responsibilities include the provision


adherence


mmendations of stan
rmation, inter-inst
the establishment o
daries. The Coordi
appointment of the
pr -whn i t h fir


rules and procedures,
ds, dissemination of
tional cooperation,
riteria for district
ing Board concurs in


.-'I


admini
nf thn


strative
IHi vi fifl nn


of programs


E l I I









Department of Education.


Section 229.75, Florida


Statutes,
tion shal
visory ag
Board of
the funct
profession
carrying
authorize
Education


pro
1 ac
ency
Educ
ions
nal
out
d by
as


V
t


ides that th


a


unaer tn
ation. T
of the D
leadershi
policies,
a law or
necessary


dm
e
he
ep
P


Department of


strative
section o
w specif
ment as
guidance
edures,


and
f th
ical
prov
e, a
and


S


Educa-
uper-
State
details
ing
in
ties


the State Board of
attain the purpose


and objectives


of the School Code.


State Commissioner of Education.


The Commissioner


of E
Educ
sibi
laws
spec
resp
shal
of t
comm


education


at
li


ion,
ties
f the
y tha
sibil
appro
Stat


h
r


, as
as b
elat
Stat
in
ties
e bu


Secre
oth ge
ing to
e, as
additi
for t
dgets


Board of


tary of


neral
commu
well a
on to
he corn
and ac
Educat


the
nds
ity
Sta
hese
unit
as
on r


State Boa
specific re
colleges.
te Board R
general
y college
executive
elating to


rd of
spon-
The
ules,


he
officer


unity college recommendations.


District Board of Trustee
created under statutes to
community college. Speci
of boards of trustees are
240.313. Florida Statutes


s is the corporate body
govern and operate the
fic duties and powers
enumerated in Section


The Division of Community Colleges


the Department of
suant to a recomm
lege Council. It
is as other divis
action 229.76, Flo
ponsible for the
lege programs and


Educ
endat
oper
ions
rida
coord
the


action i


ion
ates
of t
Stat
inat
impl


of
on
he
ute
ion
eme


n Ju
the
the
Depa
s),
of
ntat


was organized
ly 1957,
Community
same legal
rtment
and is
community
ion of


recommendations concerning the development of
community colleges in Florida.


e authority
vision of
in the fo
actions 20.
Chapter 2


y and respon
Community Co
allowing Flor
15, 228,041,
29, and 240.


sibility of
lieges are
ida Statute
applicable
301-240.379


the
contain-
s:
sections


Tn nenrdnnrnc with Rnootin 15Fi Florian


:
n






25

TABLE 1
FLORIDA COMMUNITY COLLEGES AND COUNTIES SERVED


Brevard Community College
1519 Clearlake Road
Cocoa, Florida 32922
(Brevard County)


Indian River Community College
3209 Virginia Avenue
Fort Pierce, Florida 33450


(St.Lucie,


Indian River,Martin


Okeechobee Counties)


Broward Community College
225 E. Las Olas Boulevard


Fort Lauderdale,
(Broward County)


Fl.33301


Lake City Community College
Lake City, Florida 32055
(Columbia, Baker, Dixie, Gilchris
Union, Counties)


Central Florida Community Coll.
P. O. Box 1388
Ocala, Florida 32670
(Marion, Citrus, Levy Counties)

Chipola Junior College
Marianna, Florida 32446


(Jackson,
Liberty,


Calhoun, Holmes,
Washington Counties)


Lake-Sumter Community College
Leesburg, Florida 32748
(Lake, Sumter Counties)

Manatee Junior College


5840


26th Street West


Bradenton, Florida 33507
(Manatee, Sarasota Counties)


Daytona Beach Community College
P. O. Box 1111
Daytona Beach, Florida 32015
(Volusia, Flagler Counties)

Edison Community College
Fort Meyers, Florida 33907
(Lee, Charlotte, Collier
Counties)


Florida Junior College at
Jacksonville


Miami-Dade Community College


1011 S.W.


Miami,


104th Street


Florida 33176


(Dade County)

North Florida Junior College
1000 Turner Davis Drive
Madison, Florida 32340
(Madison, Hamilton, Jefferson,


Lafayette, Suwannee,
Counties)


Taylor


District Office
210 N. Main Street


Okaloosa-Walton Junior College
Niceville, Florida 32578


Jacksonville,


Florida 32202


(Okaloosa,


Walton Counties)


(Duval, Nassau Counties)


Florida


Key West,


Keys


Community College


Florida 33040


(Monroe County)


Gulf Coast Community College
5230 W. Highway 98


Panama City. Fl.


32401


Palm Beach Junior College
4200 Congress Avenue
Lake Worth, Florida 33461
(Palm Beach County)


Pasco-Hernando Community Coll.
2401 State Highway 41, North
Dade City. Florida 33525








26

TABLE 1 (continued)
FLORIDA COMMUNITY COLLEGES AND COUNTIES SERVED

Polk Community College
999 Avenue HI. N. E.
Winter Haven, Florida 33880
(Polk County)

St. Johns River Community College
5001 St. Johns Avenue
Palatka, Florida 32077
(Putnam, Clay, St. Johns Counties)

St. Petersburg Junior College
P. O. Box 13489
St. Petersburg, Florida 33733
(Pinellas County)

Santa Fe Community College
P. O. 1530
Gainesville, Florida 32602
(Alachua, Bradford Counties)

Seminole Community College
Sanford, Florida 32771
(Seminole County)

South Florida Junior College
600 W. College Drive
Avon Park, Florida 33825
(Highlands, Hlardee Counties)

Tallahassee Community College
444 Appleyard Drive
Tallahassee, Florida 32304
(Leon, Gadsden, Wakulla Counties)

Valencia Community College
District Office 1 W. Church St.
P. O. Box 3028
Orlando, Florida 32802
(Orange, Osceola Counties)

NOTE: DeSoto, Franklin, Glades,
and Hendry Counties are not part
of any college district.
















FLORIDA


* I itl


CO


IMUNITY


ILsiD


COLLEGES


PEN:;AOLA JUNIOR COLLEGE 15. PASCO HIERNANDOO
PCfIo.odli Ft i.n COMMUNITY COLLEGE
OKAt OOSAWALTON JUNIOR COLt.EOGE deo CIty. f )mun
New ,i' l 'o wi IS6 ST. PE TARSIUJG Jlr, !O COLLEGE
CiJLF COAST COMMUNITY COLLEGE St Petoshurg, Flonsda
ParnaCn v ( t oida 17 HILtSROAOUGH COMMUNITY COt LEGE
CHIPOL.A JUNIORf COLt EGE Tanmp, Ilorid,
MIianan,& fh sIda 1. POLK COMMUNITY COLLEGE
TALLAHASSEE COMMUNITY COLLEGE WVA e* Ha.Ien,f oda
Thllthaise. fIorida 19. VALENCIA COMMUNITY Y COLLEGE
NORTH fLORtIUA JUNIOR COLLEGE Orlando, Florhda
,adlison F!ornda 20. BREVAfD.'OrM"' .ITY COL'.EGE
LAKF CITY COMMUNITY COLLEGE Cocoa. Flioria
L.ake ty., FIrtuda 21. INDIAN RIVER COMMUNITY COLLEGE
FLGnRIDA JUNIOR COLLEGE Fot Pi"ce, Fonrid
AT JACKSONVL1 t 22 SOUTH f LORIDA JUNIOR COL EGE
Ja.ck ,motvilt Plu'loea Avon Perfk Floidal
SANTA FE COMMUNITY COLLIE E 23. MANATEE JUNIOR COLLEGE
jam'esviIlr Fltoe.ta Brfdmanon, I lornila
ST. JOHNS RIVER COMMUNITY COttLLEGE 24 EDISON COMMUNITY COLLEGE
Paittk un(d Fust Myers. Flor da
',NTHAL FLOIUOA COMMUNITY COLLEGE 25. PALM BEACH JUNIOR COLLEGE
(ai3 Flh.uia tlke Woith Flonede
DAYTONA BEACH COMMUNITY COLLEGE 26. BROWARD COMMUNITY COLLEGE
rTyton., coach, Floia'd Fort LaudeI.daIte. FI.rd
SEMINOLE COMMUNITY COLLEGE 27. MIAMIDADO COMMUNITY COLLEGE


G .(lri I


0L&


olt is1


*A*w (IntlA


Uso0


Sni ,TER CITY COLLE I
! AKL SUMTER COMMUNITY COLLEGE


1- or it


MlntR, FlorKda
Pt ORIDA KIYS COMMUNITY COLLEGE
Ksy West. Flo# ia


ftst


j







28

For more explicit details the reader is referred to

the specific statutes of the state of Florida.


Community College Financing Plans


An examination of community college finance plans may

lead to an understanding of the problems resulting from


enrollment changes and associated funding cutbacks.


many finance plans exist as do community college systems.

For funding purposes many states treat community colleges


as secondary institutions.


Other states incorporate


community college funding into the established method for


funding four-year institutions of higher education.


some states community colleges are funded according to a

specific plan designated only for those two-year institu-


tions.


While it is beyond the scope of this study to deal


with the mission and philosophy of the various community


college systems,


the reader is reminded of the link between


funding and mission/philosophy since financing i

often related to the mission of the institutions


by the fund granting bodies.


very

as seen


In times of affluence many


needed community based non-credit programs may be accept-


able.


As revenue patterns change


so may the mission of


the institution with respect to needs of the community


nnnmrn fiinoinr mathnrlo arc mns'a


ne








29

that no "best" plan exists for financing of community


colleges.


They also observed


that disputes over financing policies and procedures
often disguise fundamental disagreements over the
purpose, mission and priorities of community colleges;


much of the criticism of


therefore,


current financing formulas,


misdirected because the problems are


substantive, not technical.


(p.161)


Garms (1978) identified three theoretical models for


funding community colleges.


models,


These are market economy


planned economy models, and mixed models.


examination of each model follows.


Market economy model.


All institutions under this


model would be private without any form of governmental


aid.


Free market economists favor this type of private


system.


The system would tend to be self-regulated and not


duplicate


services.


Wealthy students could purchase as much


education as wanted but less fortunate individuals could not


afford this sort of self supporting system.


Two variations


to this model would be a private system with government grants

to students and a private system with government grants to


institutions.


The private system with student grants would provide

more opportunity for needy students to attend these insti-


tutions. This strategy could be accomplished with a system

of educational vouchers to be used at the institution of








certain number of years of


education to be acquired all at


one time or throughout one's lifetime.

The private system with government grants to insti-

tutions would involve direct subsidization by the government--

this would surely mean increased government controls.


Planned economy model.


This system is patterned after


the economic system in socialist countries.


The planning


would be centralized with no profit motive as a regulator,


and with an obvious lack of diversity.


The two versions of


this model are state financing with centralized control and


state financing with some decentralization of control.


state financing with centralized control system would elim-


inate any


local autonomy either in the spending of the budget,


selection of students, or programmatic decisions.


This same system with some decentralization of


would give local autonomy to the districts.


control


This approach


allows individual colleges to respond to the needs of the

local community by providing local control over the

expenditures.


Mixed models.


The two previously discussed models


represent extremes.


The mixed models are less extreme and


thus more palatable.


The distinguishing feature is


a joint


financial/control effort between the state and local


governments.


The four types of mixed models differ by the







31

the community college to set its own budget and tends to

be disequalizing since wealthier communities can raise

greater amounts of revenue.

The flat grant system gives each college a set amount


of dollars per full-time equivalent student.


The rest of


the budget must be raised by the college through taxation


and/or tuition.


This approach may also be disequalizing


for the same reason noted for the percentage of budgets

method.

The foundation program is the method of financing most

commonly associated with elementary-secondary education.

The state determines the value of an FTE and pays that


amount times the FTE generated by the college.


The FTE


may be weighted-- that is, different dollar amounts per


FTE may be


for various programs.


Along with the FTE


funding a specific amount of local effort is required.


The state may also determine the tuition rate.


The state


supplies the amount of FTE dollars less the local effort


and/or tuition income.


equalizing,


This approach can also be d


with richer districts able to raise much greater


revenue because of their tax base,


unless some control is


placed on local effort.


The power equalizing system of


finance is used by


rN % y x t ,I.-n 4 ..- % a' .-4. 4 a n nn I....









levied.


The college can decide the tax rate to be levied


and the state guarantees the college a pre-determined


assessed


evaluation of tax base per FTE student.


drawbacks to this plan are


so severe


that no college


system has adopted it (Garms,


Garms'


1978).


plans are conceptual in nature and not limited


to methods currently in use.


However,


Wattenbarger and


Stepp (1979) used the following four categories in describing


state funding plans currently used by states.


These


categories are:


Negotiated funding.


This is a flexible plan without an


explicit formula.


Negotiations are on an individual


institution basis by college representatives with the state

legislature or a state board with granted authority for


such negotiations.


These budgets are negotiated annually


or biennually depending upon the state where used.


Unit-rate formula funding.


Allocations of funds to colleges


are on the basis of a formula specifying a fixed number of


dollars per defined unit of measure. That unit of measure

may be enrollment, student credit hours, full-time faculty


positions, or

alent student


some other defined measure.


(FTE)


Full Time Equiv-


is the most common unit used.


Minimum foundation plan.


This plan insures or guarantees


a specified level of support per student or FTE when state








Cost-based program funding.


Funds are allocated on the


basis of program costs determined by annual studies which


relate actual costs to the numbers of students served.


Each


individual college receives funds related to its own programs.

This funding method will be reviewed more thoroughly when

the Florida financing plan is analyzed.

According to Wattenbarger and Bibby (1981) only a few

major changes in financing patterns have taken place in recent


years.


The models described in Wattenbarger and Stepp (1979)


are still useful.


Although some states have altered their


formulas,


few basic differences have developed.


and Bibby (1981)


see


Wattenbarger


increased cost and assumption by the


states of funding for community colleges as hampering the


responsiveness of the institutions to community needs.


They


suggest another model


, a quality-based funding model,


as a


vehicle to allow the needed flexibility since they argue

that state funding constrains the concept of institutional

flexibility which is prerequisite to quality.


Quality-based Funding Model.


model


The quality-based funding


is gne which is program/institution specific in that


it differentiates among instructional programs and pro-

vides each community college the opportunity to set cost

levels for each program based on actual costs and insti-


tutional


priority


es.


The model is expressed as a five







34

= the number of FTE students in program p.

= a prescribed "average" dollar amount per FTE

= the cost factor for program p.

= a "quality maintenance" factor for program p


which


institution-specific


= a "quality improvement" factor for program p


which is institution-specific

The total allocation to the college is the sum of all

program allocations.

Components of the Model.


The factors F, B, and Cp.


The factor B is constant


over all programs and represents the average statewide


cost


per FTE in all programs combined.


Yearly increases


in allocations, if any, are generally incorporated into


this factor.


The program cost factor Cp is computed by


dividing the statewide average cost per FTE in program

p by the statewide average cost per FTE for all programs.

The cost factor is found by analyzing cost data from

some previous (base) year and performing the division


described,.


States in which FTE is not used as the


funding unit can still apply the model with their


established units.


Even if the program cost factor (Cp)


is not determined by base-year average costs,


the model


may still be applied.








most.


Overfunding and underfunding programs are


minimized with the inclusion of these two factors.


Relative to program funding,


the components of quality


maintenance might include sufficient funds to

1. offset declining enrollments if the formula is

enrollment-driven;

2. offset inflation and rising energy costs;

3. increase personnel salaries consistent with the

cost of living in the community college district;

4. maintain existing equipment or replace worn

equipment;


maintain a given level


of assistance to


disadvantaged students;

maintain library acquisitions at a given level;


offset higher costs for small or very


large


institutions;

8. insure that a certain percentage of courses be

conducted by full-time faculty; and

9. maintain a faculty development program.

It must be emphasized that this list is not intended to


exhaust


those elements which are considered essential


to quality maintenance.


Different elements may be


required in various settings, depending on statewide

and institutional priorities.







36

raise any of the quality maintenance

components above the maintenance level;

offer experimental courses, new degree

subprograms, or new certificate subprograms;

provide alternative learning experiences for

students;

purchase specialized equipment;

increase the ratio of full-time to part-time

faculty;

enhance program review and evaluation activities.


Again,


these elements are offered only as examples.


Specific elements must ultimately depend on state,


tutional


insti-


, and program priorities (Wattenbarger and


Bibby,


1981,


pp.35, 36, 37).


An analysis of


coQs


t-based plan employed by the


state of Florida follows.


It is expected that relationships


between this particular funding method and instructional

salaries may be revealed by this review of Florida's current

financing plan.

Florida Community CollegeFundng Formula
Florida Communit~y Collegje Funding Formula


Florida


cost-based funding method is computed from


aggregations of the individual colleges'


in~trlintinn


costs of providing


nli orronwth and ad in stment factors. and minus









with enrollments above 1,300 FTE students.


The small


colleges are those with enrollments less than 1,300 FTE


students.


The annualized total of FTE students equals the


total annual semester hours that students earned (fall

through summer) divided by 30 semester credit hours.

The program funding process for determining the state


allocation for the colleges is


follows:


An annual cost analysis is performed by each


college examining historical records of actual

expenditures for the immediate preceding year


of operation.


This cost-analysis report is


submitted to the state in October of each year.

The cost-analysis report presents the computed

unit cost per course for each course taught at


a college.


The unit cost per course includes:


A pro-rata share of


the teacher's salary


allocated in dollars per credit hour.


example, if a teacher teaches 15 credit hours,

one fifth of the salary is allocated to a


*three credit course.


(Non-credit courses


are converted to credit hour equivalents by

dividing contact hours by 27).


A pro-rata share of


costs


instructional department


allocated in dollars per credit hour.









A pro-rata share of intermediate costs,


college-wide costs, and physical plant

operations and maintenance costs allocated

in dollars per credit hour equally among

all courses taught.

The cost-analysis report exhibits course costs aggregated

into discipline costs and discipline costs are aggregated

into broad program costs.


The discipline costs and broad program


costs


are


ex-


pressed in dollars per FTE student.

B. The state separates the cost reports submitted

into two groups, a small college group and a

large college group, and aggregates the data by

group.


For each group


, the costs per FTE student by


discipline and by program are displayed, and a

"state-wide average cost" for each discipline

and program is computed.

A cost ratio for each discipline is determined by


dividing the calculated cost


for each discipline


by the state-wide average cost for all courses.


Example:


Health Professions
Average for all


$1800/FTE student


= 1.8


courses


$1000/FTE student









The base year state-wide average cost per

FTE student.

An added adjustment for economic conditions


(calculated from the consumer price index and

the wholesale price index and called the

economic lag factor).


An added adjustment for equipment.

A subtracted adjustment for student fees and

incidental college income.

A subtracted adjustment for federal funds.


The current year state-wide unitary cost


multiplied times the cost ratio for each discipline

to produce the current year projected cost per

FTE student in each discipline category.

G. The estimated FTE enrollments by discipline

category submitted by the colleges to the state

are multiplied times the current year projected

cost per FTE student in each of the discipline

categories.

H. The amounts generated in each discipline category

are summed to produce the total college allocation.

The program funding process for state allocation for


community colleges,


which exhibited general cost analyses


of discipline and broad nrogram expenditures (which include










community college district,


teacher characterization


for salary purposes, and,


to a limited extent,


the effects


of unionization on salaries.


Teacher Salary Studies


Elchanan Cohn (1971) studied the determinants of

teacher salaries utilizing data from 375 Iowa school districts.


Median teacher salaries were used


the dependent variable.


Independent variables were school size, number of


college


hours per teaching assignment, distance from nearest central


cities, and teacher salaries in the central city.

regression was the statistical method employed.


Multiple

Number of


college hours was the only statistically significant variable


related to teacher salaries.


In a sub-set of the same data


additional independent variables were available.


Two of


these variables found significant were average educational

level of the adult population in the school district and


percent of


families with income over $10,000.


Levin (1970) conducted a study of teacher salaries


utilizing data from the Coleman study.


He found that


teacher salaries were affected by teachers' verbal score,


sex,


years of schooling,


type of college from which grad-


uated, years of experience, certification level, and college


-







41

Kasper (1970) sought to find the effects of union-


ization on teacher salaries.


His sample


was


a cross-


section of the 50 states.


By using multiple regression


analysis he determined the following economic and demog-

raphic variables significantly related to teacher salaries:

extent of urbanization, sources of revenue, mix of elementary

and secondary teachers, and extent of unionization (ratio


of non-members to members).

a slight relationship to sala


Although unionization showed

ries, it should be noted that


Kasper used statewide averages which negate district variations.

King (1979) sought to determine factors which accounted

for variation in salaries among New York school districts.


After reviewing a number of


concluded,


teacher salary studies, King


"Salaries paid to teachers in a given school


district appear to be determined by characteristics of the


community and of the teachers themselves."


He hypothesized


that levels of salary paid by school districts was a function


of community


characteristics as defined by socio-economic


status (SES) variables and teacher characteristics,


including


experience (EXP) and Training (TRB)


quality (QLT)


levels


related characteristics.


as well as other


King expressed this


production function:


(SES, EXP


, TRG, QLT)


The s~~~mn1~ fn-r~K~inar'scciru UQ7Q =h









measures of


were:


the dependent variables (teacher salary


the mean salary,


level)


the mean plus one standard deviation


(high salary


level), and the mean minus one standard


deviation (low salary


level)


rather than actual amounts,


King chose these amounts


i.e.,


base, M.A., and M.A.


plus 30 levels because he believed that computed levels


would represent an assessment of actual salaries paid,


thus


reflecting a combination of hiring and staffing policy


decisions of which the salary schedule is only one.


Inde-


pendent variables included community and teacher character-


istics,


which were grouped and entered into the regression


equation based on their position in the production function.

Socio-economic factors explained a large amount of


variance in the high salary


level.


Mean family income of


district residents


was


a strong predictor of salary


levels


for all populations.

salaries two ways.


The socio-economic status affected

Economic indicators showed that high


wealth districts paid higher salaries,


but social indicators


(percent of minority students) appeared to have an inverse


relationship to salaries.


This finding can be accounted


for by the fact that most minority groups are located in

economically depressed city school districts but very often

have to pay higher salaries in order to attract teachers to

IIIC'hi 1'12* fi iovw







43

When community and teacher characteristics variances were

held constant, measures of teacher quality made a signif-


icant contribution to the remaining variance.


that socio-economic status of a community


King concluded

the strongest


predictor of salary


levels paid to public school


teachers.


Hall and Carroll


(1973) studied 118 elementary school


districts in Cook County,


Illinois,


for year


1968-1969.


Although the study emphasis was to determine unionization


effects,


the study revealed modest salary effects (salary


increase


- $613).


This finding was consistent with Kasper's


findings.


Other variables found significantly related to


salaries by Hall and Carroll were:


median family income,


percentage of white collar workers


, size


of district,


teacher


experience


, amount of state support


and class size.


Thornton's (1971) study of 83 large city school districts

found that about 50 percent of the variance in teacher


salaries


was


accounted for by union strength, average wage


rate in the city or surrounding county, and population of

the school district.


Lipsky


and Drotning (1973) sampled 696 New York school


districts in their study.


They examined four different types


of salaries.


Those used were base salary,


bachelor's degree


plus thirty college hours with seven years of


bachelor'


experience,


s nhls sixtV hniirs and 1 lven vsnrQ nf Yn








pupil-teacher ratio,


property value per pupil,


tax effort


and debt service

teacher salaries.


as significant variables in determining

Unionization appeared to have had a


rather strong effect on salary changes.


The data used in


this study were collected approximately one year after the

passage of the Taylor Law which made it legal for teachers

to bargain collectively in New York.


In a similar study Frey (1975)


districts in New Jersey.


beginning B.A.


looked at 298 school


Frey used starting salary for


teachers as the dependent variable.


independent variables were enrollment, median family income,

taxable value of property per pupil, opportunity cost, and


collective bargaining.


All of these variables had a positive


correlation explaining about 60 percent of the variance among

districts.


Size


of school district, median family or per capital


income,


size


of tax base,


and education-experience of the


teacher seem to be variables which most influence teacher

salaries.

Chambers(1977) evaluated 89 school districts in


California for his study on unionization.


He found the


presence of a union to have a significant impact on teacher


salaries.


Chambers also found that unified, unionized


districts had higher salaries than thnss diRtrintQ with









studied over a longer term.


Much of the change can be


attributed to "spillover" effects in which surrounding

districts tend to raise salaries to levels paid by districts


with collective bargaining.


size


Chambers found that district


and pupil/teacher ratio also had significant effects


on teacher salaries.

The various studies reviewed differed in scope, data


sources, and model specification.


Therefore, it is difficult


to generalize the findings.


One may, however,


identify


several


variables that were reported statistically signif-


icant in most of the research.

Comparisons between professional instructional salaries

and other industry salaries are difficult because of

differences for which there can be no true assessment.(Kershaw


and McKean,


1962).


Vacations are an example.


While there


is conclusive evidence (Cohn,


1979) that teachers'salaries


as a whole have not kept pace with private industry,


there


are causes that have not held a place in the studies leading


to this conclusion.


Many teachers are parents of children


who would ordinarily require child care if


they were employed


in an industry which operated on a twelve month basis, and

indeed this is a sizable expense borne by many employees in


other industries.


How,


then, have previous equity studies


accounted for this factor in their adjustments?


Only







46
In the next section faculty salary studies are reviewed

to determine if certain variables appear consistently.

Included in this review are studies which investigated the

relationship between unionization and faculty salaries.


Faculty Salary Studies


Cohn (1973) used 1970-1971 American Association of

University Professors (AAUP) salary and compensation data


in his study.


The method of analysis was multiple regression.


Institutions were used


individual faculty members.


;he units of analysis rather than

The sample included data on


salaries and compensation (salaries plus fringe benefits)

of faculty by rank and for all ranks combined for 204 insti-


tutions.


There were 10 dependent variables:


average


compensation by each

each of five ranks.

of institution, contr


size


of


five ranks and average salaries by


The explanatory variables included type

ol, region, quality, dynamic changes,


of institution, and per capital income of state where


institution was located.


Quality of the institution was


found to have a significant effect on salaries. Enrollment

and salaries show an inverted U curve relationship. Salaries

increased up to 30,000 enrollment--after that average


salaries appear to decline.


Also,


institutions in states


with higher per capital income nay higher salaries.


Cohn









In a study by Tuckman, B. H.


& Tuckman, H. P.


(1976)


data from the American Council on Education (ACE) were


utilized.


Using multiple regression analysis,


the authors


found significant salary related variables to be males,


blacks, some fields of


instruction, age, and experience.


Curiously, salary returns to age peak at about age forty-


nine.


The authors did further work in 1977 and found no


significant differences between the two studies.

Robert Birnbaum (1974) undertook a study matching 88

non-unionized with 88 unionized higher education institutions.

His findings show a positive relationship between salaries


and collective bargaining.


Unionized institution faculty


had a $777 salary advantage over their non-unionized


counterparts.


He also found that larger differences occurred


among four-year institutions.


However, Birnbaum failed to


control for other variables that may affect faculty salaries.


Brown and Stone (1977)


restricted their study to


four-year institutions with bargaining agreements rather


than matching institutions as Birnbaum had done.


Their


study failed to find any significant impact resulting from


collective bargaining.


Although the study revealed signif-


icant growth rates in salary among upper ranks,


these were


accounted for by several institutions with unique character-


istics.


The authors did not control for other factors.







48

Morgan and Kearney did not look at faculties by rank but


divided the institutions into three categories.


Multiple


regression analysis was the statistical method employed


to control


for other variables such


per capital personal


income,


institutional control, and institutional quality.


Although unionized faculty members showed a slight monetary


advantage over non-union faculties,


the greatest difference


found was in fringe benefits for the unionized faculties.

Morgan and Kearney believe that union growth will continue

but long range effects cannot yet be determined.

Larry Leslie and Teh-wei Hu (1977) used Morgan and


Kearney's model to look at certain financial aspects of

collective bargaining. They employed Birnbaum's sampling

method of matching institutions. The sample included 150

two-year and four-year institutions. The two dependent


variables were faculty compensation and institutional

financial data from the Higher Education General Information


Survey Files (HEGIS)


, and average compensation was used for


a comparison between 1974-1975 and 1975-1976.


Leslie and


Hu took the Morgan-Kearney model one step further and


analyzed compensation by rank.


They found that unionized


faculties received approximately $1,291 more than faculties


in non-unionized institutions during 1974-1975.


comparison was made for


When this


1975-1976 the difference was only







49

faculty members received higher benefits among unionized


institutions.


This finding does not seem unusual although


other studies reviewed revealed that age 50 is the peak and

salary differences decline after that.

Leslie and Hu found other variables affected compensation.


These included institutional control,


per capital income,


institutional quality, and percent of faculty holding a


doctoral degree.


nation could affect compensation.


Again this may be because


of the relative newness of unionization among higher education

institutions.

Finally, uses of part-time faculty are reviewed in terms


of cost control,

unionization. C


program and schedule flexibility, and non-


constructive utilization of the adjunct


faculty is also explored.


Adjunct Faculty


"Recent statistics reveal that over fifty percent of


the teachers in today


two-year colleges are part-time"


(Parsons,


1980,


p.vii).


The reasons seem to be that budget


constraints force administrators to look for alternatives

to high personnel costs and adjunct (part-time) faculty


utilization provides more budgetary flexibility.


flexibility offered by the use of adjunct faculty has


The authors suggested that union affil-









paid to full-timers according to Eliason (1980).


Fringe


benefits,


if offered at all


to adjunct faculty, are limited.


Adjunct faculty, unfettered by union contracts, allow flex-


ability in scheduling also.


important


This issue becomes increasingly


as colleges offer programs at non-campus sites


since it is often easier to hire adjunct faculty to fill

those instructional roles.

The National Center for Education Statistics (NCES)

projected a small growth through 1982 for employment of

full-time faculty in all postsecondary education and a

decline of 5.7 percent over the following three-year period


ending in 1985.


Full-time faculty members are growing older


and are less mobile than at any other time in history.


According to Tuckman, H. P.


& Tuckman, B. H.


(1981)


part-timers are neither good nor bad for academe


in their own right.


Instead they are a diverse


group with many different motives and goals.


Whether


we learn to employ them in a constructive manner will


surely be one of


the fascinating questions of the


Summary


The literature reviewed suggests the following variables


are related to instructional salaries:


educational level


of district,


teacher characteristics


(sex,


age, race), extent


of urbanization, revenue generated, measure of


income of









population of


district,


pupil-teacher ratio, enrollment,


unionization, quality of


institution,


publications,


field


of instruction, and institutional control.


Certain of


these variables would not be relevant in


study of


community colleges.


Publications and research a e


not an integral part of community college faculty qualitative


evaluations.


Institutional control would not be appropriate


to this study since all institutions studied are governed


commonly.


or racial discrimination are issues beyond


the scope of this study, and therefore


sex


and race variables


were


not used.


District revenue variables do not seem


appropriate because the institutions in this study are fully

funded by the state of Florida.

The independent variables chosen for this study are


total


student FTE,


population of district,


district per


capital Effective Buying Income, percentage of classes taught

by adjunct faculty, and unionization. The three dependent


variables for each of the twenty-eight districts are mean

daily rate of salary for faculty holding doctorates, mean


daily rate of'salary for faculty with master


s degree plus


thirty additional hours of preparation, and mean daily rate


of salary for faculty with master


s degree.


Nothing was found in the literature reviewed to indicate

-t-Hnc"h rQT ont+-nao n\f rlaooccc1ca tiirbt bn i\r rliinrt- for~iil-tn Hador
















CHAPTER III
ANALYSES OF DATA AND FINDINGS


This study sought to determine what relationships

existed between selected socio-economic variables and


salaries paid to


full-time community college instructional


personnel in the twenty-eight community college districts


in the state of Florida.


The related literature and


research were reviewed to determine what variables would


be included in the study.


chosen.


Five independent variables were


Three measures of salary were used as the


dependent variables.

The salary and full-time equivalent data were collected

from the Division of Community Colleges, Department of

Education, State of Florida, Report of Annual Salaries

1980-81, Full-Time Instructional Personnel and Annual FTE


Count by Term 1980-81.


The population data were obtained


from the Florida County Comparisons,


1981.


The population


of the counties served by the institution was aggregated


where indicated.

only one county.


In some instances an institution serves

The Effective Buying Income per capital


was obtained from Sales and Markptina Mnna-omant Mnana'in0


- --- -- -----------~r










district served by the institution.


The percentage of


classes taught by adjunct faculty was obtained from a


survey sent to all institutions.


Those institutions not


responding within the time frame allotted were surveyed


by telephone to get the percentage of


adjunct faculty.


classes taught by


This method of follow up insured 100


percent response.


Many of


the survey instruments were


not returned until after the statistical analysis was


completed; however, the crucial data had been obtained

previously by telephone. A list of Florida community


colleges with an instructional bargaining unit was ob-

tained from the Division of Community Colleges, Depart-

ment of Education, state of Florida.


The data were analyzed by Pearson correlations


partial


correlations, and standard multiple regression


utilizing the Statistical Package For Social Sciences


(SPSS).


This section presents the data analyses and


findings.


Initially,


the simple correlations are dis-


cussed and compared with partial correlation


coefficients.


The results of the regression equations are then presented

and discussed in relationship to the results of the


partial


correlation coefficients.


In presenting the findings,


the results are related to the specific questions this study


1i rit n'ofriohlac ot+urlio


clnrlnkt


~r]hln









TABLE 2
IDENTIFICATION OF VARIABLES


Dependent Variables


Abbreviation


Algebraic


mean daily rate of salary
for full-time instructional


personnel
degree


with doctorate


mean daily rate of salary
for full-time instructional
personnel with a master's
degree plus 30 additional
graduate hours preparation


mean daily rate of salary
for full-time instructional
personnel with a master's
degree


Independent Variables


Abbreviation


Algebraic


total FTE of the institution

total population of the
community college district


per capital Effective
Buying Income of the
community college district


percentage of


classes


taught


by adjunct faculty

unionization 0= No;
(dummy variable)


1= Yes







55

This study sought answers to specific questions


dealing with the relationships among salary


levels and


selected variables.


The findings


as related to the spec-


ific questions of the study follow:

The first question of the study was:

Is there a relationship between the widespread

use of adjunct faculty and salaries paid to full-

time instructional personnel?


The Pearson


s product-moment correlations between mean


average daily salaries paid to faculty holding doctoral

degrees (DOC) and the percentage of classes taught by adjunct


faculty (ADJ) was 0.18, indica

direction in the relationship.


sting only a weak positive

The mean of the mean average


daily salary was $108.43 with a standard deviation of $10.33.

The Pearson correlations between mean average daily


salaries paid to faculty with master


s degrees plus 30


additional hours of preparation (MDL) and the percentage of


classes taught by adjunct faculty (ADJ) was 0.35.


This


correlation also indicates only a weak positive relationship.


The mean of the mean average daily salaries was


103.67 and


the standard deviation


was


$11.32.


The product-moment correlation between average daily


salaries paid to faculty with master'


degrees (MAS) and the


no> ilnor1 r I' cf + fl CT 0 n i-i n o or -innl- Pan.1 +txT (A ATT\


nnr rlnni~nn














Cl
H
a)
C']
0
I


C-
0
cc
H

0


0
CD
H
0



Vt-
cc

0


C~2
CD
eq
Co
0


0
Cl
Cl
cc
0


0
cc
cc

0



0
0
0
H









The partial


correlations indicate a very weak linear


relationship between each of


the three salary


levels and


percentage of

reports that,


classes


taught by adjunct faculty.


after controlling for FTE, POP


Table 4


and EBI


third order partial correlations between ADJ and DOC, MDL,


and MAS were 0.08,


0.20, and 0.08,


respectively.


Tables 5,


6, and 7 report the results of the regression analyses.

The largest contribution made by the ADJ variable in pre-

dicting mean average daily salaries was for the master's


plus 30 hours level


(MDL).


That contribution was only 3


percent of the total


variance accounted for in the model.


The multiple regression equations seemed to verify


this weak relationship.


The percentage of classes taught


by adjunct faculty accounted for a minimal amount of the

variance when regressed with the three salary categories.










PARTIAL


TABLE 4
CORRELATION
3RD ORDER


COEFFICIENTS


Variable FTE X1 controlling for POP, EBI, ADJ


Y1 DOC 0.3906

Y2 MDL 0.5521

Y3 MAS 0.5083












Variable POP X2 controlling for FTE, EBI, ADJ


Y1 DOC -0.2897

Y2 MDL -0.4606

Y3 MAS -0.4771


4









TABLE


- CONTINUED


Variable EBI X3 controlling for FTE, POP, ADJ


Y1 DOC 0.1562

Y2 MDL 0.4123

Y3 MAS 0.4379














Variable ADJ X4 controlling for FTE, POP, EBI


Y1 DOC 0.0777

Y2 MDL 0.1998

Y3 MAS 0.0819

4










The second question of the study was:


Is there a relationship between the


population of


size


the community college district and


salaries paid to full-time instructional personnel?


The Pearson'


product-moment correlations between


size


of population of the community college district (POP) and

salaries paid to faculty holding doctoral degrees (DOC)


was 0.19,


indicating only a weak positive direction in the


relationship.


The Pearson correlation between


size


of population of


the community college district and salaries paid to faculty

with master's degrees plus 30 additional hours of preparation


was


0.33.


This correlation indicates a weak positive


relationship.


The zero order correlation between the


size


popu-


nation of the community college district and salaries paid


to faculty with master


s degrees was 0.12,


also a very weak


positive relationship.


The mean of the population of the


community college districts was 346,0


with a standard


deviation of 342,417.

The partial correlations indicate a somewhat stronger,


although negative,


relationship between each of the three


salary


levels and


size


of population of the community


nnl 1 _p' dis trict.


A n nr n 4-


*1Ty -


p- inn'


nrrplntinn inriit n


i







61

TABLE 5
REGRESSION ANALYSIS OF THE RELATIONSHIP OF THE DEPENDENT
VARIABLE DOC TO INDEPENDENT VARIABLES


Variable Multiple R R Square RSQ Change Simple R *


0.34133


0.45277


0.47918


0.53614


0.54608


0.11650


0.20501

0.22961

0.28744

0.29820


0.11650


0.08850

0.02460

0.05784

0.01076


-0.34133


0.33208

0.18906

0.17069

0.16096


* Simple R:


Standard error of


zero order correlation between the dependent
variable and each of the independent variables.


estimate $10.50







62

TABLE 6
REGRESSION ANALYSIS OF THIE RELATIONSHIP OF THE DEPENDENT
VARIABLE MDL TO INDEPENDENT VARIABLES


Variable Multiple R R Square RSQ Change Simple R *


0.21100


0.50390


0.52809


0.69431


0.71581


0.04452

0.25392

0.27888

0.48207

0.51239


0.04452

0.20939

0.02496

0.20319

0.03032


-0.21100

0.47743

0.33201

0.41990

0.34656


* Simple R:


zero order correlation between the dependent


variable and each of the independent variables.

Standard error of estimate $8.93
*









TABLE


REGRESSION


ANALYSIS


OF THE RELATIONSHIP OF THE DEPENDENT


VARIABLE MAS TO


INDEPENDENT VARIABLES


Variable Multiple R R Square RSQ Change Simple R *


0.18419


0.37503


0.42055


0.63339


0.63941


* Simple


zero order


0.03393

0.14065


0.17686


0.40119


0.40884


correlation


0.03393

0.10672


0.03621


0.22433


0.00766


between


-0.18419

0.34440


0.20113


0.35431


0.25159


the dependent


variable


each


independent


variables.


Standard


error


estimate


$7.50







64

In the standard regression analyses, POP accounted


for a minimal amount of the total


variance when regressed


with the three salary


levels.


The largest contribution


made by the POP variable in predicting mean average daily


salaries was for the master's degree


level.


That contri-


bution was only 3 percent of the total


variance accounted


for in the model.

The third question of the study was:

Is there a relationship between the size of the

institution and salaries paid to full-time

instructional personnel?


size


of the institution was indicated by the number


of Full Time Equivalent students (FTE) generated by the


institution.


The Pearson's product-moment correlations


between FTE and salaries paid to faculty holding doctoral


degrees was 0.33,


indicating a positive relationship.


The Pearson correlation between the


size


of the insti-


tution and salaries paid to faculty with master's degrees


plus 30 additional hours of preparation


was


0.48.


This


indicates a stronger positive relationship.


The Pearson's


product-moment correlations between the


size


of the insti-


tution and salaries paid to faculty with master'


was 0.26,


degrees


indicating the weakest relationship of the three


salary


levels.


The mean of the


size


of the institutions









for POP, EBI, and ADJ.


0.39,


The partial correlations were


0.55, and 0.51 for the doctoral, master's plus 30


hours and master's degree levels of salaries respectively.

This variable (FTE) showed the strongest relationship to


salary


levels.


The linear relationship was strong as well


the predictive power


as shown by the amount of variance


accounted for in the regression models.


size


of the


institution accounted for 9 percent of the variance in the


regression model when doctoral level


was the dependent


variable.


When the master


degree plus 30 hours of addi-


tional preparation was the dependent variable, FTE accounted


for 21 percent of the total


variance.


When master


degree


level salary was used as the dependent variable, FTE accounted

for 11 percent of the total variance in the model.


The fourth question of the study was:

Is there a relationship between full-time instruc-

tional personnel salaries and the Effective Buying

Income per capital of the community college district?

The Pearson's product-moment correlation between mean

average daily -salaries paid to faculty holding doctoral

degrees and the Effective Buying Income per capital of district


(EBI) was 0.16;

The mean of the


indicating only a weak positive relationship.


SEBI was $6824 with a standard deviation of


$1375.







66

The Pearson correlation between the mean average

daily salaries paid to faculty with master's degrees and


the EBI


was


0.30


, indicating a weak positive relationship.


The partial


correlations indicate about the same


relationship as found by the Pearson correlations.


third order partial correlations between EBI and DOC, MDL,


and MAS were 0.16,


0.41, and 0.44,


respectively.


Effective Buying Income accounted for 6 percent of the


variance in the regression model when using DOC


dependent variable.


When MDL was used


the dependent


variable, EBI accounted for 20 percent of the total

Effective Buying Income accounted for 22 percent of


variance.


the


total variance in the regression model using master's degree


salary


level


the dependent variable.


The fifth question of the study was:

Is there a relationship between the presence of

a faculty bargaining unit and full-time instruc-

tional salaries of the community college district?

The Pearson's product-moment correlation between mean

average daily Salaries paid to faculty holding doctoral


degrees and unionization (UNN) was -0.29.


The zero order


correlation between the mean average daily salaries paid


to faculty with master


nf nrrnnnrnft i nn nnr INN wns.c -0.21-


s degrees plus 30 additional hours


Tho Pcnrsnn pnrrelation









appear to be a significant variable in this study.


data were badly skewed because of the small


number of


unionized institutions.

The variance accounted for by UNN in each of the


regression models was .12,

and MAS respectively. Ana


.04, and .03 for DOC, MDL,


lysis of variance (ANOVA) was


run on subpopulations of unionized institutions and non-


unionized institutions with each of the three salary


levels.


The results verified the lack of a significant relationship.

The sixth question of the study was:

When using mean daily salary for individuals hold-

ing doctoral degrees as the dependent variable,

which independent variables account for the great-

est percentage of variance in the mean daily

salary?


Standard multiple regressions of salary


levels with


the five independent variables for the twenty-eight community

college districts in the state of Florida were computed


(Tables 5,6,7).


These analyses provided the multiple


correlation coefficient (Multiple R) which is the correlation

between a dependent variable and the weighted sum of the

independent variables.

The multiple regression program also provided the


R Sauare (R2) statistic.


This statistic does not imply









of the variance in salary


level for instructors with doctoral


degrees (Table 5).


The R


Square Change column gives the actual change


accounted for by each variable.


For example, unionization


accounted for about one-third of the total


variance


assoc-


iated with doctoral salaries,


and each of the remaining


variables accounted for a lesser amount of co-variation.


Percentage of


classes taught by adjunct faculty (ADJ)


accounted for the least.


It is of interest that while


unionization did not appear to correlate significantly in

either simple correlations or ANOVA of subpopulations (Table


it did, nevertheless, account for the greatest amount


of variance in regression with doctoral degree salaries.

The results of the regression analysis (Table 5)


indicate that unionization accounted for

variation followed by FTE with 9 percent


12 percent of the


EBI with 6


percent, POP with


percent


, and lastly ADJ,


1 percent.


Unionization had a negative


simple


correlation of 0.34 in the


standard regression results.


(SEE) was $1050.


The standard error of estimate


The five independent variables accounted


for approximately 30 percent of the variance.

The seventh question of the study was:

Which independent variables account for the

largest percentage of variance in the mean







69

Table 6 shows the results of the regression analysis

for dependent variable master's degree plus 30 hours (MDL).


The results showed FTE (21 percent) and EBI


(20 percent)


as accounting for the most variance in this equation as


indicated by RSQ Change Column.


Unionization accounted for


4 percent followed by ADJ


(.03) and POP (.02).


Union-


ization had


a negative Simple correlation of 0.21 with


salary


level MDL.


The combined Multiple R for the five


independent variables was 0.71581, explaining approximately


51 percent of the variance in salary


level MDL.


The eighth question of the study was:

Which independent variables account for the

largest percentage of variance in the mean

daily salary for individuals at the master's

degree level?


Effective Buying Income accounted for


22 percent


of the variance and FTE accounted for


11 percent.


Population


accounted for 4 percent followed by UNN with 3 percent and


ADJ with less than 1 percent.


Again UNN showed a negative


Simple R.


Al of


the variables together accounted for


41 percent of the variance.

Table 7 summarizes the regression results using MAS


the dependent variable.


The combined Multiple R for


the five independent variables was 0.63941.









independent


variables chosen.


Each of


the variables


chosen did hold a place in the stepwise regression analyses,

and while the relationships were extremely weak in some


instances,


they were selected for the regression.


This may


suggest that other measures of those variables need to be


considered.


Because of the relatively small number of


(28) the number of variables to be selected


was


cases


limited;


therefore,


rather than increase the number of variables,


different measures of these same variables and/or others

are suggested.

Stepwise multiple regression analysis was used on the

dependent variable MAS to determine the difference between

the results of Standard Multiple Regression and Stepwise


Multiple Regression.


Multiple R and R Square were exactly


the same


however,


in stepwise regression the variables were


entered in optimum order.


For comparative purposes


see


Tables


7, and 8.











TABLE 8
STEPWISE MULTIPLE REGRESSION
DEPENDENT VARIABLE MAS
SUMMARY TABLE


Variable


Multiple


R Square


RSQ Change


Simple


0.35431


0.44002


0.46528


0.48287


0.63941


0.12554

0.06809

0.21642

0.23316


0.40884


0.12554

0.06809

0.02287

0.01667


.17568


0.35431

-0.18419

0.25159

0.34440

0.20113


Mean


daily


rate of


salary


full-time


instructional


personnel


with


a master's


degree


(MAS)


was


used as


basis


comparison


remunerative


because


level


the most


commnu n it y


widely


college


recognized


instructional


personnel.


Stepwise multiple


regression


confirmed


results


standard


regression


analysis


that


each


1~ndepecnden~t


variable


accounted


for a


part


total


variance


regression


equation.


In stepwise multiple


regression


variable


that


explains


greatest


amount


variance


conjunction


with


first


will


enter


second


and so on.


Each


subsequent


variable will


be entered


only









TABLE 9
ANALYSIS OF VARIANCE (ANOVA)
BETWEEN SUBPOPULATIONS


Dependent Variable


Significance Level


0.1327


0.32


0.3232


Unionization did not appear to be a significant variable

probably because of the small percent of unionized insti-


tutions.


More careful scrutiny was accomplished by running


an analysis of variance (ANOVA) on the subpopulations


between the


institutions with unionization and the


institutions which are not unionized.


Each salary


level


was used as the criterion variable in an analysis of


variance.


The SPSS ANOVA program automatically computes


and prints out the significance level as shown in Table 9;


therefore,


is immediately evident that no significant


difference


exists


between the salary


levels of union and


non-union subpopulations.


This was true with all


three


levels of salary.


Di scunssi i nn







73

widespread use of adjunct faculty would show a strong


relationship with full-time instructional salaries;


this prediction was not verified by the data.


of classes taught by adjunct faculty


however,


Percentage


(ADJ) did not show


strong zero order correlations with the salary


levels


and became much weaker when partial correlations were


computed, holding FTE,


POP, and EBI


constant.


Percentage of classes taught by adjunct faculty (ADJ)

also accounted for a minimal amount of the variance when


regressed with the three salary categories.


It may be that


the use of adjunct faculty should be measured in a different


manner but

reasonable.


have


the technique chosen for this study seems

Perhaps the use of adjunct faculty does not


a relationship to full-time instructional salaries


under any conditions.


The method of


funding in Florida


(cost based) may also preclude a relationship between

instructional salaries paid and the use of adjunct faculty.

Total Full-Time Equivalent students of the institution


(FTE) appears to be related to salaries paid.


The funding


method in Florida seems to have insured this relationship.


The more FTE generated,


the more money the institution


receives from the state and, of course,


the greater effect


on salaries.

Dc-r rnonitn P ffcr tlyo nfliiin Tnrnmn r-i f- H rnrnnmuinlti7









indicates the income


level of the district, and while it


vould be expected that salaries would be higher in areas


with higher per capital income,


it does not necessarily


follow that salaries for public institutions are related


to other area salaries.


Total compensation would need tc


be considered before comparisons could be made because of


the uniqueness of fringe benefits in the field of


education.


Population does not appear to be strongly related to


the salary


levels.


Total population of the community


college district (POP) accounted for


.02,


.02, and .04 of


the total


variation in the three analyses using regression.


In terms of predictive power a widely used statistic


is the standard error of estimate


, which is simply the


standard deviation of actual Y values from the predicted


yl values.


preted


The standard error of estimate may be inter-


as a sort of "average error" in predicting Y from


the regression equation (Nie et al.


, 1975,


The multiple regression equation for master's salary


was


a fairly strong predictor as shown by the results


wherein the five independent variables accounted for 41


percent of the variance.


The standard error of estimate


was $7.50;


therefore


, 68 percent of the time this equation


would come within $7.50 of


estimating the master's level


salary.


The mean


was


$93.26 and the standard deviation









of the time would predict that salary within $8.93.


mean average daily salary was $103.67 and the standard

deviation was $11.32.


In the doctoral salary equation


, the five independent


variables accounted for 30 percent of the variation in


the dependent variable DOC.


was $10.50.


The standard error of estimate


In terms of predictive power, doctoral salaries


at those institutions could be estimated within $10.50,


around the 68 percent confidence interval.


dail


The mean average


y salary was $108.43 and the standard deviation was


$10.83.

With a cost-based funding formula it could be hypothe-

sized that the more FTE generated the higher the salaries


paid to full-time instructional personnel.


While


relationship is indicated, it is neither strong nor weak.

The negative partial correlation between the salary

levels and population indicates an inverse relationship.

Again, the relationship is not strong, but it did strengthen

when the effects of the other independent variables were


partialled out or held constant.


A negative correlation


does not indicate lack of strength, merely inversity.

The correlations for Effective Buying Income remained

about the same for the first two salary levels but in-

nrnn*<^ocnA nnl-* no-i Arar*h~l7 u thaf- m actafrt c? r*l








salary


levels and percentage of classes taught by adjunct


faculty decreased,


indicating a weak linear relationship.


A strong relationship between salary


levels and the


independent variables is not indicated; however,


the partial


correlations do strengthen the notion that each of the

chosen variables have a relationship with salaries.

The method of funding in Florida (cost based) may be

cause for the lack of relationship between full-time


instructional salaries and percentage of


classes


taught by


adjunct faculty.


Further analyses are indicated.















CHAPTER IV


SUMMARY


, CONCLUSIONS, AND RECOMMENDATIONS


Summary


The purpose of this study was to determine what rela-


tionships exist,


if any,


between selected socio-economic


variables and salaries paid to full-time community college

instructional personnel in the twenty-eight community


college districts in the state of Florida.


Further,


study evaluated the importance of other selected factors


they related to instructional salaries, specifically


the effects that utilization and use of adjunct instructors

may have on full-time professional salaries.

Specific questions were asked regarding the presence

of relationships between the three dependent variables


and each of the five independent variables.


A deter-


mination of the most significant relationships to each

dependent variable was sought.


Chapter II of this study reported a review of related


literature and research.


Chapter II


The literature reviewed in


included the growth of higher education, spe-









Chapter


III of this study presented the analyses


Tf data and findings.


A summary of the findings related


to study questions follows.


Is there a relationship between the widespread


use of adjunct faculty and salaries paid to full


time instructional person

mine what relationship,


In order to deter-


if any, exists between the


use of adjunct faculty and salaries paid to full-


time instructional personnel,


a measure for


adjunct faculty was included in the analyses


The percentage of classes taught by adjunct faculty


was chosen


as the independent variable.


This


variable (ADJ) was subjected to Pearson corre-


nations,


partial correlations, and multiple


regression procedures.


The Pearson correlations


indicate relationships between pairs of variables.


The relations


between each of the three levels


of salary and percentage of classes taught by


adjunct faculty were relatively weak,


with the weak-


est relationship being between doctoral level


salary and ADJ


In the partial


correlations the


relationships weakened considerably when FTE, POP,


and I31 were partialled out.


The multiple regression


equations seemed to verify this weak relationship.







79

procedure ADJ was included in the equation

which may indicate that adjunct faculty should

be considered but perhaps another study design


would be more appropriate.


literature,


In the review of


no study was found which dealt with


adjunct faculty as its use may affect full-time


instructional


salaries.


literature focused


more on the exploitive issue associated with


adjunct faculty.


Certainly


, consideration must


be given to this segment of the instructional

population inasmuch as Parsons (1980) reports

that over fifty percent of the teachers in today's

two-year colleges are part-time.


Rather than affecting salaries


, it may be


that class


size


full-time instructors is


affected by the use of adjunct faculty.


widespread use of


salary schedules may also


be a reason for lack of impact from adjunct faculty.

The funding formula of a state may also preclude

a relationship between instructional salaries


paid and the use of adjunct faculty.


The issue


of adjunct faculty may be better dealt with in

another study with a different design.


Ts there a relationshin hetweon th


Qim 'y


nf non-









subjected to correlation,


partial correlation,


and multiple regression analyses,


it was deter-


mined that an inverse relationship existed.


weak linear relationship exists as evidenced by


the Pearson correlations; however,


when partial


correlations were computed holding FTE, EBI, and

ADJ constant, a somewhat strengthened negative

relationship resulted indicating an inverse rela-


tionship.


In the regression analyses, POP accounted


for a minimal amount of the variance in the


equations.


In many of the salary studies reviewed


some measure of the district population was used


as an independent variable.


1977)


According to Chambers


larger districts tended to pay higher wages


than smaller districts.


The inverse relationship


indicated by partial correlation procedure does


not verify Chamber's findings.


The conflicting


results may be attributed to different controls.

Smaller districts may have to offer higher salaries

to attract qualified instructors, and larger

districts may have a larger supply of qualified


persons from which to draw.


For the purposes of


this study design no conclusive findings were re-

vJ&2 1 eri









Equivalent students (FTE)


is an indicator of


size


of the institution.


This was an


important variable in studies by Frey (1975),


and Cohn (1973).


This variable proved to be the


strongest variable in this study.


A strong


linear relationship as well as strong predictive


power was indicated.


The community college


system in Florida is funded by a cost-based


formula.


This type of funding formula would be


strongly related to institutional


size.


Is there a relationship between full-time instruc-

tional personnel salaries and the Effective Buy-

ing Income per capital (LBI) of the community


college district?


Some measure of district income


was significant in studies by Morgan ana Kearney

(1977), Leslie and Hu (1977), Cohn (1973) and


others.

chosen

study.


Effective Buying Income per capital was


the measure of district income for this


The same procedures were applied to this


independent variable with a medium linear rela-

tionship evidence by the correlation procedures.

The results of the multiple regression analyses


were conclusive for the Master


s and Master


s plus


5~Ci 1~rO~c, t~ii4+ Z'flTanr'rnn1-inr 4'n 1sran anh


~n








82

may be indicative of the wealth of that district.

Wealthier districts may make more demands on insti-

tutions to provide more costly programs which may


increase salaries.


Certainly instructors would


nrot be willing to work for a low salary if other


living costs were higher in a given district.


could be hypothesized that higher EBI indicates

wealthier districts which could indicate elevated


living expenses.


Effective Buying Income (EBI)


appears to be related to salaries to some extent;

however, there may be other measures of income which

would have a stronger relationship with salaries.


For instance


, median family income


, income per


capital or Ef[lective Buying Income per household

are measures to be considered for other studies.

Is there a relationship between the presence of

a faculty bargaining unit and full-time instruc-

tional salaries of the community college district?

Apparently, unionization was not a significant
variable for inclusion in this study design. A
variable for inclusion in this study design. A


very weak negative correlation was shown by the


Pearson correlations.


The subpopulations of


union and non-union districts were subjected to







83

accounted for a minimal amount of variance in the

regression analyses for master's and master's plus


30 levels of salary


Unionization accounted for


the greatest amount of variance in the regression


with doctoral level salaries.


There have been


numerous salary studies which have considered the


affects of unionization.


Brown and Stone (1977)


failed to find any significant impact resulting


from collective bargaining.


Leslie and Hu (1977)


found unionizati t;


to be significant.


It appears


that


for every negative result there is a positive


result relating to unionism.


Studies dealing


strictly with salaries and not total compensation


probably are not


as valid


those which consider


fringe benefits.


Bargaining is often directed


to items other than salary especially in the

field of education.

The small number of unionized institutions

in this study caused the data to be badly skewed.

Probably another design is indicated if unionization


to be dealt with properly.


When using mean daily salary for individuals hold-


ing doctoral degrees


as the dependent variable,


1,.4 K.. A -, ~ A~- .. .4,. 1,~ -. nn ~ .n 4 *A 4..4









and ADJ.


The total


variance accounted for was


approximately 30 percent.


The doctoral level


regression analysis resulted in a completely


different equation than MDL and MAS.


Union-


ization again showed a negative correlation


indicating inversity.


Further analyses are


indicated for the doctoral level salary.


year institutions have not required


Two-


a doctorate


for instructional purposes;


therefore,


it m.y


be that salaries at the beginning doctoral level


are


lower than the more experienced personnel


with master's plus 30 and master


s salaries who


have


reached


the upper


levels of salary schedules.


This may account for some of the confusion


surrounding this particular variabi


Other


compensation may need to be studied before a

definitive conclusion can be drawn.

Which independent variables account for the

largest percentage of variance in the mean daily


salary for individual

30 credit hours? Mul


with master's degree plus


tiple regression analysis


shows the amount of variance for each variable


in the equation.


Full Time Equivalent students


nnri rrfort-itera n flii ncrr Tn rn/mc na r^ r-oanfta tiroro th^a









negligible amount of the total


variance.


From


literature,


indications were that institution


size and a measure of per capital income were

significant variables and this was verified by


the results of


these analyses.


One could con-


elude that these two measures should be included


in any study of


salaries.


Which independent variables account for the largest

percentage of variance in the mean daily salary


for individuals at the master


s degree level?


Effective Buying Income and Full Time Equivalent

students, combined, accounted for the major


portion of


the variance (33 percent).


This


finding further verifies that these two variables


are strongly related to salaries.


In another


design perhaps different measures of these


but for this


variables would be more appropriate,


study design the chosen measures indicated strong

predictive power and their inclusion was justified.

A


Conclusions


A thorough review of


the relevant literature reveals


very few salary studies using data from two year institutions.

Thp denpi svi nn tn tbhn ccr th o r x o h^io l c -n +1,4<- c4.,, *nnn ^- ,/-










a combination of two and four-year institutions.


Because


public community colleges are relatively new, there is need


for more exploration in the area of wage determination


more research is completed, a clearer understanding of the

variables affecting instructional salaries will be developed.

The researcher suggests that one important factor in community

college salary variations may be the type of funding formula

utilized by the systems investigated.

This study utilized aggregated institutional data.

Different relationships may be revealed by studying salaries


within institutions.


A study of this type would allow the


use of different kinds of variables.

From the results of this study it may be concluded that

unionization was not a significant variable for inclusion


in this study as designed.


The small number of unionized


institutions in relation to the total caused the data to


be badly skewed, also.


Although unionization accounted


for a larger share of the variance in the doctoral level


equation it probably was not a significant variable

indicated by the small negative simple correlation.


It is


quite possible that no strong relationship was found

between salaries and unionization in this study because

compensation items other than salaries may be more affected


XT nn- nAA 1 1.. 4 -I A n n n 4 r' 4 n r^ n n I c, 4 p


L.







87

Factors other than salary may be affected by the


widespread use of adjunct faculty.


The flexibility


allowed by using less costly part-time staff may have


resulted in small student/instructor ratios.


Community


colleges have a propensity to serve community needs in


remote locations.


It may be that the use of adjunct


faculty allows this service.


It could be concluded that


further investigation is indicated with a different study

design.

The master's salary level equation was fairly strong


as shown by the multiple regression results.


The five


independent variables accounted for 41 percent of the


variance.


The standard error was $7.50; therefore, 68


percent of the time this equation would come within $7.50

of estimating the master's level salary.

The master's plus 30 salary level equation accounted

for 51 percent of the variation in the dependent variable

and 68 percent of the time would predict that salary

within $8.93.

The doctoral level salary equation accounted for 30

percent of the variation in the dependent variable DOC.


The standard error of estimate was $10.50.


In terms of


predictive power, salaries could be estimated within

.10O50- rnriind nercpnt tmnnfirlpncp intrxrnil








associated with two-year institutions.


The doctoral


level


equation probably was not strong because other variables


might


more closely related to that salary


level.


Two-


year institutions have


not


consistently sought instructors


with doctorate degrees.


Although the relationships between


the independent variables and salary


levels were not extremely


strong,


still relationships do exist


This may indicate


the need for a closer examination of the endogenous variables
and may be cause for a further study of disaggregated

institutional data rather than aggregated data of the insti-

tution.


Recommnendat ions


Those interested in further research of


community


college instructional salaries should consider the follow-

ing recommendations.


First,


the population for this study was limited to


one state with a cost based funding formula.


Conclusions


from this study may not be generalizable to states with


different methods of funding.


Further research may be


needed using other states with different funding formulas.


Second,


it is suggested that the widespread use of


adjunct faculty may be the basis for further study utiliz-


ing a different study design.


An average of 25 percent of









Third,


Florida


it is suggested that any salary study using

the population studied should consider the use


of subpopulations as determined by Full-Time Equivalent


students rather than grouping all sizes of


institutions


into one population.


Fourth, it is suggested that further study may be

indicated utilizing different measures of salary variables

since the doctoral level is not a salary category as


common to two-year institutions as it is to four-year

institutions.


Fifth, because of the small number of


institutions


which are unionized, further study may be indicated which

analyzes the affects of unionization within the institutions.


It may be possible to draw conclusions about the effects


of collect


ive


bargaining by comparing institutional data


before and after unionization.









APPENDIX A


1980-81 School Year


Survey Completed By:


Institution:

Contact Number:


Adjunct Faculty (Part-time) defined


as:


Non-Tenured, Part-Time


Instructional/Professionals

Percentage of classes taught by adjunct faculty in 1980-81



What is the established pay scale for adjunct faculty?


Is adjunct faculty compensated for years of experience with
institution? ___ _______.

Is there a limit to number of hours adjunct faculty may instruct?


If so, what is limit?


Do adjunct faculty receive fringe benefits? Yes
If yes, specifically define which fringe benefits


Do you have


a bargaining unit?


If yes, what is bargaining unit?


Are adjunct faculty a part of bargaining unit?


Are adjunct faculty required to have same educational training
as full-time faculty? Yes No


Are full-time faculty members paid at the adjunct Dav rates









there defined procedures
adjunct faculty? (other


and
than


criteria
student


evaluation


evaluations)


Give a


brief


explanation


have


a handbook


adjunct


faculty?


If y
this


Cs,


would


please


return


questionnaire.


a copy


the handbook with


Number of


adjunct


faculty


employees


1980-81 school


year.


Total

Total


dollars


spent


generated


on adjunct


adjunct


faculty

faculty


1980-81

1980-81














'-4


0



0

In
Cl






c';i
co
to
U?

U-)
Cl
0
a
CD

cv)




1'
Cs-
(0




CM

Ct,
(ft




(C'
S
ct~
0
'-4
'9


0,



0
r4
U?









REFERENCES


Annual FTE count by term 1980-81. Tallahassee, Florida:
State of Florida, Division of Community Colleges, 1982.


Biennial survey of education.
Printing Office, 1958.


Washington, D. C.:


Government


Birnbaum, R. Unionizati
Educational Record,


on and
Winter


acuity compensate ion.


1974, 55,


29-33.


Brenernan, D.


& Nelson,S. C.


An economic perspective.
Brookings Institution, 1


Financing cormmui ty colleges:
Washington, D.C.: The


.981


Brown,


& Stone, C. C.


Academic unions in higher


educate ion :


Imtp a c t


and promotions.
385-396.


faculty salary,


Economic Inquirl, Juy
__ __J u l


compensation
1977, 15,


Chambers, J. G.


The impact of collective bargaining for


teachers on resource allocation in public school


districts.
324-339


Clark, C.


Journal of Urbn Economics, June
a. J- n


& Schkade, L. L.


administrative decisions.
Publishing Co., 1979.


1977,


Statistical analysis for


Cincinnati:


South-Western


Clark, C.


T., & Stockton, J. R.


and economic statistics.
Publishing Co., 1971.


Introduction to business


Cincinnati:


South-Western


Cohn,


Methods of teachers'


remuner


and theoretical considerations.


action: Some empirical
1970 Proceedings of


the Business and Economic Statistics Section of the


American Statistical Association,
American Statistical Association,


Cohn, E.


1971,


Washington:


452-457.


Factors affecting variations in faculty salaries


and compensation in higher education.


Higher Education, February


1973,


Journal of


124-136.


Cohn, E.


The economics of


Massachusetts:


educate ion.


Cambridge,


Ballinger Publishing Co.,


1979.