An analysis of performance gap scores as measured by the Student Satisfaction Inventory : the
relationship to retention
by Gloria Jean Lambertz
A thesis submitted in partial fulfillment of the requirements for the degree Of Doctor of Education
Montana State University
© Copyright by Gloria Jean Lambertz (1998)
Abstract:
This study assessed whether a student’s decision to return or not return to Carroll College the following
semester could be predicted by comparing performance gap scores on 11 comprehensive scales of
institutional characteristics measured by the Student Satisfaction Inventory (SSI). Additionally, the
researcher examined eleven scales identified in the SSI (Academic Advising, Campus Climate,
Campus Life, Campus Support Services, Concern for the Individual, Instructional Effectiveness,
Recruitment and Financial Aid, Registration Effectiveness, Responsiveness to Diverse Populations,
Safety and Security, Service Excellence, Student Centeredness) for statistically significant differences
among class levels and gender.
A literature review yielded research relevant to this study in areas of retention, enrollment
management, consumer behavior principles, and student satisfaction.
The sample for the study was divided into two groups of students who responded to the SSI in the
spring of 1997. The first (n= 386) were a group of freshmen, sophomores, juniors, and seniors. The
second group (n=221) consisted of students who returned or did not return to Carroll College the
following semester and who provided a valid social security number. Data were collected and
performance gap scores were tabulated on 11 comprehensive scales of institutional characteristics.
Multiple regression and analyses of variance were the statistical methods used for the analyses of the
hypotheses. Newman-Keuls post hoc analysis was performed as a follow-up procedure for the
ANOVAS that showed significance at the .05 level.
Findings indicated that a prediction could not be made on the decision to return or not return to Carroll
College using the performance gap scores. Interaction was noted among class level and gender of the
students. Significant differences were found between male and female students on 9 of the 11 scales of
institutional characteristics. Also, significant differences were found among class level on 2 of the 11
scales of institutional characteristics. These findings provide information to Carroll College for making
institutional changes.
It was concluded that the SSI could not be used to predict whether the students’ decisions to return or
not return could be predicted by comparing performance gap scores. AN ANALYSIS OF PERFORM ANCE GAP SCORES AS MEASURED BY
THE STUDENT SATISFACTION INVENTORY: THE RELATIONSHIP
TO RETENTION
by
G loria Jean Lam bertz
A thesis subm itted in p artial fulfillm ent
of the requirem ents for the degree
Of
D octor of E ducation
; MONTANA STATE U N IV ER SIT Y -B O Z EM A N
Bozem an, M ontana
. A pril 1998
j>yre
APPROVAL
of a thesis subm itted by
G loria Jean Lam bertz
This thesis has been read by each m em ber of the graduate
com m ittee and has been found to be satisfactory regarding content,
E nglish usage, form at, citations, bibliographic style, and consistency, and
is ready for subm ission to the College of G raduate Studies.
Dr. Jqnn Kohl
C om m ittee C o-C hairperson
Date
3 -7
Dr. Ra^ph Brigham /7
C om m ittee C o-C haifperson
Date
A pproved for the M ajor D epartm ent
Approved for the C ollege of G raduate Studies
Date
iii
STATEM ENT OF PERM ISSION TO USE
In presenting this thesis in p artial fulfillm ent of the requirem ents
for a doctoral degree at M ontana State U niversity-B ozeman, I agree .
that the L ibrary shall m ake it available to borrow ers under rules of the
L ibrary. I fu rth er agree that copying of this thesis is allow able only
for scholarly purposes, consistent w ith “fair use” as prescribed in the
U. S. C opyright Law. R equests for extensive copying or reproduction
of this th esis should be referred to U niversity M icrofilm s
In tern ational, 300 N orth Zeeb Road, Ann Arbor, M ichigan 48106, to
whom I have granted “the exclusive right to reproduce and distribute
my d issertatio n for sale in and from m icroform or electronic form at,
along w ith the rig h t to reproduce and d istribute my abstract in any
form at in w hole or in p a rt.”
Signatur
D ate
iv
ACKNOW LEDGEMENTS
In m ost endeavors of this enorm ity, m any people are called upon to
o ffer th eir m uch-needed wisdom , advice, encouragem ent and support.
Such was the case w ith this effort.
The author wishes to acknow ledge w ith deepest gratitude the
follow ing people who gave of their tim e, expertise, guidance, and
encouragem ent in the com pletion of this thesis and doctoral program :
Dr. John Kohl, who served as my advisor and chair of my dissertation
com m ittee throughout the doctoral program , Dr. Ralph B righam who
stepped in to co-chair during the w riting of the thesis; Dr. Eric
Strohm eyer, for his w ealth of know ledge and patience w hile guiding
me through the research and sta tistica l process; Dr. L arry Baker, for
m aking tim e to answ er questions; Dr. Duane M elling who, when asked
to rep lace a com m ittee m em ber who retired , did so w ithout hesitating.
A big thank-you for the support of so many sta ff and faculty at
C arroll C ollege who “hung in th ere” w ith m e—too m any to name.
W ithout your caring support and reassurance, this paper would have
never been com pleted.
And fin ally to those very dear close friends who never doubted me
when I had doubts about m yself: I shall be forever grateful to you for
your unrequited love and support.
TABLE OF CONTENTS
LIST OF TABLES
ABSRACT
. Page
......vii
viii
I. IN TR O D U CTIO N .....................
Statement of the Problem..............
Significance of the Study...............
Definitions of Terms......................
General Questions to Be Answered
2. REVIEW OF LITERATURE
Introduction.........................................
Overview of Retention.......................
Overview of Enrollment Management
Overview of Smdent Satisfaction.......
3. M ETHODOLOGY
Introduction.........................................................
Theoretical Background......................................
Instrument Used..................................................
Reliability and Validity.......................................
Hypotheses..........................................................
Population Description and Sampling Procedure
Method of Data Collection..................................
Analysis of Data............ .....................................
Limitations and Dehmitations............................
39
39
39
42
46
47
48
49
49
53
4. FINDINGS AND INTERPRETATIONS
Introduction......................
Description of the Sample
Statistical Analysis..........
Hypothesis I ....................
Hypothesis 2....................
Hypothesis 3....................
Hypothesis 4 ....................
55
56
57
58
60
63
65
I
vi
TABLE OF CONTENTS - C ontinued
Page
5. SUMMARY, CONCLUSIONS, AND RECO M M EN D A TIO N S...............67
I n tro d u c tio n ................................................ .................... ...................;.......
67
Sum m ary of F in d in g s.......................................................................................... 74C o n clu sio n s..........................................................................................................••79
R ecom m endations ..........................................................
85
REFERENCES C IT E D ............... ....................................... ...............................
88
A P PE N D IC E S............................................
96
A ppendix A - Student S atisfaction In v e n to ry ............................................. 97
A ppendix B - Item s W ithin C om prehensive In stitu tio n al S c a le s ....... 103
)
Vii
LIST OF TABLES
Page
Table I
Sum m ary of Sam ple used in H ypotheses 2 Through 4 .............57
Table 2
R esults of M ultiple R egression A n a ly s is ..................................... 59
Table 3
Perform ance Gap Scores of R eturning and N on-R eturning
Students from the 11 In stitu tio n al C h a rac te ristic s.....................60
Table 4
Sum m ary of Perform ance Gap Scores Including tests for
In teractio n and M ain E ffe c ts........................................................... 61
T able 5
ANOVA for Interaction of Perform ance Gap Scores— Safety
and S e c u rity ........................................................................................
T able 6
Sum m ary of M ain E ffects T est for Campus Support Services
and R ecruitm ent and F inancial A id ...............................................64
T able 7
N ew m an-K euls Post Hoc Com parison for Campus Support
S e rv ic e s.....................................................
64
T able 8
N ew m an-K euls Post Hoc Com parison for R ecruitm ent and
F inancial A id ........................................................................................ 65
T able 9
M ain E ffects for M ales and F e m a le s.............................................. 66
I
viii
ABSTRACT
This study assessed w hether a stu d en t’s decision to return or not
retu rn to C arroll C ollege the follow ing sem ester could be predicted by
com paring perform ance gap scores on 11 com prehensive scales of
in stitu tio n a l ch aracteristics m easured by the Student S atisfaction
Inventory (SSI). A dditionally, the researcher exam ined eleven scales
id en tified in the SSI (A cadem ic A dvising, Campus C lim ate, Campus Life,
Campus Support Services, C oncern for the Individual, Instructional
E ffectiveness, R ecruitm ent and Financial Aid, R egistration E ffectiveness,
R esponsiveness to D iverse P opulations, Safety, and Security, Service
E xcellence, Student C enteredness) for statistically significant differences
among class levels and gender.
A lite ra tu re review yielded research relevant to this study in areas
of reten tio n , enrollm ent m anagem ent, consum er behavior principles, and
student satisfaction.
The sam ple for the study was divided into two groups of students
who responded to the SSI in the spring of 1997. The first (n= 386) were a
group of freshm en, sophom ores, ju n io rs, and seniors. The second group
(n=221) consisted of students who returned or did not return to C arroll
C ollege the follow ing sem ester and who provided a valid social security
num ber. D ata were collected and perform ance gap scores w ere tabulated
on 11 com prehensive scales of in stitu tio n al c h a rac teristics. M ultiple
reg ressio n and analyses of variance were the statistical m ethods used for
the analyses of the hypotheses. N ew m an-K euls post hoc analysis was
perform ed as a follow -up procedure for the ANOVAS that showed
significance at the .05 level.
F indings indicated that a p rediction.could not be m ade on the
decision to return or not return to C arroll College using the perform ance
gap scores. In teractio n was noted am ong class level and gender of the
students. S ignificant differences w ere found betw een m ale and fem ale
students on 9 of the 11 scales of in stitu tio n a l characteristics. Also,
sig n ifican t differences w ere found among class level on 2 of the 11 scales
of in stitu tio n a l characteristics. These findings provide inform ation to
C arroll C ollege for m aking in stitu tio n al changes.
It was concluded that the SSI could not be used to predict w hether
the stu d e n ts’ decisions to return or not return could be predicted by
com paring perform ance gap scores.
I
CHAPTER I
INTRODUCTION
Student retention can be studied from various perspectives,
including the benefits derived from student persistence. R etention of
students im pacts not only individuals and th eir personal success through
an educational outcom e, but also society, and the financial well being of
in stitu tio n s (Pate, 1993). W hen students becom e d issatisfied with their
educational process, they are likely to drop out of college (Bean, 1990b).
This d issatisfactio n leads to failure to m eet their educational goals.
A long w ith not m eeting th eir educational goals, students w ithout a college
degree earn a low er incom e (Bean, 1990b) than those students who
com plete a college degree.
Students are not the only ones im pacted when they do not m eet their
educational goals. Society is also affected. Society loses by not having
as m any educated individuals going into the work force (Bean, 1990b).
Finally, and very critical to those colleges w hose budgets are
tu itio n driven, not retaining students can have a significant im pact on
w hether th eir doors rem ain open (Bean, 1990b). It is im portant to look at
w hat w ill help satisfy students and increase the.likelihood of their
p ersistin g w ith a college education. Rowley, Lujan, and D olence (1997)
stated, “If a central purpose of higher education is to advance know ledge
2
w hile helping to create a b etter society, then colleges and universities
m ust respond to consum er dem ands in ways not yet perfected by present
p ractice” (p. 54).
In both student num bers and abundant budgets the 1950s through
the 1960s w ere boom years for higher education. As these boom years
gave .way to the retrenchm ent and dem ographic shifts of the 1970s and
1980s (Porter, 1990), in stitu tio n s put self-preservation as a top priority.
As the population increases in age and the pool of young college-age
persons continues to change, colleges and universities grow to be even
m ore concerned about com peting for new students. The 100-year national
average graduation rate of all college students is 50% (Boyle, 1989). As a
po in t of com parison, the overall graduation rate after five years of college
entrance is 44.6% (ACT, 1997). W ith the college dropout rate at 50%,
and w ith enrollm ent forecasting becom ing m ore d ifficu lt to predict
(H ealy, 1997), it is critical that colleges and universities look for ways to
retain th eir students. The study of student persistence, or retention, is not
new, but w ith an ever-changing student population from which to draw
new students, the subject of retention has taken on a new urgency.
C olleges and u niversities are increasingly com m itted to stabilizing
student enrollm ents by raising retention rates. The problem s with
reten tio n were com pounded by m istakes in college enrollm ent projections.
W hen unexpected econom ic and educational trends changed the projected
3
enrollm ent figures that w ere forecast by the W estern Interstate
C om m ission for H igher E ducation (1993), growth in student num bers at
fou r-y ear colleges did not m aterialize (H ealy, 1997). W hen those changes
did not m aterialize, the in stitu tio n s which are enrollm ent driven became
m ore vulnerable than others (B illson & Terry, 1987). W hen lim ited
resources are coupled w ith the cost of recruiting new students (W erth,
1988; H ossler, 1991) and the average drop-out rate of first-y ear students
at p riv ate colleges at 29.9% , private colleges and universities need to
analyze, thoughtfully, factors that w ill m axim ize retention at m inim um
expense. Since m ost p rivate college budgets are tuition driven (Townsley,
1993), reten tion of students is vital. W hile the “cost of recruiting is often
m easured in thousands of dollars, the savings from retention of a full-tim e
student can be m easured in tens of thousands of do llars” (Bean, 1990b, p.
147). One m ight argue that recruitm ent of non-trad itio n al students could
help fill this gap. H ow ever, the problem s associated w ith retention of
n o n -trad itio n al students in m any ways are the same as retention of the
trad itio n al student (H ealy, 1997).
T hree in terrelated factors have em erged over and over as predictors
of first-y ea r student persistence. These predictors are a felt sense of
com m unity, involvem ent of students in all aspects of the life at the
in stitu tio n , and academ ic/social in tegration (Tinto, 1993). Student
satisfactio n and how it relates to student p ersistence is becom ing a more
4
pow erful p red icto r that w ill aid institu tio n s in th e ir retention efforts
(Sanders & Burton, 1996).
W hen colleges address only a stu d e n t’s academ ics and fail to treat
them h o listically (Sanford, 1967), they fail to give the student a sense of
com m unity, w hich is T in to ’s first p redictor of first-y ear persistence.
Successful learning depends on the student's whole p ersonality, not ju st
in tellig en ce. It is the use of both attributes that helps to in tegrate the
student into the cam pus com m unity. Sanford (1967) said:
In m ost of our u niversities and in many of our liberal
arts colleges a m ajority of the students suffer from a
lack o f a sense of com m unity, confusion about values, a
lack of intim ate friends, a very tenuous sense of self
(including serious doubt about their personal w orth),
and the absence of a great cause, m ovem ent, service,
relig ion, b e lief system , or anything else that they m ight
see as larger than them selves and in which they could
becom e deeply involved, (p. 14)
Student involvem ent, T in to ’s second predictor of retention, is the
am ount of physical and psychological energy that a student devotes to his
or her college experience. The correlation betw een student involvem ent
in all aspects of life at the in stitu tio n is a predictor of success in the
freshm en year. Im proved retention through student involvem ent has been
researched and docum ented by many educators (Tinto, 1993; Astin, 1993,
Bean, 1982).
T into's research on departure, w hich studied the stu d en t’s social
and academ ic in tegration, a third p red icto r of retention, used the
5
fram ew ork of A rnold Van Gennep (1960), a D utch anthropologist. Van
G ennep, and now Tinto, divided the first year of college into three stages
or three "rites of passage." The first stage that Van Gennep identified is
rite of separation. In separation there is a decline or com plete cessation
of in teractio n s betw een the individual and m em bers of groups in which he
or she p reviously interacted.
Van G ennep’ s second stage is tran sitio n al rites. This stage is
ch aracterized by am biguity. In the tran sitio n al rite stage, the individual
begins to in teract w ith m em bers of the new group. H ere the student learns
the know ledge and skills necessary to function in the new group. In the
tran sitio n al rites stage the individual is challenged as to w hether he or she
w ill adjust to the dem ands o f higher education.
C erem onies or ritu als that sym bolically express m em bership into
his or her new position characterize Van G ennep1s last stage,
in co rp o ratio n. An exam ple of new m em bership may in clude becom ing a
m em ber of a fratern ity or sorority, or it may include belonging to one of
the m any clubs or organizations typically found on a college campus. The
duration of the phases differs from student to student and setting to
setting.
Students need sufficient form al m echanism s that w ill assure that
those social interactio n s with other students and faculty take place. T into
(1987) stated that there is not one grand theory to explain all the reasons
6
why a student w ould p ersist w ith his or her education, including his own
m odel. His m odel has becom e accepted as m ost useful for explaining the
causes of d eparture from college and em phasizes two conditions that help
p red ict student retention: social and academ ic integration. Tinto (1993)
reported th at for a student to persist in his or her education there needs to
be congruency betw een the student and the institution. The stu d en t’s
m otivation and academ ic ability m ust be m atched with the in stitu tio n ’s
academ ic and social characteristics in order for the student to be
com m itted to rem ain at the in stitu tio n . The in stitu tio n m ust also be
com m itted to the stu d e n t’s educational goals.
Som ew hat sim ilar to T in to ’s m odel is B ean's Student A ttrition
M odel. B ean’s m odel was developed by synthesizing research on turnover
in work organizations (Bean, 1980). He com pared reasons for students
leaving in stitu tio n s to sim ilar reasons for why em ployees leave work
organizations. As in work organizations w here intent is related to
behavioral in tentions, persistence in college is connected to behavioral
in tentions. B ehavioral intentions are those statem ents that students m ake
regarding educational goals. The processes in which b eliefs about
educational goals shape attitudes and attitudes guide behavior, are
referred to as behavioral intentions (Fishbein & Ajzen, 1975).
A stin (1984) stated that the am ount of student learning and personal
developm ent is d irectly proportional to the quality and quantity of student
7
involvem ent. For exam ple, a student who is highly involved w ill devote
energy to studying, to cam pus organizations, and to interactions with
other students and faculty. It is im portant that colleges and universities
aid the student in finding these opportunities for involvem ent.
Involvem ent can occur through social and academ ic integration. In
A stin ’s Theory of Involvem ent, rather than stressing the need for full
in teg ratio n , he believes that students can be alienated in som e areas of the
cam pus experience, but still persist because they are involved in another
area such as athletics or m usic. A stin believes that the b etter the fit
betw een the student and the in stitu tio n , the m ore likely the student is to
p ersist or achieve (A stin, 1985). A stin also believes, how ever, integration
is a m uch m ore com plex phenom enon than "fitting in." He believes
in teg ratio n requires a series of connections with various aspects of the
college environm ent, which m ake it a com m unity.
The foregoing three theories have the common them e of
involvem ent as a p redictor of student persistence. In addition to
involvem ent, there are other variables that are predictors that affect
academ ic and social in tegration. These variables include the follow ing:
high school GPA, financial aid support, and gender. Studies have been
conducted m easuring these variables (P ascarella & T ere n zin i, 1980,
Cabrera, Nora, and Castaneda, 1993; Kennedy, 1995), and these variables
have also been found to be predictors of persistence. It is essential that
8
the relatio n ships am ong the variables be established with clarity by
looking at factors unique to an individual institu tio n . Noel, Levitz, and
Saluri (1985) stated that retention studies should define the unique
ch aracteristics of the cam pus. B aselines m ust be established for
m easuring gains in reten tio n and identifying strengths of the institution.
It is from these baselines that in stitu tio n s w ill be able to m ake in telligent
decisions when assessing th eir program s.
Students enter college w ith a set of expectations, and it is from
these expectations that they are m otivated (H odgetts, 1982; Bank, Biddle,
and Slavings, 1992). Given the stu d en ts’ investm ent of tim e, energy, and
m oney, th eir perception or expectancy should be given substantial w eight.
A stin (1993) w rote: “V irtually all m easures of satisfaction w ith the
undergraduate experience are significantly related to the num ber of
undergraduate years com pleted” (p. 278). R esearch has shown (Astin,
1993; T in to , 1993; Bean, 1980) that academ ic and social integration of the
student into college life are strong indicators for retention. The findings
of B raxton, V esper, and M ossier, (1995) also indicated th at m eeting
student expectations positively influences both academ ic and social
in teg ratio n and becom es im portant to those concerned about retention.
B raxton, V esper, & M ossier (1995) said, “If such expectations are not
m et, then there is early disenchantm ent w ith these com m unities” (p. 596).
B raxton, V esper, and M ossier (1995) also stated, “Put differently, students
9
w ith unm et expectations for college are unlikely to becom e integrated into
the academ ic or social com m unities of the institution because they
perceive th at they w ere m isled by the in stitu tio n p rior to m atricu latio n ”
(p. 596).
Statem ent of the Problem
The problem of this study was to determ ine if the Student
S atisfaction Inventory (S chreiner & Ju illerat, 1994), and the “perform ance
gap scores” derived from the data, could p redict w hether or not a student
returned to C arroll C ollege the sem ester after com pleting the survey.
Perform ance gap scores were calculated. They are based upon the
stu d en t’s perception of the difference betw een his or her level of
expectation and his or her level at w hich that expectation was met, called
“sa tisfac tio n .” E leven d ifferent perform ance gap scores were calculated,
each reflectin g a com prehensive scale of in stitu tio n al characteristics.
A secondary investigation attem pted to gain a m ore thorough
understanding of perform ance gap scores. For each of the eleven
com prehensive scales, the differences betw een the m eans o f gender and
class level were studied. Tests for in teractio n were also investigated.
For purposes of this study, data regarding students at Carroll
C ollege, a four-year private lib eral arts college, were collected. There
were 12 com prehensive scales included in the instrum ent, but for purposes
of this study only 11 of the 12 scales were used in the analyses of data.
10
The 12 com prehensive scales that m easure in stitu tio n al effectiveness
include the follow ing: A cadem ic A dvising E ffectiveness, Campus
C lim ate, Campus Life, Campus Support Services, Concern for the
Individual, Instructional E ffectiveness, R ecruitm ent and F inancial Aid
E ffectiveness, R egistration E ffectiveness, R esponsiveness to D iverse
P opulations, Safety and Security, Service E xcellence, and Student
C enteredness.
The tw elfth scale, R esponsiveness to D iverse Populations was the
only scale which exclusively m easured satisfaction.
Im portance was not m easured due to the sensitive nature of the
question w ithin the scale and the im plications of asking such questions to
a m ajority population (Juillerat, 1995).
Significance of the Study
In response to the unpredictable num ber of graduating high school
seniors that attend college and the high cost of recruiting students—from
$1800 to $2400 per student (C. Cain, personal com m unication, A ugust 22,
1997) colleges and un iv ersities are increasingly com m itted to achieving
enrollm ent stab ility by increasing the student retention rate. The national
average for reten tio n of first-y ear college students at private institutions
is 70.1% (ACT, 1997). C arroll C ollege has a 75% reten tio n rate of firstyear students, w hich is slightly higher than the national average. C arroll s
fo ur-year graduation rate is 30%, and five-year graduation rate is 42% of
11
entering freshm en. In spite of the fact that C arroll C ollege s retention
rate of first-y ea r students is slightly higher than the natio n al average, the
adm inistration is striving for a goal of an 80% retention rate for first-y ear
students and a four-year graduation rate of 65% (R. Pastoor, personal
com m unication, Septem ber 5, 1997). These num bers m eet a level which
the in stitu tio n deem s financially feasible and acceptable in order to reach
its long-term goal for grow th of the in stitu tio n . This rate of retention is
also based on a com parison of colleges sim ilar in size and m ission to that
of C arroll C ollege.
C arroll C ollege depends heavily upon student tu itio n to fund its
operating budget. In a recent study of 1,000 colleges, T ow nsley (1993)
found th at 73% drew th eir revenue from students. Seventy of these
private colleges drew 94% of th eir revenue from student tu ition
(T ow nsley, 1993). C arroll C ollege is one of those enrollm ent-driven
p riv ate colleges and derives 85% of its budget from student tuition.
C arroll C o lleg e’s dependence upon student tuition, coupled w ith the high
cost of student recruiting, m eans that it is crucial that C arroll College
m aintain a high student retention rate after the freshm an year. In an era
of increasing college costs and less predictable num bers of college-bound
high school graduates, it is also very im portant that C arroll College learn
w hat expectations its students have regarding their educational
experiences and to what degree these expectations are being satisfied.
12
A ssessing student expectations and satisfaction is an im portant
m easurem ent of the needs and desires of students that in stitu tio n s m ust
address in order to understand student success, satisfaction, retention, and
in stitu tio n a l effectiveness (Schm idt & Sedlacek, 1972; A stin, 1993; Bean,
1980; Spanbauer, 1992; and T in to , 1993). Earw ood-Sm ith and Colbert
(1989) found that as student satisfaction increased, so did the rate of
student retention. R egularly assessing a broad range of student outcom es
should lead to continuous im provem ent in student and academ ic program s
w hich, in turn, w ould lead to greater student satisfaction and greater
student retention. Student satisfaction is a valid gauge of retention
(Sanders & Burton, 1996). E arw ood-Sm ith and C olbert (1989) stated, “A
m ore productive research approach is to exam ine why college students
stay instead of why they leave” (p. 14).
W hen applying for grants and other financial aw ards, institutions
are frequently required to dem onstrate that they are retaining their
students in high num bers (How ard & R ogers, 1991). As part of receiving
accred itatio n renew al, institu tio n s m ust dem onstrate that student
assessm ent is a part of the in stitu tio n ’s procedure (D olence, 1991).
C arroll C ollege is preparing for its accreditation review which will take
place in 2000.
For some years now C arroll C ollege has been concerned with its
reten tio n num bers. In the past, inform ation as to why students left C arroll
13
C ollege has for the m ost part been speculative. W ith concrete data gained
from a student expectation and satisfaction assessm ent, C arroll College
can use its lim ited resources carefully and wisely, getting the m ost out of
a strained budget. A lso, w ith current data on student expectation and
satisfactio n , C arroll C ollege can re-evaluate its short-term and long-term
planning goals based on what students expect from th eir educational
experience and how satisfied they are w ith that experience. Earw oodSm ith and C olbert (1989) suggested that looking at stu d e n ts’ “greatest
satisfactio n and greatest dissatisfactio n is ,the basis for reordering
in stitu tio n al p rio ritie s’’ (p. 14). C abrera, Nora, and C astaneda (1993)
have said: “In stitu tio n s should constantly m onitor w hether an in stitu tio n ’s
in terv en tio n plan is having an effect on the persistence p ro cess” (p. 125).
A ssessm ent data w ill allow C arroll C ollege to look system atically at ways
to m ake changes to its educational and cam pus life program s that w ill
eventually aid in the retention of its students and provide inform ation for
other in stitu tio n s to u tilize. C haffee (1990) suggested using data to
“quickly m ove from dissatisfactio n to so lu tio n ” (p. 61).
D efinitions of Terms
The follow ing definitions explain term s as they are used in this
study:
r
I.
A cadem ic In teg ratio n —a condition or perception of fitting
into and being able to keep up with the academ ic rigor of the
14
in stitu tio n .
2.
A ttritio n —the voluntary or involuntary student departure
from an in stitu tio n , not including graduation.
3.
C om prehensive Scales—found w ithin the Student S atisfaction
Inventory, there are tw elve com prehensive areas: Academ ic
A dvising E ffectiveness, Campus Clim ate,
Campus Life,
Campus Support Services, C oncern for the Individual,
Instru ctio n al E ffectiveness, R ecruitm ent and Financial Aid
E ffectiveness, R egistration E ffectiveness, R esponsiveness to
D iverse P opulations, Safety and Security, Service E xcellence,
and Student C enteredness. Each of these areas consists of a
set of related questions.
4.
Im portance Score R atings—m easurem ents of how strongly
students feel about the expectation for the given item ,
referred to as level of expectation.
5.
N egative Gap Score—the im portance score rating minus
student satisfaction score rating resulting in a negative
num ber. The negative num ber indicates the in stitu tio n is
surpassing the stu d en t’s expectation.
6.
Perform ance Gap Score—the im portance score rating m inus
student satisfaction rating.
15
7.
P ersisten ce—the flow of students through college over a sixyear period.
8.
R etention—the continuation of a stu d en t’s education from
year to year.
9.
S atisfaction Score—m easurem ent of how satisfied students
are that the in stitu tio n has m et their expectation.
10.
Social In teg ratio n — a condition or perception of fitting into
the social life of the cam pus.
11.
Student S atisfaction Inventory (SSD -d e v e lo p e d by Laurie A.
Schreiner, P b.D., and Stephanie L. Ju illerat, Ph.D. (1994),
w ith assistance from USA Group N oel-L evitz. The assessm ent
m easures expectations of students and the satisfaction of
students on tw elve com prehensive scales.
The survey also provides a m eans to m easure the gap betw een
expectations and satisfaction.
G eneral Q uestions to Be Answered
1.
Is there a relationship betw een the decision to return or not
return to C arroll C ollege and any of the perform ance gap
scores derived from the Student S atisfaction Inventory (SSI)?
2.
In w hich of the 11 perform ance gap scores can interaction
betw een gender and class level be found?
16
3.
In. which of the 11 perform ance gap scores are there
significant differences betw een the m eans am ong the
freshm en, sophom ores, ju n io rs, or seniors?
4.
In w hich of the 11 perform ance gap scores are there
significant differences betw een the m eans of m ale and fem ale
students?
;
17
CHAPTER 2
REVIEW OF LITERATURE
Introduction
T here has been m uch w ritten about retention in higher education in
the past tw enty-five years. A ccording to the N ational C enter for
E ducation S tatistics (NCES, 1997), retention continues to be an issue of
im portance for colleges and universities as the num ber of students
available to attend postsecondary schools continues to flu ctu ate (1997).
A review of the literatu re yields a w ealth of inform ation on the topic.
This lite ra tu re review w ill focus on three m ain areas: first, an overview
and review of literatu re related to the topic of retention, second, an
overview of research related to enrollm ent m anagem ent, and third, an
overview of student satisfaction.
O verview of R etention
A review of the literatu re on the im portance of the study of
reten tio n and the im pact that those resu lts have for in stitu tio n s doing
research shows that certain com m onalties reoccur w ithin the theories or
m odels o f retention. The common them es are academ ic integration and
social in teg ration. T here are various theoretical m odels th at provide a
com prehensive fram ew ork on college departure decisions (A stin, 1984;
• Bean, 1982; T in to , 1993).
18
For this review of literatu re, three of these m odels were exam ined,
T in to ’s Student Integration M odel, B ean ’s Student A ttrition M odel and
A stin ’s Theory of Involvem ent.
T into's Student Integration m odel has becom e accepted as one of
the m ost useful m odels for explaining the causes of departure from
college. Two factors that help predict student retention, social and
academ ic in tegration, are the m ain factors discussed in. the T into m odel
(B oyle, 1989; T into, 1993). These two factors becom e a m eans for
discussing and testing the m odels for college w ithdraw al (Stage, 1989).
The factors, academ ic integ ratio n and social integration, are som etim es
referred to as cognitive and affective conditions (Stodt, 1987).
T into's Student Integration M odel is based on the work of A rnold
Van Gennep (1960), a D utch anthropologist, and E m ile D urkheim (1951),
a French so ciologist. Van G ennep’s work looked at a culture s rites of
p assag e.” The c u ltu re ’s ritu als were designed to move an individual from
one developm ental stage to the next. There were three stages to Van
G ennep’s rites of passage. The first stage that Van Gennep identified was
rite of separation. In rite of separation there is a m arked reduction or a
com plete cessation of in teractio n betw een the individual and m em bers of
the group in which he or she had previously interacted. Van G ennep’s
second stage was tran sitio n al rites, w hich is characterized by am biguity.
In this stage the individual is tested to see how he or she adjusts to a new
19
environm ent, leaving behind fam iliar relatio n s and recognized patterns.
Van G ennep’s third and final stage in “rites of passage” was
incorp o ratio n. This stage is characterized by a series of rituals and
cerem onies that sym bolically express the new position in the new
“c u ltu re.”
The duration of these three stages differs relative to the individual
and the situation. W hen the in stitu tio n does not provide college students
the opportunity to com plete the three stages, they w ill m ost likely drop
out of school (T into, 1993). T in to ’s com parison was those in stitu tio n s’
function as “societal rites of passage.” He stated that the m ore integrated
students becam e in the tran sitio n stages the m ore likely they were to
p ersist w ith their education (Tinto, 1993).
A nother influence on the developm ent of T in to ’s m odel was
D urkheim ’s (1951) research on suicide. D urkheim found that when a
person shares values w ith a group, that person is less likely to commit
suicide. He also found that when an individual has friends for support
th at the person is less likely to com m it suicide. From D urkheim s
research, Spady (1970) m ade the com parison of students w ithdraw ing
from college to people w ithdraw ing from society (Bean, 1982).
T into's m odel exam ines the processes that m otivate individuals to
leave colleges and un iv ersities before graduating (C abrera, Nora, &
C astaneda, 1993). W ithin social and academ ic integration there is a range
20
of form al and inform al experiences of students. Some of these form al and
inform al experiences are determ ined by the institution, and some are
determ ined by the student. Tinto (1993) m aintained that these form al and
inform al experiences are both c ritical to long-term success and
persisten ce in college.
Tinto (1993) reported that for a student to persist in his or her
education there needs to be congruency betw een the student and the
in stitu tio n . The higher the stu d en t’s goal to com plete college the greater
the p ro b ab ility of p ersistin g in college (C abrera, Nora, & C astaneda,
1993). The student's m otivation and academ ic ability m ust be m atched
w ith the in stitu tio n ’s academ ic and social characteristics in order to help
shape two underlying com m itm ents: com m itm ent to an educational goal
and com m itm ent to rem ain with the in stitu tio n (H odgetts, 1982; Cabrera,
Nora, & C astaneda, 1993). The in stitu tio n m ust also be com m itted to the
student's educational goals.
The second m odel to be review ed that explains retention is B ean's
Student A ttrition M odel. This m odel is sim ilar to, yet d ifferen t from, the
T into m odel. B ean’s m odel was developed from the resu lts of studying
turnover in work organizations (Starr, B etz, & M enne, 1972; Bean, 1980).
Em ployee turnover is chiefly a resu lt of the work environm ent. When
em ployees are satisfied w ith the work environm ent they w ill m ost likely
not quit (Bean, 1980).
21
B ean ’s m odel asserts that p ersistin g in college is analogous to job
turnover in work organizations, and that it is im portant to look at the
stu d e n t’s behavioral intentions. B ehavioral intentions are m olded by a
process in w hich beliefs shape attitudes and attitudes influence behavior
(Fishbein & Ajzen, 1975). A process in which beliefs about educational
goals shape attitude, and attitudes guide behavior directs these behavioral
in tentions. The stu d en t’s beliefs are affected by the stu d en t’s experience
w ith the d ifferent com ponents of an in stitu tio n (C abrera, Nora, &
C astaneda, 1993). Stodt (1987) rem arked that students m ust be convinced
as consum ers to buy m ore than a sem ester at a tim e, to buy a four-year
contract. The student needs to recognize how his or her investm ent is
paying off. The level of satisfaction is expected to increase the level of
in stitu tio n a l com m itm ent. If there is not enough in stitu tio n al
com m itm ent, the student w ill drop out of school (Bean, 1980). Bean
(1986) found that in stitu tio n al com m itm ent was the prim ary variable
influencing dropout. H ossler (1994) found that goal com m itm ent and
in ten t to leave are the strongest p redictors of dropping out.
The Student A ttrition M odel recognizes those external factors, such
as fam ily support and fin an cial stab ility , can play a m ajor role in
affectin g both attitudes and decisions w hile still attending college
(C abrera, Nora, & C astaneda, 1993). The Student A ttritio n M odel also
acknow ledges that the stu d en t’s b eliefs, and thus the student s attitudes,
22
are subject to factors external to the in stitu tio n . The external factors have
been studied and tested to m easure the relevancy that they have on a
stu d en t’s p ersistin g in college.
The final m odel used to study retention is A stin ’s Theory of
Involvem ent (1977). The roots for this m odel come from a longitudinal
study of m ore than 200,000 college dropouts. A stin iden tified several
factors in the college environm ent that influenced stu d e n ts’ decisions to
stay or leave college.
A stin id entified over 80 d ifferent student outcom e m easures
(variables). His m odel focused on the effects of various types of
involvem ent including w here the student lived; honors program s;
academ ic involvem ent; m ajor; athletic p articipation; student-faculty;
student-student; and involvem ent in student governm ent. A stin (1984)
found th at the variable that contributed to persistence suggested
involvem ent in the cam pus com m unity w hereas the variable that
contributed to dropping out im plied a lack of involvem ent.
Common them es that occur in T into s and Bean s m odels that are
closely related to retention are academ ic integration or perform ance and
social in teg ration. T in to ’s concept of integration is the m ore
com prehensive of the two (H ossler, 1994). R etention research shows a
high co rrelation betw een academ ic perform ance and p ersistence (Spady,
1970). A cadem ic integ ratio n occurs through faculty contact such as
23
teaching, academ ic advising, or in stitu tio n al involvem ent such as
providing social events. M ost student-faculty contact occurs w ithin
academ ics. H ow ever, evidence shows that student-faculty contact outside
the classroom is also im portant (P ascarella & T erenzini, 1991; T into,
1993). C ontact w ith faculty comes about in form al and/or inform al
contacts (Stodt, 1987; Tinto, 1993), Form al contact occurs through
teaching and academ ic advising. In stitu tio n s m ust support teaching and
academ ic advising so th at academ ic integ ratio n of the student occurs.
C urricula and courses m ust provide m aterials and inform ation relevant to
the stu d en t’s educational goals. One way colleges can help students
become aware of what their educational goals are is by offering some type
of course, such as a freshm an sem inar course, that w ill aid the student in
gaining th at know ledge (U pcraft & G ardiner, 1989).
Since m ost form al contacts supporting academ ic integration are
ones th at occur in the classroom , in stitu tio n s m ust place a prio rity on the
quality of student-faculty interactions (P ascarella & T erenzini, 1991;
T into, 1993). Institu tio n s should m aintain quality by challenging their
facu lty to exam ine th eir current teaching m ethodology to see how the
m ethodology fits the current needs of to d ay ’s student (Stodt, 1987,
P ascarella & T erenzini, 1991; Sines & D uckw orth, 1994). Faculty m ust
consider various teaching techniques and m ethods for delivery of
inform ation (G ardiner & N azri-R obati, 1983). Rowley, L ujan, and
24
D olence (1997) stated, “C olleges and u niversities are no longer the
gatekeepers— they no longer hold the only way of creating and
dissem inating inform ation, let alone learning from it” (p. 23). An active
mode of delivering m aterial to students is considered appropriate. It is
also im portant for faculty to teach students to becom e responsible for
th eir learning. Students m ake judgm ents about their academ ic experience
on the basis of factors such as quality of instruction, freedom to contact
faculty for consultation, availab ility of faculty for consultation, and
faculty involvem ent outside the classroom (Noel, 1994).
A nother type of form al contact m ade betw een student and faculty
th at supports academ ic in tegration is academ ic advising of students by
faculty m em bers. Faculty advising has long been considered a part of the
facu lty m em ber’s job description. A cadem ic advising is also considered
an in teg ral part of the higher education process (C rockett, 1978). Good
academ ic advising by faculty has a strong link to student persistence
(C rockett, 1978; Stod't, 1987). W hen a student feels good about the
academ ic advising he or she receives, the student also feels, good about
the in stitu tio n . A cadem ic advising can stim ulate in te lle ctu al growth and
developm ent of the student and can also enrich the educational curricula
of the school. It has been shown that frequent and high quality contact by
faculty w ith advisees results in student persistence (C rockett, 1978).
Inform al interactio n w ith faculty also plays an im portant role in
25
reten tio n . Inform al contacts w ith faculty have a role in the total quality
educational experience. Exam ples of inform al interaction include casual
conversation betw een classes and conversation at school events. These
types of social contacts w ere consistently found to be linked positively to
freshm an year persistence (T erenzini & Pascarella, 1980; A stin, 1984;
T into, 1993). Not only has inform al contact been found im portant,
frequency and quality of inform al contact played a role in attrition
decisions. Sines and D uckw orth (1994) noted, “W hat determ ines a quality
education is m ore than ju st w hat takes place in the classroom ” (p. 3).
There is also evidence that inform al contact is som ewhat m ore im portant
for women students than men students (T erenzini & P ascarella, 1980).
W ith the a ttritio n rate for first-y ear students at 29.9% (ACT, 1997)
for private four-year in stitu tio n s, the beginning of the freshm an year is
critical to retention (ACT, 1997). C olleges and u niversities need to
strongly encourage contact by faculty and staff with students in the
freshm an year. M any colleges offer freshm en sem inar courses as one
m ethod of m aking this investm ent. T hese courses help the student to
in teg rate academ ically and socially into the institution. The m ore we
concentrate resources on a freshm an's beginning experience, the greater
the likelihood that the freshm an w ill p ersist with his or her education
(Dunphy, 1987). W hen freshm en take orien tatio n courses, the student
experiences a greater sense of com m unity, becom es m ore com fortable
26
talking w ith his or her professors, and gains a greater understanding of
w here and how to get help (G ordon, 1989; U pcraft & G ardiner, 1989).
Further, frequent absences from the freshm an course serve as an early
w arning of likely a ttritio n (Dunphy, 1987).
R esearch has shown that student participation in orientation courses
resu lts in increased retention rates, increased student know ledge, and
increased p articip atio n in extracurricular activities (Perigo & U pcraft,
1989; U pcraft & G ardiner, 1989).
V ariables in the T into and Bean m odel serve as a standard against
w hich other research can be m easured. M any research efforts currently
are iso latin g one variable and m easuring that variable against one of the
m odels.
P ascarella and T erenzini (1991) looked at the variables of inform al
faculty and student contacts. W hen com paring their resu lts to the Tinto
m odel, we see a significant difference betw een persisters and leavers
(P ascarella & te re n z in i, 1991). This difference betw een persisters and
leavers held true when students discussed intellectual or course-related
subjects w ith faculty outside of the classroom (P ascarella & T erenzini,
1991).
In other studies done to test the valid ity of T in to ’s m odel, T erenzini
and P ascarella (1980) developed a m ultidim ensional instrum ent. T heir
resu lts indicated strong support for both the instrum ent and the m odel,
27
especially validating the constructs of academ ic and social integration
(Boyle, 1989). G etzlaf, Sedlacek, K earney, and B lackw ell (1984) also
tested against T in to ’s m odel, and in their findings found GPA to be a
sig n ifican t p redictor of persisten ce (Boyle, 1989).
R etention strategies that work for som e students may not work for
other students. A dditionally, strategies that work for som e institutions
may not work for other in stitu tio n s. P atterns of results from studies of
reten tio n can lead in stitu tio n s to various solutions. It is im portant,
how ever, for the in stitu tio n to question the student about academ ic
in teg ratio n and social integration concerns before the student leaves. Any
in terv en tio n for retention should occur w hile the student is enrolled, not
after he or she leaves.
O verview of E nrollm ent M anagem ent
W hile studying theories of retention is useful to understanding why
students drop out of college, they do not explain attritio n in its entirety.
E nrollm ent M anagem ent has two aspects. The first aspect studies
recru itin g students; the second aspect is the study of retention. For
purposes of this paper, I w ill look at only the retention aspect of
enrollm ent m anagem ent.
As p art of the study of reten tion, it is im portant to exam ine why
students rem ain in college. To do so we m ust consider the reasons
associated w ith why students return to college year after year. These
28
reasons involve consum er behavior principles. First, the problem s related
to retention as a part of enrollm ent m anagem ent with special em phasis on
consum er behavior w ill be addressed and second, student satisfaction w ill
be addressed.
E nrollm ent practices becam e m ore form alized starting back in the
early 1970s (Pike, 1991) when com petition for students began to increase
as student num bers available for college w ere predicted to drop. At the
sam e tim e, there was also an increased em phasis on quality instruction
and service to students. This trend came about in part as a result of
students becom ing m ore vocal in expressing their dissatisfactio n with
som e aspects of cam pus policies such as student life and adm inistrative
p o licies. Student protests becam e m ore common on college cam puses,
and the demand for student services and quality education becam e more
p rev alen t (Schm idt & Sedlacek, 1972).
As a resu lt of dem ands for increased accountability and outcom e
m easures, in stitu tio n s needed m ethods for m easuring these factors.
E nrollm ent m anagers in higher education borrow ed consum er principles
from the business industry in order to aid in this task. W hile institutions
can not to tally m ove from a “p ro v id er” m odel to a “consum er” model, the
in stitu tio n m ust respond to consum er dem and (Rowley, L ujan,
1997).
Sc
D olence,
29
One industry theory that enrollm ent m anagem ent borrow ed is the
expectancy theory (Cum m ings & Schwab, 1973). E xpectancy theory
explains m otivation behind w orkers in industry, which can be com pared to
student m otivation. Expectancy theory predicts that a w orker will be a
high perform er when he or she sees that work effort w ill have some kind
of an effect, w hether from a “pat on the back” or a pay increase. In
expectancy theory, job perform ance is also related to job satisfaction.
W hen w orkers are satisfied w ith their jo b s, they “perform
at a higher
level. E xpectancy theory can be com pared to stu d en ts’ satisfaction w ith
th eir education. The m ore satisfied students are with the educational
experience, the m ore likely it is that students w ill continue w ith their
education.
By the late 1980s, dem ographers told us that there w ould be a
decline in the num ber of high school seniors going on to post-secondary
schools (Porter, 1990). In some states in fact enrollm ent decrease at fouryear colleges did not m aterialize, and in some states there has been a
greater demand for higher education (Healy, 1997), Next came the
p red ictio n that there w ould be dram atic increases in student num bers.
W ith the recent release from NCES (1997), the new forecast for p o st­
secondary schools takes on yet a d ifferen t look. E nrollm ent forecasting
has become more complicated by several trends. These trends include the
follow ing: high school students are taking college classes w hile still in
30
high school; there is a heavy demand for classes at com m unity colleges by
n o n -trad itio n al students; and there have been changes in federal and state
fin an cial aid policies. O ther trends that affect enrollm ent are growth in
distance learning and a good econom y w hich attracts prospective students
to jo b s instead of going to school (H ealy, 1997). Students are m aking
differen t choices about th eir post-secondary education from those
anticipated in earlier projections. W ith so many factors influencing
enrollm ent trends, it is nearly im possible to generalize national trends to
ind iv id u al states, or state trends to the nation as a w hole (W estern
In terstate C om m ission for H igher E ducation, 1993). Som ewhat easier to
p red ict than enrollm ent are the benefits of a stu d en t’s staying in college
and the cost of attritio n when a student decides to leave school:
It is m uch easier in tim e, m oney, and effort to retain students than
to recru it them . If a student drops out after the first year, it means the
loss of three years of tu itio n (Bean, 1990b). It would take four freshm en
who quit after one year to equal the incom e of one student who stays for
fo u r'y ears. Concern about the retention rate of students is not lim ited to
freshm en. The cost of losing students can not and should not be m easured
only in dollars. A ttritio n of students is a problem for m ore than ju st the
in stitu tio n that loses the student.
A ttritio n of college students is also a problem for society. W hen a
student chooses not to fu rth er his or her education, this decision creates a
31
reduction of educated individuals for our society. The individual loses by
dropping out of school is the sam e in that there is personal failure to
achieve his or her educational objectives. The student who drops out of
college is also at risk of receiving an incom e at 15% below that of
contem poraries who stay in college (Bean, 1990b). R etaining students
w ith p o sitiv e attitudes tow ard the college or university m eans that the
student w ill re-enroll. Upon graduation “alum ni” will speak favorably
about the in stitu tio n w hich w ill aid in fu rth er recruiting new students
(Bean, 1990a).
Society and the student lose out when the student drops out of
college, and the college or university fails to collect tu itio n money. The
in stitu tio n surrenders m ore than ju st dollars. Faculty m orale is affected
by a ttritio n (Bean, 1986) as is the m orale of adm issions counselors who
work hard to recru it a new students. A dm ission counselors expend m uch
tim e, money, and energy recruiting students, and at tim es th eir jobs are in
jeo p ard y when they do not m eet th eir quotas (Ferguson, 1990).
In addition to theories on student departure, are two concepts
related to student retention: enrollm ent m anagem ent, w ith the construct of
consum er behavior and student satisfaction. These two concepts are an
in teg ral p art of T in to ’s, A stin ’s, and B ean ’s work.
The study of consum er behavior comes from m arketing principles,
inclu d in g expectancy theory. Consum er behavior studies show how to
32
get, and keep, custom ers. W hile som e educators argue that m arketing is
incom patible w ith the educational m ission and does not fit in higher
education, in stitu tio n s can use m arketing know ledge to explain student
satisfactio n and attritio n concerns. Row ley, Lujan, and D olence (1997)
w rote: “If a central purpose of higher education is to advance know ledge
w hile helping to create a b etter society, then colleges and universities
m ust respond to consum er dem and in ways not yet perfected by present
p rac tic e ” (p. 54).
Consum er expectancy can .be explained through V room ’s Theory of
E xpectancy (1960). V room m ade the assertion that consum ers form
expectations about product perform ance before they buy. In purchasing
an education, w hich is considered to be an “intangible good,
(Levitt,
1981) students “purchase” their college education based on w hat they see
and hear about the in stitu tio n . From the b u y er’s (stu d en t’s) point of view ,
the product (education) is a prom ise, and the custom er (student), although
sold, can ju st as quickly be unsold if expectations are not m et (Levitt,
1981).
D epending on the level of satisfactio n with the purchase,
called
post-purchase action, this satisfaction can have many p ositive benefits for
the in stitu tio n as w ell as som e harm ful effects. W hen the consum er is
satisfied he or she w ill “repurchase” the service (take another sem ester of
classes), he or she w ill recom m end the school to other custom ers (children
33
and frien d s), and they w ill contribute to alum ni associations (Stodt, 1987;
Pate, 1993). If on the other hand, the consum er is d issatisfied with the
service, he or she may speak poorly about the product (Pate, 1993).
O verview of Student S atisfaction
W hile m odels from the business industry can explain expectancy
theories, they do not entirely explain student satisfaction, the other
p rin cip le related to retention. We know from the literatu re on student
reten tio n th at recruiting students is expensive (Rowley, H erm an, &
D olence, 1997). It was not u ntil the 1980s when studies show ed that
increasing student satisfaction had a po sitiv e effect on retention. Astin
(1993) found in his studies on retention that there “appears to be a direct
association betw een student satisfactio n and retention (p. 278).
G ardiner and N azari-R obati (1983) and Earw ood-Sm ith and C olbert
(1989) suggested that researchers look at attritio n in a d ifferent light.
R ather than studying why students drop out of college, research must
focus on the p ersister and retention rates. The quality of student life m ust
be a cam pus-w ide concern. Sanders and B urton (1996) said researchers
should be asking about w hat kind of experience the in stitu tio n provides to
its students, and w hat the level of satisfaction is with th eir experience.
Of p a rticu la r im portance, faculty need to believe that retention is part of
th eir jo b (Toy, 1985). In a highly com petitive m arket, in stitu tio n s m ust
continually assess how satisfied students are if they hope to retain them .
34
Sanders and B urton (1996) stated, “A lthough strongly connected to
reten tio n , student satisfactio n is a m ore pow erful m easure that can
continue to be im proved even in institu tio n s with high retention and
graduation ra te s” (p. 556).
In the 1970s, when the num ber of high school graduates was
projected to decrease dram atically, along w ith the rising costs for higher
education, the com petition for students began to increase at institutions of
higher education. At this same tim e dem ands for greater accountability
were put on in stitu tio n s by leg islato rs, m edia, parents and the general
public (H artm an & Schm idt, 1995). In addition, to the dem ands for
quality in stru ctio n and service to students, there were public policy shifts,
and continued reduction in state and federal aid to in stitu tio n s (H ossler,
1994).
Some educators argued that m arketing is incom patible w ith the
educational m ission and does not fit in higher education institu tio n s (Pate,
1993). H ow ever, Sines and D uckw orth (1994) stated, “I t ’s tim e for
educational institu tio n s to face two facts: they are in a com petitive battle
for students, and students are custom ers” (p. 2). W ith the continued
risin g costs of education and unpredictable pools from w hich to draw
students, it is d ifficu lt to consider students as less than consum ers.
Students should be view ed as consum ers of educational services and not
products (H artm an & Schm idt, 1995). H ossler (1994) said, “Suppliers of
35
educational services need to know w hat c riteria w ill influence students’
choices, w hat services w ill help students achieve their goals, and what
factors determ ine overall student satisfaction or d issa tisfac tio n ” (p. 198).
B usinesses are suppliers of two types of products, those that are
tan g ib le and those that are not (L evitt, 1981). Education is an exam ple of
an in tan g ib le product (L evitt, 1981; Sines & D uckw orth, 1994).
P rospective students are consum ers looking to “purchase” the intangible
p roduct— an education. W hen prospective custom ers, students in this
case, can not experience the product they are purchasing in advance, they
are essen tially being asked to buy prom ises, prom ises of satisfaction
(L evitt, 1981). Students are “purchasing” their college education based on
w hat they see and hear to m ake a judgem ent about the realities. From the
b u y er’s po int of view , the product is a prom ise, and although the
custom ers are sold, they can ju st as easily be unsold if th eir expectations
are not m et. Further, Row ley, L ujan, and D olence (1997) said,
B u tif the
student finds that the in stitu tio n al environm ent is coarse, unhelpful, and
unforgiving that investm ent is w asted as the student looks elsew here”
(p. 225).
W hen it comes to retaining custom ers, intangible products, like
education, pose very special problem s. Unique to intan g ib le products is
the fact th at the custom er is seldom aw are of being served well. If
everything is going w ell, the custom er is oblivious to w hat he or she is
36
getting (L evitt, 1981). It is only when som ething does not go well that
the custom er (student) becom es aware of what he or she is not getting. It
is on these satisfactions that he or she dw ells. For this reason it is
im portant to rem ind students regularly of the presence and value of their
investm ent (Stodt, 1987).
Students m ust be convinced th at attending in stitu tio n s of higher
learning w ill help them attain th eir educational goals. R etaining students
is related to another concept called “post-purchase m arketing” (Levitt,
1981). C onsum ers’ post-purchase evaluation is a crucial issue for two
reasons. F irst, a college education is costly, m aybe the m ost costly
purchase an individual w ill m ake. The second reason post-purchase
evaluation is im portant is from the benefits an in stitu tio n w ill gain when
it has “satisfied custom ers” graduate. These graduates may not only
contribute m oney to the in stitu tio n , but the in stitu tio n w ill also receive
free w ord-of-m outh advertising, and they w ill serve as resources for jobs
for new graduates (Chadw ick & W ard, 1987). Students choose their
schools on one set of factors, which are intangible, and later evaluate
th eir experience on another set of factors, which are tangible (Chadwick
& W ard, 1987). A stin (1993) found th at the degree of satisfaction with the
college experience is less dependent on the entering ch aracteristics of the
student than on the experience he or she has after arriving on campus.
37
Students are becom ing m ore and m ore expressive of their
d issatisfactio n s w ith certain aspects of th eir educational experience
(H artm an & Schm idt, 1995). Bean and B radley (1986) suggested, “A
substantive area of investigation (student satisfaction) has received scant
attention over the last decade” (p. 393). In a study of colleges and
u n iv ersities, Pike (1991) reported that v irtually all of the adm inistrators
surveyed id en tified satisfaction as a key elem ent in in stitu tio n al
effectiveness. Pike (1991) noted that w ith the continued in terest in
assessm ent of higher education program s w ould also come increased
assessm ent of student satisfaction. N oel and L evitz (1996) stated that
student assessm ent would continue to be used to pinpoint and prio ritize
action steps to im prove in stitu tio n al effectiveness. A stin, Korn, and
Green (1987) found as a result of studying the results from C ooperative
In stitu tio n al R esearch Program s (CIRP) studies that one in four students
reports overall dissatisfactio n w ith his or her school. A stin, Korn, and
Green (1987), along w ith H artm an and Schm idt (1995), found that
upperclassm en would m ore often than not offer very different sentim ents
about th eir college experience than the low er-division students.
Spady (1970) suggested that academ ic satisfaction was critical in
explaining college persistence. A stin (1993) supported Spady and wrote,
“V irtu ally all m easures of satisfaction w ith the undergraduate experience
are sig n ificantly related to the num ber of undergraduate years com pleted”
38
(p. 278) and “one prom ising way to reduce an in stitu tio n ’s dropout rate is
to focus m ore attention on student satisfaction as an ‘interm ediate
outcom e’” (p. 278). An im portant aspect of focusing on student
satisfactio n instead of a ttritio n is that the difference betw een sub-groups
such as gender, ethnicity, or p a re n ts’ educational background can be
studied (Sanders & B urton, 1996).
39
CHAPTER 3
METHODOLOGY
Introduction
The literatu re review in C hapter 2 has provided the foundation from
w hich this research was conducted. C hapter 3 described the procedures
and m ethods that the researcher used to co llect and analyze the data.
Included in this chapter are the follow ing sections: theoretical
background; instrum ent used; re liab ility and validity; hypotheses;
population description and sam pling procedure; m ethod of data collection;
analysis of data; and lim itations and delim itations.
T heoretical Background
This study was based upon several theories of student retention.
The theories include the follow ing: T into's Theory of D eparture, B ean’s
Student A ttrition M odel, and A stin ’s Theory of Involvem ent. Also
discussed w ere the concepts of enrollm ent m anagem ent, w hich includes
consum er behavior and student satisfaction.
T into (1975, 1987, & 1993) form ulated a theory, T into's Theory of
D eparture, w hich explains the process that m otivates individuals to leave
college before graduating. T in to ’s theory hypothesized that the match
betw een an in d iv id u a l’s m otivation and academ ic ability and the
40
in stitu tio n ’s academ ic and social characteristics determ ines student
p ersisten ce in college.
A stin ’s Theory of Involvem ent has at its base inform ation from a
lo n g itu d in al study of college dropouts. His study iden tified factors in the
college environm ent that influenced stu d e n ts’ decisions to stay or leave
the college.
B ean's (1982) Student A ttrition M odel builds upon process m odels
o f o rganizational turnover. Bean hypothesized that student attritio n is
analogous to em ployee turnover in em ploym ent situations and stressed the
im portance of behavioral intentions as predictors of persistence behavior.
As a m ethod of studying the various process m odels and o th er theories
related to student behaviors, enrollm ent m anagem ent m odels were
in stitu ted and the people in charge of enrollm ent, enrollm ent m anagers,
were hired.
E nrollm ent m anagers have as one of their tasks to becom e
know ledgeable about both recruiting and retaining students. T oday’s
students are in a position to be m ore selective about w here they w ill go to
college and w hether or not they w ill stay at the in stitu tio n that they
choose.
As a m eans of staying abreast of enrollm ent trends, enrollm ent
m anagers study m arketing trends and issues. Since the study of consum er
behavior is a part of the study of m arketing trends, it is critical that
41
enrollm ent m anagers pay p articu lar attention to the theories encom passing
college stu d e n ts’ level of satisfaction w ith their educational experience.
Included in these consum er behavior theories is the theory of expectancy.
This theory purports to explain why som e w orkers in industry perform at a
higher level than others. E xpectancy theory can be used to m ake the
com parison to student satisfaction w ith his or her educational experience.
The m ore satisfied a student is w ith his or her educational process the
m ore likely he or she w ill return sem ester after sem ester.
R elated to custom er behavior is student satisfaction. The study of
student satisfaction parallels research conducted on em ployee satisfaction
(Bean, 1980; Bean & B radley, 1986). Sanders and B urton wrote:
“A lthough strongly connected to retention, student satisfactio n is a m ore
pow erful m easure that can continue to be im proved even in institutions
w ith high retention and graduation ra te s” (p. 556). The difference
betw een expectation and w hat actually takes place (perception) leads to
satisfactio n . Looking at student expectations and student satisfaction w ill
enable in stitu tio n s to m ake decisions about how to use th eir resources to
m eet the expectation of the student. W hen students are satisfied with the
services they are given, it is m ore likely that they w ill continue with th eir
education (A stin, 1993).
42
Instrum ent Used
U sing the Student S atisfaction Inventory (SSI), students from
C arroll C ollege w ere surveyed regarding th eir level of expectation and
level of satisfaction on 11 com prehensive scales of in stitu tio n al
ch aracteristics. The SSI was developed by Drs. Stephanie Ju illerat and
L aurie Schreiner. The SSI is used by four-year public and private
in stitu tio n s and tw o-year vocational schools across the country. The SSI
is used to m easure level of expectation and level of satisfaction of
students on com prehensive scales of in stitu tio n al ch aracteristics.
The difference betw een level of expectation and level of
satisfactio n , which is called the perform ance gap score, was com puted.
The instrum ent chosen to gather data was the Student S atisfaction
Inventory (h ereafter referred to as "SSI") (Schreiner
Sc
Ju illerat, 1994).
The instrum ent consists of 73 individual questions that m ake up the 12
com prehensive scales of in stitu tio n al ch aracteristics (see appendices A
and B). Of the 12 com prehensive scales, 11 scales were used in the study.
The follow ing provides a description of each of the com prehensive scales:
I.
A cadem ic A dvising E ffectiveness assesses the
com prehensiveness of the academ ic advising program .
A cadem ic advisors are evaluated on the basis of their
know ledge, com petence and personal concern for student
success, as well as on th eir approachability.
43
2.
Campus C lim ate assesses the extent to w hich the institution
provides experiences which prom ote a sense of campus pride
and feelings of belonging. This scale also assesses the
effectiveness of the in stitu tio n ’s channels of com m unication
for students.
3.
Campus L ife assesses the effectiveness of student life
program s offered by the in stitu tio n , covering issues ranging
from athletics to residence life. This scale also assesses
cam pus policies and procedures to determ ine stu d en ts’
perceptions of their rights and resp o n sib ilities.
4.
Campus Support Services assesses the quality of the support
program s and services w hich students u tilize in order to make
th eir educational experiences m ore m eaningful and
productive.
5.
Concern for the Individual assesses the in stitu tio n ’s
com m itm ent to treating each student as an individual.
6.
In stru ctio n al E ffectiveness assesses the stu d e n ts’ academ ic
experience, the curriculum , and cam pus’s overriding
com m itm ent to academ ic excellence.
7.
R ecruitm ent and F inancial Aid E ffectiveness assesses the
in stitu tio n ’s ability to enroll students in an effective m anner.
44
8.
R egistration E ffectiveness assesses issues associated with
reg istratio n and billing. This scale also m easures the
in stitu tio n ’s com m itm ent to m aking this process as smooth
and effective as possible.
9.
R esponsiveness to D iverse P opulations assesses the
in stitu tio n ’s com m itm ent to specific groups of students
enrolled at the in stitu tio n , e.g., under-represented
populations, students w ith disab ilities, com m uters, part-tim e
students, and older returning students.
10.
Safety and Security assesses the in stitu tio n ’s responsiveness
to stu d e n ts’ personal safety and security on the campus. This
scale m easures the effectiveness of both security personnel
and cam pus facilities.
11.
Service E xcellence assesses the perceived attitu d e of the sta ff
tow ard students, especially fro n t-lin e staff. This scale pinpoints the areas of the cam pus where quality service and
personal concern for students are rated m ost favorably and
least favorably.
12.
Student C enteredness assesses the cam pus’s efforts to convey
to students that they are im portant to the in stitu tio n . This
scale m easures the extent to w hich students feel welcome and
valued. (A dapted from N oel-L evitz, 1996 (p. 1-4)
45
The scale, R esponsiveness to D iverse P opulations, was the only
scale th at m easured satisfaction only and not im portance. A ccording to
Ju ille rat (1995):
The SSI does not include an im portance score for the item s on the
R esponsiveness to D iverse P opulations Scale because of the
sensitive nature of the questions and because of the statistical
im plications of asking a m ajority population about the im portance
of predom inantly m inority-related issues. T herefore, the authors
decided to consider all six of these questions to be im portant to
students (or at least the in stitu tio n ) and avoid any com plications
w ith trying to analyze responses based on m ajority vs. m inority
populations, (p. 83)
Each of the 73 item s is expressed as a statem ent of expectation or
satisfactio n . Students w ere asked, first, to rate how im portant the
expectation was to th eir overall satisfactio n with college, using a sevenp oint L ikert rating scale ranging from 1-7, w ith one indicatin g “not at all
im p o rtan t” to seven indicating “very im p o rtan t.” Next, students were
requested to rate th eir level of satisfaction that the school has met the
expectation using a seven-point L ikert scale ranging from one indicating
“not at all sa tisfie d ” to seven in dicating “very sa tisfie d .” The difference
betw een the two scores is then com puted and is called the perform ance
gap score. The SSI is a self explanatory and easily adm inistered
questionnaire, w hich takes approxim ately 20 m inutes to com plete. The
SSI was used to m easure the overall satisfaction and p rio rities of students.
46
Surveys were sent to USA N oel-L evitz for in itial scoring. A data
disk with the raw data was purchased from N oel-L evitz and was used to
provide the data for the study.
R eliab ility and V alidity
There are 12 com prehensive scales on the SSI, each with
established reliab ility . C ronbach1s coefficient alpha, which was used to
determ ine how the internal consistency of the instrum ent was correlated
fo r each item , was .97 for the set of im portance scores and .98 for the set
of satisfactio n scores. T hese scores support the internal consistency of
the instrum ent. Ju illerat stated “that each item was im portant to the over­
all instrum ent and that no item should be deleted ” (telephone interview ,
O ctober 12, 1997). F urther she said, “The instrum ent w ould not be
im proved by deleting any of the q u estio n s.”
The SSI also dem onstrates good score reliab ility over tim e. The
re liab ility co efficien t was found to be .85 for im portance scores and .84
for satisfaction scores for the 3-week retest that was conducted. Ju illerat
(1995) also found there w ould be no im provem ent in the instrum ent, based
on alpha, if any of the item s were rem oved.
T here is also evidence to support the validity of the Student
S atisfactio n Inventory. C onvergent valid ity was assessed by correlating
satisfactio n scores from the SSI w ith satisfaction scores from the C ollege
Student S atisfaction Q uestionnaire (CSSQ), another statistica lly reliable
47
satisfactio n instrum ent. The Pearson correlation betw een these two
instrum ents (r=.71; pc.OOOOl) was m oderately high, enough to indicate
adequate agreem ent that the SS I’s satisfactio n scores m easure the same
satisfactio n construct as the C SSQ ’s scores, and yet the correlation is low
enough to in dicate that there are d istin ct differences betw een the two
instrum ents.
H ypotheses
The follow ing hypotheses were tested in this study at an alpha of
There is no significant relationship betw een the dependent
variable (decision to return to college or not to return to
college) and the independent variables (the perform ance gap
scores) of the students surveyed.
Class level and gender of students in this study do not
in teract on the 11 perform ance gap scores of the
com prehensive scales of in stitu tio n al ch aracteristics.
There are no significant differences in perform ance gap
scores among freshm en, sophom ore, ju n io r, and senior
students in the survey population on the 11 com prehensive
scales of in stitu tio n al ch aracteristics.
There are no significant differences in perform ance gap
scores betw een m ale and fem ale students in the survey group
48
on the 11 com prehensive scales of in stitu tio n al
characteristics.
P opulation D escription and Sam pling Procedure
The population from which the sam ple was drawn consisted of
students enrolled at C arroll C ollege, a private four-year lib eral arts
college located in the west central part of M ontana. C arroll C ollege had a
student population of 1,263 students when the data were collected for this
study. D ata for this study were collected in the spring sem ester, 1997.
B ased on class level, the sam ple used in the study consisted of a
p roportional stratified sam ple draw n from the freshm an, sophom ore,
ju n io r, and senior classes. At the tim e of the survey the population
consisted of 35% freshm en, 22% sophom ores, 18% ju n io rs, and 25%
seniors. N oel-L evitz, the com pany from which the SSI was purchased,
recom m ended surveying h a lf of the student population. C arroll College
surveyed 600 students.
From the 600 surveys distributed, 210 (35%) freshm en, 132 (22%)
sophom ores, 108 (18%) ju n io rs, and 150 (25%) seniors, 409 (68%)
returned. From the 409 returned questionnaires, 386 had responses that
m et the criteria to in vestigate the hypotheses.
49
The results from the SSI were divided into two groups. The first
group consisted of all of the returned useable surveys.
The second group consisted of those students who supplied a valid
social security num ber. The social security num ber was the only means
by which student records could be searched to find out the status of the
student.
M ethod of D ata C ollection
The survey was adm inistered through the O ffice of A dm issions and
R etention. C lassroom d istribution by faculty based on class level was
selected as the m eans by w hich students were surveyed. The O ffice of the
R eg istrar provided a com puter-generated course listin g w ith student
classificatio n s of freshm an, sophom ore, ju n io r, and senior students.
C lasses surveyed were selected based on classes that had discrete
populations of freshm en, sophom ore, ju n io r, and senior students.
A nalysis of D ata
T hree sta tistica l m ethods w ere used to analyze the data from this
study, M ultiple R egression, T w o-Factor A nalysis of V ariance, and
N ew m an-K euls post hoc com parison. The first sta tistica l m ethod,
m ultiple regression, is a m ethod used for analyzing the separate and
co llectiv e contributions of two or m ore independent variables to explain
the v ariab ility in the dependent variable. As K erlinger and Pedhazur
(1973) w rote: “M ultiple regression analysis is nicely suited to studying
50
the influence of several independent variables, including experim ental
(m anipulated) variables, on a dependent variab le” (p. 4).
M ultiple regression has a num ber of strengths to be considered for
use. M ultiple re g re ssio n ’s biggest strength is its close relatio n to the
purpose of scien tific investigations, and that is to explain natural
phenom ena (K erlinger & Pedhazur, 1973). A second strength of m ultiple
reg ressio n is that this statistical m ethod can handle any num ber and type
of independent variables. Yet another strength of m ultiple regression is
th at it is appropriate for use when there are m ultiple independent
v ariab les and one dependent variable, or one dependent variable at a tim e
(K erlinger & Pedhazur, 1973). The sam e authors noted a final strength of
m ultiple regression “is the yield of sta tistic s” (p. 445) found for use in
interp retin g the data.
M ultiple regression analysis was used to test H ypothesis I. This
analysis was used to determ ine if there was a significant relationship
betw een the decision to return or not to return (dependent variable) to
C arroll C ollege and the 11 perform ance gap scores (independent
variab les).
To analyze H ypotheses 2 through 4 the second sta tistica l m ethod,
tw o-factor analysis of variance, was used. A nalysis of variance is used to
in v estig ate the differences betw een m ean scores of the variables being
analyzed. In this study, a tw o-by-four facto rial design w ith tests for
51
in teractio n was used. The tw o-com ponent factors were gender, and the
four-com ponent factors were class level.
T w o-factor analysis of variance is a precise and inform ative
sta tistica l m ethod (K erlinger & Pedhazur, 1973). The researcher exam ined
w hether gender and class level interacted on the 11 perform ance gap
scores from the 11 com prehensive scales of expectation and the 11
com prehensive scales of satisfaction. T w o-factor analysis of variance is
often used in the field of education. Popham and Sirotnik (1992) wrote,
“B ecause education is one of the m ost com plex behavioral fields,
in v estig atio n s conducted m ust em ploy data analysis techniques that take
into consideration not only m ore than one variable but also extrem ely
subtle in teractions among v ariab les” (p. 180).
For each of the 11 perform ance gap scores, analysis of H ypotheses
2 through 4 were tested at the same tim e. First, the p o ssib ility of
in teractio n was exam ined. W hen in teractio n was found, it was
in terp reted . If no interactio n was found, the researcher went on to test for
m ain effects. Second, differences betw een the m eans among class levels
were studied. Finally, differences betw een the m eans of m ales and
fem ales were exam ined. If one of the m eans was found to be significantly
differen t, the null hypothesis was rejected, and N ew m an-K euls post-hoc
com parison was used to identify w hich m eans were different.
N ew m an-K euls was chosen for the post-hoc com parison as it
52
provided the best balance betw een m aking a Type I or a Type II error. As
G lass and H opkins (1996) stated, “The N ew m an-K euls [significance]
levels provide a high degree of protection for the entire [om nibus] null
hypothesis, and this is the m ultiple range this author favors. M oreover, it
does not suffer from the conservatism of the Tukey test caused by
u tilizin g ju st a single c ritical value” (p. 452).
T here are many attitudes (Popham & Sirotnik, 1992) on what
sig n ifican ce level should be used for testing the null hypothesis. Popham
and S irotnik (1992) w rote, “It has been conventional in behavioral science
research to use the 0.05 level of sig n ifican ce” (p. 50). Further, Skipper,
G uenther and Nass (1970) wrote: “C asual exam ination of the literature
discloses th at the common, arbitrary, and virtually sacred levels of .05,
.01, and .001 are alm ost universally selected regardless of the nature and
type of problem . Of these three, .05 is perhaps the m ost sacred” (p.155).
By setting alpha at 0.05 the researcher w anted to balance against m aking
either a Type I error or Type II error. A Type I error would occur if the
null hypothesis was rejected when in fact there really was a difference
betw een students who did not return to C arroll and the students who did
retu rn to C arroll C ollege based on th eir perform ance gap scores on the 11
in stitu tio n a l ch aracteristics, when in fact there was no difference.
Problem s encountered by rejecting the null hypothesis may include
changes to in stitu tio n a l policies based on reasons why students left, which
53
would be invalid. If the null hypothesis were retained, which is a Type II
error, the researcher would be saying that there were no differences in
perform ance gap scores betw een the students who returned to Carroll
C ollege and those students who did not return to C arroll C ollege. This
m ight suggest that no changes were needed to im prove student satisfaction
or that in ap p ropriate changes were m ade to im prove student satisfaction,
when in fact some changes to im prove student satisfaction should have
been m ade. In this study, the researcher choose .05 alpha, as the
consequences from either a Type I or a Type II error are about the same.
A nalyses for each hypothesis were perform ed using
P a c k a g e fo r Social Sciences
7.5
(SPSS,
Statistical
version 7.5).
T-im itations and D elim itations
As w ith many studies there were lim itations and delim itations. The
lim itatio n s in this study included the follow ing:
1.
The sam ple used in the study was not a random sam ple.
2.
The findings from this study should be generalized only to
sim ilar populations.
3.
The students surveyed w ere selected on class level to get an
equal rep resentation of students from each class level.
4.
The sam ple of non-returning students used in the m ultiple
regression was not random . Only students who provided a
valid social security num ber could be tracked.
54
5.
Faculty who did not em phasize the need to record social
security num bers as a m eans of tracking returning and non­
returning students distributed the instrum ent.
D elim itations of the study included the follow ing:
1.
The data for this study were collected from one institution.
2.
The study was delim ited to those students who provided an
accurate social security num ber as a m ethod of tracking
students for the study.
55
CHAPTER 4
FINDINGS AND INTERPRETATIONS
Introduction
The m ajor goal of this study was to investigate perform ance gap
scores from 11 com prehensive scales of in stitu tio n al ch aracteristics. Of
those perform ance gap scores where significant differences were found,
in terp retatio n s w ere m ade. In Chapter 3, inform ation was provided that
described how the data w ere collected, the instrum ent used, and a
d escription of the sam ple. This chapter provides results of the tests of
hypotheses and analyses of those results.
The problem of this study was to determ ine if the difference
betw een level of expectation and level of satisfaction, called the
perform ance gap score, could predict w hether or not a student would or
w ould not return to C arroll C ollege the follow ing sem ester. The
follow ing questions w ere also answ ered: did perform ance gap scores
in teract on class level and gender; w hether or not there was a significant
difference in the perform ance gap scores among freshm en, sophom ore,
ju n io r, and senior students; and were there differences in the perform ance
gap scores betw een m ale and fem ale students. Those scales for which the
null hypotheses of no in teractio n were retained were fu rth er tested for
m ain effects. W hen a significant m ain effect was found, a Newm an-K euls
56
post hoc com parison was run. W hen the hypothesis for no interaction was
rejected , m ain effects were not tested and the interaction was interpreted.
This chapter is arranged in the follow ing order, description of the
sam ple, statistical analyses of the data, and tests of the null hypotheses.
D escription of the Sam ple
A population of 1,263 students attended C arroll C ollege when the
q u estionnaire was distributed spring sem ester, 1997. Of the students
attending C arroll C ollege, 600 were targeted to be surveyed for the study.
C lassroom d istrib u tio n by faculty was the m ethod by w hich the surveys
were distrib uted. Of the 600 questionnaires that were handed out, 409
surveys were returned. From the 409 returned surveys, 386 (64%) had
response results that were valid for the c riteria of class level and gender.
D epending on the hypotheses to be tested, responses to the survey were
fu rth er classified.
For H ypothesis I the sam ple (n=214) consisted of those students
who returned (n=183) or did not return (n=31) the follow ing sem ester who
provided a valid social security num ber. The social security num ber was
the only m eans by w hich a student could be tracked and categorized as a
retu rn in g or non-returning student. Students who graduated from C arroll
C ollege at the end of spring sem ester, 1997, were not u tilized in this
phase of the study.
57
To test H ypotheses 2 through 4, the sam ple consisted of all students
who responded to the questions on the SSI and classified them selves as
freshm en, sophom ores, ju n io rs, or seniors. The dem ographic summary of
the sam ple used in H ypotheses 2 through 4 is presented in Table I. The
sam ple used for the analyses in H ypothesis 2 through 4 included the
follow ing: 142 (37%) freshm en, 88 (23%) sophom ores, 52 (14%) juniors,
and 104 (27%) seniors. This group was com prised o f.225 (58%) m ale and
161 (42%) fem ale.
T able I. Sum m ary of sam ple used in H ypotheses 2 through 4.
Source
Freshm en
Sophom ores
Juniors
Seniors
Total
M ales
78
50
27
70
225
Fem ales
64
38
25
34
161
Total
142
88
52
104
386
S tatistical A nalysis
The hypotheses were tested using one of two analytical statistics,
m ultiple regression, or analysis of variance (ANOVA). Four hypotheses
were tested w ith the resu lts presented in Tables 3 through 9. Main effects
were tested for any interaction that was not found to be statistically
sig n ifican t. The null hypothesis for H ypothesis I was tested using
m ultiple reg ression. The null hypothesis for H ypotheses 2 through 4 was
tested using ANOVA.
58
All hypotheses were tested at an alpha level of .05. The p-values
are listed in the tables for the analyses that pertain to each hypothesis.
R esults for the analyses are presented in tables to follow .
A ll scores used in the analysis of each hypothesis were derived
from the raw data scores. Perform ance gap scores were com puted from
the raw data by subtracting the m eans from level of satisfaction scores
from m eans of level of expectation scores for each scale. The
perform ance gap score represents the difference betw een what the
stu d en t’s expectation was and his or her perceived level at which that
expectation was m et.
H ypothesis I
T here is no significant relationship betw een the dependent
v ariab le (decision to return to college or not return to
college) and the independent variables (the perform ance gap
scores).
The relationship betw een the decision to return or not return and the
11 perform ance gap scores was determ ined using m ultiple regression
analysis. Table 2 presents the results of the m ultiple regression analysis.
The null hypothesis, w hich stated that there was no significant
relatio n sh ip betw een the decision to return and the decision not to return
and the 11 perform ance gap scores, was retained, since there was not
enough evidence to reject it. Less than 3% of the v ariab ility in the
decision to return can be explained by the 11 independent variables.
59
Table 2. R esults of M ultiple R egression A nalysis.
SS
df
MS
F-R atio
P-V alue
.692
11
6.287E-02
.49
.91
R esidual
25.818
202
T otal
26.0509
M odel
R egression
.128 .
' 213
R2= .026
The perform ance gap scores of students who did or did not return to
C arroll C ollege the follow ing sem ester were com puted for each of the 11
com prehensive scales of in stitu tio n al characteristics. T hese results are
not p art of the m ultiple regression analysis, but it is in terestin g to note
that of those students who did not return their perform ance gap scores
w ere sm aller, on 10 of the 11 com prehensive scales then those students
who returned. These sm aller perform ance gap scores in dicated .that
students who did not return were m ore satisfied, on 10 of the 11
com prehensive scales, than those who did. It should be noted however,
th at the perform ance gap scores have not been analyzed for statistical
significance. Those data are presented in Table 3 on the follow ing page:
60
T able 3. Perform ance gap scores of returning and non-returning students
from the 11 in stitu tio n al ch aracteristics.
N on-R eturn
Return
Mean
Std. Dev.
M ean
Std. Dev.
A cadem ic A dvising
.97
1.11
.93
1.61
Campus C lim ate
.80
.78
.66
1.01
Campus Life
.97
.95
.84
1.16
Campus Support
Service
.90
.89
.67
1.03
C oncern for the
Individual
1.02
.92
.95
1.46
In stru ctio n al
E ffectiveness
.86
.68
.81
1.00
R ecruitm ent and
F inancial Aid
1.07
1.03
.79
1.17
R eg istratio n
E ffectiveness
.89
.97
.71
1.18
Safety and Security
1.53
1.34
1.59
1.45
Service E xcellence
.87
.86
.64
1.05
Student C enteredness
.78
.83
.72
1.29
Source
H ypothesis 2
Class level and gender of students in this study do not interact
on 11 perform ance gap scores of the com prehensive scales of
in stitu tio n a l characteristics.
A tw o-factor ANOVA tested for interaction betw een class level and
gender of student on the perform ance gap scores of the 11 com prehensive
scales of in stitu tio n a l characteristics. Table 4 provides a summary of the
tw o -facto r ANOVAS. The sum m ary for each of the 11 perform ance gap
scores included tests for in teractio n and m ain effects.
61
Table 4. Sum m ary of Perform ance Gap Scores Including Test for
In teractio n and M ain E ffects.
V ariable
A cadem ic A dvising
Gender
Class
Interactio n
Campus C lim ate
Gender
Class
In teractio n
Campus Life
Gender
Class
In teractio n
Campus Support Services
Gender
Class
In teractio n
C oncern for the Individual
Gender
Class
Interaction.
In stru ctio n al E ffectiveness
Gender
Class
In teractio n
R ecruitm ent Sc F inancial Aid
Gender
Class
In teractio n
R eg istratio n E ffectiveness
G ender
Class
In teractio n
Safety & Security
G ender
Class
Interaction
df
MS
F-R atio
P-V alue
.05
I
3
3
31.077
0.975
0.968
18.849
0.591
0.587
.000
.621
.624
I
3
3
11.448
1.144
0.944
15.123
1.511
1.247
.000
.211
.292
I
3
3
0.051
0.708
1.120
0.048
0.669
1.058
.027
.072
.367
I
3
3
11.976
4.863
0.746
14.028
5.696
0.874
.000
.001
.455
I
3
3
19.204
1.090
0.258
16.623
0.943
0.223
.000
.420
.880
I
3
3
9.408
1.246
0.946
14.167
1.876
1.425
.000
.133
.225
12.175
I
5.808
3
3 . 1.943
10.177
4.855
1.624
.002
.003
.183
I
3
3
7.165
1.196
0.927
7.073
1.181
0.915
.008
.317
.434
I
3
3
8.295
18.230
7.578
4.765
10.473
4.353
.030
.000
.005
62
Table 4. (Continued from previous page).
Service E xcellence
Gender
Class
Interactio n
Student C enteredness
Gender
Class
Interaction
I
3
3
21.847
1.628
0.624
25.64
1.19
0.732
.000
.127
.533
I
3
3
9.553
0.334
0.555
9.256
0.324
0.538
.003
.808
.657
=hBold indicates significance
The resu lts from the tw o-factor ANOVA showed that the null
hypothesis of no in teractio n was retained for all scales w ith the exception
of one— Safety and Security. There was significant interactio n on the
scale Safety and Security (p= .005). The researcher found the follow ing:
freshm an and sophom ore fem ales have higher mean perform ance gap
scores than th eir m ale counterparts, but ju n io r and senior fem ales have
low er m ean perform ance gap scores than their m ale Counterparts. The
m ean perform ance gap scores for freshm en are low er than the mean
perform ance gap scores for ju n io rs and seniors; sophom ore and junior
m ean perform ance gap scores are higher than freshm en, but low er than
seniors; and seniors are higher than the rest of the classes.
T able 5 shows the m eans of perform ance gap scores by class level
and gender.
63
T able 5. ANQVA for interactio n of perform ance gap scores— Safety and
Security.
Freshm en
Sophom ores
Juniors
Seniors
M ales
1.25
1.60
2,00
2.60
Fem ales
1.32
1.81
1.66
1.56
T hese resu lts im plied that there was interaction among class level
and gender on the scale Safety and Security. For those scales where no
sig n ifican t in teractio n was found and the null hypothesis was retained,
they were fu rther tested for m ain effects.
H ypothesis 3
T here are no significant differences in perform ance gap
scores among freshm an, sophom ore, ju n io r, and senior
students on the 11 com prehensive scales of in stitu tio n al
ch aracteristics.
The null hypothesis as stated in H ypothesis 3 was rejected for two
scales— Campus Support Services and R ecruitm ent and F inancial Aid.
A ccording to the analyses, there were significant differences among the
m eans of perform ance gap scores for freshm en, sophom ores, ju n io rs, and
seniors on the scale Campus Support Services (p= .001) and R ecruitm ent
and Financial Aid (p= .003).
T his im plied that there were significant differences among class
levels on two scales. The researcher used a N ew m an-K euls post hoc
procedure to determ ine which p air-w ise differences were significant.
T able 6 contains the results of m ain effects among freshm en, sophom ores,
64
ju n io rs, and seniors on Campus Support Services and R ecruitm ent and
Financial Aid.
Table 6. Summary of m ain effects test for Campus Support Services and
R ecruitm ent and Financial Aid.
P Value
F R atio
MS
df
SS
Source
Campus Support Services
14.588
3
4.863
5.696
.001
R ecruitm ent & F inancial Aid
17.424
3
5.808
4.855
.003
The results from N ew m an-K euls (Table 7) post hoc procedure for
Campus Support Services indicated that the m ean perform ance gap scores
fo r seniors is higher than the mean of freshm en, sophom ores, and juniors.
The perform ance gap score is significantly higher for seniors, which
im plies seniors were m ore d issatisfied with Campus Support Services than
w ere freshm en, sophom ores, or juniors.
T able 7. N ew m an-K euls post hoc com parison for Campus Support
Services.
Subset for alpha= .05
I
2
N
Class level
.834699
143
Freshm en
.928355
88
Sophom ores
1.026099
52
Juniors
1.357279
105
Seniors
The results from the N ew m an-K euls (Table 8) post hoc com parison
fo r R ecruitm ent and F inancial Aid indicated the perform ance gap score of
freshm en w ere sign ifican tly sm aller than that of sophom ores, juniors, and
65
seniors. This result indicated freshm en were m ore satisfied than
sophom ores, ju n io rs, and seniors on the item s m easured w ithin the scale
o f R ecruitm ent and F inancial Aid. A dditionally the resu lts from the
N ew m an-K euls indicated that the m ean perform ance gap scores for
sophom ores, were not different from freshm en, ju n io rs, or seniors. This
resu lt indicated that sophom ores are no m ore or less satisfied than the
other three class level on the scale of R ecruitm ent and F inancial Aid.
Table 8. N ew m an-K euls post hoc com parison for R ecruitm ent and
F inancial Aid.
Subset for alpha= .05
2
N
I
Freshm en
143
.919930
Sophom ores
88
1.135227
Juniors
52
1.334936
Seniors
105
1.439048
Class Level
1.135227
H ypothesis 4
There are no significant differences in the perform ance gap
scores betw een m ale and fem ale students on the 11
com prehensive scales of in stitu tio n a l ch arac teristics.
H ypothesis 4 was tested using ANOVA. The m ean perform ance gap
scores w ere com puted and used to test for significant differences betw een
m ale and fem ale students on 10 com prehensive scales w here an analysis
was perform ed when no interaction was present. R esults from the
ANOVA show ed sig n ifican t differences betw een m ale and fem ale students
66
on 9 of the 10 com prehensive scales tested. The results are docum ented
in Table 9 on the next page.
The differences betw een m ales and fem ales indicated that m ales are
sig n ifican tly less satisfied than are fem ales. The com prehensive scales
w hich show sig n ifican t differences include the follow ing: Academ ic
A dvising, Campus C lim ate, Campus Support Services, C oncern for the
Individual, Instru ctio n al E ffectiveness, R ecruitm ent and Financial Aid,
R eg istratio n E ffectiveness, Service E xcellence, and Student C enteredness.
Campus L ife was the only com prehensive scale where no significant
difference was found betw een m ale or fem ale students.
T able 9. M ain effects for m ales and fem ales.
Source
M ale
Eem ala
F Ratio
P Value
A cadem ic A dvising
1.26
.68
18.849
.000
Campus C lim ate
1.06
.72
11.448
.000
Campus Support Services
1.17
.81
14.028
.000
C oncern for the Individual
1.29
.84
16.623
.000
In stru ctio n al E ffectiveness
1.10
.78
14.167
.000
R ecruitm ent and F inancial Aid
1.31
.95
10.177
.002
R eg istratio n E ffectiveness
1.11
.83
7.073
.008
Service E xcellence
1.19
.71
25.640
.000
Student C enteredness
1.04
.72
9.256
.003
67
CHAPTER 5
SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
Introduction
T o d ay ’s college student is becom ing m ore selective not only in
choosing a college but also when looking at the student services the
in stitu tio n provides. W ith the increased cost of college and a more
consum er-oriented applicant, colleges and universities are finding it m ore
com petitive to recru it and retain students.
W ith a less-th an -p red ictab le dem ographic shift of eligible entering
college students and strained operating funds at many institu tio n s, there is
a critical need to use each budget dollar in the m ost productive m anner.
In order th at in stitu tio n s use their m onies well, students are frequently
asked for th eir input through surveys, to find out what services they deem
m ost im portant or useful to them .
In addition to surveying students to find out w hich services are
m ost im portant, in stitu tio n s try to assess the level of satisfaction of these
services. There are a num ber of reasons for assessing student satisfaction
including the follow ing: im provem ent of student retention; docum entation
of assessm ent for accreditation purposes; feedback to adm inistrators who
are charged w ith m aking in stitu tio n al changes; setting the retention
agenda; and for inform ation to be used in m arketing the in stitu tio n to
68
pro sp ectiv e students (Ju illerat & Schreiner, presentation, July, 1997,
N ational C onference on Student R etention, W ashington, D. C.)The problem of this study was to determ ine if the difference
betw een level of expectation and level of satisfaction on 11
com prehensive scales of in stitu tio n al characteristics, called the
perform ance gap score, could be used to predict w hether or not a student
would retu rn to C arroll C ollege the follow ing sem ester. The study sought
also to determ ine if com bining the variables of class level and gender
produced statistica lly sig n ifican t resu lts on the com prehensive scales of
in stitu tio n al ch aracteristics. Further, the study tested a hypothesis to
determ ine if there were significant differences in perform ance gap scores
among freshm en, sophom ores, ju n io rs, or seniors on any of the 11
com prehensive scales. Finally, the study tested a hypothesis to determ ine
if there w ere any sig n ifican t differences in perform ance gap scores
betw een m ale and fem ale students, on the 11 com prehensive scales of
in stitu tio n a l characteristics.
The data for this study were collected spring sem ester, 1997, at
C arroll C ollege, a p rivate four-year lib eral arts college located in west
central M ontana. The study exam ined perform ance gap scores for 11 of
12 com prehensive in stitu tio n a l scales. The tw elfth scale, R esponsiveness
to D iverse P opulations, was the only scale which exclusively m easured
satisfactio n . Im portance was not m easured due to the sensitive nature of
69
the questions w ithin the scale and the im plications of asking such
questions to a m ajority population (Juillerat, 1995).
The follow ing 11 scales were used in this study. The questions that
m ake up each scale can be found in appendix B:
1.
A cadem ic A dvising E ffectiveness assesses the
com prehensiveness of the academ ic advising program .
A cadem ic advisors are evaluated on the basis of their
know ledge, com petence and personal concern for student
success, as w ell as on th eir approachability.
2.
Campus C lim ate assesses the extent to w hich the institution
provides experiences w hich prom ote a sense of cam pus pride
and feelings of belonging. This scale also assesses the
effectiveness of the in stitu tio n ’s channels of com m unication
for students.
3.
Campus L ife assesses the effectiveness of student life
program s offered by the in stitu tio n , covering issues ranging
from athletics to residence life. This scale also assesses
cam pus policies and procedures to determ ine stu d en ts’
perceptions of their rights and resp o n sib ilities.
4.
Campus Support Services assesses the quality of the support
program s and services w hich students utilize in order to m ake
th eir educational experiences m ore m eaningful and
70
productive.
C oncern for the Individual assesses the in stitu tio n ’s
com m itm ent to treating each student as an individual.
Instru ctio n al E ffectiveness assesses the stu d e n ts’ academ ic
experience, the curriculum , and cam pus’s overriding
com m itm ent to academ ic excellence.
R ecruitm ent and F inancial Aid E ffectiveness assesses the
in stitu tio n ’s ability to enroll students in an effective m anner.
R egistration E ffectiveness assesses issues associated with
reg istratio n and billing. This scale also m easures the
in stitu tio n ’s com m itm ent to m aking this process as sm ooth
and effective as possible.
Safety and Security assesses the in stitu tio n ’s responsiveness
to stu d e n ts’ personal safety and security on the campus. This
scale m easures the effectiveness of both security personnel
and cam pus facilities.
Service E xcellence assesses the perceived attitude of the sta ff
tow ard students, especially front-line staff. This scale
pinpoints the areas of the cam pus where quality service and
personal concern for students are rated m ost favorably and
least favorably.
Student C enteredness assesses the cam pus’s efforts to convey
71
to students that they are im portant to the in stitu tio n . This
scale m easures the extent to w hich students feel welcom e and
valued. (A dapted from N oel-L evitz, 1996 (p. 1-4)
The instrum ent used to collect the data, the Student S atisfaction
Inventory (SSI) (USAGroup N oel-L evitz, 1993), was form ulated to
m easure the changing trends in higher education to a m ore consum eroriented focus (Ju illerat and Schreiner, 1995). The SSI is also a more
com prehensive survey when it comes to m easuring student desires and
satisfactio n . C urrently, other available tools, such as the C ollege Student
S atisfactio n Q uestionnaire (CSSQ), are used to collect inform ation on
student satisfaction. The CSSQ, how ever, is less than satisfactory for the
purpose of m easuring student satisfaction today as it was first published
in 1970 (Ju illerat, 1995).
Of the 600 students who were targeted for data collection, 409
responded w ith 386 useable responses. To get the highest return rate
possible, faculty were asked to distrib u te the questionnaires in their
classroom s. T hose classes targeted for surveying were selected with the
aid of the O ffice of the R egistrar.
R esults from the questionnaires w ere sent to USAGroup NoelL evitz for in itia l scoring. A data disk containing the raw scores from
questionnaires the C arroll C ollege students had com pleted was purchased
from USAGroup N oel-L evitz by the researcher. The perform ance gap
72
scores were com puted from the raw data using
S o c ia l S c ie n c e s (S P S S,
S ta tis tic a l P a c k a g e f o r the
version 7.5, 1996).
Students w ere asked to respond to questions using a seven-point
L ikert rating scale of 1-7, w ith one in dicating “not at all im portant” to
seven in d icating “very im p o rtan t.” Three scores were com puted for each
question. The first score calculated, the expectation score, m easured the
stu d e n t’s response to the level of im portance he or she placed on the
question. The second score m easured the stu d en t’s level at which the
expectation was m et, again using a seven-point L ikert rating scale. This
score is called the satisfaction score. The third score, perform ance gap
score, was calculated by subtracting the satisfaction score from the
expectation score. The perform ance gap scores were used to depict the
difference betw een the stu d en t’s level of expectation and his or tier
perceived level that the satisfaction was being met. The perform ance gap
scores w ere used for the analyses of the data throughout the study.
From the returned surveys, two sam ple groups w ere form ulated
depending on the criteria o f the hypotheses. The first sam ple group
(n=386) consisted all of the returned useable surveys.
The second sam ple group (n=214) consisted of those students who
provided a valid social security num ber. The social security num ber was
the only m eans by which student records could be searched to find out
w hether or not the student returned to C arroll College the follow ing
73
sem ester. O f the 214 responses there were 183 students who returned to
C arroll C ollege the follow ing sem ester, and 31 responses were from those
students who did not return.
H ypothesis I was tested using m ultiple regression. M ultiple
reg ressio n analysis is used to determ ine if there is a significant
relatio n sh ip betw een the dependent variable and independent variable (s).
In this study, m ultiple regression was used to determ ine if there was a
sig n ifican t relationship betw een the decision to return or not return to
C arroll C ollege and the 11 perform ance gap scores.
H ypotheses 2 through 4 were tested using analysis of variance
(A N O V A ). A nalysis of variance is used to investigate the differences
betw een m ean scores. For purposes of this study, ANOVA was used to
in v estig ate the differences in m ean scores among class levels, and
differences in m ean scores betw een m ales and fem ales.
N ew m an-K euls post hoc procedures were perform ed as a follow -up
procedure for the ANOVAS that were found to be significant at the .05
level. Post hoc procedures are done to identify specifically w hich of the
m eans found in the analyses of variance is the statistically different m ean.
In this chapter the researcher presents a review of the study,
sum m arizes the resu lts, draw s conclusions from the results, and makes
recom m endations for fu rth er studies.
74
Sum m ary of Findings
Student success and satisfactio n have been determ ined to be
im portant m easures of student retention and institu tio n al success. Just as
im portant as m easuring success and satisfaction for retention purposes is
assessing the stu d e n ts’ level of expectations of their college experience.
R etention studies have shifted from an in stitu tio n al perspective to one
that looks at what is im portant and satisfying to students. In stitutional
and student ch aracteristics, as highlighted in the review of literature, were
u tilized in this study to assess and exam ine perform ance gap scores. The
follow ing four hypotheses were form ulated and tested.
H ypothesis I stated that there was no sta tistically significant
relatio n sh ip betw een the decision to return or not return to Carroll
C ollege the follow ing sem ester (dependent variable) and the perform ance
gap scores (independent variables). The analysis show ed perform ance gap
scores could not be used to predict a stu d e n t’s decision to return with any
reliab ility . The perform ance gap scores at best could only account for
2.6% of the variation in the decision.
H ypothesis 2 stated that class standing and gender of students do
not in teract with perform ance gap scores on the com prehensive scales of
in stitu tio n a l ch aracteristics. The null hypothesis for in teractio n was
retain ed on one of the 11 scales, Safety and Security. Com bining the
variables class level w ith gender, the researcher found a statistically
75
sig n ifican t interaction. This interactio n indicated that the com bination of
class level with gender showed a statistica lly significant difference.
R esults from H ypothesis 2 showed that freshm en and sophom ore
fem ales have higher m ean perform ance gap scores than freshm en and
sophom ore m ales. This higher m ean perform ance gap score indicated that
freshm en and sophom ore fem ales w ere less satisfied on the scale Safety
and Security than th eir m ale counterparts. This scale considers the degree
of feeling safe and secure on cam pus, am ount of parking available,
lig h tin g on cam pus, and how quickly security responds to em ergencies.
Junior and senior fem ales had low er m ean perform ance gap scores than
ju n io r or senior m ales. This difference in m ean perform ance gap scores
betw een m ale and fem ale ju n io rs and seniors indicated that ju n io r and
senior fem ales were m ore satisfied than th eir m ale counterparts on the
scale Safety and Security.
It was also found when analyzing H ypothesis 2 that the mean
perform ance gap scores of freshm en were low er than the m ean scores of
ju n io rs or seniors. Once again, these low er m ean perform ance gap scores
indicated th at freshm en were m ore satisfied than ju n io rs or seniors on the
scale Safety and Security. Sophom ore and ju n io r m ean perform ance gap
scores were larger than the m ean perform ance gap scores for freshm en,
but low er than those of seniors. The m ean perform ance gap score for
seniors was larger than the m ean perform ance gap scores of the other
16
three class levels. Seniors were the least satisfied as revealed by the
questions w ithin the scale Safety and Security. As a point of com parison
on this scale w ith those students who did not return, results indicated that
the m ean perform ance gap scores of students who did not return were
higher than the mean perform ance gap scores of those who did. W hile
this resu lt was not analyzed for significance, it appears that students who
did not retu rn w ere less satisfied on the scale Safety and Security than
those students who did return.
H ypothesis 3 stated that no statistica lly significant differences
existed among freshm en, sophom ores, ju n io rs, or seniors on the 11
com prehensive scales of the SSI. The findings of H ypothesis 3 verified
the null hypothesis on 8 of the com prehensive in stitu tio n al scales. A
statistica lly significant difference among class levels was found on the
analyzed scales, Campus Support S ervices and R ecruitm ent and Financial
Aid.
U sing a N ew m an-K euls post hoc procedure on the scale Campus
Support Services, findings showed the m ean perform ance gap score for
seniors was significantly higher than the m ean perform ance gap score of
the other three class levels. The resu lts from the ANOVA indicated that
seniors w ere m ore dissatisfied w ith Campus Support Services than were
freshm en, sophom ores, or ju n io rs.
77
The other scale for which a sta tistica lly significant difference was
found. R ecruitm ent and Financial Aid, show ed that the m ean perform ance
gap scores of freshm en were significantly sm aller than the mean
perform ance gap scores of sophom ores, ju n io rs, and seniors. This finding
indicated th at freshm en w ere m ore satisfied than sophom ores, juniors, and
seniors on item s m easured w ithin the scale. R ecruitm ent and Financial
Aid.
Further, the m ean perform ance gap score for sophom ores is not
differen t from the m ean perform ance gap scores for sophom ores, juniors,
or seniors. This resu lt indicated that sophom ores were no m ore or less
satisfied on the scale R ecruitm ent and Financial Aid than are freshm en,
sophom ores, or ju n io rs.
C arroll C ollege needs to exam ine ways in which to support
sophom ores, ju n io rs, and seniors to allev iate their d issatisfactio n in the
area of R ecruitm ent and F inancial Aid. The college needs to examine
reallo catio n of funds for sophom ores, ju n io rs and seniors. They may also
consider a tu ition freeze for those students who are ju n io rs and seniors.
H ypothesis 4 stated that no statistica lly significant differences
existed betw een m ale and fem ale students on the 11 com prehensive scales
of in stitu tio n al characteristics. The findings of this hypothesis resulted in
retain in g the null hypothesis on only one of the scales, Campus Life.
S ig n ifican t differences were found on the follow ing scales: Academ ic
78
A dvising, Campus C lim ate, Campus Support Services, Concern for the
Individual, Instru ctio n al E ffectiveness, R ecruitm ent and F inancial Aid,
R eg istratio n E ffectiveness, Service E xcellence, and Student C enteredness.
An exam ination of these statistica lly significant findings showed
that the m eans of perform ance gap scores for m ales were significantly
higher com pared w ith the m eans of perform ance gap scores for fem ales.
The higher perform ance gap scores indicated that m ale students were less
satisfied than fem ale students on all of the com prehensive scales of
in stitu tio n a l ch aracteristics, except the scale Campus Support Services.
C arroll C ollege m ust exam ine m ore closely in what ways the m ale
students are m ore d issatisfied than fem ale students and in stitu te changes
that w ill correct the situation.
C om paring resu lts from this study to results on a n ational level was
not p art o f this study. H ow ever, a com parison of the resu lts can provide
C arroll C ollege w ith inform ation that could be used for recruitm ent and
reten tio n purposes. O verall findings of perform ance gap scores, as
m easured by the SSI, w ere com pared to results of other four-year private
in stitu tio n s which included the follow ing: C arroll C o lleg e’s results
indicated low er perform ance gap scores on 6 of 11 com prehensive
in stitu tio n a l scales tested. These low er perform ance gap scores indicated
that when com pared to students on a national level, the levels of
expectation and satisfactio n of C arroll C ollege students were being met to
79
a greater degree. The three scales w here C arroll C ollege perform ance gap
scores were higher than the com parison group included Academ ic
A dvising, C oncern for the Individual, and Safety and Security. This
resu lt provides C arroll C ollege w ith additional inform ation that can be
used to m ake suitable in stitu tio n al changes. C arroll C ollege should look
at the item s/questions w ithin each of the three scales and assess how the
college can best m ake in stitu tio n al changes that w ill prom ote higher
satisfactio n for students.
C onclusions
This study attem pted to dem onstrate both theo retically and
statistica lly the usefulness of assessing stu d e n ts’ expectations when
com pared to their level of satisfaction. The SSI provided a m eans by
w hich C arroll C ollege could assess its students to determ ine how satisfied
they w ere regarding various in stitu tio n a l ch aracteristics. A prediction
regarding w hether or not a student w ould or would not return to college
the next sem ester based on a com parison of perform ance gap scores was
weak. H ow ever, other useful inform ation was gathered from the analyses
in this study.
The results o f analyzing perform ance gap scores provides
inform ation for recom m endations to C arroll College and for further
studies. It should be noted, how ever, th at results from the analyses of the
perform ance gap scores provide an overview of the scales of
80
com prehensive in stitu tio n a l characteristics. To obtain a m ore detailed
sum m ary from the findings of the analyses, one would need to look at the
perform ance gap scores for each question w ithin the com prehensive
scales.
The perform ance gap scores provide inform ation as a starting point
for m aking in stitu tio n a l changes. C arroll C ollege should note that the
larg er the perform ance gap score, the. greater the level of d issatisfaction
expressed by the student. In this-study, som e perform ance gap scores
produced m ean scores that were found to be statistically sig n ifican t when
tested at an alpha level of .05. These findings may lead one to believe
th at for those scales w here significant differences were found changes
should be m ade to those areas first. H ow ever, depending on the scale
being evaluated, it may prove to be m ore beneficial in term s of retention
to m ake in stitu tio n al changes based on the level of expectation that
students placed on the individual item s w ithin the com prehensive scale,
rath er than m aking changes to the in stitu tio n based on the findings related
to the scale.
The results from H ypothesis 2, w hich stated there was no
in teractio n among class level and gender of students on the 11
com prehensive scales, indicated that there was interactio n among class
level and gender on only one of the com prehensive scales, Safety and
Security. It can be concluded from the results of the tests for interaction
81
found in H ypothesis 2 that perform ance gap scores are not strong enough
in d icato rs to support the hypothesis that class level and gender together
affect the satisfactio n that students place on the item s w ithin the
com prehensive scales. In the absence of interaction, the ability to
generalize the results is strengthened. T hese results could be very
im portant C arroll C ollege.
H ypothesis 3, w hich stated that there was no significant difference
among the class levels, was rejected on two of the com prehensive scales.
It can be concluded from these findings that freshm en are generally m ore
satisfied than students from the upper division classes on only two scales
Campus Support Services and R ecruitm ent and F inancial Aid. The result
th at freshm en are m ore satisfied than upper-class level students is
consistent, how ever, w ith what is found in literatu re concerning freshm en
and satisfaction.
R esults from studies show trends in satisfaction research are m ore
co n sisten t when students are separated by class level. Several researchers
have found that first-y ear students are significantly m ore satisfied than
students in the upper classes. Schm idt and Sedlacek s 1972 findings
indicated th at students becom e m ore d issatisfied the longer they are
enrolled in school. Part of this discrepancy is explained by the optim ism
and/or idealism that are held by freshm en in anticipation of starting
college (Schm idt & Sedlacek, 1972). A stin et al. (1987) concluded that
82
“upperclassm en offer very different (frequently m ore critical) assessm ents
of th eir college experience than low er division students” (p. 40).
In this study, findings showed sig n ifican t differences among class
levels on two of the com prehensive scales, R ecruitm ent and Financial Aid
and Campus Support Services. W hen these two scales w ere analyzed,
findings showed that freshm en were m ore satisfied than sophom ores,
ju n io rs, or seniors on the scale R ecruitm ent and Financial Aid. On the
other scale w here a significant difference was found, Campus Support
Services, findings show ed that seniors w ere the least satisfied. R elating
the resu lts of this post hoc com parison to results, found in a study by
Schm idt and Sedlacek (1972) and A stin et al. (1987) show ed that the
longer a student stayed in college the m ore d issatisfied he or she becam e
on som e aspect of his or her educational experience.
H ypothesis 4, w hich stated there w ere no significant differences
betw een m ale and fem ales, was analyzed. On the scales w here no
in teractio n had been found, significant differences w ere noted between
m ale and fem ales on 9 of 10 scales tested. In this study m ales were found
to be less satisfied than fem ales.
In previous research findings, w here level of satisfactio n was
com pared to gender (Pennington et al., 1989; Okum, 1981; Betz et al.,
1972; Knox et al., 1992), results show ed differing conclusions. It can be
concluded th at the resu lts from this study are consistent w ith the results
83
of the Okum study (1981), w hich found fem ales had higher satisfaction
levels than m ales. Pennington et al. (1989) found that m ale college
students had significantly higher overall satisfaction scores than women,
w hich is contrary to the results of this study. Betz et al. (1972) and Knox
et al. (1992) found no significant differences in satisfaction betw een
m ales and fem ales. C arroll C ollege, how ever, should consider how to use
the findings on gender differences that w ere found in this study as it
sp ecifically addresses the opinion of the student population.
Findings from this study and a review of the lite ra tu re and research
indicated that the 11 in stitu tio n al ch aracteristics iden tified in the SSI and
assessed by the respondents are im portant at various sta tistic a l levels in
the educational experience of C arroll C ollege students. Sim ilar
conclusions w ere supported by the works of Tinto (1993), A stin (1977),
Bean (1980), A nthrop (1996), and Bank, B iddle, and Slavings (1992), in
th at the m ore academ ically and socially integrated students becom e in the
various aspects of the cam pus culture the greater the likelihood of being
m ore satisfied and thus rem aining in college.
Further, it was concluded that m easuring the difference betw een
student expectation and student satisfaction concerning the educational
process can be a useful m easure for assessing the needs and desires of
students. H ossler and Bean (1990) and Spanbauer (1992), who concluded
that students should be considered as consum ers, support this point of
84
view . In stitu tio n s should in vestigate the expectations, needs and wants of
th eir students, and look at students as custom ers as a m ethod that leads
tow ards m eeting the needs and desires of students, which should lead to
higher levels of retention.
Using satisfactio n as a m eans of com parison to retention and noting
w hether or not an in stitu tio n is successful has been supported by the
works o f A stin (1977, 1993), A stin, Korn and Green (1987), Chadwick
and W ard (1987), and E arw ood-Sm ith and C olbert (1989). They
concluded th at a stu d e n t’s decision to continue with his or her education
is dependent on how satisfied he or she is. They also b elieve that student
success and level of satisfaction is a useful m easure of retention. H ossler
and Bean (1990), and Spanbauer (1992) found that overall m easures
betw een expectation and satisfaction regarding the educational experience
were an im portant m eans by which to assess retention. T he use of student
satisfactio n as a m easure related to reten tio n and to in stitu tio n al success
has been show n to be an indicator of retention.
B ased on these findings and conclusions, this study reaffirm s the
com plexity betw een student expectations, institu tio n al facto rs, and class
levels and/or gender of student in term s of predicting and describing
factors th at influence retention. The findings from this study do not
verify th at a pred ictio n can be m ade on w hether or not a student will
retu rn based on perform ance gap scores. The study does how ever provide
85
useful inform ation for C arroll C ollege in various in stitu tio n al areas where
changes can be in itiated that may result in greater student satisfaction.
R ecom m endations can be m ade from the results found in this study for
fu rth er consideration.
R ecom m endations
B ased on the resu lts of this study, the follow ing recom m endations
are m ade. Some of the recom m endations are for C arroll C ollege, and
som e of the recom m endations are general suggestions for further study.
F irst, C arroll C ollege should conduct m ore in-depth research on the
com prehensive scales. The com prehensive scales provide a broad
overview of how satisfied students are on 11 in stitu tio n al scales. To gain
a m ore com prehensive picture from those scales, it is im portant to look at
the in d iv id u al questions of which each scale is com prised. From those
in d iv id u al item s, a m ore com plete rep resentation of student satisfaction
can be form ulated. For exam ple, the scale for Safety and Security could
benefit from further exam ination of why freshm en and sophom ore fem ales
are m ore d issatisfied than freshm en and sophom ore m ales. All freshm en
and sophom ores are required to live on cam pus. If there is a concern by
fem ale students that they do not feel safe w alking across cam pus, Carroll
C ollege needs to do som ething to rectify this perception.
C om parisons among class levels should be m ade based on
expectations on the com prehensive scales. A dditionally, com parisons
86
among class levels should be m ade based on satisfaction on the
com prehensive scales. A nalyzing expectations and satisfaction separately
m ight provide in sig h t that was not noted from the perform ance gap scores
as to why students do not return.
Second, research should be com pleted to determ ine where
differences occur among class levels on the individual questions w ithin
the com prehensive scales. K now ledge of w here the differences in
satisfactio n are among class level w ill provide m ore specific inform ation
to C arroll C ollege.
A th ird recom m endation for fu rth er study includes analyzing
perform ance gap scores betw een m ales and fem ales using other variables.
O ther variables to be studied may include GPA, involvem ent in campus
organizations, involvem ent on athletic team s, and the students declared
m ajor.
A fourth recom m endation for fu rth er study w ould be to assess
carefu lly the m eans by w hich to distrib u te the questionnaire. In order to
receive com plete responses and detailed inform ation, surveyors need to
stress the im portance of recording accurate social security num bers as a
m ethod of tracking students who return or do not return.
A fin al suggestion for future research includes com paring
perform ance gap scores of students who are currently enrolled with
students who have graduated. The fact th at students rem ained at Carroll
87
C ollege long enough to graduate, does not necessarily im ply that they were
com pletely satisfied with their educational experience. Surveying
students who graduated from C arroll C ollege may provide insight for
adm inistrators charged w ith m aking in stitu tio n al changes. M aking
changes based on feedback from students who graduated m ay ju stify the
changes that are needed to influence those students who may be thinking
of leaving C arroll C ollege before th eir graduation.
88
REFERENCES CITED
89
ACT (1997). N ational D ropout R a te s. The A m erican C ollege Testing
Program . In stitu tio n al D ata File.
A n th ro p , P. S. (1996). E xpectations and satisfaction of freshm en in
h ig h er education. U npublished doctoral dissertation, U niversity of
C entral Florida.
A stin, A. W. (1977). Four critical years: E ffects of college on beliefs,
attitu d es, and know ledge. San Francisco: Jossey-B ass.
A stin, A. W. (1984). Student involvem ent: A developm ental theory for
higher education. Tournal of C ollege Student P erso n n el, 25, 297308.
A stin, A. W ., Korn, W., Green, K. (1987). R etaining and satisfying
students. E ducational R ecord, w inter, 36-42.
A stin, A.W. (1993). W hat m atters in college: Four critic a l years,
re v isite d . San Francisco: Jossey-B ass.
Bank, B. J., B iddie, B. J., Slavings, R. L. (1992). W hat do college
students want? E xpectations and undergraduate persistence. The
Sociological Q uarterly, 33(2), 321-355.
Bean, J. P. (1980). D ropouts and turnover: The synthesis and test of a
’ casual m odel of student attritio n . R esearch in H igher E ducati o i l .
12(2), 155-187.
Bean, J. P. (1982). C onceptual m odels of student attritio n : How theory
’ can help the in stitu tio n al researcher. New D im ensions for
In stitu tio n al R esearch: Studying Student A ttrition, 36, Decem ber,
17-33.
Bean, J. P. (1986). A s s e s s i n g and reducing attrition. New D irections fo r
H igher E ducation. 53, M arch, 47-61.
Bean, J. P. (1990a). U sing retention research in enrollm ent m anagem ent.
’ In D. H ossler, J. P. Bean (E ds.). The Strategic M anagem ent of
C ollege E n ro llm en ts. San Francisco: Jossey-B ass.
90
Bean, J. P. (1990b). Why students leave: Insights from research.
T heS trategic M anagem ent of C ollege E nrollm ents, (p p .147-169).
San Francisco: Jossey-B ass.
Bean, J. P., B radley, R. K. (1986). U ntangling the satisfactionperform ance relationship for college students. Journal of H igher
Education, 57(4)., July-A ugust, 393-412.
Betz, E. L., Starr, A. M., & M enne, J. W. (1972). C ollege student
satisfaction in ten public and p riv ate colleges and universities.
Journal of C ollege Student P erso n n el, 13. 456-461.
B illson, J. M., Terry, M. B. (1987). A Student retention m odel for higher
education. C ollege and U niversity. 62. summer, 290-305.
B oyle, T. P. (1989). An E xam ination of the Tinto m odel of retention in
. higher education. NASPA. 26(4). sum m er, 288-294.
B raxton, J.M ., V estper, N., H ossler, D. (1995). E xpectations for college
and student persistence. R esearch in H igher E ducation. 36(5), 595612.
C abrera, A. F., Nora, A., C astaneda, M .B. (1993). C ollege persistence
stru ctu ral equations m odeling test of an integrated m odel of student
reten tion. Journal of H igher E ducation. 64(2). M arch/A pril, 123139.
Chadw ick, K., W ard, J. (1987). D eterm inants of consum er satisfaction
w ith education: Im plications for college and u niversity
adm inistrators. C ollege and U niversity. 62(3), spring, 236-246.
C haffee, E.E. (1990). S trategies for the 1990s. New D irections for
H igher E ducation,7 0 , sum m er, 59-65.
C rockett, D.S. (1978). Academ ic advising: A cornerstone of student
reten tion. New D irection for Student Services. 3, 29-35.
Cum m ings, L.L., Schwab, D.P. (1973). Perform ance in organization:
D eterm inants and appraisal. G lenview , IL : Scott and Foresm an.
D olence, M.G. (1991). Setting the context for evaluation of recruitm ent
and retention program s. New D irections for In stitu tio n al R esearch,
70, sum m er, 5-19. San Francisco: Jossey-B ass.
91
Dunphy, L. (1987). E xem plary retention strategies for the freshm an year.
New D irections for H igher E ducation. 60. 15(4), 39-60.
D urkheim , E. (1951). Suicide: A Study in sociology. Trans. J.A.
Spalding and G. Sim pson, G lencoe, N.J: The Free Press.
E arw ood-Sm ith, G., C olbert, M. (1989). Student satisfaction: A key
facto r in retention. The C ollege Student A ffairs Journal. 9(1).
spring, 14-20.
Ferguson, M. (1990). The R ole of faculty in increasing student retention.
C ollege and U niversity. 65(2). w inter, 127-135.
Fishbein, M., Axjen, I. (1975). B elief, attitude, intention and behavior:
An Introduction to theory and re se a rc h . Reading, MA: AddisonW esley.
G ardiner, J.J., N azari-R obati, A. (1983). .Student retention research:
Im plications for retention strategies. NASPA, 20(3). w inter, 25-33.
G etzlaf, S.B ., Sedlacek, G.M ., K earney, K .A ., B lackw ell, J.M . (1984).
Two types of voluntary undergraduate attrition: A pplication of
T in to ’s m odel. R esearch in H igher E ducation, 20(3). 257-268.
G lass, G. V., H iopkins, K. D. (1996). S tatistical m ethods in education
and psychology (3rd ed.). N eedham H eights, MA: A llyn and Bacon.
Gordon, V. N. (1989). O rigins and purposes of the freshm an sem inar.
The Freshm an Y ear E xperience, (pp. 183-215). San Francisco:
Jossey-B ass.
H artm an, D.E., Schm idt, S.L. (1995). U nderstanding student/alum ni
satisfaction from a consum er’s perspective: The E ffects of
in stitu tio n a l perform ance and program outcom es. R esearch in
H igher E ducation. 36(2). 197-217.
H ealy , P. (1997). The W idely predicted grow th in enrollm ent hasn’t
m aterialized in some states. The C hronicle of H igher Education,
A ugust 15. 1997. pp. A23-24.
H earn, J.C ., (1985). D eterm inants of college stu d en ts’ overall evaluation
of th eir academ ic program s. R esearch in H igher E ducation, 23(4),
413-437.
92
H odgetts, R. M. (1982). M anagem ent: Theory, process and practice
(3rd ed.). Chicago: The D ryden Press CBS C ollege Publishing.
H ossler, D. (1991). E valuating recruitm ent and retention program s. New
D irections for In stitu tio n al R esearch. 7 0 , 95-99.
H ossler, D. (1994). E nrollm ent m anagem ent an integrated approach.
New York: The C ollege E ntrance E xam ination Board.
How ard, R. D., R ogers, B. H. (1991). T racking academ ic progress w ithin
a com plex academ ic environm ent. New D irections for Institutional
R esearch. 7 0 . sum m er, 61-72.
Ju illerat, S. (1995). Investigating a tw o-dim ensional approach to the
assessm ent of student satisfaction: V alidation of the.student
satisfaction inventory. U npublished doctoral dissertation, Tem ple
U niversity.
K ennedy, G .J., et al. (1995). Changes in social and academ ic integration
in freshm en of high and average ability: Im plications for retention.
NACADA journal. 15(2), fall, 9-19
K erlinger, F.N ., Pedhazur, E.J. (1973). M ultiple regression in behavioral
re se a rc h . New York: Holt, R inehart and W inston.
Knox, W. E., Lindsay, P., & Kolb, M. N. (1992). H igher education,
college characteristics, and student experiences: L ong-term effects
on educational satisfactions and perceptions. Journal of Higher
E d u catio n . 63(3), 303-328.
L evitt, T. (1981). M arketing intangible products and product intangibles.
H arvard B usiness R eview , M ay-June, 94-102.
N ational C enter for S tatistics (1997). P rojections of e d u cation statistics
to 2 0 0 7. U.S. D epartm ent of E ducation.
Noel, L. (1994). Student satisfaction: A breakthrough in designing a
cam pus w ide retention program . Paper presented at the N oel/L evitz
C onference on Student R etention, W ashington, D.C.
N oel, L., L evitz, R., Saluri, D., & A ssociates (1985). Increasing student
reten tion: E ffective program s and practices for reducing the
dropout ra te . San Francisco: Jossey-B ass.
93
N oel, L., L evitz, R. (1996). Key lessons learned from retention and
recru itm ent research. Iowa City, IA : USA Group N oel-L evitz.
Okun, M. a., K ardash, C. A., Stock, W. A., Sandler, I. N., & Baumann, D.
I. (1986). M easuring perceptions of the quality of academ ic life
among college students. Journal of C ollege Student P ersonnel, 27,
447-451.
Pantages, T .J., Creedon, C.F. (1978). Studies of college attrition: 19501975. Review of E ducational R esearch, 4 8 , 49-101.
P ascarella, E .T., T erenzini, P.T. (1980). Predicting freshm an persistence
and voluntary dropout decisions from a theoretical m odel. Journal
of H igher E ducation. 51(1). 60-75.
P ascarella, E .T., T erezini, P.T. (1991). How college affects stu d en ts.
San Francisco: Jossey-B ass.
Pate, W .S. (1993). C onsum er satisfaction, determ inants, and postpurchase actions in higher education. College and U n iv ersity ,
spring and sum m er, 100-107.
Pennington, D. C., Zvonkovic, A. M., & W ilson, S. L. (1989). Changes
in college satisfaction across an academ ic term . Journal of C ollege
Student Personnel. 30(6), 528-535.
Perigo, D. J., U pcraft, M. L. (1989). O rientation program s. The
Freshm an Y ear E xperience, (pp. 82-94). San Francisco: JosseyBass.
Pike, G.R., (1991). The E ffects of background, coursew ork and
involvem ent on stu d en ts’grades and satisfaction. R esearch in
H igher E ducation, 32(1),
15-30.
Popham , W .J., Sirotnik, K.A. (1992). U nderstanding statistics in
ed u catio n . Itasca, IL: F.E. Peacock Pub, Inc.
Porter, O. F. (1990). U ndergraduate com pletion and p ersistence at fouryear colleges and u n iv e rsitie s. W ashington, D.C: The N ational
In stitu te of Independent C olleges and U niversities.
94
Row ley, D. J., L ujan, H .D ., D olence, M. G. (1997). S trategic change in
colleges and universities: Planning to survive and prosper. San
Francisco: Jossey-B ass.
Sanders, L., B urton, J. D. (1996). From retention to satisfaction: New
outcom es or assessing the freshm an experience. R esearch in H igher
E ducation. 37(5), 555-567.
Sanford, N. (1967). W here colleges fail: A study of the student as a
person. San Francisco: Jossey-B ass.
Schreiner, L. A., Ju illerat, S. L. (1994). Student satisfactio n inventory.
Iow a City, IA fN o e l-L e v itz C enters, Inc.
Schm idt, L. K., Sedlacek, W. E. (1972). V ariables related to university
student satisfaction. The Journal of C ollege Student Personnel,
13(2), May, 233-238.
Sines, R. G. Jr., D uckw orth, E. A. (1994). Custom er service in higher
education. Journal of M arketing for H igher education, 5(2), 1-15.
Skipper, G uenther, Nass (1970). The Significance test controversy.
M orrison, D. E., H enkel, R. E., (E ds.). Chicago: A ld rin e.
Spady,W . (1970). D ropouts from higher education: Tow ard an em pirical
m odel. Interchange, 2, 38-62.
Spanbauer, S. J. (1992). A Q uality system for education: Using quality
and p roductivity techniques to save our schools. M ilw aukee, WI:
ASQC Q uality Press.
Stage, F. K. (1989). R eciprocal effects betw een the academ ic and social
in teg ration of college students. R esearch in Higher Education,
30.(5), 517-530.
Starr, A., B etz, E. L., M enne, J. (1972). D ifferences in college students
satisfaction: A cadem ic dropouts, nonacadem ic dropouts, and
nondropouts. Journal ofC ounseling Psychology, 19(4), July, 318322.
SPSS, Inc. (1996). S tatistical package for social sciences for W indows,
version 7.5 [Com puter softw are], Chicago, IL : A uthor.
95
Stodt, M. M. (1987). E ducational excellence as a prescrip tio n for
reten tion. New D irections for H igher E ducation. 60. w inter,
5-13.
T eren zin i, P. T., P ascarella, E. T. 1980). Toward the validation of
T in to ’s m odel of college student attrition: A review recent studies.
R esearch in H igher E d u catio n .12 (3), 271-282.
T into, V. (1975). D ropout from higher education: A th eo retical synthesis
o f recent research. Review of E ducational R esearch , 45., 89-125.
T into, V. (1987). The P rinciples of effective retention. Fall Conference
o f the M aryland C ollege Personnel A ssociation (ERIC D ocum ent
R eproduction Service No. ED 301 267.)
T into, V. (1993). L eaving C ollege: R ethinking the causes and cures_of
student attritio n (2n^ed.). C hicago: The U niversity of Chicago
Press.
Tow nsley, M. K. (1993). A S trategic m odel for enrollm ent driven private
colleges. Journal for H igher E ducation Managem ent, 8(2),
w in ter/spring, 57-66.
Toy, T. J. (1985). Increasing faculty involvem ent in retention efforts.
Increasing student retention. San Francisco: Jossey-B ass.
U pcraft, M. L „ G ardiner, J. N., and associates. (1989). The Freshman,
year experience: H elping students survive and succeed in college.
San Francisco: Jossey-B ass.
Van G ennep, A. (1960). The Rites of passage. Trans. By M onika B.
V izedom , G abrielle L. C haffee. Chicago: U niversity of Chicago
Press.
W erth, B. (1988). Why is college so expensive? M aybe A m erica wants it
that way. Change, M arch/A pril, 13-25.
W estern In terstate C om m ission for H igher Education, T eachers Insurance
and A nnuity A ssociation, and The C ollege Board. (1993). H igh
school graduates: Projections by state 1992-2009. B oulder, CO:
A uthor.
96
APPENDICES
APPENDIX A
STUDENT SATISFACTION INVENTORY
U S A G f O U p Noel-Levitz
Strategicsolutionstoreoucation
98
Januonr 20, 1998
Ms. nioria Lambertz
Carroll College
1601 North Benton Ave.
Helena. MT 59625-0002
Fax: 406-447-4533
Dear Glona:
I am glad to hear that your dissertation work is progressing! Keep up the good work.
ou have permission to reproduce a copy of the Student Satisfaction Inventon ™to be
included in your dissertation appendix.
r i :=-e'
Please let me know if you need any additional information. I am glad that vou and
Stephanie Juulerat were able to connect to discuss her research! I will look forward to
continuing to work with you in the future.
Sincerely,
(6
Julie Honschman
Program Consultant
99
U S A G fO U p IMoel-Levitz
;
T
•
-
.,tor
-
------------
X•'- I - —
D e a r S tu d e n t.
H E E 5 E ^E ~ ~ = =
To p re s e rv e c o n fid e n tia lity , y o u r n am e is n ot re q u e ste d .
— Thunk yon for your purricipahon.
-------------------------------------------------------------
Instructions:
• U se a N o. 2 p e n c il only. P le a s e d o n o t u se in k o r b a llp o in t p en .
h r a s e c h a n g e s c o m p le te ly a n d c le a n ly .
C o m p le te !) d a rk e n the o v al th a t c o rre s p o n d s to y o u r re sp o n s e .
-.....-Mylevel ofsatisfaction:
iportant at all
: very im p o rta n t
jgSBW&HS
som ew hat u n im p o rta n t
4 - n eu tral
______
_
5 - somew h at i m p o r t a n t ____
6 - im p o rta n t
7 ' veT im p o rta n t
smsTyrirrit.;
"
I
m m
■- --.vfcsaSSSSSSSSS
JXXX
I . Most students feel a sense of belonging here.
The campus staff are canng and helpful
3. Faculty care about me as an individual.'
4. Admissions staff are knowledgeable.
D X X X
5. Financial aid counselors are helpful.
D X X X
D X X X
6. My academic advisor is approachable
7. The campus is safe and secure for all students.
I The content ot the courses within my major is valuable.
OXXX
) X X X
D X X X
9. A variety of intramural activities are offered
0. Administrators are approachable to students'
11. Billing policies are reasonable.
DXXX
D X X X
)XXX
13. Library staff are helpful and approachable
IX X X
1X X X
IX X X
X
.X
X
X
X
X
X
X
X
X
X
X
X X X X X X X
X X IX X X X X :
X X X X X X X
X X X X X X X
X X X X X X X
,xxxxxxx
X X X X X X^ X :
,X X X X X X X :
|X j X X X X X X
xxxxxxx
x
16
1S
. The instruction m my major field is excellent.
11 l i q u a t e financial aid is available for most students
18. Library resources and services are adequate
19. My academic advisor helps me set goals to work toward
The au^ness 0 # « 's open during hours which are convenient for most students.
V_/
x:x:x:xxxx G
xxx.xx xx:
x
x
x
x
x
x
xxxxxxx
xxxxxxx
xxixxxxjx:
X X X X X X X
xxxxxxx
X X X X X X X !
jQ o m
5 CL CDCD
JCLCDCD
J CD X CD
-C -
I OO
Importance-to me.....
*'■- ■■'-'V*-
L -jio t im p o rta n t a t ail
■- v; 2 - not very im p o rta n t_____
3 - so m ew hat u n im p o rtan
4 - n eu tral
S - som ew hat im p o rta n t,
6 - im p o rta n t____
7 - very im p o rtan t
... My level of satisfaction_______
Ii.. . noCavailahle/i
_ _ _ _ _ _ _ _ _ _ _ satisfied":--7"
11,
e
-
L
*
D
D
D
D
® ffi© © ®
C S ® © ® ®
® ® © ® ®
® ® © ® ®
D ® ®
D ® ®
Dffi®
Df fi f fi
©
®
©
©
3
©
©
©
©
Dffiffi© ® ® ®
D C S ® © ® ® ©
D C S ® © © ® ©
DCS C D © ® ® ©
Dffiffi®
Dffiffi®
DCS® ®
Dffiffi®
ffiffi®
ffiffi®
® CD©
ffiffi®
D C S ® © © ® ©
Dffiffi® ©CD©
©ffiffi®®®©
©ffiffi®®® ®
D
D
D
D
ffiffi®
ffiffi®
ffiffi®
ffiffi®
®
®
®
®
®
®
®
©
©
©
©
®
D
D
D
D
ffiffi®
ffiffi®
ffiffi®
ffiffi®
®
®
®
®
®
®
®
®
©
®
®
®
Dffiffi®®®®
D C S ® ® ® ® ©
Dffiffi® ®®©
Dffiffi®®®®
D C S ® ® ® ® ©
O ffiffi® ffiffi®
D CS® ® ® ® ©
I T- T ' -Tl
ts
i
I
l
S
S
i
=
.
.
.
.
.
.
.
.
.
©ffiffi®®®©
®
®
®
®
®
®
®
®
©
©
®
©
©
©
©
©
ff
ff
ff
ff
©
®
®
®
®
®
®
®
©
©
©
®
©
©
©
©
ffiffi®®
ffiffi®®
ffiffi©®
®©ffiffi
®
®
®
®
©
®
©
©
©
©
©
©
©ffiffi
ffiffi®
ffiffi®
ffiffi©
®
©
®
®
®
©
®
®
©
©
®
®
- '“ »>” » a o ' u S £ ' s d^
©
©
©
©
ffiffi©©
©ffiffi©
ffiffi®©
ffiffi®©
®
®
®
®
©
©
©
©
S i V nn^ t1I11Inon has a So o d Jeputation within the community,
smdent center is a comfortable place for students to spend their leisure time.
©
©
©
©
f
f
f
f
®
®
®
®
®
©
®
®
©
©
©
©
ffiffi© © © ©
ffiffi® © ® ©
ffiffi® ® ffiffi
ffiffi® ® ffiffi
m
”
r
z
g
a
g
s
a
z
s
I
M
a
-
^
-
45. Students are made to feel welcome on this campus.
1P nenif^sl y Sf ' ' " ' oiYe^ ln campus organizations.
47
” *"*
©©©©ffiffi©
©ffiffi®®®©
©©ffiffi®®©
ffiffi®
ffiffi©
ffiffi®
ffiffi®
37. I k e I a sense or pride about mv campus
j j »
^r
©
©
©
©
3
F
[
F■
Fg
:.§ •
®
©
®
©
" den'’ -
a
J
T
6
®
®
®
®
°f
8 S ® t t F 5 S t t S S 5 5 S 3 a S S .
j j 32." Totonng s e n tc e f a ^ T e X a^ailaWe"1" ” “ panidpate m -ntercolkSiate athietics-
a*
M tl s f ie ri .
©ffiffi®®
©© ffiffi®
©© ffiffi®
ffiffi®®®
1
DCS C D © © © ©
D C S ® © ffiffi®
D C S ® ® ffiffi©
Dffiffi®©®®
D C S ® ® ffiffi©
DCS C D © © ® ®
Dffiffi®®®©
D C S ® ® ® ® ©
i
ercollegiate athletic programs contribute to a strong sense of school spirit.
Il
a
® ® ® ©
© ® ® ©
© © © ©
©ffiffi©
DCS C D © © ©
D C S ® © © ®
D C S ® © ® ®
D © © @ ® ®
i
■ —
"
T a a a g g f f i g g i--t I :
# i I
# I '
D CD CD CD CD CD CD
D CD CS © CD CD CD
DCD©® © ® ©
DCD© @ ® ® ®
:•
uim ewhat satisfied - S i
____________________ neutral - 4
I
somewhat dissatisfied - 3
: 54: B % s m r ^ t l r i ^ % i p f u l " ° " StUdem differences as theV ,each a course■ 35. Major requirements are clear and reasonable
I - 6- The SIudenI ^ n d b o o k provides helpful information about cam pus life.
f
f
f
f
iff
iff
iff
iff
if
if
if
if
f
f
f
f
i®
i©
i©
i®
i®
i®
i®
i©
®
©
©
©
■
I E |E = B H E E E » ’-
© ffiffi® © ® ©
ffiffi© ® © © ©
© f f iff i® © ffiffi
© f f if f i® ffiffi©
J
61. Adjunct faculty are competent as classroom instructors.
© ffiffi® ® ® ©
© ffiffiffiffiffiffi
© ffiffi® © © ©
ffiffiffiffiffiffi©
"
j
1
generally know what s happening on campus.
« : S S n d iS
^
T
S
dS
ta o n y d d
64' New student onem atton services help students adjust to college.
™ £
du^ s
■ % Freedom of expression is protected on campus.
"
I 68- N e a r , y al1 ot the tacuIty are knowledgeable in their field
*
5 m ' S f f u 'S ar f ? 0.d. vanetY of e°urses provided on this campus.
I
r h f n n f i Irfr hmg as5istantS are com petent as classroom instructors.
P ■ - I ' Channels tor expressing student com plaints are readilv available.
I
ffiffiffiffiffiffi©
© © © © ffiffi© :
© ffiffi© © ® ®
© ffiffi© © ® ©
ffiffi© © © ® ©
© © ® ® ® ® ®
© © ® ® ® ® ®
I 01
My level of satisfaction
LZZr^iofovaila-
D (B d )C D
DCBCDCD
DCBCBCD
D dK B CD
DCBCBCD
D C B dK D
D CB dK D
DCBCBCD
D CE'CB CD
DCBCBCD
CD CD'CB CD.CB CB CD
CD®:®!®:®:®!®
® : ® : ® '® : ® : ® i c D
p!® .® :® :® ® ,®
p ® :® ® !® ® ®
Pan-lime students?
Evening smdents ?
Older, returning learners?
84.
85.
86.
87.
89.
88.
Students with disabilities?
p® :® :® :® ® :®
p ® .® ® ® :® ® ]
p ® ® ® ® :® :®
I I I I I I I I
89.
C D ® :® :® ® !® ®
i® :® :® :® ® ® ®
m ® :® :® ® :® :®
0 f ,h e follow inS factors in your
)® ® ®
)® ® ®
.® ® o
7® ® ®
®®®
®® ®
® ® ®
j ® ® CD
"® ® ®
LhS
90.
91
92.
93.
94.
Cost
~
Financial aid
Academic reputation
Size of institution
Opponum ty to piav spons
"
--------------
96:
97. Campus appearance
98. Personalized attention prior to enrollment
resp0nse 'hn‘ heSt “PPliK
. how has y o u r college experience
m r expectations?
ich worse than I expected
ite a bit worse than I expected
>rse than I expected
out what I expected
tier than I expected
ite a bit better than I expected
ich better than I expected
“arken lhe ,„rr=sp„„dm8 ova, tor each of ,he
100. R ate y o u r overall satisfaction
with
your experience h ere th u s far.
’ Mot satisfied at all
2 Mot very satisfied
3 Somewhat dissatisfied
4 Neutral
5 Somewhat satisfied
6 Satisfied
7 Very satisfied
CVivrrxrr-Tr -rn
101. All in all. if you had it to do over
again, w ould you enroll here?
' Definitely not
2 Probably not
3 Maybe not
4 I donI know
5 Maybe yes
6 Probably yes
7 Definitely yes
IIIIIIIIIIIIIIIIIIIIIIIIIIIIII
84.
85.
86.
se the one response that best describes you and darken the corresponding oval for each of the items below.
G e n d e r:
L F em ale
2 M ale
109. Educational Goal:
C Associate degree
2D Bachelor’s degree
£> M aster's degree
<2 Doctorate o r professional degree
C Certification (initial or renew al)
Self-im provem ent/pleasure
t Job-related training
>
'■Other
A ge:
v 18 and under
2
19 to 24
2 25 to 34
4 35 to 44
s 4 5 and over
HO. E m p lo y m en t:
I Full-tim e o ff cam pus
2j P an-tim e o ff cam pus
32 Full-tim e on cam pus
C P an-tim e on cam pus
5 Not em ployed
E th n ic ity /R a c e :
' A frican -A m encan
2 A m en ean Indian or A laskan Native
3 A sian or Pacific Islander
4 C aucasianA V hite
s H ispanic
s O ther
7 P refer not to respond
111. C u r r e n t R esidence:
v Residence hall
2 " Fraternity / Sorority
3 Own house
4 Rent room o r apartm ent o ff cam pus
5 Parent s hom e
e Other
C u r r e n t E n ro llm e n t S ta tu s:
1 Day
2 Evening
3 W eekend
. C u r r e n t C lass L o ad :
> Full-tim e
2 P an -tim e
112. R esidence C lassification:
<- In-state
2
O ut-of-state
3 International (not U.S. citizen!
. C la ss L evel:
' Freshm an
2 S ophom ore
113. D isabilities:
Physical disability or a diagnosed learning disability?
1 Yes
2
No
3 Junior
4 Senior
s S pecial Student
s G raduate:P rofessional
7 O ther
114. W h en I e n te re d th is in stitu tio n , it w a s m y :
’. 1st choice
2
2nd choice
3 3rd choice o r lower
. C u rren t G PA :
' No credits earned
2
1.99 o r beiow
3 2 .0 - 2 .4 9
4 2 .5 - 2 .9 9
s 3.0 - 3.49
6 3.5 or above
S ocial S e c u rity N u m b er:
Write your Social Security
number in the nine spaces of
the box provided.
Completely darken the
corresponding oval.
Mmr 'nviri :-v.uri:' XtimiHf \ ivuuvMeti fur research
purfw.es .inti will not .ppcar mi :.nx report.
I
j
T
0
0
!
0
o
2 I
3 I
2
2
2
3
3
3
4
4
4
4 ,
5
5
S
5
6
6
6
8 1
7
7
7
7I
8
8
8
8 i
9
9
9
1
2
3
I
TluiiiIx you for taking the time to complete this inventory.
i%«r* I zs
^*rxI ri
'
0.
0
0
0
0
0.
1
I
I"
I
f
I
I
I
2
2
2
2
2
2
2
2
I
3
3
3
3
3
3
4
4
I
4
4"
5'
5.
3
<
S
6
6
6
4
| 5
i 6
I 7
I8
Q
3
4
5
5
5
<
s;
6
6.
6
6
7
7
7
7J
7
7
7
3
3
9
3
8
8
8
8
3
9
9
9
9
9
9
9
116.i Item re q u e ste d by y o u r in s titu tio n :
|
'
Fill in m ajor code
from list provided
by your institution.
!
0
9
5- M a jo r:
i
0
5'
S
7
103
APPENDIX B
ITEMS W ITHIN COM PREHENSIVE INSTITUTIONAL SCALES
104
Student S atisfaction Inventory
ACADEMIC ADVISING
6.
My academ ic advisor is approachable.
14.
My academ ic advisor is concerned about my success as an
individual.
19.
My academ ic advisor helps me set goals to work tow ard.
33.
My academ ic advisor is know ledgeable about requirem ents in
my m ajor.
55.
M ajor requirem ents are clear and reasonable.
105
Student Satisfaction Inventory
CAMPUS CLIMATE
1.
M ost students feel a sense of belonging here.
2.
The cam pus sta ff are caring and helpful.
3.
Faculty care about me as an individual.
7.
The cam pus is safe and secure for all students.
10.
A dm inistrators are approachable to students.
29.
It is an enjoyable experience to be a student on this cam pus.
37.
I feel a sense of pride about my cam pus.
41.
There is a com m itm ent to academ ic excellence on this
cam pus.
45.
Students are m ade to feel welcom e on this cam pus.
51.
This in stitu tio n has a good reputation w ithin the com m unity.
57.
I seldom get the “run-around” when seeking inform ation on
this cam pus.
59.
This in stitu tio n shows concern for students as individuals.
60.
I generally know w hat’s happening on cam pus.
62.
There is a strong com m itm ent to racial harm ony on this
cam pus.
66.
T uition paid is a w orthw hile investm ent.
67.
Freedom of expression is protected on cam pus.
71.
C hannels for expressing student com plaints are readily
available.
106
Student S atisfaction Inventory
CAMPUS LIFE
9.
A variety of intram ural activ ities are offered.
23.
L iving conditions in the residence halls are com fortable
(adequate space, lighting, heat, air, etc.)
24,
The in terco lleg iate athletic program s co n trib u te to a strong
sense of school spirit.
30.
R esidence hall sta ff are concerned about me as an individual.
31.
M ales and fem ales have equal Opportunities to p articip ate in
in terco lleg iate athletics.
38.
There is an adequate selection of food available in the
cafeteria.
40.
R esidence hall regulations are reasonable.
41.
There are a sufficient num ber of weekend activ ities for
students.
46.
I can easily get involved in cam pus organizations.
52.
The student center is a com fortable place for students to
spend th eir leisure tim e.
56.
The student handbook provides helpful inform ation about
. cam pus life.
63.
Student disciplinary procedures are fair.
64.
New student orientation services help students adjust to
college.
67.
Freedom of expression is protected on cam pus.
73.
Student a c tiv itie s fees are put to good use.
107
Student S atisfaction Inventory
CAMPUS SUPPORT SERVICES
13.
L ibrary sta ff are helpful and approachable.
18.
L ibrary resources and services are adequate.
26.
Com puter labs are adequate and accessible.
32.
T utoring services are readily available.
44.
A cadem ic support services adequately m eet the needs of
students.
49.
There are adequate services to help me decide upon a career.
54.
B ookstore sta ff are helpful.
108
Student S atisfaction Inventory
CONCERN FOR THE INDIVIDUAL
3.
Faculty care about me as an individual.
14.
My academ ic advisor is concerned about my success as an
individual.
22.
C ounseling sta ff care about students as individuals.
25.
Faculty are fair and unbiased in their treatm ent of individual
students.
30.
R esidence hall sta ff are concerned about me as an individual.
59.
This in stitu tio n shows concern for students as individuals.
I
109
Student S atisfaction Inventory
INSTRUCTIONAL EFFECTIVENESS
3.
Faculty care about me as an individual.
8.
The content of the courses w ithin my m ajor is valuable.
16.
The instru ctio n in my m ajor field is excellent.
25.
Faculty are fair and unbiased in their treatm ent of individual
students.
39.
I am able to experience in tellectu al growth here.
41.
There is a com m itm ent to academ ic excellence on this
cam pus.
47.
Faculty provide tim ely feedback about student progress in a
course.
53.
Faculty take into consideration student differences as they
teach a course.
58.
The quality of instruction I receive in m ost of my classes is
excellent.
61.
A djunct faculty are com petent as classroom instructors.
65.
Faculty are usually available after class and during office
hours.
68.
N early all of the faculty are know ledgeable in th eir field.
69.
T here is a good variety of courses provided on this cam pus.
70.
G raduate teaching assistants are com petent as classroom
instructors.
no
Student S atisfaction Inventory
RECRUITM ENT AND FINANCIAL AID
4.
A dm issions sta ff are know ledgeable.
5.
F inancial aid counselors are helpful.
12.
F inancial aid aw ards are announced to students in tim e to be
helpful in college planning.
17.
A dequate financial aid is available for m ost students.
43.
A dm issions counselors respond to prospective stu d en ts’
unique needs and requests.
48.
A dm issions counselors accurately portray the cam pus in their
recruiting p ractices.
I ll
Student S atisfaction Inventory
REGISTRATION EFFECTIVENESS
11.
B illing policies are reasonable.
20.
The business office is open, during hours w hich are
convenient for m ost students.
27.
The personnel involved in reg istratio n are helpful.
34.
I am able to reg ister for classes I need w ith few conflicts.
50.
Class change (drop/add) policies are reasonable.
112
Student S atisfaction Inventory
. SAFETY AND SECURITY
7.
The campus is safe and secure for all students.
21.
The am ount of student parking space on cam pus is adequate.
28.
Parking lots are w ell-lighted and secure.
36.
Security sta ff respond quickly in em ergencies.
113
Student S atisfaction Inventory
SERVICE EXCELLENCE
2.
The cam pus sta ff are caring and helpful.
13.
L ibrary sta ff are helpful and approachable.
14.
The sta ff in the health services area are com petent.
22.
. C ounseling sta ff care about students as individuals.
27.
The personnel involved in reg istratio n are helpful.
57.
I seldom get the “run-around” when seeking inform ation on
this cam pus.
60.
I generally know w hat’s happening on cam pus.
71.
C hannels for expressing student com plaints are readily
available.
114
Student S atisfaction Inventory
STUDENT CENTEREDNESS
1.
M ost students feel a sense of belonging here.
2.
The campus sta ff are caring and helpful.
10.
A dm inistrators are approachable to students.
29.
It is an enjoyable experience to be a student on this cam pus.
45.
Students are m ade to feel w elcom e on this cam pus.
59.
This in stitu tio n shows concern for students as individuals.
MONTANA STATE UNIVERSITY LIBRARIES
3 1762 10280430 7
>-