|Illlllllllil‘llllI

V..-

1 Er-h.!-u Al. l‘- a!”

f

t- __.:; 7-1 I

l .9 I .
. 3...._.. f1--. 1
“39;.--” "-1 QI. I-._.'.'.

[1. ’ - -31
la‘S?¥L-iliiiJ;—L_'!

 

 

 

 

This is to certify that the

dissertation entitled

UNDERGRADUATE WOMEN, OVER 24:
SOCIODEMOGRAPHIC, EDUCATIONAL AND ADJUSTMENT FACTORS
IN PERSISTING OR DROPPING OUT
AND PREDICTIONS OF THOSE AT RISK FOR NOT GRADUATING

presented by

Dorothy LucilIe Mercer
has been accepted-towards fulfillment
of the requirements for

Ph .D . degree in Counsellng Psychology

 

 

fitn‘kg, (194.41; aft

 

 

Major professor

Date November 11, 1986

 

MS U i: an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

 

MSU

LIBRARIES
”-

 

 

RETURNING MATERIALS:

ace n o rop to
remove this checkout from
your record. FINES will
be charged if 50°F is
returned after the date
stamped below.

 

UNDERGRADUATE WOMEN, OVER 24:
SOCIODEMOGRAPHIC, EDUCATIONAL AND ADJUSTMENT FACTORS
IN PERSISTING OR DROPPING OUT
AND PREDICTIONS OF THOSE AT RISK FOR NOT GRADUATING
By

Dorothy Lucille Mercer

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling Psychology

1986

ABSTRACT
UNDERGRADUATE WOMEN , OVER 24:
SOCIODEMOGRAPHIC , EDUCATIONAL AND ADJUSTMENT FACTORS
IN PERSISTING OR DROPPING OUT
AND PREDICTIONS OF THOSE AT RISK FOR NOT GRADUATING
By

Dorothy Lucille Mercer

Undergraduate women who were at least twenty-five years old at
Michigan State University's September, 1984, registration were
contacted in a blind three-wave mailing over two weeks in early
summer, 1986, in an attempt to predict who will persist to graduation
and who will drop out.

The 363 returned questionnaires of 584 delivered represent a
62.22 return rate.

The mean age of the women in this study is 34.55 years.
Twenty—eight percent are single, 50.42 married and 17.72 divorced.
Fifty-six percent are mothers. Mean income is $20,000-$24,999.

During their last school year, 40% worked half- to full-time; 28% were
unemployed; one-half were full-time students. Mean grade point
averages (GPA's) are: 3.01 prior to reentry, 3.17 cumulative, and 3.42
for the last term.

More withdrew earlier from lack of purpose or goals and returned
now for their personal satisfaction.

The 194 graduates and 41 current withdrawers from the university
are compared using chi squares, t-tests and ANOVA's.

Discriminant function analysis produced an equation accurately

classifying between 82.41 and 86.22 the total group of dropouts and

Dorothy Lucille Mercer

graduates and each split—half combination. These predictors, in order
of strength in the direction of predicting dropouts, are:

Less expectation of completing the degree without further breaks

Fewer total reasons given for last entry into college

Greater happiness during school

Possession of a previous degree

Lower rating of how well she is getting wanted things out of life

Mere hours per week of employment

Finances seen as less of a problem

Lower cumulative GPA

Lower (more external) Duttweiler's Locus of Control Scale score

Single status

Dropouts also are likely to have higher family incomes, handle
crises less well, have more previous education, he part—time students
during their last year, and have lower most recent term GPA's than do
graduates but do not differ in age, marital or parental status.

With marital status as a separate independent variable, life
satisfaction during and since school, achieving what is wanted in life
and self-esteem all affirm the prediction of graduates being better
adjusted.

A combination of sociodemographic, educational and psychological

adjustment variables appears to be an appropriate, useful predictor of

potential graduates and potential dropouts.

ACKNOWLEDGEMENTS

My doctoral committee members are a group of special people whom
I value as individuals and to whom particular appreciation is due: to
my chairwoman, Dr. Linda Forrest, for her role modeling and for
raising the cogent issues, focusing my attention and challenging me to
see a better picture; to Dr. Sam Plyler for his steady presence and
his firm and contagious beliefs in my ultimate success and that many
possibilities remain within reach; to Dr. Mildred Erickson for
breaking the ground at Michigan State for older students and for her
unfailingly cheerful availability to think through the next steps on
the path with me, and to Dr. Albert Aniskiewicz for broadening my
understanding of people and for remaining so helpful as the path
twisted and turned.

I am most grateful to Diana Algra who made this project possible
by so willingly giving her sponsorship and labor to the reproduction
and mailing of the survey itself.

My understanding of the mysteries of computerized statistical
analysis has been greatly enhanced due to Dr. David Solomon's
assistance.

Many thanks to Karen Firos and Elizabeth Spata for long hours of
reading and rating the open-ended responses, to Betsy Bricker for

editing when the draft was rough and time was short, and to Ruby Freed

ii

iii

and Elizabeth Spata and many other friends for concrete assistance and
for quiet cheering when the journey appeared bleak.

Especially I value those who lived through the process with me
and who made major accomodations for many years to make it possible:
Todd and Andrew Mercer. And to my mother, Ida Dewey, who was always

helpful and forever proud, thank you.

TABLE OF CONTENTS

LIST OF TABLES viii
CHAPTER 1 STATEMENT OF THE PROBLEM I
A. The Historical Picture 2
B. More Benign Views 4
C. The Older Student 6
D. Statistical Trends 9
E. Purpose of the Study 12
F. Hypotheses 14
CHAPTER 2 REVIEW OF THE LITERATURE 16
A. Age Definition 17
B. Terminology 18
C. Characteristics 19
D. Ability and Achievement 21
E. Non-Academic Comparisons of Older Versus Younger Students 25
F. Reasons for Having Left in the Past 28
G. Reasons for Return 28
H. Problems and Barriers 34
I. Attrition 40
J. Counseling Needs 47
K. Rational and Scope 49
L. Theory 52
M. Components of Psychological Adjustment 56
1. Life Satisfaction and Happiness S6
2. Self-Esteem 58
3. Locus of Control 59
N. Summary 61
CHAPTER 3 DESIGN AND METHODOLOGY 62
A. Subjects 62
B. Sample Size 63
C. Subject Identification and Contact 64
D. Procedures to Provide Subject Identification and
Confidentiality 66

iv

E. Method of Data Gathering 67
F. Instrumentation 69
1. Objective Self—Assessment 69
2. Subjective Self-Assessment 70
3. Self-Esteem Instrument 71
4. Locus of Control Instrument 73
5. Single Item Assessments of Psychological Adjustment 75
6. Grade Point Averages 76
G. Responses to Open-Ended Question: Further Topics
Suggested by Respondents 77
H. Analysis of Data 78
CHAPTER 4 RESULTS OF THE STUDY 84
A. Description of the Population 84
l. Sociodemographic Variables 84
a. Current Student Status 84
b. Age 86
c. Marital Status 87
d. Parental Status 88
e. Income 89
f. Weekly Employment Rate 91
2. Educational Variables 91
a. Previous Education 91
b. Rate of Recent College Enrollment 93
c. Dropouts' Attitudes 93
d. Spouses' Education 94
3. Subjective Self-Assessment of Income and Educational
Variables 95
3. Income 95
b. Reasons for Previous Withdrawal 97
c. Reasons for Return to School Now 102
d. Change or Crises Contributing to the Decision
for the Recent College Entry 105
e. Change or Crises Resulting from the Last Entry
to College 107
f. Reasons for Recent or Contemplated Withdrawal 109
4. Academic Achievement 113
5. Self-Assessed Single Indicators of Adjustment 115
a. Expectation of Degree Completion 115
b. Life Satisfaction 116
c. Happiness 117
d. Getting What Is Wanted Out of Life 118
e. Coping With/Handling Crises 118
f. Self-Assessment of Degree of Control
Over Own Lives 119
6. Formal Test Batteries 119
a. Locus of Control 119
b. Self-Esteem 120
7. Responses to Open-Ended Question 121
8. Summary of Frequency Data 125

B. Findings Across All Student Status Categories 128

C.

D.
E.
F

vi

Sociodemographic Variables

Perceptions About Income

Educational Variables

Relationships Between Selected Sociodemographic and
Educational Variables

Causes of Earlier Withdrawal

Changes or Crises

Reasons for Current Education

Reasons for Withdrawal Now

Academic Achievement

Psychological Adjustment

11. Comparisons by Marital Status

Comparisons of Dropouts and Persisters
1. Sociodemographic Variables

Educational Variables

Causes for Previous Non-Completion of Education

Crises/Changes

Reasons for Reentry

Recent Reasons for Withdrawal

Academic Achievement

Psychological Adjustment

Comparisons by Marital Status

Analyses by Marital Status and Persister/Withdrawer
Status

10. TwoéWay Analysis of Variance by Marital Status and

Criterion Groups

Correlations of Psychological Adjustment Variables

Discriminant Function Analysis

Summary

H
0000qu kWNI-i
o o o o

OQNO‘U‘kNUN
0

CHAPTER 5 DISCUSSION AND RECOMMENDATIONS

A.

Limitations of the Study
1. Retrospective Data
. Setting of the Study
. Unknown Makeup of Population
. Use of Undergraduate WOmen Only
. Non-Objective Nature of Survey
. Instrumentation
Discussion of Major Hypotheses
1. H1: Psychological Adjustment
a. Possible Explanations of Findings
on Hypothesis 1
2. H2: Expectation of a Degree
3. H3: Total Grade Point Average
Discussion of Secondary Hypotheses
1. H1: Correlation of Adjustment Measures
2. H2: Sociodemographic Variables
3. H3: Discriminant Function Equation
Conclusions
1. Theory
2. Implications

autumn

129
130
130

131
133
133
133
134
134
135
135
138
139
139
140
140
141
141
142
142
147

147

149
152
153
156

157

157
157
158
159
161
162
163
166
166

167
175
176
178
178
179
181
185
186
190

vii

E. Recommendations
BIBLIOGRAPHY

APPENCICES
Appendix A Cover Letter
Appendix B Survey Questionnaire
Appendix C Post Card
Appendix D Follow—up Letter

194
198

215
215
216
222
223

Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table

\OQNO‘Ul‘kUNr-i

HHHHHHHHHH
\OGJNO‘U!#UNHO

LIST OF TABLES

Current Student Status

Age By Five-Year Increments
Marital Status
Number of Children Per Mather By Age Category
Number of Children Per Mather

Annual Household Income During Last Three Terms
Weekly Employment Hours During Last Three Terms
Years of Education Prior to Last College Entrance
Level of Married Spouses' Education

Adequacy of Family Income During College
Finances During College: Rating As a Problem
Reasons for Previous Withdrawal

Main Reason for Previous Withdrawal

Total Reasons for Previous Withdrawal

Reasons for Return to College

Total Number of Reasons for Last College Entry
Main Reason for Return to College

Primary Type of Main Reason to Return

Type and Severity of Changes/Crises Contributing

to Return

viii

85
87
87
89
89
90
91
92
95
96
96
98
100
101
103
104
104
105

106

Table

Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table

Table

20

21
22
23
24
25
26
27
28
29

31
32
33

35
36
37
38
39
40
41

42

ix

Type and Severity of Changes/Crises Resulting
from Return

Current Reasons for Withdrawal

Primary Current Reason for Withdrawal

Grade Point Averages (GPA's)

Expectation of Completing Degree Without Further Breaks

Life Satisfaction During and Since School

Happiness During School

Getting What Is Wanted Out of Life

How Well I Cope With/Handle Crises

Self-Assessment of Control Over Own Life

Open-Ended Responses by Categories

Non-Overlapping Open-Ended Responses by Categories

Frequent Data (n a 361)

Literature Findings Compared to This Study

Sociodemographic Variables by Student Status

Income by Marital Status

All Subjects by Marital Status

Psychological Adjustment by Graduates and Withdrawers

Findings on All Subjects and on Criterion Groups

Dropouts Compared to Persisters Within Marital Status

ANOVA Using Graduates/Withdrawers and Marital Status

Correlation Matrix of Significantly Related
Psychological Adjustment Variables

Classification Results of Discriminant Function

Analysis

108
111
112
114
115
116
117
118
118
119
123
124
126
127
129
132
137
145
146
149
151

152

183

CHAPTER 1
STATEMENT OF THE PROBLEM

A "quiet revolution" is taking place on college and university
campuses across this country (Apps, 1981). It began in the 1960's and
1970's and continues to build. This revolution is the return of
thousands of adults to higher education. The increased numbers raise
a new dilemna, "...that colleges and universities can not continue
with business-as-usual" (Apps, 1981, p. 11). This change is coming
about, according to Bishop and Van Dyk (1977), because of the
increased numbers of convenient colleges offering courses tailored to
adults, the need to learn new skills to replace those made obsolescent
by technological progress, and increased desire for training to enable
professional advancement.

Cross (1981) sees the reasons for this change as a mixture of
declining traditional female roles and rising new roles coming about
from children entering school earlier and staying in school longer, an
increase in divorce, and technological changes reducing time needed
for family and home care. A further reason is the resulting social
change including the explosion in the need of women to work made
neccessary as a consequence of these other changes. The problem is
that wife and mother roles have not changed as rapidly and continue to
exert their pull on the women who are going into the educational and
working world. Social and technological change push women out of the
home while labor market and educational opportunities pull women out.

Education then often becomes the mode of transition from home to work.

Lipman-Blumen (1975) cites recent trends for women: They are
"living longer, postponing maternity, having fewer children, more
often having no children, heading more households, and living alone
more often" (p. 680). Clarke (1975) points out that while women once
had two stages of life, pre-marital and child-bearing, the increased
life-span now gives an equivalent of a second life after the last
birth which generally occurs around age 30. It is this "second life"

that is frequently used for education and a career.

THE HISTORICAL PICTURE

Historically older women in higher education have either not been
valued or have been rejected or neglected. McDonald (1979) cites that
in the mid 1800's the debate on women was whether or not to even allow
them to enter college. She quotes a college president in the 1850's
who wrote, "Of what use degrees are to be to girls, I don't see,
unless they addict themselves to professional life" (p. 10). For
years the idea that the women belong in the home was emphasized. Back
in 1909, Wells said, "It is for the advantage for mankind that
superior women should become mothers" (p. 733). Citing the decreased
birthrate for college women, he feared that the idea of higher
education for women could both make them feel too superior to men to
marry them and could "deplete the stock." He worried, "If women's
interests become materialized women will surely be degraded to the
base level of all material competitions" (p. 739). However, he

recognized that "Training of women is both natural and inevitable" (p.

731).

There was little change by the time Goodsell wrote in 1924:

"The ideal of womanhood for the ages has been a modest, docile,

clinging creature, trained in home-keeping arts, with physical

charms sufficient to compensate for an empty mind, and with

unlimited capacity for self-immolation" (p. 7).
He asked if higher education is "unfitting" women for marriage and
motherhood and if those women are leaving maintenance of the race to
the lower classes. In more recent times the psychiatrist, Monsour
(1963), takes up the strain saying, "A man's task is to learn things;
a woman's is to train" (p. 17). He further asserts that females are
hard to educate at certain times of their lives: during adolescence,
when in love, and when having a baby. In their late twenties to the
mid—thirties after cessation of child bearing, he says women are at a
"critical age" when education again is possible and there is a "real
force toward independence" in women.

The President's Commission on the Status of Women in 1963 says,

"The means of aquiring or continuing an education must be

available to every adult at whatever point he or she broke off

traditional normal schooling. The structure of adult education

must be drastically revised. It must provide practicable and

accessible opportunities, developed with regard to the needs of

women..." (American Women..., 1963, p. 13).

However, ten years later Mitchell (1974) reports,

"We may be relatively sure that he (the over-35 student) is not

being identified as a particular student and is not really being

encouraged and supported in his college program" (p. 22).

Harrington (1977) confirms this view:

"When asked if his institution should teach adults and help them

with their problems, the president of a prestigious midwestern

university snapped to his colleagues, 'I'm too busy for that. I

have too much to do. I can't do everything. Let somebody else
take care of the grownups'" (p. 8).

Watkins (1974) asserts, "A second women's liberation movement is
underway, this one on the nation's campuses....The college and
university response has been both positive and negative--and generally
slow." While she affirms that some schools began to adjust in the
sixties, "The vast majority of institutions appear not to notice" (p.
6).

As recently as 1974, even the title of the article used by Class
and Harshburger of "The full—time, middle aged adult student in higher
education" refers only to male students, not just a generic "he" used
for all students. And it is not by chance that a title used in 1979
by Daniels begins, "Welcome and neglect: The ambiguous reception of
re-entry women...." Fischer-Thompson (1981) points out that when the
last adult surge came to higher education, the World War II veterans
were welcomed, had wives to take care of their children if there were
any, received federal financial support, and gained much praise and
encouragement for their educational efforts. The women in today's
surge too frequently go without that encouragement and often must cope
with discouragement from family and friends as well as from the
institutional response.

"They may be told that they are too old and too late, and that

they should be home taking care of their family, and they often

have problems even getting financial assistance and childcare"

(Fischer—Thompson, p. 1).

MORE BENIGN VIEWS

The popular literature has perhaps been more encouraging and more

responsive to the phenomenon of the older student going to campus than

the institutions themselves have been. The Farm Journal in 1962

 

raises the issue with an article whose title asks, "Should Mom Go Back
To School?" (Gillies, p. 59). In a 1963 article called "Women:
Second Wind", lim§_says, "Nobody's more noisely dissatisfied than that
symbol of stability--the 40'ish housewife with teenage children and a
reasonably successful husband," and suggests education as a good
alternative (p. 56). Articles in the mid—seventies were in such a

variety of magazines as Working Woman (Bestor, 1973), Weight Watchers

 

(Westin, 1975), and Dynamic Maturity (Carlson, 1977) which all

 

encourage older women to return to school. Articles in the eighties
have appeared in Psychology Today (Cottle, 1980), Newsweek
("Grownups...", 1981), and Essence (Smalley, 1982) which urge that if
you are thinking of more higher education, "Do it now" (Smalley, p.
40). The emphasis on higher education in these articles has been very
positive.

Beginning in the mid-70's, major newspapers began touting older
students. Early emphases were on successful older students as in the
Wall Street Journal's front page article, "Mrs. Suzy Coed: More Older
Women Return to College; Most Do Very Well" (Elsner, 1972). Later
front pages emphasize the economic impact of the "Graying of the
Campus" (Graulich, The Wall Street Journal, 1977) and colleges' need
for these students (Maeroff, New York Times, 1978).

Perhaps more attention will be paid to the older student as more
and more schools agree with people like Milton Stern, Dean of
University Extension at Berkeley who asserts that the new older
consumers of education,

"...are not automatically ours to educate. We do not get them to

enter merely by opening the doors. Adults who pay for their

courses——or whose companies pay-—will look carefully at what we
have to offer." (Hechinger, 1975, p. 18)

Weinstock (1978), speaking of returnees over 55, reminds us to pay
attention to those older of the older learners because they have the
expendable dollars for school. And the Carnegie Commission (1973)
warns,

"Higher education will no longer be a growth industry unless an

entirely new constituency can be attracted to its institutions,

and unless continuing education becomes an accepted pattern in

our society." (p.6)

THE OLDER STUDENT

The older—than-traditional students are the hope of the future
and the oft-neglected resource of the present for higher education.
They are increasingly the more flexible source of new members of the
student\body as the traditional 18-22 year old population is
shrinking. Older students show vast heterogeneity relative to age,

amount of education previously completed, experiential and work

backgrounds, financial autonomy or dependence, and personal, family

and social commitments. Each comes as a result of a deliberate action

out of a different context and perhaps having a different motivation

than that which may bring a traditional-aged student to college. When

students complete high school, they and their peers are considering

and discussing what to do with their lives and whether and where to go

to college or to get a job or to marry. At that stage, going to

college is a normal developmental choice made by approximately half of

all high school graduates (U. S. Bureau of the Census, 1983b). The
college experience is then their way of establishing identity and

autonomy.

However, to come to college after being employed, raising a
family, being a full-time homemaker and/or volunteer, or combining
some of these activities is a very non-traditional decision. It is
made with fewer knowledgeable supporters (counselors, teachers, peers)
readily available at a time when a person has already filled his or
her available time with other activities. Something has to shift
within their lives to permit time to go to school. Roles must be
changed or added. Finances have to be stretched or changed to
accomodate paying for college. This decision at a later stage of life
is to become a non-traditionalist--to do something less expected, less
usual, and sometimes less supported by family, friends, and/or
colleagues.

Despite the fact that older students come with often
unacknowledged strengths such as academic abilities, the breadth added
by life experience, and rather high motivation, this is still clearly
a time of change and transition (Greenfieg and Goldberg, 1984).

Muskat (1978b) believes that returning women are in a state of
psychological transition and of conflicting pressures, but that they
can gain self-esteem and autonomy through this transition.

Although Muskat thus sees college as a period of growth,
Wertheimer and Nelson (1977) suggest women need to already be
psychologically strong before coming to college: "A woman must
overcome conditioning from childhood before she decides that she is
important enough to ask the family to rearrange its schedule to
accommodate her need for further education" (p. 65). Despite this,

Gray (1975) asserts,

"The hardest part of going back to school is the actual
decision to do it. Once past that the rest is much easier.
Seriously committing your energies to school, possibly as a
full-time student, is a sobering thought. Essentially you have
got to begin believing you are not so dull that you can not pick
up new ideas" (p. 13).

She also claims that although greater earnings and more opportunities
are considered to be the traditional rewards for college effort, for
older women students the most sweeping changes will be psychological
changes which stem from growth and general consciousness.

It takes courage to believe in one's self enough to take the
risks of attending school when one is older. In fact, Douvan (1981)
suggests that older students who have already established their
identities may even be risking their own identity now by this
transition into college life. Batdorf (1976) and Douvan (1981) also
point to the relative safety of college for the traditional student to
learn adult roles and responsibilities while adjusting to a somewhat
dependent (safe) world but note that this same adjustment backwards to
dependency is a difficult step for older students.

Hopson (1981) gives a much more optimistic view of transitions:

"A transition simultaneously carries the seeds of our
yesterdays, hopes and fears of our futures, the pressing
sensations of the present which is our confirmation of being
alive. There is danger and there is opportunity, ecstacy

and despair, development and stagnation, but above all there

is movement. Nothing and no one stays the same. Nature

abhors a vaccuum and stability. A stable state is merely a

stopping point on a journey from one place to another. Stop

too long and your journey is ended. Stay and enjoy but with

the realization that more is to come. We may not be able to

stop the journey, but we can fly the plane" (p. 39).

So what is this transition that is going on? Eric Erikson (1959)

would call this the task of integrity, the task of accepting our own

responsibility for our lives and making what we can of them. Astin

(1976c) agrees that adult women are searching for both integrity and
identity. "Although men become more affiliative as they mature, women
show a great need for independence, become more outward and assertive,
and remove themselves somewhat from the role of nurturer" (p. 55).
More and more older persons and women in particular are carrying out

this transition at least in part in an educational setting.

STATISTICAL TRENDS

The data shows the college age population to be in the midst of
three escalating trends. The traditional age college population is
shrinking, the older college student population in growing, and the
rate of growth of the older woman student population is outstripping
all other growth.

According to national statistics, the traditional college age
population has increased nearly 20% from 1970 to 1980 but is expected
to decrease 15% from 1980 to 1990. In the decade until 1980 the 25-34
year old population increased 43% and is expected to gain another 14%
by 1990. The most radical change is in the 35-44 year old
population, which increased 11% from 1970 to 1980 and is expected to
increase 42% by 1990 (Trends in Higher Education, 1982). The 18-24
year old group is eXpected to have peaked in 1981 and to have declined
by five million by 1990 (U. S. Bureau of the Census, 1983b). By 1981,
students 21 and under at all levels became a minority, and women 25
and older had become one fifth of the student body (U. S. Bureau of
the Census, 1983a). The sagging birthrate which led to a drop of
600,000 first graders in each of the years 1973, 1974, and 1975 is

causing a corresponding drop in the traditional college age population

10

from 17 million in 1978 to 13.5 million in 1988 (Hechinger, 1975). As
the median population age changes from 27 in 1970 to a predicted 34 in
1990 and to over 37 in the year 2000, colleges and universities will
have to be prepared to respond (Ostar, 1981). A Carnegie Council
(1980) expects that by the year 2000, at least 50% of undergraduates
alone will be over 22 years old.

Our total higher education enrollment grew 41% from 1970 to 1980
but is expected to have no further growth by 1990 (Trends..., 1982).
In fact it is projected that while those age 25 to 64 will increase
their enrollment by nearly one million by then, those 14 to 24 will
decline in enrollment by approximately 800,000. Because many older
students will be part-timers, their increase will offset about 70% of
the decline (Magarrell, 1981).

While from 1972 to 1981 college enrollment increased 33%, it
increased 63% for women (U. S. Bureau of the Census, 19838).
Furthermore, while in 1972 (which was the first year college data was
kept on all ages) there were 100 men to 74 women, by 1981 the
preponderance had switched and there were 108 college women to 100
college men (U. S. Bureau of the Census, 19833). At about the same
time that students 21 and under became a minority, so did men. Women
increasingly hold an edge in each age bracket until those over 35 have
a nearly 9 to 5 edge over men (Magarrell, 1981). Women passed men in
1979 in college as a whole, in 1978 in undergraduate programs, and in
the 25-34 year age bracket in 1980. They had already passed men in
the over—35 age bracket before 1970 when data was not kept on the

highest age brackets. Women passed men in Michigan in college in

11

1980 (U. S. Bureau of the Census, 1983b). While men slacked off on
completion of college in the later 1970's, women steadily increased in
the proportion of completion. Twenty-three per cent of 25-29 year old
males are now college graduates and 20% of similar women are.

Not all students who enter college, however, remain to complete
their education. The cumulative data on females shows that many drop
out after the freshman year and even more after the sophomore year.
There is then a major slacking of dropouts following the junior year.
Approximately half of those who begin college do graduate near
schedule.

Many who drop out at an earlier age come back later to complete
what they began. 0f women 25 and older who begin college, singles are
more likely to get a degree, marrieds next, divorced women third, and
widows least likely to get a bachelor's degree (U. S. Bureau of the
Census, 1984). Over one third of those with four years of education
continue their education (U. S. Bureau of the Census, 1984).

All this data shows the increasing importance to higher education
of older women and shows that they may have increasing clout in the
academic world simply because of their swelling numbers. Although a
large share of the growth in higher education has gone to two year
community colleges, four year institutions are now getting a larger
share of the increase.

In a time of shrinking enrollments, the competition between
colleges to get and retain students is intensifying, for to lose
enrollees is to lose the financial support which allows colleges to

retain the personnel, courses, programs, and services they see as

12

neccessary to their existence. As the traditional age student
population declines, the attractiveness to the colleges of the adult
student increases. Older students can fill the empty chairs and
coffers-~if they can be attracted and retained. It may become
increasingly important to attract their interest, to woo them into
entering college and then to design their experiences in ways to keep

them as students in what may be a difficult and taxing experience.

PURPOSE OF THE STUDY

Although it is useful to understand older women students to know
who are likely candidates to be wooed into colleges, it is equally
important to know what causes them to do better or worse while in
college and know why some remain to graduate and others drop out.

Only with this knowledge can a college which truly wants to attract
and keep older students plan which external or situational variables
need adjustment to facilitate older students' acquisition of an
education. Educational institutions need to discover how to identify
and offer support to women at risk for doing less well and/or dropping
out. If some common features can be identified, then we can turn to
designing ways to target and help subgroups of older women to remain
in and to succeed in school.

The question is, who copes and stays and who doesn't? Anecdotal
data began to accumulate which conflicted with an early hypothesis
that women in the midst of situational/personal stress would have
extra difficulty achieving or staying in school. While this appears

to be true for some women, others with apparently equal stresses and

13

demands are staying and apparently succeeding. In fact, the anecdotal
data appears to be saying that part of the reason for better grades in
older women is that they set and achieve high standards for gauging
success. These succeeding women still repeatedly cite financial
problems, adequate time, balancing of many roles, family demands, and
personal expectations as being burdens. Continuing discussions and
inconsistency of concrete indicators lead to a new approach.

The hypotheses of the present work are based on the theoretical
premise that people in good psychological health are better able to
cope with all aspects of their lives. If so, well-adjusted older
woman students will persist and get better grades than will those with
poorer psychological health despite variable stressors or
socio-demographic variables. Specifically, people who feel they do
have power in their own lives and feel confident and generally happy
and satisfied should manage many situations including college better
than those who feel much of their lives are controlled by luck or by
external persons or events or those who feel unsure about themselves
and unhappy or dissatisfied with how their lives are going.

Because it is difficult to differentiate who among current
students are "doing better" than others (Grades have been the primary
criterion but some question whether that criterion is adequate), this
study attempts to identify variables which differentiate recent older
women college graduates at the baccalaureate level from recent college
dropouts. Using these variables, we can make predictions as to which
current students may be more at risk for not completing their studies.

It is hoped that some day educational institutions will attempt to

14

identify these students early in their college careers and design
special ways to improve their chances of staying in school.

Several demographic and situational variables are studied and
used first, to describe the population, second, to find any
correlations with persisters or dropouts and with higher/lower grade
point averages (GPA's), and third, to use in discriminate analysis.
The hypothesis that selected psychological components are able to
differentiate students who persist and/or get better GPA's is tested.
All variables are used to find those which best differentiate the
persisters/drop out and higher/lower GPA groups and those which may
indicate higher chances of either success or risk for older women

undergraduates.

HYPOTHESES

The primary hypotheses are:

H1: Among women undergraduates recently in college, those who
persist to graduation are more likely to show better psychological
adjustment, as measured by having higher self-esteem, internal locus
of control, reporting more happiness and life satisfaction, coping
better with crises, and feeling they are getting more of what they
want out of life than those who drop out.

H2: Among women undergraduates recently in college, those who
persist to graduation are more likely to have believed that they would
graduate than those who drop out.

H3a: Among women undergraduates recently in college, those who
persist to graduation are more likely to have higher total Grade Point

Averages than those who drop out.

15

H3b: Among women undergraduates recently in college, those who
persist to graduation are more likely to have higher recent Grade
Point Averages than those who drop out.

Secondary hypotheses are:

H1: The measures of psychological adjustment will be
significantly correlated with each other.

H2: The women who persist to graduation and the women who
withdraw from college will not be significantly different on the
sociodemographic variables of age, marital or parental status, income,
or employment.

H3: An equation of variables can be formed which significantly
distinguishes dropouts from graduates and distinguishes groups of
current students who will be predicted to graduate from those who can

be expected to drop out.

CHAPTER 2
REVIEW OF THE LITERATURE

A review of the literature reveals that many articles and some
books are based on simple theorizing or on data primarily informally
gathered or on limited data from, for example, only one or two support
groups. In selecting material for review here, articles without basis
in research were rejected unless they offer some useful theoretical
strength. Findings are used only from studies which have been
adequately done and are not used, for example, if results are based on
an informal group or if conclusions were drawn from a few interviews
or if the methods of data gathering and analysis appear to be missing
or inadequate.

The newer literature is becoming increasingly sophisticated
statistically and therefore is often somewhat stronger. Much of the
older cited literature is primarily or purely descriptive, since
frequencies were then the primary tool used to assess data. More
recently many more and varied tools are used. However, throughout
this review, only the better studies and literature reviews of the
time or of the subject are cited. This critical selection of
literature means a majority of the citations found are omitted here
primarily due to their weaknesses.

The literature is vast since many topics are included in this
study of older women undergraduates. Much of what is written is on
the traditional-aged student, much is on students of all ages and both
sexes, some is on older students in general, and some is on older

women. Although some issues such as reasons for return to college or

16

17

barriers to return appear to have been overstudied and information has
become quite redundant (and therefore not further cited), a breadth of
background is needed on several topics to appropriately understand the
complexity of sociodemographic, educational, experiential and
psychological variables of interest in this study.
AGE DEFINITION

Unfortunately there has been absolutely no uniform definition of
the older student. Some of the literature simply calls them "adult
students" with no definition as to age (Burgess, 1971; Morstain and
Smart, 1977). _Others give no definition but at least speak only of
women (Marple, 1976) or give some non-age definitions. Ancheta (1980)
calls full-time day students "traditional" students and considers that
part-time evening students are non-traditional and presumed to be
"mature". Farmer and Fyans (1983) study women who have been out of
school two years and are married. Tittle and Denker's (1977) major
literature review of re—entry women covers those who have been out of
school at least a few years and are in credit courses currently.
Other definitions give either just a minimal age or add a requirement
of having been out of school for at least two or more years. These
minimal ages begin with 20 year old freshmen (Reed and Murphy, 1975)
or 21 year olds (Perkins, 1961) with these labelled by Roelfs (1975)
as "late bloomers, lifelong learners, retrainees, job upgraders,
veterans, housewives, senior citizens, and dropouts" (p. 5). Others
begin at 23 (Doty, 1966; Hiltunen, 1965; Ice, 1971; Myers, 1964), 24
(Stephan and Wheeler, 1969) or 25 (Aanstad, 1972; Adelstein, Sedlacek,

and Martinez, 1979; Kimmell, 1976; Roehl and Okun, 1984; Sensor, 1964;

18

Smallwood, 1980; White, 1984). A few begin in the later 20's
(Erickson, 1968: 26; Kasworm, 1982: 26; Badenhoop and Johansen, 1980:
28; Erdwins, Tyer and Mellinger, 1980: 29). Others start at 30
(Kaplan, 1982; Kahnweiller and Johnson, 1980; Rawlins and Davies,
1981; Reehling, 1980) and 35 (Hooper, 1979; DeWolf and Lunneborg,
1972) or at higher ages (Bross, 1967: 40; Hooper and Traupmann, 1984:
50; HOOper and March, 1978: 62). Age variations make comparison of
studies difficult.

Because 25 appears to be the modal age definition and because
the census data on students uses an age bracket beginning at 25, that
minimum age will be used in this study. Much useful data is missed by
beginning a study at a much older age than 25, and those in their
mid-twenties have lived enough adult years to have probably
established some non-student life experiences and life style

variations which are of value in most studies.

TERMINOLOGY

The terminology to describe the students who are older is
diverse. "Non—traditional" has often been used (Anchata, 1980;
Dwinell, 1980; and Warchall and Southern, 1986). This language,
though, seems to be confounded because the same term is used for women
students in fields which are or have been primarily male dominated
(engineering, medical or law school, etc).

"Returning" or "re-entry" is commonly used (Adelstein, Sedlacek,
and Martinez, 1979; Hooper, 1979; Roehl and Okun, 1984; Tittle and

Denker, 1977). However, here the emphasis is on absence from school

19

rather than age, and the term seems as apprOpriate for a 20 year old
as it is for an older student.

"Adult student" is probably the most prevalent term in recent
research and is likely to become the normative term (Burgess, 1971;
Marple, 1976; Morstain and Smallwood, 1980; Rawlins and Davies, 1981;
Smart, 1977; White, 1984). This term seems to connote that those
younger than the defined age are not really adults, a perception which
could further widen the rapport gap some perceive between the students
of traditional and of non-traditional ages.

"Mature" adult or student is also used (Aanstad, 1972; Doty,
1966; Erdwins et. al., 1980; Hooper and Traupmann, 1980; Kaplan, 1982;
Myers, 1964; Perkins, 1961; Reed and Murphy, 1975; Sensor, 1964;
Stephan and Wheeler, 1969). Again, this leaves a stigma of the
traditional age student being seen as immature.-

Therefore this research will use the term "older student," since
specifically an age factor is used to define the population and since
this term is less likely to be confounded than "non-traditional," to

be less exclusive than "returning,"

and to be less alienating than
"adult" or "mature" students. This term has been used previously

(Bross, 1967; DeWolf and Lunneborg, 1972; Hooper and March, 1978;

Kasworm, 1982; Roelfs, 1975). (Kimmel, 1976, uses the term

"older-than—average").

CHARACTERISTICS
What do we know about the characteristics of the older student
population? Many of the studies have been done on mixed—sex samples,

some of which separate the sexes in the analysis and some of which do

20

not. The married rate varies from a low of 58% (Geisler and Thrush,
1975) to a norm of 75-77% (Erickson, 1968; Folland, Pickett and
Hoeflin, 1977; Johnson, Wallace, and Sedlacek, 1979; Steele, 1974).
Some studies lump those widowed, separated, and divorced together
while others give these as three categories. The divorced rate is
blurred because of the lumping, but varies from approximately 8%
(Erickson, 1968) to 15% (Magill and Cirksena, 1978) or possibly more.
Sands and Richardson, 1984, have 29% in the non-single non-married
category. Singleness ranges from frequent single digit figures
beginning at 5% (Sands, et.al., 1954) up to 25% (Geisler and Thrush,
1975). The rate of having children varies from approximately 50%
(Johnson, et, al., 1979) to 90% (Steele, 1974). One study of 303
"nontraditional" women in a college of education shows frequencies
quite different than these normal frequencies. Therefore, her
findings are reported separately: Dwinell (1980) found only 48%
married, 20% divorced, and 29% single while only 43% had children.

Although the various beginning ages confound the data, the
majority of the studies show an average age in the early to the mid
30's and a modal age often in the upper 20's to near 30. The mean
number of years out of school varies from three or four to fifteen or
more. The majority have previous college education varying from 61%
(Steele, 1974) to 97% (Geisler and Thrush, 1975). Although a few
studies cite that a minority are working (Magill and Cirksena, 1978:
23%), most show at least 50% of the older students work.

In a major summary study, Cross (1981) cites Solmon, Gordon, and

Ochsner's analysis of 172,400 freshmen over 21 from 1966-1978.

21

Results confirmed an earlier study in which, compared to college
students of traditional age, the older students were more likely to:
1. be socio—economically disadvantaged, 2. be more concerned about
financing education, 3. have made lower grades in high school and four
year colleges in their major field, 4. see college's benefit as
monetary, 5. have lower educational aspirations, 6. have more
difficulty in universities and least difficulty in community college,
and 7. be more prone to dropping out and dropping in. Since these
were freshmen, though, it is highly likely they are different from the
majority of older students who had some college education previously.
Ostar (1981) summarizes adult learners differently saying they
are: better educated than non-learners, better off financially, more
likely to be employed and to be professional or technical workers,
more likely to be single or divorced, and slightly more likely to be
urban. This description seems to better fit the total population of

older students.

ABILITY AND ACHIEVEMENT

Pertinent questions about these older clientele of academia
include: What is the ability they bring with them when they come to
education? And how well can they achieve academically since that is
the standard measure used in higher education?

It has long been assumed 1. that ability declines with increasing
age and 2. that older students simply do not have as much intelligence
or are not as capable of learning as younger students and are,
therefore, more susceptible to failure. Owens (1966) studied Army

veterans who were first tested in 1919 and were retested in 1950 and

22

1961. The longitudinal data shows "...relative constancy in mental
ability test performance" in the decade between their 50's to the 60's
(p. 311). Several studies show that verbal abilities either do not
decline with advancing age (Lunneborg, Olch and DeWolf literature
review, 1974) or may actually increase (Honzik and MacFarlane, 1973),
particularly with activity in educational activities (Lunneborg,
et.al, 1974). This is supported by Stephan and Wheeler (1969) who add
that each increasing age category does better than the younger age
category (with the oldest category being 40 and over). Cagiano,
Geisler, and Wilcox (1977—1978) show a significant GPA difference for
those whose education had been interrupted with the GPA increasingly
significantly higher with longer interruptions. They interpret this
finding to be a result of increased maturity, motivation, and purpose.
Knox and Sjogren (1964) find no difference between older and younger
students on the WAIS or in experiments to test learning ability. It
appears that Lunneborg, et.al. (1974) are correct in saying the
longitudinal data is more important and does not support a decline of
intellectual functioning with age.

Older students do better than younger students academically in
studies of welfare mothers (Young, 1977), of women over 40 (Halfter,
1962), of community college students over 21 (Preston, n.d.) and of
university freshmen (Hull, 1970). Reed and Murphy (1975) find sex
differences with women doing better than men. Hansen and Lenning
(1963) find older students do better than younger, and women of all
ages have better GPA's both in high school and in college than do men.

In studies comparing older students to freshmen norms, Lunnenborg,

23

et.al. (1974) discover that for both sexes, the older students average
at the 75%ile of freshmen norms on verbal and 25%ile of freshman norms
on quantitative scores. Hiltunen (1965) finds older women to be at
the 80%i1e verbal and 33%ile quantitative. Although Hull's (1970)
older and younger students have nearly equal ability test scores and
high school ranks, the older students perform significantly higher
than predicted but the younger do not.

On qualifiying examinations or entrance tests, older students
frequently do more poorly than traditional-aged students (Perkins,
1971). But once they actually enroll, the evidence is relatively
uniform that older students will have significantly better GPA's
across both sexes than the traditional age students (Ice, 1971; Byrne,
1974) and that women get better grades than men (Ice, 1971; Carnegie
Commission, 1973). Byrne's (1974) less expected finding in a
well-done dissertation is that married older students do better than
single older students. This dimension is rarely studied.

It thus appears that older students are quite capable although
they may have some difficulties with quantitative subjects. They do
achieve academically, even better than younger or male students. The
next question is: Can we predict which older students or specifically
older women will have better academic performances?

Lavin's (1965) literature review of attempts to predict academic
performance is representative of studies of traditional-age students.
He finds the best single predictor for traditional students was high
school GPA or rank, but men are more predictable than women. He finds

no consistent findings relative to anxiety, mixed results about

24

motivation or need for achievement, 8 positive relationship between
introversion and academic performance, some negative correlation
between lack of impulse control and school achievement, and a
correlation of self-image of intellectual abilities with increased
GPA. In an attempt to find appropriate predictors for older students'
grades, Lunneborg, et.al. (1974) find no individual significant
correlations with background characteristics or tested abilities and
find that high school GPA is as good a predictor as recent test
scores. Their regression formula does show the background variables
of high school grade point average, activities engaged in in high
school, years since having been a full—time student, having an
advanced degree goal, and expected years to the bachelor's degree all
are more important to multiple prediction than is test performance.
Metz (1966) reports the high school GPA is not as valuable a predictor
for returnees as it is for traditional students. This agrees with
Cagiano, Geisler, and Wilcox (1977-1978) who say that the GPA is a
less valuable predictor if people are out at least three semesters.
Apparently we have not yet found good predictors for academic
achievement of older students. Since such predictors are not found in
the more easily studied concrete variables, it appears to be time to
look for such predictors in the internal makeup of the students rather

than in their social or academic histories and abilities.

 

 

25

NON-ACADEMIC COMPARISONS OF OLDER VERSUS YOUNGER STUDENTS

What do we know about older women students besides their
demographic and academic characteristics which could help understand
why some remain in school and others leave?

Several studies compare older versus younger college students on
the dimensions of personhood, values, psychological make-up, needs,
and orientations. As Badenhoop and Johansen (1980) cite, older
undergraduates have more going for themselves, are not just being
pushed by parents, and have therefore higher educational goals and
consequently higher GPA's. Several studies simply note that their
wealth of life experience makes more well rounded persons and more
experienced persons who come back to school, which is both an asset
and a detriment in a university or college setting (Apps, 1981;
Krings, 1976). They know why they have come back and are very
purposeful and are willing to work to achieve it (Apps, 1981).
Although they are adding a role (compared to those younger persons for
whom being a student is the primary role), their advantage is that
they have already demonstrated they can handle responsibility and are
more likely to know what they want out of life and are willing to make
the sacrifices to get it.

Older undergraduate students' value priorities are different
than those of their non-student peers but not different than those of
traditional—age students (Pirnot and Dunn, 1983). Women students over
30 have increased academic, intellectual, and aesthetic orientation;
more liberal attitudes, less dogmatic religious beliefs, and are less

interested in material possessions and social interaction while

26

showing more "attitudes of emotional stability" on the Omnibus
Personality Inventory (Espersson, 1975). Older women students score
significantly higher on the California Personality Inventory scale of
achievement via independence and on psychological—mindedness than do
their younger peers (Marple, 1976). They show less hostility,
anxiety, and depression than younger students, in contrast to the
classic picture of the depressed mid-life woman (Clements, 1974). The
fear of success in women undergraduates at Rutgers is significantly
higher in the younger group and for those unmarried or with no
children or unemployed according to Tomlinson-Keasey (1974). She
concludes that fear of success may be more connected to anxiety about
one's role than to be an accurate predictor of achievement.

In a theoretical paper, Datan and Hughes (1985) describe
returning women students as no longer needing to conform to the
adolescent norm of non-intellectual femininity, less bounded by
parental expectations, decreasingly relationship-oriented, and less
bound to the stereotyped female path of finding identity through
relationships. However, they see such women as more bound by
relational commitments made earlier and therefore perhaps hampered now
from full achievement.

While older women students are highly motivated and responsible,
they are more likely to have problems in academic behaviour:
"...adjusting to university life, including learning academic
procedures, rusty study skills, inability to concentrate, and
adjusting to problems associated with unlearning" and have

"unrealistic goals, poor self image, social-family problems, and

 

 

27

sometimes an excessive practical orientation" (Apps, 1981, p. 51).
This agrees with Krings (1976) and Knowles (1969). Krings and Knowles
further believe the mature person learns best in a setting of mutual
responsibility but not one of dependency and also wants learning that
can be applied immediately. Older students are generally found to
have less desire for forming social relationships or for attending
school because someone else wants them to and to have a higher
cognitive interest or internal drive for the pursuit of knowledge
(Wolfgang and Dowling, 1981).

Most of these authors suggest that because of the differences
between older and younger students, older students need some different
design in the educational process. However, a few recent authors
question this premise and also whether counseling needs to be any
different. Chandler and Gaelerstein (1982), in looking at upper
division undergraduates, conclude that since neither age nor sex
predicted students' satisfaction or dissatisfaction with their
academic experience, no special classes were needed. However, these
investigators did not have a clear youth group since their youngest
group was defined as being ages 20 to 29. Warchal and Southern (1986)
find no difference in perception of counseling needs by age or sex and
discover that the oldest have most increased concern particularly with
parenting but otherwise need little special counseling. These
findings are probably the result of not differentiating among older
students in any way other than by age and sex and therefore perhaps

disregarding possibly large minority concerns.

 

 

28

Johnson, Wallace, and Sedlacek (1979) find no differing needs
when comparing returning women to returning men and to traditional age
students of both sexes. They conclude therefore that students need
to be clarified by types rather than simply by ages. Suchinsky (1981)
suggests a possible framework for looking at older woman students. He
describes four groups who have different needs and purposes as being
the empty nester, the displaced homemaker, blue collar wife who needs
income, and the oldest students who use education to center their
lives and who are often recent widows (p. 31). Perhaps a new ways of
categorizing these people is needed, but using such purely social or

economic definitions seems simplistic.

REASONS FOR HAVING LEFT IN THE PAST

Most writers agree that the returning or older woman student did
not complete her education at an earlier age primarily because of
marriage (Doty, 1966; Steele, 1974; Brandenberg, 1974. Sewall, 1984,
describes this as "family responsibilities". Geisler and Thrush
(1975) claim that marriage and children came third with getting a job
and needing money coming first and second. Most others agree that
these are the top three reasons, although Brandenberg (1974) cited
lack of interest as being a distant third. Steele (1974) adds that
40% of woman who left and later returned have never given up their

desire to return.

REASONS FOR RETURN
Voluminous literature explains why older students, particularly

women, return for an education. However, it is rather hard to

 

 

29

generalize across the studies because the variety of options offered
in survey instruments and different ways of categorizing answers.
Cross (1981), in a good summary and review, cites that the reasons
still primarily fit the format used by the Commission on
Non-Traditional Study (CNS) national survey. Many of this survey's
items had been first formulated by Burgess (1971) although he
significantly omits any personal fulfillment dimension. Cross cites
the reasons in terms of desires: to know, to reach a personal or a
social or a religious goal, to escape, to take part in social
activity, to comply with formal requirements, for personal
fulfillment, and for cultural knowledge. Her review notes that
personal satisfaction is typically given as a single main reason by
one-third of returnees and as one of several reasons by one-half.
Knowledge for its own sake is chosen by 10 to 39%, obtaining a degree
or certificate by 8 to 28%, and over one-third say that escape is a
reason. She sees the main trend of the seventies as being toward an
increasing proportion of personal satisfaction or recreational
reasons, exceeded only by job related reasons. However, her study
goes across both sexes, and older women students differ from this
generalized picture.

In the sixties, a typical study asserts women return because of
boredom (Bross, 1967). They are tired of volunteering and seek a
refuge from a too-empty house or from the unhappiness of family or
health problems and need to forge a new pattern for life. Another
study adds they have an unfulfilled desire for knowledge (Doty, 1966).

Erickson's comprehensive dissertation study (1968) learns that getting

 

 

30

a degree is the primary reason with intellectual stimulation second
and utilitarianism reasons following. Surprisingly a majority of
college student wives rate a degree of more importance after marriage
than they had seen it to be before they had married (Hildebrand,
1967).

In the early 70's, the literature emphasizes personal growth
reasons for returning to school. Letchworth (1970) sees older women
students as having a second identity crisis after having resolved the
first identity crisis by becoming a wife and mother. Now the crisis
is whether to add a career and, if so, which career. They are also in
an integrity crisis, wondering what is the meaning of life. College
can help them try to understand and resolve both of these crises. In
fact, half of the woman undergraduates 35 and older in one study came
because of a vocational family or personal crisis (Carlsen, 1973).
One-fourth in another study came ngg_because of a crisis of self or
family (Durcholz and O'Conner, 1973). While Carlsen (1973) finds
equal vocational and personal goals for women over 35, many others
find that woman came first for personal growth reasons and only
secondarily to gain knowledge or to get interesting jobs (Brandenburg,
1974; Durcholz and O'Conner, 1973; Espersson, 1975; Markus, 1973;
Steele, 1974).

By the later 70's the issues of personal satisfaction or desire
for intellectual stimulation are still the primary motivators but
career oriented reasons increase in the frequency of being indicated
(Magill and Cirksena, 1978; McCrea, 1979). Folland, Pickett, and

Hoeflin (1977) still say people return from a sense of emptyness, and

 

31

Lenz and Hansen (1977) describe the motivations as first being
psychological, including a need to gain identity and a need to know or
to try to fill the deprivation that they perceive out of their past.
Secondarily, an economic impetus is escalating with the increase of
divorced mothers in school. In this period Truax (1975) describes
woman students over 35 as being of four types: the housewife with
children now in school who is looking for challenges, the recent
divorcee or widow who is preparing to work, the clerical person
seeking better opportunities, and the woman who simply wants to learn.

By the 80's a switch begins. Dwinell (1980), whose women are
much less married than most studies show, sees that career reasons
bring back 55% of the women, personal satisfaction brings only 15%,
and economic benefits bring another 9%. In a major survey of both
sexes, Apps (1981) discovers nearly every woman checked both career
and self-esteem reasons, including feeling that it is now socially
acceptable to go back to school. Women's fourth-ranked reason in his
study is that a change in life situation brought them back.

Career goals now frequently lead the list of reasons for return
(Rawlins and Davies, 1981; Hooper and Traupman, 1984; Sewall, 1984).
These more recent studies are also discussing more why are they coming
£23. In a major summary article, Cross (1982) says adults with low
educational attainment and low status jobs are motivated by the
external rewards of better pay and better jobs while adults who have
life's necessities cite more internal rewards such as satisfactiOn and
desire to learn. Blocker and Rapoza (1981) list persons returning to

college as those needing a vertical move, those seeking satisfaction

 

 

 

32

by a career change, and those forced to change by divorce, death of
spouse, unemployment, or physical or emotional handicaps.

A recurring theme is that older students of both sexes often
return because of life transitions (Greenfieg and Goldberg, 1984).
They claim the separated or divorced need more earning power, many
need a mid-life career change, others wish to re-enter the job market,
and the laid-off or resigned may want to upgrade their abilities.
Aslanian and Brickell's (1980) major study hypothesizes that people
return to education because of life transitions such as a job change,
marriage, divorce, arrival or leaving of children, and retirement.
With questions designed specifically to tap into life changes, they
find that of the 83% who say transition is a motivator, 56% identify
job or career motivators, 16% identify family life changes, and 13%
identify transitions in leisure life patterns.

Astin (1976a) helps understand these motivations better in
similarly patterned returnees by explaining that those who return with
a goal of career or employment are the ones who must support
themselves and the family, who need a degree to advance in a present
job, who want career changes, or who are ready to return to a job
because of fewer family demands. Those who return for interest or
enjoyment reasons are more likely to be satisfying general interest or
curiousity, to be bored, or to be finishing a degree they always
wanted.

In a study solely of returning women over 50, Hildreth,
Dilworth—Anderson and Rabe (1983) find nearly 70% come specifically to

get the degree or the certificate, 53% to gain independence in the

33

sense of identity (which they called psychological reasons), and 43%
come for employment preparation for a better job. These numbers
closely parallel Astin's (1976b) findings from a study of 15 colleges
in 1976 where the women are primarily under 50.

These reasons for returning can be placed within Houle's (1961)
framework of three types of adult learners: goal-oriented learners
who begin later in life and have specific objectives,
activity-oriented learners who want to make social/relational contacts
and are least likely to reveal this true reason for education, and
learning-oriented learners who are thirsty for the pleasure of gaining
knowledge and who never stop their thirst. When women began to return
to school in significant numbers in the 50's and the 60's, they were
probably more activity oriented and partially learning oriented, but
they have become increasingly goal oriented over the years, although
the other two reasons remain important.

It becomes apparent that although the reasons for return to
college at a later age are as varied as those doing the research, some
trends are visible. While women two decades ago may have been the
more stereotypical homemakers who wanted to fill a void or avoid the
pain of their lives, this negative impetus soon switched to an
emphasis on college as a method of personal growth. Later the
emphasis began to move toward increased career direction for women
which is somewhat more like the classic reason for male education.
Recently studies have begun to emphasize more the specific transitions

or triggers which bring older women to college. This seems to be an

34

important addition to our thinking, since the very transitions which
bring them can determine the strength of the desire to remain and
complete an education but can also be troublesome enough in themselves

to make achieving that goal problematic.

PROBLEMS AND BARRIERS

What are the problems that older women do have when they consider
or do take on the student role? Are the problems concrete? Who has
the power to change them? How severe are these barriers and which are
the more salient barriers? How complete is our knowledge of the
problems these women encounter?

Probably more is written on the barriers to reentry or the
problems encountered upon reentry for the older woman student than any
other one topic pertaining specifically to this group. An increasing
uniformity of language has come from a format used in a literature
review by Ekstrom (1972). She cites the barriers as being
"institutional" which are those barriers found specifically in the
bureaucracy, practices and expectations of the school; "situational"
which arise from one's current life situation, and "dispositional"
which are attitudes about learning and self-perceptions. It is
possible that dispositional barriers determine how much of a barrier
the institutional and situational problems truly are. If so, perhaps
it is more appropriate to speak of "roadblocks" or "speed bumps" than
to speak of barriers, since some people find ways around them or
tolerate barriers when other people are defeated and do not attempt

school or drop out because of them.

35

Institutional barriers include scheduling, a requirement to attend
full-time, lack of information-giving, stiff admission or attendance
requirements and bureaucratic red tape. Situational barriers include
costs (books, commuting, child care, etc.); lack of time; home, family
or job responsibilities; availability of child care, transportation,
or study space, and support or lack of it from family and friends.
Dispositional barriers include lack of self-confidence or self-esteem,
fear or being too old, lack of energy, dislike of studying or
indecisiveness (Cross, 1981, 1982). Almost all surveys indicate that
the situational barriers, specifically lack of time or cost, top the
list (Sensor, 1964; Espersson, 1975; Young, 1977; Wertheimer and
Nelson, 1977; Dwinell, 1981; Richter and Witten, 1984). Other
situational problems are the hours of classes and difficulties of
parking and of library use (Markus, 1973).

Even though cost and time were expected to be the highest
barriers, Richter and Witten (1984) learn these are experienced as
even more of a problem than expected, and home responsibilities are
also somewhat higher barriers than expected. Smallwood (1980) finds
coordinating studies with child care and family or with a job is the
primary difficulty and that knowing how to study effectively is also a
problem. When Erickson (1968) allows older respondents to mark more
than one problem, women mark more problems than do men. She discovers
the ranking of problems to be different for women than for men, after
both agree that scheduling is the primary issue. For women, family
arrangements are secondary followed by time for study and exam fear.

Across sexes, family arrangements are fifth. Her findings emphasize

36

the problematic character for women of trying to balance schooling and
their societally expected role. Across age groups, Markus (1973)
finds a steady 30% who do not feel a part of the university. In her
study, the primary problem is less time to spend on family and friends
and on housework. Looking at women's finances, Steiger and Kimball
(1978) conclude women have more difficulty earning the necessary
money. Further, financial aid which is designed for the full-time
student excludes many older women, and current financial processes
close out women whose husbands cannot or will not pay.

Cross (1982), however, raises an interesting question:

"The major barriers reported by survey respondents are socially

acceptable barriers such as lack of time and cost. But what is

the real role of less socially acceptable reasons for lack of

participation?" (p. 136).
Although dispositional barriers are cited by only five to fifteen per
cent of respondents, "The 'real' importance of dispositional barriers
is probably underestimated in survey data" (Cross, 1981, p. 106).
When Wilcox, et. al., (1975, quoted in Cross, 1981) ask current
students why their friends are not participating, they get a much
higher expression of dispositional reasons. They assume that people
are more willing to give such reasons about other people than about
themselves. Another reason for underciting dispositional barriers may
be that those people for whom these barriers are truly impenetrable
barriers are not further analyzed because they choose to not attempt
the return to school. Also, some barriers may be simply "convenient
rationalizations" (Aslanian and Brickell, 1980).

Several studies cite more dispositional issues such as

accomodating all roles, management of guilt/selfish feelings, shame in

 

 

37

relation to intellectual aptitudes of younger classmates, and the
reality of being alone in a college culture (Letchworth, 1970).
Dispositional factors are primary, according to Ryan (1979), and
include in order: competition and inability to learn, a "sore-thumb
complex," guilt about husband, home and family, the "I'm not worth it"
complex, spouse disapproval and then time constraints. Krings (1976)
points to mixed problems of poor educational background, inadequate
study skills, lack of confidence, unrealistic expectations, and a
sense of irrelevancy of requirements and conflict over having to meet
them. Brandenberg (1974) cites almost entirely dispositional issues
including uncertainness of the ability to achieve, resistance in
spouse, family and friends, and guilt about and/or competition with
the children. There are many more general problems if the financial
resources are fewer and for those who are not married, and there is no
evidence that a large network of supportive people is related to the
number of problems (Magill and Cirksena, 1977).

Geisler and Thrush (1975) organize the problems differently.
They say some problems are university related such as scheduling and
age/sex discrimination; some are related to self and family such as
finances, husband's attitude and helpfulness, and the expense and
convenience of child care; and some are self-related only, with most
being internal conflicts including time pressure, self-confidence,
role definition and sense of direction.

Many problems are significantly worse for older women than for
older men. These greater difficulties are time issues; admission

procedures; fear of not being smart enough, of failing, or of dulled

38

memory; difficulty with children or spouse over the reentry; and guilt
for spending money or pursuing their own goals (Lance, Lourie and
Mayo, 1979). The longer the interruption, the significantly more
difficulties were experienced. Dwinell (1981), in looking at problems
of women students by five-year age groupings from "below 26" to "over
50," finds the demands on time and energy generally increase to age
40, drop until 50 and then increase. Financial concerns generally
drop until after age 50. "Keeping up with the family" peaks as a
problem during the later 30's while social life is increasingly a
problem until the later 30's and then lessens. Financial concerns are
less for the married and the widowed, but time for study is
increasingly difficult with increasing age for the married and
divorced.

Mardoyan, Alleman, and Cochran's review (1983) cites the
problems of life disruption because of the time, attention, and
finances diverted to college, the adjustments to being few and
dependent in the younger world, the press of multiple commitments, the
clash of the pragmatic experience of the world and the
non-adaptibility of the university environment, and the lack of
awareness of services available.

Suchinsky (1981), taking an entirely different tactic in
describing problems, calls some "ostensibly environmentally derived"
and others "developmentally based." In the first category is the
family:

"By and large the response of her family will tend to be

inhibitory, either subtly or overtly. This will not infrequently
occur despite overt expressions of support for her endeavor, and

 

39

there will generally be stresses which will range from mild to

severe" (p. 31).
This includes issues such as who stays home with a sick child or
covers the absent sitter. Faculty and administrators may react to her
age and marital or parental status or react more personally to a
student who is their contemporary or their senior and whose problems
may "...resonate with the kind of struggle the faculty person or
administrator is experiencing in his own emotional life" and express

their feelings "

...in reactions, either positive or negative, which
will be stronger and more inhibitory of effective education than those
generally seen with younger students" (Suchinsky, p. 31). In the aging
process she may lose some intuitive grasps or innovative brilliance
but add insight and wisdom unavailable to the younger student. The
institution may then need to deal with women's loss of capacity to
grasp abstract concepts and loss of some memory acuity by developing
faculty who have greater skills of making the abstract comprehensible
and who have patience. Relative to ambivalent feelings toward younger
classmates, inferiority "tinged with hostile admiration and envy" is
possible as well as perhaps feeling amusement, shock or impatience
with the younger students (Suchinsky, p. 32).

In the context of adult developmental theory, Suchinsky sees that
the adaptable autonomous adult who adjusts to the environment in ways
that are satisfying and productive does have hope. If this adaptation
is not made, there is a problem. Returning students are in a growth

period, and growth often comes with pain. They are more often

depressed, have self-esteem problems, are anxious over ability, and

 

40

have more real physical problems as well as having a more mature
perspective (Suchinsky, p. 41).

Perhaps this helps explain why barriers may be different to
different people. To some, they're a definite stopping point. Others
see the barriers only as a challenge to figure out how to go on
anyway. To yet others they are a temporary setback or a permanent
problem that does not stop them but either slows them or costs them
extra energy while they continue toward their goal. Often it is the
person's perception of an impenetrable barrier or minor stumbling
block that determines its power to deter. Since barriers are most
often cited in studies of current students, those students somehow
have found their ways to circumvent or cope with the barriers enough
to remain in school. So what keeps others out? And which of the
barriers are only annoying and which are really true problems that may
cause people to leave? When is a barrier a brick wall so one does not
begin, or a mine field that blows up part way across or at least
causes one to maintain anxiety as one threads one's way, and when is
it simply an irritation? The attrition literature may help us

understand at least a partial answer.

ATTRITION
Attrition information shows a rapid advance in the twenty years
since a literature review found that only five studies had been done
on college students and none on four-year college students (Verner and
Davis, 1964). At that time the reviewers found persistence related to
increased age, increased years of education, marital status for women,

higher occupational status, higher income and higher active social

41

participation. Since then researchers find persisters are most
satisfied with the college, non-academic dropouts who were passing are
next, and academic failures are the least satisfied (Starr, Betz, and
Menne, 1972). Persistence is not related to the amount of financial
aid or to race (Selby, 1973) but is related to coming from towns of
over 50,000 people (Cope, 1972). Traditional—age students more likely
to complete a bachelor's degree in four years are Jewish, have higher
grade point averages, are white, come from private high schools, plan
to teach secondary school, and are on scholarships; but those
supported heavily by family are dropout-prone (Astin, 1973). Baier
(1974) agrees, adding that a low first quarter grade point average,
low ACT scores and high school rank, and being black are most
predictive of withdrawal.

Kowalski (1977) discovers a highly significant difference between
dropouts and persisters on home environment factors including father's
educational level, health or personal problemsflat home, satisfaction
with the general attitude at school and with the faculty and academic
advisor and residence halls, and on many personal, emotional and
academic characteristics including plans about completing education,
having good study habits and having good interest in their work and
attendance at class and in general happiness with college. However,
voluntary withdrawers among freshmen, who score higher on ability
tests and have greater verbal ability than persisters and higher
ability in general than failures, leave early because they want more
independence;, and women voluntary withdrawers are less

practical-minded than failures (Rossman and Kirk, 1970). If, in fact,

42

those younger students who value independence are more likely to
leave, what does that say about the congeniality of the college
atmosphere for autonomous older students?

The findings that dropouts and dropins at Lansing Community
College have equal hours of work per week, more returnees are married
than are dropouts, and returnees get higher grade point averages
(Byrne, 1974) suggest it is the reaction rather than the stresses that
determines educational behavior. But colleges have still lost two
older students for each younger student for non-academic reasons
(Kimmel, 1976). These losses may include the "stop-out" who goes one
or more terms without enrolling but intends to continue soon (Mishler,
Frederick, Hogan and Woody, 1982). The frequency of stopping out once
enrolled seems to be greater between the ages of 25 and 35 than either
side of that.

Much work has been done in attempts to predict who will drop out
in the younger or general student populations. Among students in
general persistence is best predicted by ability, followed by the
level of educational plans or the goal commitment to college
completion, and then by higher family factors and socioeconomic status
with supportive democratic families whose parents encourage their
children and expect them to be successful according to a major
theoretically oriented literature review, (Tinto, 1975) and by
religious background and religious preference (Astin, 1975).

When studying motivational levels, Marks (1967) learns that
expectations of dropping out correlate with the level of aspiration,

fear of failure, and parental attitudes and that those who expect to

 

43

drop out do so in significant percentages although there is no
correlation between scholastic aptitude and expectation to drop out.
He believes that those who are more likely to drop out are less
committed to college and are more concerned with their parent's
expectations than with their own.

In a major literature review in 1978, Pantages and Creedon see a
stable ratio in research going back to 1913 showing approximately
37-40% graduate in four years, 50% graduate after continuous
education, and 20% more graduate sometime, somewhere. They find more
attrition at public state-supported universities than at private
colleges and see similar attrition rates for older and usual-age
enterers. They claim there is no significant difference by sex
although several studies showed more women and others show more men
among the dropouts, but that seems to depend on the type of
institution. Women drop out more frequently in heavily male
institutions, for example (Astin, 1964). Once a female drops out,
she's less likely to re-enroll than a male (Astin, 1972).

Pantages and Creedon's (1978) summary of reasons for dropping out
begins with academic concerns closely followed by financial problems.
Next are motivation including uncertainty of goals and lack of
studying; personal considerations including emotional problems,
marriage and personal or family illness; dissatisfaction with college;
military service; and getting a job. Females drop out more for
personal reasons and males for curricular reasons. However, Pantages

and Creedon emphasize that poor grades alone are not the cause of

44

attrition but are more likely to cause attrition when coupled with
other non-intellectual factors.

In a six-year followup on women over 30, Reehling (1980) learns
that persisters plan a higher level of education and work for a
degree. She finds that discriminate function can predict persisters
96.2% of the time but is very poor at predicting dropouts. These
dropouts are distinguished by their education being financed by
others. In a study of older student withdrawers from Michigan State
University, DelDin (1980) shows a slightly higher rate of women
withdrawing than expected, and a higher proportion of divorced women
and married men doing so. In looking at reasons to withdraw she
explains they are academic, including needing a temporary break or
having inadequate study skills or lack of interest; employment;
finances and personal issues including time crunches; home
responsibilities; personal problems and personal interests. Reimal
(1976), in a study of re-entry women in nineteen community colleges,
finds that persisters among reentry women are older, without children
or previous college experiences, and with lower incomes. They are
also more likely to have participated in a re-entry program.

Among single parent university students the most pervasive
problem is having sole responsibility in decision making for children,
coupled with social disapproval of their lives and financial
difficulties (Hooper and March, 1980). "Many female single students
who entered the university to solve financial problems find that the

financial and emotional cost of student life is too great to be borne,

45

even in the short run" (Hooper and March, 1980, p. 142), and many fail
because of isolation and lack of supports.

Astin (1972) suggests that the problem of research on attrition
is its too heavy reliance on demographics and that moderator variables
of motivation, attitudes or personality are needed. However, when
Sharp and Chason (1978) look for moderator variables, they find that
scales on the Minnesota Multiphasic Personality Inventory do not work.
The problem of citing reasons for attrition, say Pantages and Creedon
(1978), is that the problems withdrawers have are also problems for
persisters and that while surveys may be simple, the reasons are
complex and there is a danger of generalizing too much from one survey
or one institution. They also believe that research looks at far too
few predictors and recommend a shift from predicting attrition to
preventing it.

The attrition literature may teach us that we may not be looking
at the right questions. It may also be important to note the
difference between those who drop out because they are failing and
those who choose to withdraw but are not failing, because the second
group may look more like the persisters than the failing dropouts.
Dollar's (1985) examination of studies of attrition over time realizes
that they have become more multivariate, including more combinations
of academic and nonacademic measures. An important conclusion drawn
by Kowalski (1977) is that academic problems and personal pressures
will likely cause dropouts. He says since these issues can be
identified, perhaps they should be. He believes that having a

definite educational goal also makes a difference in persistence.

 

46

Lunneborg, Olch and DeWolf (1974) say, "Some subtle combination
of personal questions in regard to expectations of dropping out or
getting an advanced degree could provide a good noncognitive indicator
of middle-aged school success" (p. 220). They suggest that one needs
to look at expectations of continuing or dropping out as that may have
as much validity as any other single predictor. Again, it seems clear
that the moderation caused by one's self-regard and one's experience
is probably as critical in the attrition event as is any external or
concrete event.

The literature also raises the question of whether married,
single or divorced persons handle things differently or are more
likely to drop out or to react to the same events differently and
whether those who deal confidently with life in general and who feel
good about themselves are able to cope better with the problems that
they may face.

It may be that those who drop out are those who do not cope with
what Lenz and Hansen (1977) call the "traps and traumas" for mature
students: lack of confidence, "that out-of—phase feeling," youth-age
competition, home related guilt and the subsequent attempt to be
Supermom, school related guilt, unrealistic expectations, the drag of
the past and the painful memories of previous education and the
adjustment to carrying various roles simultaneously. "Adults going
back to school experience change, and most people deal ambivalently
with change. It is bittersweet, bringing both excitement and

apprehension" (p. 154).

47

COUNSELING NEEDS

Perhaps as colleges become better able to provide for some of the
needs of older students, the attrition rate will go down. The
provisions necessary for removing many of the institutional barriers
are relatively self-evident in the descriptions of the barriers. And
colleges can alleviate at least part of the situational barriers if
they so choose.

But if students do have as many dispositional problems as the
literature suggests, it may be that colleges will need to become more
attuned to meeting the psychological needs of older students. The
literature suggests that what is needed to maintain them is counseling
regarding aptitude, emotional problems, and vocations (Sensor, 1964)
or counseling to understand "the extent to which they prevent their
own successes" and how much family resistance is real and how much is
projected as an excuse; to analyze and understand the reactions of
spouse and family; and to explore guilt feelings and
self-gratification versus sacrifice for others (Brandenberg, 1964).
Many writers recommend supports such as a "Guide for Returnees"
brochure (Rawlins, 1979) or dealing with such psychological issues as
the need for identity, low confidence and self-esteem, and the stress
of adding a role (Greenfieg and Goldberg, 1984).

Porter (1970) believes counseling should be aimed toward helping
students in dealing with a lack of confidence, adjusting to the press
of time, choosing appropriate long-range goals, budgeting time,
dealing with the pressure of family life, blending real world

experience into the ivory tower framework, dealing with the tolerance

 

48

of education by those at work, dealing with the bad memories of early
educational experience, and gaining flexibility in order to avoid
discouragement. Erickson (1970) suggests that counselors of adults
need to deal with fears of: inadequacy, examinations, inability to
study and read rapidly, class discussion, being out of date,
competition from adolescents, and failure and also need to deal with
individual reactions to stress, the search for identity and role
problems, personal adjustment, social problems and other issues.

Tittle and Denker (1977) in their literature review see the main
problems for returning women as being:

1. "Stereotyped attitudes (held by counselors and reentry
women) about psychological adjustment as students, as
wives/mothers, as employees, and as decision makers.

2. Management of home and academic responsibilities

3. Asserting the rights of the woman, as an individual, in
home, social, work and academic settings." (p. 544)

They further assert that while women need help in handling these
problems, sex and age bias of counselors has been documented and must
be attended to in the university.

Kelman and Staley (1974) add social skills training to a list of
what women need to cope in college. And in addition to these forms of
assistance, others suggest the colleges provide for: preschool
nurseries and a club for married women (Sensor, 1964), finding better
ways of selecting for admission and lifting financial aid restrictions
(Brandenberg, 1974), having women's centers for reentry information
and training faculty in listening to adult women's concerns (Folland,
Pickett and Hoeflin, 1974), waiving prerequisites for suitable

experience, encouraging "testing-out" for employment/experiential

learning, keeping university services open later and at noon (Rawlins,

49

1974), offering child care and peer support groups (Greenfieg and
Goldberg, 1984) or a place to stay overnight in bad weather, and
counseling and workshops on assertiveness, values clarification,

relaxation and study skills (Rawlins and Davies, 1981).

RATIONALE AND SCOPE OF THE STUDY

Problems inherent in previous research have been that often women
have not been looked at separately from men, the older student has not
been separated from the traditional age student and attrition study
has focused heavily on academic achievement on the assumption that
achievement is related to persistence which doesn't explain the
attrition of the more able student. Either the dropout or the
persister is investigated but rarely are both, which leaves no control
group for assumption of results. Focus is frequently narrow, on only
one or two reasons for attrition or only on reasons for leaving or
reasons for return or barriers or other singular items. The dropout
is rarely defined in ways that exclude or account for the stopout or
the transfer or those who have gained all they want from college and
therefore we don't know whether they are temporary or permanent
dropouts (Pentages and Creedon, 1978). The problem is that we want
direct reasons for poor grades when they may be only consequences of
some other perhaps less direct event. Furthermore, research on
motivation and psychological factors has been hampered by a lack of
measures, and researchers have therefore often chosen to stay only
with demographic variables or with simple prechosen lists of external
reasons rather than investigating those which pertain to one's

psychological health and coping ability (Cross, 1982).

50

"Sociodemographic variables such as age, sex, income
and schooling appear to play a relatively modest role in
influencing the educational participation and persistence
behavior of American adults....Variables such as age and
schooling...are of relatively modest importance if one
wished to explain and predict participation and persistence
in adult education. Future research needs to employ more
sophisticated conceptions of the participation process that
include personal and situational variables (e.g. attitudes
toward education, life change events such as marriage, job
loss and retirement...." (Anderson and Darkenwald, 1977, pp.
5-6).

Mandel, Butcher and Maurer (1974) believe we need to pay
attention specifically to the difference between persisters, transfers
and dropouts because personality variables more strongly predict
transfer than do academic variables but academic variables predict
group membership well, especially during the most recent term.
Therefore, they suggest, researchers should use the personality
variables which can be collected at the beginning of college since
academic variables are unavailable until later.

Roehl and Okun (1984) discover higher depression rates than
expected among first semester returnees and wonder whether that was
because they were having trouble with the experience or if they really
had more negative life events than other women their age. They
suggest looking at life satisfaction, GPA's and persistence in
addition to depression.

Campbell, Converse and Rodgers (1976) find factors that relate to
mental health in middle life are that married women are more satisfied
than single, but single women have increased emotional health. While
men and women both describe their lives as equally satisfied, mental

health surveys find women less mentally healthy than men although the

51

married of both sexes, with or without children, experience life more
satisfying than do single, widowed and divorced persons.

In a study of coping styles for stressful situations of graduate
students, Kjerulff and Wiggin (1976) learn that the students
progressing well respond to all stressful situations with anxiety but
without blame and then mobilize themselves to cope with the
situations. The students considered less competent, both academically
and interpersonally, are extremely anxious when confronted with
academic problems, not anxious at all in stressful situations where
there is no clear source of blame, highly self-punitive in academic
failure situations and blaming in interpersonal situations.

After studying the literature and meeting with many older female
students, it seems appropriate to look at the sociodemographic and
educational variables such as age, parental and marital status, amount
of finances available and how adequate their finances are perceived to
be, number of children, spouses's educational level, years of
education achieved before, and whether people were working while going
to school as well as reasons to withdraw previously and to return now.
However, it seems even more important to look at some of the
psychological variables to see whether those who feel life is
satisfactory, who find they are primarily happy, who cope well with
crises, who expect to stay in school and who have an internal locus of
control are more likely to stay in school and to do better in school.
It would seem that the psychological variables are more important and
will help define which people will find the experience too much and

will leave while those who are more healthy psychologically may

52

persist despite the fact that they may have as many problems or
pressures as those who drop out. It will also be interesting to look
at the triggers for return and at whether those who are handling major
life change either before or after entering college are able to cope
with the change adequately or whether that is a primary reason for

failure to continue.

THEORY

What then are the underlying components for those who handle the
transitions well, who cope with stressful situations well, and who
manage to balance the several roles adequately while attempting to be
a student at an older stage of life? These questions seem to be best
answered by looking at some of the theory of adult development, of
adult behavior and of adult change. Neugarten's (1968) comment on the
state of knowledge about human adulthood that there is both a paucity
of data and, more importantly, a lack of a useful theory leaves us
still floundering. Unfortunately, most of the research on adult
development has been done on men, and even some of that research has
been claimed to be true for all adults. Despite even the popular
literature which describes stages of life, knowledge of what does
happen in adult women is thin.

Erikson (1959) describes adult development as proceeding linearly
through a series of stages with certain tasks for the resolution of
each stage. He believes that women often cannot finish their task of
developing personal identity until child-bearing responsibilities have
decreased. His resolution for this suggests that women's identity

crises appear after a choice of mate. This seems to imply that not

53

marrying may mean a woman can never establish her own identity.
Levinson's model (1978) reflecting male experience focuses on
chronological age rather than experience as the key variable. However
women do not appear to have a sequential mode which they follow
through appropriate ages. A major flaw of these theories is their
lack of attention to the varying role patterns for women who show
varying combinations of career, marriage, children and the degree of
commitment to each. In their different attempt to understand adult
female develoment, Barnett and Baruch (1978) suggest that locus of
control and attributions are areas that may warrant investigation.

Theory building within the field of adult education has been
difficult because of the marketplace orientation of adult educators,
the fact that the field of adult education has produced few true
scholars, and the multidisciplinary nature of the field (Cross, 1981).
One can postulate that those persons who return to college to further
their education have already filled the lower three levels (the more
basic needs) of Maslow's (1970) hierarchy adequately. Education can
be a method of filling the needs for esteem from one's self
(achievement, competence, independence) and esteem from others
(recognition, status, pay, jobs and respect). The paradox is that
while education can be a route to achieving esteem, adequate
self-esteem is needed to begin the pursuit of education. Therefore
Maslow's theory is only partially helpful.

One theory with some validity is the congruence model of Boshier
(1973) in New Zealand, continuing in a vein begun by Pervin and Rubin

(1967). He sees that the motivation for learning is a function of the

 

54

interaction between internal psychological factors and external
environmental variables. His model claims the older participant and
the dropout can both be understood by looking at the magnitude of the
discrepancy between the person's self-concept and key aspects
(principally persons) in the educational environment. He claims the
number of incongruencies are additive with the greater sum meaning the
greater likelihood of nonparticipation or dropping out. The theory's
examination of internal and psychological determinants of
participation and persistence seems to be a good beginning. But his
model explains neither why some students remain to get their education
despite the incongruencies nor what role is played by the lives of the
students external to the learning situation.

A further paradox is that to return to school and to achieve may
be precisely what causes problems elsewhere in a woman's life.
Balance theory as proposed by Newcomb (1961) asserts that a change in
one part of a social system or a relationship between persons which
puts the relationship into imbalance results in a state of tension
that leads people to attempt to reduce the strain and to restore the
balance. Thus reentering college can be considered an imbalancing of
the woman's entire life system. If the rest of the system can adapt
to the addition of education to a previously balanced system, there
will be fewer conflicting demands of the woman and more tolerance and
acceptance resulting in a milieu in which she can continue her
education. If this imbalance is seen as threatening to others, the

rest of the system may begin to pull away from her or to reject her

55

and she may then be in even more distress than that which is caused by
adding the role of student to the rest of her full life.

"Predictably, it seems that if the reentry woman begins to
achieve success on the campus, grows in self-confidence and
self—esteem, broadens her perceptions and values, and becomes
more accepting of others, those persons closest and most
important to her become less accepting of her, and they react
negatively or they withdraw" (Roach, 1976, p. 87).

If the family can accept and applaud her increased feelings of worth,
the system remains in balance. If they are threated by this, the
imbalance of the system can cause her to withdraw or to make major
adjustments in the rest of her life. It is precisely this possible
threat to the preexisting closed system that prevents many women from
attempting a return to college, and it is supposed that imbalanced
systems will cause many women to drop out.

Balance theory combined with Maslow's hierarchy of
self-actualization seem to be a beginning but incomplete model of
determining persistence in education. Bandura (1977, 1982) suggests a
theory of self-efficacy in which a person's self-perception of the
ability to cope and deal with a given situation controls the
probability of adaptive coping responses being initiated and
maintained during stressful circumstances. Although self-efficacy is
becoming a better known subject, it appears to need examination in the
light of a specific situation rather than as a generic description
(Brown and Heath, 1985). These writers believe people who deal well
with a life event can appropriately answer the questions, "Why did it
happen? What was my role in it? What do I do now?" (p. 462). Those

who are more prepared for an event will be more appropriate in their

attribution of cause, but deficits of self-esteem will come about if

56

the attribution is experienced as being internal. Perception of
control will correlate with depression and this perception will be the
moderator of their attribution rather than depression being the
specific and direct result of the attribution itself.

Thus we have some of the linkages of theory which can give us
direction. Underlying these linkages is the premise that good
psychological health generally enables much adaptive behavior. It is
hypothesized in this study that those who have good psychological
adjustment will be those more likely to persist until graduation and
those who earn higher grade point averages. It is anticipated that
those with good psychological adjustment will cope with problems, will
better manage time and multiple roles, will have appropriate goals,
and will be instrumentally appropriate in achieving their goals and
fulfilling their needs.

Because it is important according to self—efficacy theory to
couch expressions of self-efficacy within the setting and time period
of interest, this investigation looks at specific components of
psychological adjustment as they effect a woman in the student role or
to be effected by a woman in the student role. Rather than taking one
particular and generic measure of psychological health, several

specific facets of adjustment are used.

COMPONENTS OF PSYCHOLOGICAL ADJUSTMENT
LIFE SATISFACTION AND HAPPINESS
In an attempt to look at the breadth of psychological adjustment,
several issues need to be examined. The quality of life, as described

in some national studies, examines both life satisfaction in general,

57

which is conceived to be a cognitive measure, and happiness, which is
considered an affective measure (Campbell, et.al, 1976). The
literature repeatedly finds 5-15% of peOple consider themselves to be
not too happy (Bradburn, 1969; Robinson and Shaver, 1973; Black and
Hill, 1984). When changes occur over time, the shift is likely to be
from very happy or very satisfied to mostly happy or mostly satisfied.
Increased satisfaction has been found by some to be correlated with
socioeconomic status (Bradburn, 1969; Edwards and Klemmack, 1973) but
not by others (Lowenthal, Thurver and Chiriboga, 1975).

Sands and Richardson (1984) find female returning students are
very satisfied with their personal lives and that academic but not
demographic variables contributed to their satisfaction with school.
The younger of their over-30's are more depressed and stressed than
the older women. Kirk and Dorfman (1983) discover that the strongest
predictor of satisfaction as a reentry woman student over 34 is the
helpful attitudes of professors. Many studies show married persons
are the most happy and widowed and divorced persons are the least
happy. However, Kaplan says this does not hold for married women
(1986).

Assessments of both happiness and satisfaction have relatively
stable test-retest reliability. One of many such studies is Sears and
Barbee's (1977) followup on Terman's gifted women fifty years after
the initial study. While saying "one-shot" measures of satisfaction
often don't show conventional reliability, they claim the satisfaction
reports remain quite consistent over a recent ten-year period.

Robinson and Shaver (1973) assert, "One of the most impressive

58

features of (these) questions...is the stable test-retest
reliabilities they exhibit" (p. 17). Andrews and Crandall (1976)
compare several measures and conclude that perceptions of well-being
can be measured with some validity. Several measures of
life-as-adwhole showed validities of single-item scales of 0.70 to
0.79. They suggest this validity can be raised by using several
measures together.
SELF—ESTEEM

Another important component to psychological adjustment appears
to be self-esteem. This is variously described as "...liking and
respect for oneself which has some realistic basis" with

self-acceptance being necessary but not sufficient for self-esteem

(Robinson and Shaver, 1973, p.45), a set of attitudes a person holds
about him- or her-self (Martin and Coley, 1984) or simply how a person
feels about oneself which is consided to be a component of well-being
and whose integral parts include locus of control (Campbell, et.al.,
1976).

The return to school itself increases women's self-esteem (Astin,
1976c), while among married womem, self-esteem is increased in women
who percieve themselves as supported and respected by their family and
peers and who are therefore confident of their ability to handle
academic tasks (Farmer and Fyans, 1983). Self-esteem in returning
women increases with increasing semesters of schooling, although the
more guilty the woman, the higher the husband's support (Hooper,

1979).

59

LOCUS OF CONTROL

Locus of control, a further dimension of psychological
adjustment, is defined as "...the extent to which persons perceive
contingency relationships between their actions and their outcomes"
(Robinson and Shaver, 1973, p. 169) or the extent to which a person
perceives events as being a consequence of his or her own behavior and
therefore potentially under personal control (Lefcourt, (1982).
Lefcourt believes persons must come to perceive themselves as
determiners of their own fate in order to live comfortably with
themselves. Those who believe that some control rests with them are
considered internals (versus externals who see their lives being
controlled by outside persons and/or circumstances).

Burgaighis, Schumm, Bollman and Jurich (1983), using a one—item
question as well as a scale for locus of control, find the one item
correlates well with the scale. Their internals show increased
marital satisfaction. In a wide age range of undergraduates, locus of
control correlates with intelligence, self-esteem, self—concept and
age (Martin and Coley, 1984). But for women over thirty, locus of
control in the direction of having external social validation is the
most important component of a sense of competency (Feldman, 1980).
This measure highly correlates with intellectual ability and with
measures of general mental health (Powell and Vega, 1972). A group of
traditional role women in graduate school see reinforcement from
family and friends as necessary and are rated as externals while
nontraditional role women feel they have more personal control of

their lives (Brown, 1983).

60

A number of studies examining the correlation of locus of control
with academic achievement find such a correlation (Foster and Gade,
1973; Hudesman, Avramides and Loveday, 1985; Otten, 1977; Powell,
1971; Traub, 1982). All of these studies either are on both sexes
together, with or without separate reporting, or on males. In a
mixed-sex college sample, Prociuk and Breen (1973) do not find a
correlation with academic success while Traub's findings hold for
women separately but not for men alone, the direct opposite of
Nowicki's (1973) findings on younger subjects by sex. With males,
Otten (1977) goes so far as to say that locus of control better
predicts graduation than do ability tests. In a predominantly female
British college sample, Brewin and Shapiro (1984) do find internal
responsibility for academic performance while Prociuk and Breen (1977)
find internals are more accurate in predicting their levels of
academic achievement. This then does not explain why Muskat (1978a),
who looks at interrupters versus persisters among college freshmen,
finds no difference in locus of control but does find that the
internals had previously increased GPA's. She also interestingly
discovers that women are more sure they will do well and graduate than
are men.

This leaves a large area of unsureness relative to locus of
control. Very little has been done with women alone on locus of
control. There is some suspicion that women are more likely to have
an external locus of control than men and that those who have an
internal locus find it less societally valued. However, women who

turn to education are not doing the necessarily socially correct

61

thing. They therefore may be more likely to have an internal locus of

control which enables them to return to school at.

SUMMARY
Taken together, the subjects of self-efficacy, life satisfaction,

happiness, self-esteem and locus of control should give a relatively
good combined indicator of psychological adjustment. It is the
hypothesis of this study that better psychological adjustment will
differentiate those who persist to graduation from those who do not.
It is also predicted that these separate measures of psychological
adjustment will be intercorrelated, and that it will be possible to
use one or two of the separate items to stand for the whole group of

items predicting persistance versus withdrawal.

CHAPTER 3
DESIGN AND METHODOLOGY
SUBJECTS

Subjects in this study are women who registered for undergraduate
credit in fall term, 1984, at Michigan State University who have
birthdates before September 15, 1959. These women, therefore, are
those who were 25 and older at the beginning of fall term, 1984. They
fall into three distinct major groups and two minor groups. Major
groups include:
1. Students who graduated since September, 1984 (some of whom are
current graduate students).
2. Students who dropped out since September, 1984. Dropouts were
expected to differ in that those who achieved all they wanted in
school were assumed to be more like the persisters than like the
dropouts who didn't achieve their goals. These goal—achievers were to
be analyzed separately. However, since only three dropouts fit this
category, they were too few for analysis. Therefore, they are summed
with the totals of all students but are excluded from analysis when
persisters are compared to dropouts.
3. Current students still enrolled as undergraduates after spring
term, 1986.
Minor groups include:
4. Transferees who left Michigan State University and enrolled

elsewhere.

62

63

5. "Stopouts" who were enrolled winter term, 1986, did not enroll
spring term, 1986, for whatever reason, but plan to reenroll fall
term, 1986.

It was assumed that transfers would look more like persisters
than like dropouts since they do continue their education somewhere.
And those who "stop—out" briefly to cope with a particular life
situation and very quickly pick up the academic threads again were
assumed to differ from other dropouts. Because of their small
numbers, these assumptions could not be tested for the minor groups.
These minor groups were included in all the general frequency counts
and are described as part of the population but were too few to be

analyzed separately in most statistics.

SAMPLE SIZE

Two assumptions were made before beginning to survey this
population. One is that, as older students, they are more likely to
be part-time students who would take a longer time than usual to
graduate. The second is that they are more likely to have had some
undergraduate education previous to this most recent enrollment,
either recently in a community college or a four year institution or
previously in the period immediately after high school. According to
this second assumption, the population might be expected to be
somewhat farther along in their undergraduate education than the
average population of students.

In deciding on an appropriate sample size to allow an adequate
number of subjects within the target groups of graduates and

withdrawers, it was assumed that approximately fifty per cent of the

64

students in the population would be still currently enrolled at the
time of the survey, twenty percent would have graduated, and thirty
per cent would have left Michigan State University. These assumptions
were made in part by having consulted with staff members in the
Student Affairs office, the Withdrawals office, the Registrar's
office, Lifelong Education, and Adult Services and other departments
on campus where people make educated guesses about this population.
Because it was expected that this design would yield criterion
groups of graduates and withdrawers of unequal size, an important
requirement of sample size was that the smaller of these groups
contain at least thirty subjects to satisfy the assumption of

normality of the underlying distribution despite unequal group sizes.

SUBJECT IDENTIFICATION AND CONTACT

The registrar's office of Michigan State University identified
the subjects. The registrar's computer generated three sets of
mailing labels for a random sample of 450 of the 1247 undergraduate
women who were born before September 15, 1959, from the fall term,
1984, university registration records. Because of the method of
record-keeping, these labels used the addresses which were current in
September, 1984, without access to updated addresses. The Division of
WOmen's Programs of the Department of Human Relations attached a set
of these mailing labels to previously prepared envelopes which
contained a cover letter, the questionnaire, and a stamped,
preaddressed return envelope and mailed these envelopes on May 30,
1986. Thus, the subjects remained anonymous to the researcher until

, 9
the subjects identified themselves.

 

l!".‘
I (All C III.‘ ‘

.t‘ [J ‘1
‘ a

. ,iJlld ls

,Tll?‘

11‘. .)a

 

65

One week arcer tne first mailing, a post card expressing thanks
for participation was sent by the same process to all who received the
first letter. A followup mailing including a second letter, another
copy of the questionnaire and another stamped preaddressed envelope
was mailed in the same manner to all who did not respond or who had
responded anonymously two weeks after the initial mailing. This meant
anonymous responders received all three mailings. Because few
dropouts had responded within twelve days of the original mailing, the
second envelope was personalized by writing a request to open and
reply soon in a bright colored ink on the envelope. It was hoped that
those who may have thrown the first mailing away unopened would be
curious enough because of a handwritten note on the envelope to open
the letter. Another brief paragraph was also then handwritten in
bright ink across the top of the letter saying more responses were
needed in all categories and especially from those who had left
Michigan State University.

Due to not receiving responses from the thirty dropouts deemed
necessary for assuming normality of the distribution and for
replicated discriminate analysis from the first sample, a second
complete set of mailings was completed during July, 1986, by an
identical process including the handwritten notes, to 202 more women
from the original population.

0f the original sample of 450 women requested from the registrar,
426 sets of labels were actually provided. That office explained that
the remaining 24 would be those persons who requested confidential

status, who did not give a usable address, or whose records were in

66

some such way unavailable after they had been identified as being in
the appropriate pool. From the pool of 426 subjects surveyed, letters
to 23 were undeliverable leaving 403 deliverable surveys. The 255
returns yield a 63.3% return rate from the first sample. Of these,
two were under 25 years old at the time of the survey and had been
erroneously sampled. They are not included in the data set.

0f the second sample of 202 women requested from the registrar,
202 sets of labels were actually provided. Questions of why the first
sample decreased from the requested number and the second sample did
not decrease were never answered by the registrar's office. From the
pool of 202 subjects surveyed, letters to 21 were undeliverable. This
left 181 deliverable surveys. The 109 returns yield a 60.2% return
rate from the second sample.

Taken together, the 584 deliverable surveys yield 363 responses
or an overall response rate of 62.2%. Of these, 361 responses are
included in the data set. It is interesting to note that in both
samples, the majority of those who left Michigan State University by
dropping out, stopping out or transferring only responded after
receiving the full set of the three-wave mailings although a majority
of the'£g§§l_replies from each sample were received within less than

two weeks from the original mailing.

PROCEDURES TO PROVIDE
SUBJECT IDENTIFICATION AND CONFIDENTIALITY
Subjects were asked to provide names and contact information so
that followup mailings could be sent only to nonresponders or to

unidentified responders and to make possible a future followup survey.

67

A future longitudinal study could ask the present students whether
they graduated or left the university and could examine whether they
did indeed fall into the categories which this research would
postulate as probable.

Measures taken to assure anonymity of data and to protect the
confidentiality of subjects included assigning a number consecutively
to each person's materials upon arrival. Approximately 90% of
respondents provided names and addresses on a separate page from
survey data which were immediately separated from surveys. Any
reports made from this information will be without identifying data,
including but not limited to this research report. The surveys
include a personal code number which subjects would be able to
reproduce in the future. This code will allow one-to—one comparison
of future survey information with that on the present survey even for
those completing the survey who do not give their names. This
personal code number consists of six digits representing the month,
day and year of their birth plus four more digits representing the

month and day of their mother's birth.

METHOD OF DATA GATHERING

A packet originally mailed to all subjects contained:

1. A letter explaining the purpose of the study, why their input
is important, how issues of confidentiality will be handled, that
their participation is voluntary, and how to raise questions if they
desire;

2. A survey instrument;

68

3. A form for the name, address, telephone number and student
number;

4. A stamped, self-addressed envelope (to the Division of
Women's Programs using a format not currently used there for any other
purpose to identify survey responses).

This research was conducted in a mailed survey format for several
reasons. The literature suggests that this population is extremely
busy, and that too many demands on time is one of the major (if not
the primary) problems older women students face. A written survey
instrument which can be completed at their convenience offers a method
to gather a large amount of information with the least amount of
intrusion into their busy schedules. Furthermore, the options for
response are visible at one time, and respondents can most efficiently
decide on the best response. They are the most knowledgeable persons
to make their own choices in cases of unsure answers.

A checklist format is used to gain specific information quickly
and with the least amount of effort for the subjects while gaining the
greatest amount of information. It is assumed that subjects are more
likely to respond to something which will take only a brief time and
that this consideration will increase the response rate. Since access
to the subject population is restricted by the university prior to
each subject giving consent to participate (because they are being
identified by age which is deemed to be a confidential piece of
information), each subject must, at some point, be asked whether she
is willing to participate in this research. It made the most sense

for them to have the research instrument at the time of making this

69

decision and to be asked to respond only once, giving the requested
information at that time, so we do not have the added attrition of
those wno agree to be surveyed but then do not complete the survey.
Dillman's book on surveys (1978) uses social exchange theory to
underline some of these decisions. He says survey results are
maximized by minimizing the costs of responding, maximizing the
rewards, and establishing trust. He sees mailed surveys as having the
least social desirability bias of all types of information gathering.
Dillman's formula for ordering the survey items, for using
contrasting upper and lower case letters, for controlling survey
length, for pretesting (piloting) and for the design and timing of
mailing waves were primary guides in the design and implementation of

this research.

INSTRUMENTATION

This survey elicits information on both factual/objective
indicators and a set of subjective indicators which reveal how these
women evaluate their own lives. This is in accordance with
indications that information on the quality of life needs to be both
objective and subjective (Gitter and Mastofsky, 1974).
OBJECTIVE SELF-ASSESSMENT

The survey first establishes the respondent's category as
graduate, withdrawer, transfer or current student. Those who are out
of school are asked their future educational plans to differentiate
stopouts from dropouts. Stapouts are identified by the date when they
last attended,Michigan State University compared to when they intend

to return to Michigan State University.

70

Sociodemographic variables (age, marital status, children,
income, employment while a student) and education-specific concrete
variables (part- or full—time student, years of education completed
prior to last entry to college, number of times of college entry,
current student status) are included. Married students are asked the
level of their husbands' education since some literature leads one to
expect higher attrition due to less spousal support for her education
when a woman is working toward a higher educational level than her
husband possesses.

SUBJECTIVE SELF-ASSESSMENT

The subjects' self-assessment of several factors is questioned.
These factors include the perceived adequacy of income since different
people see different amounts of income as adequate, an assessment as
to whether money is a problem in student life, reasons for having not
completed college at a traditional age, reasons for return to
education, personal changes or crises which contributed to their
return and personal changes or crises which occurred after their
return and an assessment of the severity of these, reasons for
accomplished or possible future withdrawal from the university, and
expectation of completing a degree.

The subjective questions pertaining to income are included
because an amount of income which one woman may see as adequate could
be a cause of withdrawal for another woman. It is also interesting to
see what income ranges are perceived as adequate or inadequate.

The reasons for original non-completion, reasons for current

return and reasons for withdrawal are heavily researched questions.

71

Most of these items are taken from previous studies with the only item
not previously found being the reason for leaving because a person
"Achieved all I wanted from school." This item is included to
separate out those who leave because they have accomplished their
goals, even though they may not have graduated, although surprisingly
few marked this option.

The questions on personal change or crisis are based on
frequently published observations that many older women return because
of life changes or crises or that the change or return to school
initiates other life changes or crises. However, since some women
cope and persist and others do not, these items are constructed (with
a partial base in Aslanian and Brickell, 1980, and in part from
personal reports) to assess both the types of events and the severity
of their effect (DSM 111, 1980).

Since the literature suggests that a single question assessing
one's expectation of completing a degree will be one of the most
important predictors of graduation, such a question is included.
SELF-ESTEEM INSTRUMENT

The final two categories of questions pertain to psychological
adjustment factors. In one category, two instruments are used intact
from the literature. At this time the literature is relatively weak
on measures of self-esteem or self-concept although Campbell, Converse
and Rodgers (1976) see self-esteem as being "relatively fixed early in
life and thus...to be causative of...satisfaction" (p. 59). Three

measures merited consideration: Rosenberg's Self-Esteem Scale (1965),

72

the Tennessee Self—Concept Scale (Fitts, 1965) and Coopersmith's
Self-Esteem Inventory (1967).

The Tennessee Scale has the advantages of several hundred usages,
several subscales, and (varying according to the author) possible good
validity and reliability. It has two major deficits. It consists of
one hundred items which taken alone seemed to make it too bulky, given
the concern for brevity in this instrument. And the very fact that it
generates several subscales underlines the fact that it is much more
complex than needed or desired as well as being repetitive. Also, as
a separate published scale, it is prohibitively costly, especially for
repeated mailings.

Rosenberg's Self-Esteem Scale (1965) consists of ten items, each
of which can have two levels of agreement and two levels of
disagreement. Rosenberg's and Coopersmith's scales appear to have
about equal validity (Demo, 1985) and reliability and about equal
support and detraction in the literature. Both have been frequently
used in research. Coopersmith's scale has a drawback of being rather
highly correlated with social desirability (Ryden, 1978). One study
comparing several measures of self-esteem indicated about equal
usefulness but said the older subjects preferred the Rosenberg
instrument over four others, in part for its brevity. Therefore, the
final choice between approximately equal instruments was made of the
basis of this published preference and of length, the Coopersmith
being a 58 item scale and the Rosenberg being only ten items. The

Rosenberg scale is reputed to have a reproducibility coefficient of

73

.92 and scalability of .72, a rather high level of reliability and
construct validity for a brief scale (Kernaleguen and Conrad, 1980).
LOCUS OF CONTROL INSTRUMENT

The locus of control instruments have somewhat the same problem
as self—esteem instruments. In both cases an ideal instrument has not
yet been substantiated, probably in part because of the complex nature
of the construct being tapped.

The original and most popular instrument is Rotter's (1966).
However, for all its popularity and usage, it is heavily and rather
uniformly criticized for such weaknesses as low point-biserial item
total-score correlations, inclusion of non-scored filler items, heavy
reliance on items pertaining to control in the political realm,
multidimensional aspects particularly in the external portions of what
is claimed to be a unidimensional scale, and variation in referents
between beliefs about people in general and beliefs about oneself.

The forced-choice format has both the problem that its items too often
are not parallel and the problem that the format makes it more
susceptible to responses deemed to be socially more desirable.

The problem then remains of selecting a better instrument.
Several have been devised over the years with four or five getting
some repeated use. Although few of the other measures have received
the scrutiny of Rotter's instrument, two were worth consideration.
Levenson's (1974) scale has frequent use and is probably the best of
those with much attendant research. It avoids the forced choice
format but falls into the trap of obvious response bias since all

questions are scored in one direction. This scale also has the more

74

prominent difficulty of not having factors which account for a very
large amount of the variance.

The scale used here is quite new, appears to have very good face
validity, but is too new to have had much use in research. Duttweiler
(1984) developed an instrument focusing on aspects of personal choice,
belief in one's self and independent action. She used a four-phase
system of instrument development including defining the network
involved in locus of control, pretesting in several educational
settings (junior college, continuing education, college and university
students), field testing, and supplemental administration with Mirels'
Factor 1 (Mirels, 1970) of Rotter's scale to look for replicability
and convergent validity. She finds small but significant differences
between subject categories in that a slight increase in the Internal
Control index is found to correlate with increasing age, educational
level and socio—economic level during her larger field administration.
However, Duttweiler considers that difference to be small and to have
appeared only due to large sample size.

The coefficient alpha estimates of reliability for the two
primary tastings were .84 and .85. Discriminate analysis shows one

' accounts for 68.7% of the variance. A

factor, "self-confidence,'
second factor labeled "autonomous behavior" accounts for 31.3% of the
variance. Convergent validity is claimed due to significant

correlation with a portion of Rotter's scale. Further research is in

progress at several sites using this scale, but no further studies

have yet been published (Duttweiler, 1986).

75

SINGLE ITEM ASSESSMENTS OF PSYCHOLOGICAL ADJUSTMENT

The final category of questions also pertains to psychological
adjustment but consists of several single items rather than complete
scales:

A single-item asessment of locus of control is used, testing
whether one item can assess this aspect nearly as well as an entire
battery of items and whether it has criterion validity. (Bailey,
1978). One previous study using both a battery and a one-item
assessment showed a significant moderate correlation between them
(Bugaighis, et. al., 1983).

Since "happiness" seems to assess an emotional state while
"satisfaction" assesses a more cognitive state (Campbell, et. al.,
1976), both are studied here. Life satisfaction while in school (and
afterwards for those no longer in school) and happiness during school
are eaCh assessed by an item adapted from a national survey of the
quality of American life and originally developed by Bradburn (1969).
These two topics have a correlation of .50 in Campbell's study,
although they find satisfaction to be much lower among single than
married adults. These two measures together appear to be an
appropriate expression of global or overall well—being (McKennell and
Andrews, 1983).

One item assesses a self-perception of coping with crisis. This
item is included because it is considered that equal crises have
different meaning for different women and that an assessment of crises
alone is imcomplete. Further, one's self evaluation of her ability to

cope may be correlated with other measures of adjustment and may

76

account for why some women withdraw because of changes or crises
assessed previously while other women continue in school while having
the same experience. This assessment is one method of assessing
self-efficacy.

Another measure of self-efficacy is a general question asking how
well tne women are doing in getting the things they want out of life.

Two items asked only of those who withdrew assess how good they
feel about their decision to leave and how much they wish they had
remained in college.

In his complex statistical analysis of surveys used nationally
which include several of the single items in this survey, Andrews
(1984) finds validity is improved by having four or more response
categories, allowing comparative rather than absolute answers, briefer
battery length for embedded batteries, and omitting topical headings.
He found average variances for questions such as those used here of
happiness and satisfaction to average .81. Using his formula for
estimating the validity of single items, those used here assessing
psychological adjustment would have validities in the high .80's to
the low .90's.

GRADE POINT AVERAGES

The survey asks for the most recent cumulative Grade Point
Average (GPA) and the most recent one-term GPA. It also asks for as
accurate an assessment as the subjects can make of the GPA attained
'bgfggg the last entry to college. Obtaining GPA's allows testing of
the hypothesis that those with higher GPA's are more likely to persist

and to be better adjusted. The recent GPA, the cumulative GPA (the

77

usual measure in the literature) and the entering GPA are all studied
since a woman's GPA may have changed markedly after a return to
education later in life from what she had accumulated earlier. The
recent one-term GPA is thought to be a good indicator of what to
expect next in a woman's academic career.

RESPONSES TO OPEN—ENDED QUESTION: FURTHER TOPICS SUGGESTED BY
RESPONDENTS

Because the checklist format may miss interesting and useful
information which may not fit neatly into analyzable configurations or
which may not have been considered at the time of instrument
construction, a space is provided for unstructured response to the
question, "What else do you think is important to know in a study of
older women who do or do not complete an undergraduate degree?" Three
lines are provided with an invitation to use the back if desired.

More than two-thirds Of the women have more to say, many of them using
much more space than the lines provided.

Some of them use this opportunity to tell of their own
-experience at Michigan State University. Others see the question as a
request to indicate areas that the questionnaire could have covered
but omitted. A smaller minority both suggests further questions and
tells of their own experiences.

Because of the volume of the responses and apparent grouping of
some topics, two raters read the responses on two separate occasions.
First they categorized the responses and agreed on a set of
categories. Then they classified the responses by the agreed-upon

categories.

78

Although it is likely that a woman would not mention a topic
unless she had found it to be important to her, a conservative
approach is taken here: Responses are reported in two groups. The
first group of responses is a summary of topics mentioned which are
expressed in the first person as a stated comment on experience of the
respondent. The second group of responses is a summary of topics
which the respondents suggest should be raised as questions in such a
study as this. It is noteworthy that these two categories primarily
are the same. A summary is then provided of the total number of
different women who refer to each topic, making comments and/or
suggesting questions. This summary may give a better indication of

the relative importance of each topic.

ANALYSIS OF DATA

The population is first described using frequencies. The program
Crosstabs is then used in the Statistical Package for the Social
Sciences (SPSS) to find chi squares. Chi squares are used to answer
the question of whether there are equal proportions between groups on
several nominal or categorical variables. This statistic informs as
to whether two classification variables (which may occupy several
cells each) are statistically independent or are significantly
related. Since the level of significance is set a priori at p-.05,
all results reported as being significant are at least at the .05
level. Ine chi square test merely tells whether there is an
association but does not measure the strength of the association.

Further tests (which vary depending upon whether the data is nominal,

79

ordinal or interval) are used to test the strength or weakness of the
associations which are found to be significant.

T-tests are used to test the differences between means of two
variables. They are of particular use in understanding ordered or
interval data. Primarily, t-tests are used to compare the two
criterion groups of those who persisted to graduation (persisters) and
those who dropped out without having achieved their goal (dropouts) on
several variables although some other means are also compared. Many
t-tests are two-tailed, meaning that while a difference is expected
between means, the direction is not predictable a priori. The t-tests
which are one-tailed are those for which the working hypotheses of the
study suggest the direction expected.

Comparisons are made between the criterion groups and the items
or scales of psychological adjustment to see if these measures do
actually discriminate between groups. Other tests examine
relationships between expectations of graduating and the criterion-
groups, between the sociodemographic variables and the criterion
groups, between the education-specific variables and the criterion
groups, between subjective questions about income and criterion
groups, between crises and criterion groups and between the single
item and multiple item scales of locus of control. Other comparisons
are made using single marital status categories and several of the
other variables of interest.

Analyses of variance are used for testing appropriate data which
lies on an underlying continuum. Several variables based on GPA's are

tested by analyses of variance. Additionally, the sum scores on the

8O

self-esteem and the locus of control instruments are compared by this
statistical method.

Pearson product moment correlations are made between the measures
of psychological adjustment, between financial questions, and between
other variables which have been found to be significantly different
between the criterion groups, in part to decide which of these items
or measures should be used in discriminate analysis. Factor analysis
is used to test the psychological adjustment measures for
unidimensionality and scalability and to derive psychological factors
for inclusion in discriminant analysis.

Discriminant analysis is then used. This is a nonparametric
statistical technique similar to multiple regression which allows a
researcher to engage in interpretation to determine the best-fitting
set of variables to discriminate between groups by determining which
characteristics.gg_discriminate between groups and then assessing how
gggggthey discriminate and which of the characteristics discriminate
most powerfully. Additionally, discriminate analysis allows one to
use a technique called classification to form equations or
"discriminate functions which combine the group characteristics in a
way that will allow one to identify the group which a case most
closely resembles" (Klecka, 1984, p. 9).

Discriminate analysis requires at least two distinct groups with
at least two cases per group and any number of discriminating
variables as long as the number of subjects is at least two greater
than the number of variables. It does not allow use of two highly

correlated variables in the same equation nor does it allow both

81

summed variables and the items within that summation to be used
together. The restriction against use of highly correlated variables
explains in part why simple correlations must first be drawn.
However, correlations which are not significant while standing alone
often become significant once some other variable is controlled.
Therefore, a variable which is not significant on its own can account
for a significant amount of the remaining variance once another
variable has first accounted for its share of the variance.

Specifically, discriminate analysis is used to develop the
equation which combines the most effective set of predictors of group
membership and to then find the most efficient (briefest) combination
of variables which will still quite accurately predict group
membership. Variables to use in trial equations are selected from an
examination of group means and standard deviations, a series of
Pearson correlations and examination of the correlation matrices found
during factor analysis. The SPSS program uses a stepwise method of
selecting, in order, those variables which together provide the
greatest separation of the group means. When no more of the variables
selected for use in a particular analysis have enough more
discriminating power, the process stops and no more variables are
entered into the equation. Multiple equations were used in attempts
to find the "best fitting" equation.

A set of standardized discriminant analysis coefficients are
produced with each equation. These numbers tell of the relative
importance of each item used in the equation. A set of unstandardized

discriminant analysis coefficients are generated as well. Each

82

variables' unstandardized coeffiecient multiplied by the actual value
of that variable in a given case gives a weighted value for that
variable. The weighted values are then summed to produce a weighted
equation which can then be tested for the significance of its
discriminating power.

One-half of the members of each criterion group, dropout and
graduates, are randomly separated into two samples which then each
contain half of the graduates and half of the dropouts. This allows
first, finding equations which predict one of the two samples and
second, testing the equation found on one sample to assess its
predictive ability on the second sample. A double prediction format
is used which allows a separate equation to be derived on‘gggh_sample,
each of which is then tested on the opposite sample to assess its
predictability. Furthermore, an equation is derived from the summed
group to find what set of variables would have best predicted this
total dropout and graduate sample.

Once these equations are derived, the current students' responses
are analyzed to see to what extent they fit the responses of the
criterion groups, to see how their proportions compare to the
criterion groups, and to predict the likelihood of persisting or
dropping out for current students. Responses of the current student
group are analyzed on those items which significantly discriminate
between criterion groups to assess whether the current student group
can be divided by the clustering of the significant items into groups
for whom likelihood of persistence to graduation or of leaving can be

predicted.

83

The level of significance for tests used in this study is set a
priori at .05.

Although it is not part of the current study, it would be very
useful and interesting to resurvey the current students in two or
three years to assess to what extent they were accurately predicted to

persist or to drop out.

CHAPTER 4
RESULTS OF THE STUDY
DESCRIPTION OF THE POPULATION

It seems appropriate, in looking at the entire sample from which
we are drawing conclusions about the population of undergraduate women
at Michigan State University who were over 25 years old in September,
1984, to look first at their student status, since that is the primary
issue of interest. The sample is then described in terms of
sociodemographic variables. The next appropriate descriptions are the
education-specific variables such as full- or part-time school
attendance, previous educational information, and expectations of
return to college for those who have withdrawn.

The more subjective assessment factors including self-perceptions
about income, reasons for dropping out of and returning to school,
personal changes or crises that have affected schooling and reasons
ror recent or possible withdrawal from the university. Academic
achievement is described.

Finally, the sample is described in terms of the single items and
the scales which make up the assessments of psychological adjustment.

Responses to a concluding open-ended question are summarized.

SOCIODEMOGRAPHIC VARIABLES
CURRENT STUDENT STATUS
The most critical variable used in describing the women in this

study is the assessment of their current student status at a point

84

85

which is two full academic years after the stage when they were
identified as having been registered as undergraduate students.

0f the 361 who returned the survey, 194 or 53.7% have graduated
as of June, 1900. Forty-one women or 11.4% of the sample have
withdrawn from the university. Three of the forty-one dropouts
indicate that they withdrew because they had achieved all they wanted.
Because so few withdrew because they met their goals, they are simply
not analyzed separately nor in the same criterion group with the
others who withdrew.

Thirteen students or 3.6% have transferred to another school.

Six or 1.7% are stopouts who had attended winter term, were not
enrolled spring term, but indicate that they intend to return fall
term.

Finally, 107 women or 29.6% are still current undergraduate
enrollees at Michigan State University. Because there is no
permissible way to sample only those who are no longer enrolled at the
university or to know who would be receiving the survey prior to its

mailing,

Table 1 Current Student Status

Relative Cumulative
Student Status Number Frequency Frequency
Graduated 194 53.7% 53.7%
Current Enrollee, MSU 107 29.6% 83.4%
Withdrew 38 1U.Jb 93.95
Transferred Out 13 3.6% 97.5%
Stopped Out 6 1.7% 99.2%
Withdrew Satisfied 3 0.8% 100.0%

Total 361 100.0%

86

undergraduate women who are still enrolled are included in the general
survey and in the frequency counts of the total group They are not
included in many of the statistical analyses since they were not in
the identified subject pools of interest.

AGE

The women's ages are requested in two different ways. A survey
question brackets ages into five-year interval beginning with 25 to
29. The code number each generates includes her birthdate, which is
used by a computer program to generate an exact age. By categories,
the women are, as might be expected, weighted toward the younger end
of the age range. However, they are not as heavily weighted in that
direction as one might predict. Furthermore, all five-year age
brackets up to the age of 60+ have subjects in them. Although the
youngest age bracket is listed as ages 25 to 29, it can, by definition
of the study, only begin with those who were at least 26 3/4 years old
at the time of the survey. If this age category included a full
five-year range of women, the frequencies would be adjusted somewhat
in favor of the younger brackets. This is one reason it is useful to
compute exact ages and to run specific and separate analyses using the
actual ages as drawn from the code number. Analyses using exact ages
are also more accurate and powerful.

The mean age is 34.55 years with a minimum of 26.8 years and a
maximum of 60.19 years. Ine modal age is 29 years, and the median age
’is 32.09 years. It is interesting to note that one quarter of the
population is at least 38 years old. This is noticeably older than

many of the previous studies have shown.

87

Table 2 Age By Five-Year Increments

Relative Cumulative
Age Bracket Number Frequency Frequency
25-29 122 33.8% 33.8%
30-34 99 27.4% 61.2%
35-39 78 21.6% 82.8%
40-44 35 9.7% 92.5%
45-49 16 4.4% 97.0%
50-54 8 2.2% 99.2%
60+ 1 0.3% 100.%
Total 361 100.0%

'3 age - 34.55 Median age a 32.09 Modal age = 29

MARITAL STATUS

A remarkably large number of women are‘gg§_married in this
sample. Those who have never married are 28.0% of the sample (n-lOl).
Divorced women make up 17.7% of the sample (n-64). Categories allow
for separated women (n=5 or 1.4%) and widows (n-2 or 0.6%), but the
smallness of their numbers make separate analyses rather useless.
Those who.g£g_married number 182 or 50.4% of the sample. Seven

respondents do not give their marital status.

Table 3 Marital Status

Relative Cumulative

Marital Status Number Frequency Frequency

(1) (1)
Always Single 101 28.0 28.0
Married 182 50.4 78.4
Separated 5 1.4 79.8
Divorced 65 18.0 97.8
wIOOwen 2 0.6 98.4
not Given 6 1.7 100.0
Total 361 100.0

88

PARENTAL STATUS

WOmen with children are asked to write in numbers indicating
their number of children in each age bracket. The youngest category
represents preschoolers with an age range of zero to five years. The
second category represents school-age children with an age range of
six to eighteen years. The last category represents grown children
with an age range upward from nineteen years. Most respondents comply
with the request to use a number within each category, but some only
check a given category rather than writing in a number. Because the
preponderance of these non-specific responses are in the 0-5 category
while ages of those mothers varies widely, it is unclear whether some
mistakenly are marking what they consider to be a category
representing a total number of children (and who may have zero to five
children) rather than having children in the zero to five year age
category. As a result of that unsureness, it is decided that at least
those twenty-three checks in the 0-5 category could be used in a tally
of the total number of women who are parents but must be excluded from
analyses of data on specific numbers of children.

A total of 202 women or 56% indicate they have children. Of
those, we know the age categories of children for 173 women.
Seventy—one (excluding the 23 unclear checks described above) have
children under six. One hundred seventeen have school-age children
from ages six to eighteen, while 43 mothers have children over
eighteen years old. The 56% rate of parenthood is lower than most of

the literature would suggest.

19

Table 4 Number Of Children Per Mether By Age Category

Children
Per Mother

CDNO‘UIbUONI—A

Unknown

Total

The mean number of children per mother is 2.33.

94

Ages Adjusted

0-5 Frequency
(1)
48 51.1
18 19.1
5 5.3
23 24.5
100.0

Ages Adjusted
6-18 Frequency

Ages Adjusted
19+ Frequency

 

 

(Z) (Z)

49 41.9 12 27.9
43 36.8 11 25.6
17 14.5 7 16.3
3 2.6 5 11.6
1 0.9 1 2.3
1 2.3

l 2.3

1 2.3

4 3.4 4 9.3
117 00.0 43 100.0
The median

number of children is 2.09, and the mode is two children per mother.

Table 5 Number Of Children Per Mether

Number

mNOHJtJ-‘LONH

Total

INCOME

Absolute
Frequency

Relative
Frequency

Cumulative
Frequency

(1)
26.0
67.1
87.3
94.2
96.5
97.1
98.8
100.0

Respondents are asked which category best characterizes their

annual household income during the last three terms that they have

been in school.

These categories are arranged in increasing order by

$5,000 brackets, since test construction specialists suggest that

90

people are more likely to omit answering the question when they are
asked to provide their precise income and are more likely to answer a
question which allows them to only indicate a less explicit income
range. The modal annual family income is the highest category
offered: $40,000 or more. If more higher categories had been
provided, this top income group would be more dispersed and better
understood. As it is, the mean income falls into the $20,000 to
$24,999 category, although this could be too low, given the clustering
in the highest category. The second largest numbers of women (after
the large group at over $40,000) fall into the lowest two categories
of income. It is interesting that these two categories are of nearly
equal size and together make up just over one—third of the entire

group. Only three people do not indicate their income.

Table 6 Annual Household Income During Last Three Terms

Absolute Adjusted Cumulative
Income Range Frequency Frequency Frequency
($) (Z) (Z)
O-4,999 64 17.9 17.9
5,000-9,999 58 16.2 34.1
10,000-14,999 28 7.8 41.9
15,000-19,999 24 6.7 48.6
20,000-24,999 33 9.2 57.8
25,000-29,999 27 7.5 65.4
30,000-34,999 33 9.2 74.6
35,000-39,999 24 6.7 81.3
40,000+ 67 18.7 100.0

E
C

Total 358

91

WEEKLY EMPLOYMENT RATE

WOmen report their employment rates for the last three terms they
were in school by checking one of the categories which range from "not
at all" by ten—hour brackets through "over 40 hours per week." The
largest group consists of those who did not work at all (n=101 or
28.1%). Of those who did work, the distributions are remarkably even
between those who worked up to quarter-time, up to half time, up to
three-quarter time, and up to full-time. Mere than one-tenth worked

over forty hours weekly.

Table 7 Weekly Employment Hours During Last Three Terms

Hours Absolute Adjusted Cumulative
WOrked Frequency Frequency Frequency

(1) (1)
None 101 28.1 28.1
1-10 49 13.6 41.7
11-20 61 16.9 58.6
21—30 52 14.4 73.1
31-40 57 15.8 88.9
Over 40 40 11.1 100.0
Total 360 100.0

EDUCATIONAL VARIABLES

PREVIOUS EDUCATION

Approximately half of the women (n-181 or 50.1%) have entered
college three or more times. The second largest group (n-140 or
38.8%) have entered college twice. Only 37 (10.2%) are in their first
college experience, and three did not respond to this question.

A parallel question asks how many years of education each had

accumulated before this last college entrance. A plurality of 120

92

(33.2%) indicate they had two years of college before they entered
this time. The second highest category consists of the 81 (22.4%) who
had a remarkable four years or more of college before they entered
this time. For at least fifty women this means they had a previous
degree, since that many people indicated elsewhere in the
questionnaire that they do have a previous degree (although the
question was not specifically asked). Many cite this previous degree
when asked for reasons they did not complete a degree before age 25.
Others give a non-recent year in response to a request for their
graduation date. We can only speculate about the 31 others who mark a
response of four or more years of previous education. They could have
changed majors or lost credits along the way or may be counting the
total number of years they have attended school without having

achieved four full years of credit toward a degree.

Table 8 Years of Education Prior to Last College Entrance

Absolute Adjusted Cumulative

Prior Education Frequency Frequency Frequency
High School or GED 38 10.6 10.6
1 Year College 54 15.0 25.6
2 Years College 120 33.3 58.9
3 Years College 67 18.6 77.5
4 Or Mere Years College 81 22.5 100.0
Total 360 100.0

Given that 212 women (58.7%) have two years or less of previous
education, including the 38 women with only a high school diploma or a

General Education Diploma (GED), it is all the more impressive that

93

194 of the 361 respondents have already graduated in this two year
period.
RATE OF RECENT COLLEGE ENROLLMENT

Almost exactly half of the respondents have been full-time
students during the most recent three terms of education (n-179 or
49.3%). One hundred thirty—five (37.3%) have been part-time students
and forty-five (12.2%) have been both over the last three terms. Two
do not respond.

DROPOUTS' ATTITUDES

Three questions are asked only of those who have withdrawn to
assess their attitudes about having withdrawn. The first question
inquires about future educational plans. Of those who have been out
more than one term, fourteen say they will return to Michigan State
University at some time in the future. Half of that number (7) report
that they will continue their education sometime, somewhere, but
apparently have no certain plans now. Two indicate they will return
to Michigan State University in the Fall term, 1986. One will attend
elsewhere in the fall, 1986. Fourteen of the withdrawers do not
respond to this question.

When asked whether they feel good about their decision to leave
Michigan State University, twenty-seven answer. Twelve (44.4%) say
they rarely or never feel good about this decision. Two indicate they
occasionally feel good about it. Seven usually feel good about this
decision and six always feel good about the decision. Some of those
six who always feel good wrote messages saying it was the right

decision considering family or job responsibilities or lack of

94

flexibility of the university at the time, but that when other
circumstances change, they do want to continue their education.
Therefore, a total of thirteen primarily feel good about this decision
and fourteen primarily do not.

These responses parallel quite well the answers to the next
question: "Do you wish you had stayed in college?" Of the
twenty-nine replying, thirteen or 44.8% say they frequently wish they
had stayed in college. Nine sometimes wish they had stayed. Sixteen
have that wish occasionally. Only one rarely or never wishes she had
stayed in college.

These responses also match well with their indications of why
they left the university. Contrary to expectations, few left because
they had achieved all they wanted from education. Primarily they left
for other sorts of reasons which apparently caused them to feel they
could not go on at this time in this place, but they wish they had
been able to do so. (See Table 21.)

. SPOUSES ' EDUCATION

Married women are queried about the highest level of their
spouses' education. Interestingly, there are 189 replies to that
question although only 182 are married. Some who respond list
themselves as divorced. Some single women indicate they are
cohabiting and include their partner's information. 0f the 182 who
are married, 94.5% are married to men with at least some college

education.

95

The most common amount of education is a Bachelor's degree, held
by 63 of the husbands (34.6%). Fifty-six (30.8%) have some college

education. Twenty-five others (13.7%) have a Master's degree or its

Table 9 Level Of Married Spouses' Education

Prior Education Absolute Adjusted Cumulative
Achieved by Spouse Frequency Frequency Frequency

(1) (2)
Some/All High School 10 5.5 5.5
Some College 56 30.8 36.3
Bachelor's Degree 63 34.6 70.9
Master's Degree 25 13.7 84.6
Doctorate 28 15.4 100.0
Total 182 100.0

equivalent (ex.: law degrees were written in). Twenty-eight (15.4%)
have a Doctorate. Ten mark the space indicating their spouses have

some or all of a high school education.

SUBJECTIVE SELF-ASSESSMENT OF INCOME AND EDUCATIONAL VARIABLES

INCOME

Asked to assess the perceived adequacy of their family income
while they were in school this last time, the largest group say that
their income was adequate (n-162 or 45.0%). However, a disturbingly
large group report their income as being inadequate (n-99 or 27.4%) or
very inadequate (n-49 or 13.6%). Only one did not respond to this
question while 50 (13.9%) call their income very adequate. This means
that 41% of the women feel their family income was inadequate while
attending school according to their individual assessments of adequacy

or inadequacy.

96

Table 10 Adequacy 0f Family Income During College

Income Absolute Adjusted Cumulative
Adequacy Frequency Frequency Frequency

(1) (1)
Very Adequate 50 13.9 13.9
Adequate 162 45.0 58.9
Inadequate 99 27.5 86.4
Very Inadequate 49 13.6 100.0
Total 360 100.0

A separate question asks whether finances during college were a
problem. It is interesting to note that while a majority (59%)
believe their income was "adequate", a majority (64.5%) also believe
finances have been a major or a moderate problem. Only seventy-one
(19.7%) find finances to have been no problem at all, and 56 (15.5%)

say they were a minor problem.

Table 11 Finances During College: Rating As a Problem

Problem Absolute Adjusted Cumulative
Rating Frequency Frequency Frequency

(1) (1)
Major Problem 97 27.1 27.1
Moderate Problem 134 37.4 64.5
Minor Problem 56 15.6 80.2
Not a Problem 71 19.8 100.0
Total 358 100.0

While only 13.6% describe their family income as very inadequate,
a larger group or 17.7% had an annual income of under $5,000, and
26.9% indicate that finances were a major problem. Therefore,
apparently inadequacy of income and whether or not finances are

perceived as a problem are not as clearly the same issue as might be

97

presumed. It thus appears appropriate to ask about both adequacy and
whether or not finances are a problem rather than to use only one of
those questions as a complete substitute for the other.

REASONS FOR PREVIOUS WITHDRAWAL

Several possible responses are offered to a question about why
they had not completed their baccalaureate education before they were
twenty-five years old. Subjects mark as many responses as-each
considers appropriate. They are then asked to indicate the one main
reason for that previous noncompletion.

A total of 324 women check one or more reasons from the fourteen
choices offered. Several use this opportunity to indicate that they
jggg receive a degree before age twenty-five but came back now for
undergraduate credit for another reason.

Although it would not be surprising if marriage was the most
frequent reason for previous withdrawal as it has been on other
previously reported surveys, a lack of purpose or goal is instead the
most common single reason marked for earlier dropping out (n-157).
This is the most frequently marked category by both current students
and transfers although it is the second most frequent response for
graduates and ties for second place for dropouts. A very close second
reason is the lack of money. This ranks as the first reason for
stop-outs and for graduates while being the second reason for current
enrollees and ties for second place among those who withdrew (n-154).

The third most commonly marked reason for dropping out previously
is the first reason among those who have currently withdrawn:

marriage. This is given as a reason by 144 respondents, which means

98

Table 12 Reasons For Previous Withdrawal

Current Student Status

Reason for Previous Grad- Cur- With— Trans— Stop- Achieved

 

Withdrawal uate rent drawn fer out All Total
Marriage 75 43 19 3 4 0 144
Having Children 73 39 8 2 4 0 126
Family

Responsibilities 41 22 6 1 3 0 73
Personal Problems 25 24 2 3 4 0 55
Lack Family Support 44 17 6 0 1 1 69
Lack Friend Support 9 5 0 0 O 0 14
Lack Time 33 20 7 1 1 0 62
Academic Difficulty 17 12 3 2 0 0 34
Lack Interest 47 32 7 3 0 1 90
Lack Purpose, Goal 82 54 13 1 5 1 157
Lack MOney 87 47 13 1 5 1 154
Wanted to Work 50 22 11 2 1 1 87
Job/School Conflict 45 15 11 4 1 0 76
Other 30 15 7 3 1 1 57
Total 658 367 113 26 30 6 1198

99

it is only the third ranking response for graduates and current
enrollees and ties for third among transfers and for second among
stop-outs. The other rather heavily indicated reason is "having
children" which ties for second among current stop-outs, is in fourth
place for graduates and current students, and ranks sixth among those
who have now again withdrawn.

Several responses cluster next. In order these are: lack of
interest, wanting to work, conflict of job with school, family
responsibilities, lack of family support, lack of time and other
nonspecific reasons.

What is probably the most surprising of the responses is that
academic difficulties are second to last in terms of reasons for
having previously dropped out. The least frequent reason is lack of
friends' support. In fact, it is only marked by a few graduated and
current students and not by anyone in other categories.

When limited to indicating one mgig_reason for previous
noncompletion of college, 299 reply. Here the differences are even
more striking. Lack of purpose or goal is the primary reason for
sixty-six or 22.1% of those responding. The next most frequent
responses are a cluster marked by only half as many women as the first
reason: lack of money, cited by only thirty-two or 10.7% of those
responding; having children, selected by thirty-one or 10.4% of those
responding; having a previous degree selected by thirty or 10.0%.
Other responses heavily clustering around twelve responders: family
responsibilities (nulS), wanting to work (n-13), job/school conflict

(n-12), lack of family support (n-12) and personal problems (n-lO).

100

Table 13 Main Reason For Previous Withdrawal

Main Reason for Absolute Adjusted Cumulative
Previous Withdrawal Frequency Frequency Frequency

(2) (1)
Marriage 22 7.4 7.4
Having Children 31 10.4 17.7
Family Responsibilities 15 5.0 22.7
Personal Problems 10 3.3 26.1
Lack Family Support 12 4.0 30.1
Lack Friend Support 1 0.3 30.4
Lack Time 6 2.0 32.4
Academic Difficulty 4 1.3 33.8
Lack Interest 23 7.7 41.5
Lack Purpose, Goal 66 22.1 63.5
Lack Meney 32 10.7 74.2
Wanted to WOrk 13 4.3 78.6
Job/School Conflict 12 4.0 82.6
Other 22 7.4 90.0
Previous Degree 30 10.0 100.0
Total 299 100.0

Note: "Previous Degree" includes those who marked "Other" and then
wrote in that they checked that to indicate an earlier degree and
those who marked nothing but wrote in such a degree instead.
Marriage is selected by twenty-two or 7.4% as the primary factor in
earlier departure and academic difficulty by only 4 or 1.3%.

If we combine the sixty-six who indicate lack of purpose or goal
with those who lacked interest (23), a clear one-third (99) of the
respondents say they did not earlier complete college simply because
it was not attracting them or helping them focus at that time. Their
presence back in college now appears to say that something important
has changed for them in the interim.

Scores are also summed across what are labeled "family related"

reasons which consists of marriage, having children, and family

101

responsibilities. This category is chosen by 179 women. One of these
is marked by 65 women (18%), two by 62 persons (17.2%), and all three
by 52 persons (14.4%). Another summed category is "personal and
support problems" which encompasses personal problems, lacking support
from family, and lacking support from friends. A total of 116 persons
use this category with 94 marking one of these responses, 21 marking
two of them, and one marking all three. When combining those who
wanted to work with those who had a job/school conflict and therefore
had to leave school, 131 responders use this category. Ninety-eight
mark one of these "job related" reasons and 33 mark two. The final
combined category consists of those who "lack focus". This includes
both those who say they did not have a goal or a purpose as well as
those who expressed a lack of interest in school. A total of 183
women cite this category. One hundred eighteen indicate one of the

reasons and 65 mark two.

Table 14 Total Reasons For Previous Withdrawal

Total Number Absolute Adjusted Cumulative
of Reasons Frequency Frequency Frequency
(1) (Z)

1 24 7.4 7.4

2 55 17.0 24.4

3 79 24.4 48.8

4 75 23.1 71.9

5 52 16.0 88.0

6 18 5.6 93.5

7 12 3.7 97.2

8 3 0.9 98.1

9 4 1.2 99.4

10 2 0.6 100.0
Total 324 100.0

102

This now clarifies a real closeness of frequencies for the most
prominent type of reasons for earlier withdrawal. A lack of focus
(n=103) and family issues (n-l79) are clearly primary.

It seems useful to total the number of reasons for previous
dropping out. Of the 324 who give any reason, the most common number
of reasons marked is three, selected by seventy-nine women (24,4% of
those marking any reasons). The mean number of reasons marked is 3.7
and the median is 3.55. Twenty-four indicate only one reason (7.4%),
four report nine reasons (1.2%) and ten mark only two reasons (2.6%).
REASONS FOR RETURN TO SCHOOL NOW

All 361 women report their reasons for this most recent college
entrance. The strikingly large number of 323 women indicate that at
least one reason for return is personal satisfaction. This category
is chosen by the most persons in every student status category with
the exception of being tied for first with "getting a degree" among
stopouts. Uniformly across categories getting a degree is the next
most frequently chosen response, marked by a total of 259 women. One
hundred ninety-one say they came to prepare for a career change, the
third ranked reason among current enrollees, transfers and dropouts.
Career change is the fourth-ranked reason for graduates and stop-outs.
"Prepare for a job" is marked by 182 women. This is the third ranked
reason among graduates and stop-outs and the fourth-ranked reason for
all others. Fifth is the need to be seen as a success.

A different way to understand this question is to categorize
these seven different reasons into three summary groups. The first

category is called "career reasons" for returning and includes to

103

prepare for a job, to keep or to upgrade a job, and make a career
change. The second category, "personal reasons", includes personal
satisfaction and the choice to get a degree. The third category is
labeled "social reasons" and includes pressure from others and to be

seen as a SUCCESS.

Table 15 Reasons For Return To College

Current Student Status

Reason for Grad- Current With- Trans- Stop— Achieved
Return uate Student drawer fer Out All Total
Prepare For Job 108 57 9 4 3 1 182
Keep, Upgrade Job 23 13 7 3 1 1 48
Prepare For Career

Change 99 61 20 7 2 2 191
Personal Satisfaction 174 98 31 12 5 3 323
Get a Degree 139 81 23 11 5 0 259
Pressure from Others 21 7 2 1 1 0 32
Be Seen As a Success 70 33 8 3 2 0 116
Other 23 5 1 1 0 0 30
Total 657 355 101 42 19 7 1181

 

 

 

Personal reasons for return are cited by 328 women. Both
personal choices are marked by 233 (71.9%) and 95 (28.1%) give one
personal reason. A total of 309 claim career related reasons. Of
these, 202 (65.4%) give one such reason, 99 (32.0%) give two of these
reasons, and eight (2.6%) give all three career related reasons for
returning now. Social reasons are indicated by 130 women with
eighteen (13.8%) choosing both such responses and 112 (86.2%) giving

only one of the social reasons.

104

Table 16 Total Number Of Reasons For Last College Entry

Total Number Absolute Adjusted Cumulative
of Reasons Frequency Frequency Frequency
(1) (Z)

1 22 6.1 6.1

2 73 20.2 26.3

3 115 31.9 58.2

4 86 23.8 82.0

5 55 15.2 97.2

6 5 1.4 98.6

7 3 0.8 99.4

8 2 0.6 100.0
Total 361 100.0

Again, the most frequent number marked is three reasons. The
mean number of reasons is 3.3, and the median point on the list of
reasons is 3.2. One hundred fifteen or 31.9% give three reasons and
two check all eight possible reasons. Twenty—two (6.1%) give only one

reason a

Table 17 Main Reason For Return To College

Absolute Adjusted Cumulative
Main Reason Frequency Frequency Frequency
(1) (2)

Prepare For Job 54 17.4 17.4
Keep, Upgrade Job 11 3.5 21.0
Prepare For Career

Change 72 23.2 44.2
Personal Satisfaction 112 36.1 80.3
Get A Degree 45 14.5 94.8
Pressure From Others 5 1.6 95.5
Be Seen As A Success 2 0.6 97.1
Other 9 2.9 100.0
Total 310 100.0

 

When only one main reason can be given, the one most important

reason for returning to college is cited as "personal satisfaction" by

105

112 or 36.1% of the 310 responders. This choice widely leads the
second reason, career change, which 72 (23.2%) select. In order, the
next most popular responses are to prepare for a job, (n-54, 17.4%)
and to get a degree (n-45, 14.5%). Only a few indicate the primary
motivator as being to keep or to upgrade a job (nall, 3.5%), pressure
from others (n=5, 1.6%) or to be seen as a success (n-2, 0.6%). Nine
mark "other".

Another way to summarize reasons for return is to note that of
those women who indicate the one main reason for return, 157 (50.6%)
persons give a personal reason, 137 (44.2%) indicate a career reason,
and only seven (2.3%) give a social reason as the primary reason to

return now to college.

Table 18 Primary Type Of Main Reason To Return

Type of Reason Absolute Adjusted Cumulative

 

to Return Frequency Frequency Frequency

(1) (1)
Career 137 44.2 44.2
Personal 157 50.6 94.8
Social 7 2.3 97.1
Other 9 2.9 100.0
Total 310 100.0

CHANGE 0R CRISES CONTRIBUTING TO THE DECISION FOR THE RECENT COLLEGE
ENTRY

While sixty-five people say no changes or crises contributed to
their decision to return to school and nine do not respond at all to
this question, 287 people (79.5%) say such effects occurred.

Respondents are asked to both indicate which crises or changes had

106

occurred and to rank each by a scale of 1 (little) to 4 (severe),
indicating their assessment of the severity of the change or crisis.
The eight possible topics are marriage, family, health, personal
problems, job, move, family member status and an unspecified other
topic. The largest number of responses fall in the "severe" category
for all of eight possible topics except "family" where the largest
number of responses are in the "moderate" category and the "severe"
category has the second largest group of responses.

Nearly half of the sample indicate that a job change or crisis
contributed to their schooling decision. Of these 146, 84 (57.5%) say
their change or crisis was severe. Approximately 20% (ns75) list the
second most common category as "other". Here, however, a surprisingly

high 84% say the change or crisis was severe.

Table 19 Type And Severity Of Changes/Crises Contributing To Return

Severity
Little Mild Mederate Severe
Change/Crisis In: (#) (%) (#) (%) (#) (%) (#) (%) Total
Marriage 1 1.8 2 3.6 15 27.3 37 67.3 55
Family 1 2.1 5 10.6 24 51.1 17 36.2 47
Health 1 5.9 1 5.9 5 29.4 10 58.8 17
Personal Problems 0 0.0 2 3.8 25 47.2 26 99.1 53
Job 5 3.4 16 11.0 41 28.1 84 57.5 146
Move 2 6.1 3 9.1 9 27.3 19 57.6. 33
Family Member Status 0 0.0 1 2.9 10 28.6 24 68.6 35
Other 2 2.7 1 1.3 9 12.0 63 84.0 75
Total 12 31 138 280 461

Several categories hover near 14% of the sample: Marriage

(n-55), personal problems (n-53) and family issues (n-47). Less than

107

10% indicate a change of family member status (n=35) or a move (n-33)
contribute to this decision and half that number (n-17) say that
health made a difference. However, despite being chosen by small
total percentages, the total numbers of people effected by such
changes are still important.

It is curious that only a very few choose the category "little"
and only approximately 3% to 11% use the category "mild". For most
people, either the change or crisis was at least moderate or it seemed
to have little to do with the educational decision. This could
suggest that these were truly experienced as "crises" rather than or
in addition to being "changes" that occurred.

Of the 287 women responding, 150 or 52% of those indicating any
changes or crises at all say they experienced one crisis. Eighty-five
(30%) tell of two crises, 42 (15%) indicate 3 crises; eight (3%) give
four crises; and two (1%) list five crises. Thus the mean among these
287 women is 1.7 different kinds of crises and the median is 1.46
different types of crises.

CHANGE OR CRISES RESULTING FROM THE LAST ENTRANCE T0 COLLEGE

One hundred twenty-two persons say that no changes or crises
resulted from their return to college. This is nearly twice as many
as indicated no precipitant changes or crises‘bgfggg_their return. Of
the rest, 212 (58.7%) mark one or more resultant changes or crises.
They use the same categories as in the previous set of changes or
crises, although here the distribution is somewhat different. While
more still mark the "severe" category (n-127), nearly as large numbers

are in the "moderate" category (n-103). Furthermore, the proportions

of marks in the "little" and "mild" columns have increased in this

section.

108

The most striking difference is that only half as many people

indicate a job change or crisis as in the previous section (n-77 as

compared to the previous n-146), although it is still the leading

change or crisis (positive or negative direction is not indicated).

Here family (n=54), personal problems (n-53) and marriage (n-51)

nearly tie for being the next most common.

Table 20 Type And Severity Of Changes/Crises Resulting From Return

Change/Crisis In:

Marriage

Family

Health

Personal Problems

Job
Move

Family Member Status
Other

Total

Little

NHwar-IHN

26

(#) (%)

13.7
1.9

Severity

Mild Mederate
(#) (Z) (#) (Z)
11 21.6 12 23.5
8 14.8 26 48.1
6 25.0 5 20.8
9 17.0 22 41.5
9 11.7 23 29.9
6 15.8 11 28.9
1 10.0 3 30.0
3 6.7 13 28.9
53 115

Severe

(f) (%) Total

41.2
35.2
50.0
35.8
53.2
36.8
50.0
60.0

Another major difference is that in only one of these leading

51
54
24
53
77
38
10
45

352

categories did a majority of persons mark that the problem is severe.

It is interesting to note that exactly the same number and, in fact,

many of the same people, cite personal problems here as did in crises

or changes before reentry, although the severity seems to have

decreased somewhat.

Nearly the same number report having moved. The

109

major decreases are in jobs, the category labelled "other", and in
changed family member status.

Looking again at the total number of recent changes or crises
marked we find that, of the 212 who list any, slightly over half
(n-109 or 51%) cite only one change crisis, which is a similar
percentage to the previous category but is only two-thirds as many
people. Fifty—nine (28%) mark two changes crises, thirty-one (15%)
check three, eight (4%) claim four, four (2%) give five, and one says
she experienced six categories of crises or changes as a result of her
return to college. The mean at 1.78 is somewhat higher than for
previous changes or crises although the mode remains at one. When the
scores of their crises or changes are summed, the mean is 5.66 which
is slightly lower than that for changes or crises before return.

It is clear that these students have been experiencing stresses
in their lives and that many of these stresses are perceived as having
been quite severe.

REASONS FOR RECENT OR CONTEMPLATED WITHDRAWAL

Those respondents who have left the university for any reason are
asked to check all reasons which caused them to leave. Current
students are asked to mark the reasons which may still cause them to
leave. Each marks as many as apply from a list of twenty-one possible
reasons and are then also asked to select which would be the one main
reason for withdrawal.

When multiple reasons are marked, overall the most frequently
suggested possible or actual reasons for withdrawal, in order, are:

1) finances, 2) a tie of family responsibilities and not enough time,

110

4) too many roles, 5) scheduling problems, 6) job responsibilities and
7) needing a break or a change.

Several items are selected by only a few: lack of family support
or others' support, academic difficulties, problems with faculty and a
move out of the area. Despite some suggestions in the literature to
the contrary, lack of self-confidence is only indicated by twenty
women, eight of whom graduated and seven who are still in school.

However, for those who actually have withdrawn, finances is only
the fourth-ranked reason. The most frequently marked response for
withdrawers is family responsibilities. Not enough time and job
responsibilities tie for second among those who actually have
withdrawn.

Interestingly, several who graduated completed this section of
the survey. The largest group of them (n-20) say finances could have
caused them to leave. The second most frequently marked reason for
graduates is needing a break or a change.

Transfer students say lack of classes or program needed and then
scheduling problems are the most frequent reasons for having left.

When asked for the‘ggig.reason for withdrawal, finances become
clearly the single most important reason, cited by 25.6% of the 133
who actually mark a primary reason. Family responsibilities and the
general category of "other" come next, followed by scheduling
difficulties which is chosen by 6.8%. There is little uniformity to
other responses. Graduates (n-34) most frequently cite finances

(n-ll). No other category is chosen by more than three graduates.

111

Table 21 Current Reasons For Withdrawal

Current Student Status

Grad- Current With- Trans— Stop-

Reason for Withdrawal uate Student drawer fer ‘Out
Achieved All Desired 7 8 1 1 0
Family Responsibilities 11 19 17 O 3
Personal Health 5 9 5 0 1
Personal Problems 9 14 7 O 1
Family Illness/Death 4 9 3 0 1
Job Responsibilities 9 15 14 3 1
Too Many Roles 8 26 8 2 2
Not Enough Time 8 25 14 2 1
Lack Self-Confidence 8 7 3 2 0
Lack Family Support 1 8 1 0 0
Lack Others' Support 2 3 1 0 O
Classes/Program Lacking 4 7 3 6 1
Need Break 0r Change 17 13 5 O 3
Finances 20 37 13 l 4
Academic Difficulties 8 6 2 O 1
Scheduling Problems 7 23 7 5 1
Problems With Faculty 3 3 1 3 1
Problems With Univer-

sity Bureaucracy 10 ll 4 2 0
Lack Interest 9 7 3 0 2
MOved Away 3 4 3 0 O
Other 4 10 6 5 2
Total 157 264 121 32 25

Total

17
50
20
31

112

Table 22 Primary Current Reason For Withdrawal

Absolute Adjusted Cumulative
Reason for Withdrawal Frequency Frequency Frequency
(1) (Z)

Achieved All Desired 5 3.8 3.8
Family Responsibilities 13 9.8 13.5
Personal Health 5 3.8 17.3
Personal Problems 3 2.3 19.5
Family Illness/Death 3 2.3 21.8
Job Responsibilities 6 4.5 26.3
Too Many Roles 6 4.5 30.8
Not Enough Time 4 3.0 33.8
Lack Self—Confidence 5 3.8 37.6
Lack Family Support O 0.0 37.6
Lack Others' Support O 0.0 37.6
Classes/Program Lacking 3 2.3 39.8
Need break Or Change 3 2.3 42.1
Finances 34 25.6 67.7
Academic Difficulties 5 3.8 71.4
Scheduling Problems 9 6.8 78.2
Problems With Faculty 0 0.0 78.2
Problems With Univer-

sity Bureaucracy 6 4.5 82.7
Lack Interest In Classes 6 4.5 87.2
Moved Away 2 1.5 88.7
Other 15 11.3 100.0
Total 133 100.0

113

Among withdrawers, three cite family responsibilities and only
three cite finances as the main reason for withdrawal.

Again, it is interesting to attend to the total number of reasons
marked. The modal number is two, marked by 34 (20.72 of the 164
responders), followed closely by three responses. Fewer choose from
six to sixteen of the twenty-one choices.

Given that we now understand something about who these students
are sociodemographically, educationally, and in terms of what changes
or crises have gone on before and after their return to college, it
appears appropriate to look next at the academic record these women
have accumulated and then to assess the psychological adjustment they

have come to along the way.

ACADEMIC ACHIEVEMENT

Prior to this last entry to college, the mean grade point average
(GPA) of the 288 who both had previous credits and could remember
their previous GPA is 3.01, and their median GPA of 3.10. The most
common previous grade point average was the 3.5 achieved by thirty
women. The lowest is a 1.0 held by four subjects. Nine women had a
4.0 GPA prior to this last educational experience. Twenty-six
responders indicate they cannot remember their previous GPA, although
most of those do answer the question of whether their present GPA is
higher, lower or approximately the same as it had been before.
Therefore, apparently they know the range but not the exact number.
The forty-seven who say they have no previous GPA is slightly more
than the number who have not previously attended college: Some write

in that they were not previously on a numerical grading system.

114

All except two are able to provide a cumulative GPA, and by far
the most students provide at 1£§§£_one digit past the decimal point.
The mean total GPA to date is 3.17. The GPA for the most recent
single term is 3.42.

One way of understanding these grade points is to notice the
difference between the current cumulated GPA (3.17) and the previous
GPA (3.01) by subtracting the latter from the cumulated GPA. The mean
difference is +.16, meaning that the cumulative GPA is higher than the
previous GPA, even though several speak of the increased difficulty of
classes here compared to previous schools, especially community

colleges.

Table 23 Grade Point Averages (GPA's)

Type of Grade Point Average Mean s.d Max. Min. Median Mode
Cumulative GPA (Cum.) 3.167 .462 4.00 1.80 3.200 3.00
Last One-Term GPA (Recent) 3.417 .568 4.00 1.00 3.501 4.00
Previous GPA (Prev.) 3.014 .651 4.00 1.00 3.096 3.50
Difference: Cum. Minus Prev. GPA .158 .569 2.10 -l.75 .0833
Difference: Recent Minus Cum. GPA .249 .524 1.80 -2.00 .2000

Another way to look at differences is to subtract the cumulative
GPA from the last one-term GPA to assess how they are doing currently
compared to the total weight of the GPA they carry with them. This
difference is +.25, meaning that the last one-term GPA is a quarter of
a point above their total cumulative GPA. This difference is computed

with only four cases missing because of missing data.

115

The question is raised of whether the present GPA is individually
higher than before, approximately the same as before, or lower than
before they returned to college. One hundred ninety-seven or 59.32
claim an increased GPA, while only a quarter of that number (n-52 or
15.72) say they currently have a lower GPA than before. Eighty-three
(25%) claim no clear change in their GPA's. It is interesting to
note a source of variance here, however: More people answer this
question than claim any previous college experience. It is likely
that those extra people are making comparisons to high school

achievement.
SELF-REPORTED SINGLE INDICATORS 0F Anmsm

EXPECTATION OF DEGREE COMPLETION

Subjects are asked whether, while in college this last time, they
expected to complete a degree without any further breaks. 0f the 313
responses, a relatively large 74.81 or 234 claim they have been quite
sure of achieving their degrees without further breaks. Another

forty-five (14.4%) were sure much of the time. That leaves only 11%

Table 24 Expectation Of Completing Degree Without Further Breaks

Absolute Adjusted Cumulative
Degree of Expectation Frequency Frquency Frquency
()
Yes, Quite Sure 234 74.8 74.8
Sure Much Of The Time 45 14.4 89.1
Unsure Much Of The Time 26 8.3 97.4
No 8 2.6 100.0

Total 313 100.0

116

who were less than sure, including the twenty-six who claim to have
been unsure much of the time and the eight who acknowledge they did
not expect to remain and complete a degree this time. Of those who
were less sure, six actual withdrawers and ten graduates were unsure
much of the time and four dropouts did not expect to complete their
degrees.
LIFE SATISFACTION

Of the 316 assessing their life satisfaction as a whole during
school, twenty-five percent were very satisfied and 58.9% say they
were mostly satisfied. This represents a satisfied group of an
accumulated 842 of the subjects. A smaller group of 14.6% claim to
have been mostly dissatisfied and 1.6% report being very dissatisfied.
Withdrawers make up only four of the dissatisfied and one of the very

dissatisfied groups. Graduates are more heavily dissatisfied (n=29).

Table 25 Life Satisfaction During And Since School

 

During School Since School
Degree of Adjusted Cumulative Adjusted Cumulative
Satisfaction Number Freq. Freq. Number Freq. Freq.
(Z) (Z) (Z) (1)

Very Satisfied 79 25.0 25.0 59 27.6 27.6
Mostly Satisfied 186 58.9 83.9 116 54.2 81.8
Mostly Dissatisfied 46 14.6 98.4 35 16.4 98.1
Very Dissatisfied 5 1.6 100.0 4 1.9 100.0
Total 316 100.0 214 100.0

This does not differ greatly from the responses to the query
about life satisfaction as a whole since leaving school. Obviously,

this question excludes current students so yields smaller response

117

rates. Nevertheless, the proportions are quite similar with 81.8%
(n-175) being primarily satisfied and 18.22 (n-39) being primarily
dissatisfied, which is a two percent increase in the dissatisfaction
group. Of these, withdrawers account for seven (one very
dissatisfied) and graduates for 25 (three very dissatisfied).
HAPPINESS

The frequencies on the foregoing more cognitive satisfaction
measures are relatively close to the percentages found in answer to a
separate question about their happiness during school, which is a more
emotional state according to the literature. Of the three hundred
sixty replies to this question compared to the life satisfaction
responses, a somewhat smaller 19% report they have been very happy, a
larger 66% mark that they have been mostly happy (or an accumulated
852 being happy), fourteen percent report being not very happy and one
percent claim to be very unhappy. Graduates account for 34 of the
fifty who were not very happy and for one very unhappy student.

Dropouts include four who were not very happy during school.

Table 26 Happiness During School

Degree of Absolute Adjusted Cumulative

Happiness Frequency Frequency Frequency
(2) (1)

Very Happy 67 19 19

Pretty Happy 239 66 85

Not Very Unhappy 50 14 99

Very Unhappy 4 1 100

Total 360 100

118

GETTING WHAT IS WANTED OUT OF LIFE

One-third say they are able to get the things they wanted out of
life very well. Another 55% say they can at least adequately get the
things they want out of life. Only ten percent report doing poorly at
getting what they want including 19 graduates and three withdrawers,
and two percent claim to be doing very poorly at achieving what they

want including one dropout and three graduates.

Table 27 Getting What Is Wanted Out Of Life

Absolute Adjusted Cumulative
Frequency Frequency Frequency
(Z) (1)
Very Well 118 33 33
Adequately ' 200 55 88
Poorly 35 10 98
Very Poorly 8 2 100
Total 361 100

COPING WITH/HANDLING CRISES
Thirty-five percent (n-125) claim to cope with or handle crises

in their lives very well; fifty-four percent (n=196) handle crises at

Table 28 How Well I Cope With/Handle Crises

Absolute Adjusted Cumulative
Frequency Frequency Frequency
(2) (2)
Very Well 125 35 35
Adequately 196 54 89
Not Very Well 36 19 99
I Am Overwhelmed 4 1 100

Total 361 100

119

least adequately; ten percent (n-36) say they do not do this very
well, and one percent (n-4) acknowledge being overwhelmed by crises.
These frequencies are very similar to those in response to the
question on getting the things wanted out of life. Sixteen graduates
and eight dropouts feel negatively about their own crisis-handling
ability.
SELF-ASSESSMENT OF DEGREE OF CONTROL OVER OWN LIVES
Each woman assesses her degree of control over her own life.
Seventeen percent claim to have complete control, seventy-one percent
say they have much control, eleven percent acknowledge having little
control and one percent say they have no control over their own lives.
Twenty-three graduates and five dropouts see their lives as not being

much under their own control.

Table 29 Self-Assessment Of Control Over Own Life

Absolute Adjusted Cumulative
Frequency Frequency Frequency
(1) (1)
Complete Control 61 17 17
Much Control 255 71 88
Little Control 41 11 99
No Control 2 1 100
Total 359 100

FORMAL TEST BATTERIES

LOCUS OF CONTROL
In Duttweiler's twenty-eight item Locus of Control scale,
respondents write in a number representing one of five possible

responses to each question. This allows a median choice as well as

120

moderate and strong positive and negative responses. After recoding
responses to allow all so-called internal responses to be in one
direction, a higher score indicates a more internal locus of control
and a lower score indicates a more external locus of control. The
possible range of scores is from the extreme external response of 28
to the extreme internal response of 140. Two subjects skip this scale
entirely although completing the rest of the survey. Nine omit one or
more responses and their responses are therefore not summed into a
total score. Of the 350 who do respond fully, the mean is 106.20 and
the mode is 106. All of the scores achieved by at least ten persons
are between 97 and 117 which covers the range between 21% and 832 of
the respondents. Although the scale allows scores as low as 28, the
lowest score for anyone in this study is 70. Five percent score below
86 and five percent score above 126. (3:10)
SELF ESTEEM

Rosenberg's Self-Esteem scale consists of ten items and allows
one of four possible responses to each statement about oneself. When
the responses are recoded so they are all in the same direction, the
lowest possible score of ten represents extremely low self-esteem and
the highest possible score of forty represents very high self-esteem.
The range of scale scores is from sixteen to forty, with forty also
being the modal response given by 11.22 of the women out of the 357
complete self-esteem batteries. The mean is 33.9 and the median is
34.9. Three—quarters of the responses lie between 30 and 40, while

the highest quartile of the responses is above 37. This seems to

121

indicate a relatively high rate of self-esteem in this sample although
the lower quartile of responses are quite widely distributed.
RESPONSES TO OPEN-ENDED QUESTION

A remarkable number of the respondents, 245 or 67.92 choose to
make individualized responses to the open—ended question which elicits
information beyond that asked on the questionnaire.

When the open-ended responses are categorized and ranked, the two
raters are relatively consistent on all except three categories of
responses of women's comments and are in somewhat closer agreement
about their suggestions of questions to raise. One rater finds a few
more individual responses to code than the other, especially when
categorizing the comments made about respondents' own situations.

Comments are made most often about support. Some indicate that
support was important to them; others say support was needed and may
have been missing. Comments here include the desire for a social
life, generally positive but sometimes negative reactions to the
support of family and friends, reflections on the support offered by
the university or the lack thereof, and expressed desires for support
groups and programs.

Scheduling receives the second highest number of comments,
particularly complaints about the lack of night classes or lack of
much choice of night classes and also about required classes which are
never offered at night, the rigidity of scheduling (classes always
offered at the same impossible-to-go—then time) and the desire for
more classes which have at least one section offered for longer blocks

of time and on fewer days per week.

122

Finances and financial aid (and the difficulty of getting or the
unavailability of aid for older students) tie for second on the list
of comments. Fourth are the many comments about the multirole
balancing these women often do. Included here are issues of overload,
time conflict, prioritizing, role conflicts, etc.

Attitudes and discrimination toward the older woman student, on
the basis of sex and/or age rank fifth. These include in largest
numbers perceptions of bias from faculty but also from other students
and staff.

Sixth ranked are attitudes held by the respondent about her
experience as a whole as an older student and how she perceives of
herself as an older student in a younger university setting.

Four categories tie for seventh place: 1) Expressions of
positive or negative attitudes about Michigan State University,
including several by women who clearly state they are highly unhappy
alumni, 2) Parking problems, 3) Statements (mostly negative) about
the quality of education specifically including instructors and/or
curriculum, and 4) Postgraduate goals and experiences including
assessments of whether the degree was worth the effort. The last
category of comments with enough responses to rank is academic, career
and other counseling and advising as done in many departments on
campus.

The many responses which do not fit well into other categories
are summed under the heading of "miscellaneous". Each of the raters

went one step further in comsidering these to be personal (largest

123

subcategory for both raters), university-related (second for both) or
situational (last for both) issues.

The women raise questions which mostly fit into the same
categories although these rank somewhat differently. The largest
number of women suggest that questions should be asked about their
postgraduate experience. Next, they want questions about the support
they had, lacked, or wanted and the sources of that support. They
would like to answer questions about attitudes toward older women
students in the university, multiple role balancing, sources and
adequacy of finances, attitudes about themselves as older students,
child care, and counseling. Again, responses having too little

similarity to others to be categorized are labelled "miscellaneous"

Table 30 Open-Ended Responses by Categories

Category '£_Comments Ranking Overall'£_ggestions Ranking
Overall
Rater Rater Ranking Rater Rater
Ranking
A B A B A B A B
Support 60 52 1 1 1 32 28 2 2 2
Scheduling 44 35 2 3 2 5 10 9
Finances 35 36 3 2 2 14 16 4 6 5
Role/Time Load 34 34 4 4 4 l4 l7 4 4 4
Attitudes/Discrim.24 22 7 6 5 18 20 3 3 3
Self-Attitudes 10 27 10 5 6 8 l7 7 4 6
Parking 16 15 8 8 7 0 2
Quality of Educ. 13 19 9 7 7 2 5
Pos/Neg Experience24 14 6 10 7 1 O
Post-Graduate
Goals/Experience 23 10 5 11 7 49 31 1 l 1
Counseling 7 15 11 8 11 6 14 8 7 8
Child Care 4 6 12 14 13 4 8 7
Miscellaneous 43 7O 56 53

Totals 337 366 224 236

124

The first accompanying table (Table 30) presents the total number
of responses in each category, the ranking of that category by each
rater, and an overall ranking resulting from summing the individual
rankings. Only categories receiving at least six different responses
are contained in the rankings.

Some women raise questions, some make comments, and some do both.
The redundancy is eliminated in Table 31 so no one is counted twice on
one topic. The total number of non-overlapping items gives a more
generalized picture of the frequency that each particular subject is
mentioned by the different women writing replies. This overall
ranking by topic without reference to whether questions are raised or
commemts are given is more similar to the ranking of comments than to

the ranking of questions.

Table 31 Non—Overlapping Open—Ended Responses by Categories

Category Rater A Rater B Overall
Total Ranking Total Ranking Ranking
Support 87 l 74 1 1
Role/Time Load 47 4 49 2 2
Scheduling 49 3 44 4 3
Finances 45 5 49 2 3
Post-Graduate
Goals/Experience 68 2 4O 6 5
Attitudes/Discrimination 4O 6 38 7 6
Self-Attitudes 18 8 43 5 6
Child Care 18 8 19 10 8
Pos/Neg Experience 25 7 14 12 9
Counseling 11 12 26 8 10
Quality of Education 15 11 21 9 10
Parking 16 10 16 11 12

Miscellaneous 89 117
Totals 528 550

125

SUMMARY OF FREQUENCY DATA

Although the separate findings of the frequency data are
important to note topic by topic, a summary of some of the more
important findings is useful to review in one place before going on to
further statistical analyses of the data. Such a summary is presented
in Table 32.

It is interesting to compare findings in the literature to the
findings in this study. Although the women in this study are not
particularly different from those in the literature on age, they are
less frequently married, are at the low end of the spectrum in whether
they have children, and differ from the literature in main reasons to
withdraw from college previously, reasons to return now, and, among

current withdrawers, in reasons to leave now. Such comparisons are

charted in Table 33.

Table 32

Variable
Student Status

Age
Marital Status

Have Children
Income

Employment
Prior Education

Enrollment Rate
Income Adequacy
Finances - Problem?
Previous Dropout

Return Reasons
Prior Crises/Changes

Resulting Crises/Ch
Dropout Reasons Now

Previous GPA
Cumulative GPA

Last Term GPA
Expect Degree,
Without Break

Life Satisfaction
in School

Happiness in School

Getting What One
Wants in Life

Cope with Crises
Control of Own Life

126

Frequency Data (n a 361)

Frequency

53.72 graduates, 11.52 dropouts, 3.62 transfers
29.62 current students, 1.7% stopouts

Mean . 34.55 Mode - 29 Median - 32.09

28.02 single, 50.4% married, 17.7% divorced,
22 widowed and separated

56%

Mean - $20,000 - $25,000

Mode - Over $40,000, next largest brackets -
under $5,000 and $5,000 - $10,000.

28% not employed; others equally distributed
across 1/4, 1/2, 3/4 and full-time employment

50 - previous degree; 212 - up to two years;
102 . no college; 50% - 3+ college entrances

502 full—time, 382 part-time, 132 both

412 - Somewhat or very inadequate

64.5% . yes, major or moderate problem

1. Lack of purpose/goal 2. Lack of money

3. Marriage 4. Having children

Least: Lack friends' support; Acad. difficulty

1. Personal satisfaction 2. Get a degree

3. Career change

802 - Yes. 1/2 a Job Issue; 842 - severe

58.72 a Yes. Mbst - job. Split of mod., severe

Overall: 1. Finances 2.(tie) Lack of time
and Job responsibilities

Dropouts: 1. Family responsibilities 2. Lack
of time 3. Job responsibilities 4. Finances

Mean - 3.01 Mode - 3.4 Median - 3.10

Mean - 3.17. Higher than before - 591. Lower
than before - 162

Mean - 3.42

(g - 812, w - 412)*

142 - Mostly sure.
(g s 131, w - 282)

252 - Very satisfied 59% - Mostly satisfied

(g - 23%, w - 262) (g - 592, w - 582)
192 - Very happy 66% - Mostly happy
(g - 162, w - 272) (g - 662, w - 621)

332 - Very well

(g - 402, w - 242)
252 - Very well

(3 - 371, w - 292)
172 - Complete

(3 - 212, w - 202)

* g - graduates, w - withdrawers

552 . Adequately

(g - 482, w - 662)
542 - Adequately

(g - 551, w - 502)
712 - Much

(g - 682, w - 67X)

Table 33

Variable
Age
Mean
MOde

Marital Status

Single 5-252
Married Normal 75-772, low 58%
Divorced 8-152

Having Children 50-902

Prior College Education 61—971

Employed
Spouse's Education

Previous Dropout Reasons
Primary Reasons
Secondary Reasons

Reasons for Return
1.
2.
3.
Transitions

Problems/Reasons to Leave
1.

b21019

127

Literature

early to mid-30's
upper 20's to near 30

Most: at least 502

Less ed: More dropouts

Marriage/ family

Job/money

Career (1980's)
Personal Satisfaction

May contribute

Finances

Scheduling
Family
Time

Literature Findings Compared to This Study

This study (n . 361)

34.55
29

28%
50.4%
17.72

562
89.42
721

Not significant

Lack of Purpose/Goal
Money, marriage,
family

Personal Satisfaction
Get a Degree
Career Change
802 report their role

Family
Responsibilities

Lack Time

Job Responsibilities

Finances

128

FINDINGS ACROSS ALL STUDENT STATUS CATEGORIES

We already know a great deal about this population from these
multiple observations of the frequencies with which various
experiences, attributes, attitudes and accomplishments have occurred
for these women. It is next important to go beyond studying
frequencies to find the significant differences within this
population.

SOCIODEMOGRAPHIC VARIABLES

It seems useful to first assess the broadest comparisons in this
study by using all categories of student status. While the primary
hypotheses of this study are based on comparisons of dropouts and
persisters, it is also valuable to look at‘gll_the subjects including
the more than one-third who are omitted from other analyses. These
total analyses help to understand first, whether either or both
dropouts and graduates differ from students as a whole and second,
whether more or fewer associations are significant across all student
status categories compared to the two criterion groups of our
hypotheses.

The variables of age, marital status, income and whether or not a
woman is a parent are not found to be significant, although a
relationship exists between the number of children per mother and
student status (p-.036). Current student mothers have slightly more
and dropouts slightly fewer children per mother than the average in
the study. When looking at children by age groupings, no relationship
exists across all student status categories of how many have

preschoolers or have children beyond high school age, but the relative

129

proportions of those having school age children do differ (p-.014).
Graduates are somewhat more likely to have more school age children
than expected. None of these differences are significant using only
the two criterion groups.

These associations on children which are significant are,
however, rather weak, which means that they do not very reliably
predict group membership nor do they predict a very large share of the
variance. In fact, although many associations are found to be
significant, most are weak associations. Therefore, only the few
strong associations will be reported from hence forward.

The number of hours of employment per week while a student differ
(p=.029). One third of current students and one—quarter of graduates
and of withdrawers say they have not worked in their last three terms
of school. Transfers mostly worked half-time to over full-time;
graduates averaged about eleven to twenty hours of work if they
worked; working withdrawers averaged 31 to 40 hours weekly work, and

employed current students average 21 to 30 hours of employment weekly.

Table 34 Sociodemographic Variables by Student Status

Sociodemographic Variable X2: p-

Age .925 (ns)
Marital Status .958 (ns)
Having Children .184 (ns)
Number of Children .036
Ages of Children

0-5 years .684 (ns)

6-18 years .014

19+ years .051 (ns)
Income .082 (ns)
Income Adequacy .105 (ns)
Financial Problem? .243 (ns)

Hours Worked .029

130

PERCEPTIONS ABOUT INCOME

No relationships are found across categories of student status
about the women's perceptions of the adequacy of their income nor
about whether finances have been a problem during school. However,
some other comparisons are quite revealing. Self-perceived adequacy
of income compared to actual income and also compared to whether
finances were seen as a problem during school are significant beyond
the p-.0001 level and show quite strong associations, as might be
expected. Despite this rather predictable finding, not all cells are
that predictable. Some women with incomes under $5,000 see their
income as adequate, and some with incomes of $40,000 or more see that
as inadequate.

Adequacy of income compared to the number of hours worked also
shows a very significant relationship at the p-.OOOl level, with those
not being employed at all finding their income to be the most
adequate. Forty-four percent of those saying their income is very
adequate are not employed. Interestingly, those who experienced their
income as being very inadequate are primarily also not employed at all
or are employed up to twenty hours per week. Those who found their
income to be inadequate are very evenly scattered across the
categories of number of hours worked. Those working eleven to twenty
hours or over thirty-one hours generally found their income to be
adequate.

EDUCATIONAL VARIABLES
The enrollment status of students (whether they attended

full-time, part-time, or both during their last three terms) is

131

significant at the p-.OOO level. Dropouts, stopouts and transfers all
are at least twice as likely to attend part-time than full time while
graduates are more than twice as likely to attend full-time as
part-time. No significant relationship is evident between student
status and whether or not a previous degree is possessed or the amoung
of previous education.

A large majority of all groups except dropouts were very sure
they would get a degree without further breaks, including stopouts.
Only forty percent of dropouts had such an expectation. The
relationship of this expectation to student status is highly
significant (p=.0003).

The level of spouses' education for married students across all
status categories is significantly related (p-.015) with graduates'
spouses being the most likely to have some college education, current
enrollees' spouses are most likely to have Bachelor's degrees, and
dropouts' spouses being most likely to have Bachelor's degrees.
Graduates are more likely than any other group to have spouses with
advanced degrees, although they also have the largest proportion of
spouses with a high school education only.

RELATIONSHIPS BETWEEN SELECTED SOCIODEMOGRAPHIC AND EDUCATIONAL
VARIABLES

No relationship exists between level of income and the amount of
spouses' education. In fact, there is a noticeable similarity of
income spread across all categories of spouses' educatioal level for

married women. There also is no relationship between adequacy of

132

income and a woman's number of years of education prior to the present

schooling.

A highly significant and strongly associated relationship
(p-.OOOO) exists between annual income and marital status. 0f the
never marrieds (to be called "singles" hereafter), 38.6% had incomes
under $5,000 and 61.2% had incomes under $10,000 while in school.
Incomes at or above $40,000 are reported by 33.52 of the marrieds,
6.22 of the divorced and 3.02 of the singles. The most frequent
income among divorcees is $5,000 to $9,999, followed by income under
$5,000. The second most frequent income for the marrieds is $30,000
to $34,999, with the next higher bracket of $35,000 to $39,999 being
in third place.

Table 35 Income by Marital Status

Marital Status

Annual Income Never Married Married Separated Divorced Widowed
$ n row I n row I n row 2 n row I n row 1
O - 4,999 39 62.9 8 12.9 0 14 22.6 1 1.6
5,000 - 9,999 23 41.8 10 18.2 1 1.8 21 32.3 0
10,000 — 14,999 13 46.4 11 39.3 0 4 14.3 0
20,000 — 24,999 3 9.1 18 54.5 1 3.0 11 33.3 0
25,000 - 29,999 5 19.2 14 53.8 2 7.7 4 15.4 1 3.8
30,000 - 34,999 4 11.8 26 76.5 1 2.9 3 8.8 0
35,000 - 39,999 1 4.3 22 95.7 0 O 0
40,000+ 3 4.5 60 89.6 0 4 6.0 0
Total ' "101' 179 "'5" 65 ‘7'
Chi Square: p - .OOOO

133

CAUSES OF EARLIER WITHDRAWAL

No association is found between any of the reasons peOple left
college before age twenty-five and the variable of student status
except for the one association of having left due to a lack of money
(p-.Ol7). Stop—outs are the group marking this response in the
highest proportions. Approximately 442 of both graduates and current
enrollees cite the lack of money as having previously contributed to
their dropping out, but it is only cited by 342 of present withdrawers
and eight percent of transferees.

CHANGES OR CRISES

A comparison across groups on changes or crises which contributed
to women's return to school reveals a relationship on the crises or
changes of family (p-.003) and personal problems (pa.026). Whether
these crises or changes are perceived as being positive or negative is
generally not known, although a few subjects comment on that in the
margin.

On changes or crises occurring'gfgg£_returning to school, a very
strong association is found across groups on the issue of health
(p=.0002). A change or crisis of family member status is also
significant (p-.Ol9). although this relationship is curvilinear. No
other changes or crises are significant across all groups.

REASONS FOR CURRENT EDUCATION

Only one of the reasons for having returned this time is
significant across categories of student status: to prepare for a job
(p-.003). Fifty-three to fifty-six percent of current enrollees and

graduates returned to prepare for a job, and 312 of transferees, 24%

134

of dropouts and 502 of stop-outs returned for that purpose. No other
reasons for return nor clustered summaries of reasons are significant
across all groups.

REASONS FOR WITHDRAWAL NOW

An interesting difference from the other previously cited sets of
responses where one or several applicable reasons can be indicated is
found by statistical analysis when looking at the reasons people have
withdrawn or contemplate(d) withdrawing: Of the twenty-one possible
reasons given for withdrawal, nearly all are significantly related to
the variable of student status.

Highly significant associations (p-.OOOO) are found for the
reasons of family responsibilities, job responsibilities, too many
roles, not enough time, lack of classes or programs wanted, finances,
scheduling problems, and a category of other unspecified reasons.
Other significant associations are in problems with faculty (p-.OOOl),
lack of family support, needing a break or change and in personal
problems (all at p-.01). Other reasons showing a significant
association are personal health and lack of interest (both at p-.05).
It is noteworthy that this is the only set of chi square analyses
where the majority of items show a significant level of association
across all categories of student status.

ACADEMIC ACHIEVEMENT
No academic achievement variables are significantly related

across categories of student status.

135

PSYCHOLOGICAL ADJUSTMENT

The only relationship across all categories of student status
with any of the measures of psychological adjustment is how well the
student sees herself as getting the things she wants out of life
(p-.OO6). Approximately forty percent of graduates and transfers and
23% of current enrollees and dropouts say they accomplish this very
well. Transfers and stopouts are the two groups who have larger
percentages who do not think they are doing very well on this
dimension (approximately 162 each).

COMPARISONS BY MARITAL STATUS

Because a visual scan of the data seems to reveal major
differences by whether women are single, married or divorced, each of
these categories is treated separately by creating dummy variables.
This allows testing of significance on separate sets of groups which
now include the singles as compared to all of the rest of the
respondents who are here called the non-singles, the married compared
to all of those not married, and the divorced compared to all who are
not divorced. Fewer significant relationships are found when
comparing the divorced with the not divorced categories using all
responders than are found for married/not married and single/not
single comparisons.

Divorced students show a bulge in the age range between thirty
and thirty-four years old. Single women compared to non-single women
in the study (non—single includes divorced, widowed and separated as
well as married women) are younger than non-singles (p-.OOOO) and

married women are older than non-married women (p-.OOO4).

136

The singles and divorcees find their income less adequate than
non-singles and non-divorcees do while marrieds consider their income
to be more adequate than do non-marrieds (p-.OOO for all three). This
finding parallels the finding that income is lower for single and
divorced women than for married women in the study. Both singles and
divorcees experienced financial problems more often than non-singles
and non-divorcees and the married women experienced fewer financial
problems than their counterparts (all: p=.000). Both single and
divorced women worked more hours per week than their opposite groups
(p=.05) while married women are much less likely to work as many hours
weekly as do non-marrieds (p-.OOO).

Single women are more often full-time students than all
categories of non-single women (p-.Ol). Married women are less often
full-time students than are non-married women (p=.01). Singles
accumulated more education prior to this entry to school (p-.05) and
give fewer reasons for previous dropping out than non-singles (p-.01).
Divorced women give more reasons for withdrawal the previous time than
do those who have never divorced (p-.OOO). Married women list fewer
reasons for recent return (p=.OSO) than do non-married women.

The cumulative GPA is lower for singles than for non—singles
(p-.05), higher for marrieds than for non-marrieds (p-.Ol) and almost
exactly equal for divorcees and non-divorcees. Divorced women claim a
comparatively poorer GPA now than they had in the past (p-.05).

On the psychological adjustment factors, several are significant
when examined by the discrete marital groups. Singles were less

satisfied during school than non-singles (p-.05) and marrieds claim

137

Table 36 All Subjects by Marital Status
Variable Single/ Married/ Divorced/
Not Single Not Married Not Divorced
1:: p- p- p-
Age .000 .000 .069 (ns)
Having Children .000 .000 .068 (ns)
Number of Children .000 .000 .337 (ns)
Income .000 .000 .000
Income Adequacy .000 .000 .000
Financial Problem? .000 .000 .000
Hours Employed .037 .000 .047
Enrollment Status .002 .008 .772 (ns)
Years of Pre-entry Education .048 .196 (ns) .829 (ns)
# Reasons for Previous Withdrawal .003 .613 (ns) .000
# Reasons for Return .243 (ns) .050 .323 (ns)
Cumulative GPA .015 .006 .948 (ns)
Present GPA Minus Previous GPA .713 (ns) .349 (ns) .038
Life Satisfaction During School .012 .002 .496 (ns)
Life Satisfaction Since School .251 (ns) .002 .045
Happiness During School .005 .000 .715 (ns)
Getting What Is Wanted In Life .005 .000 .163 (ns)
Self-Esteem .006 .079 (ns) .040
Locus of Control .197 (ns) .981 (ns) .217 (ns)
Expect Degree Without Break? .366 (ns) .642 (ns) .945 (ns)

138

more satisfaction both then and since leaving school than all
non—marrieds (both p=.Ol). Divorced women have been less satisfied
since leaving school (p-.05) but have higher self-esteem (p-.05) than
non-divorced women. Single women found less happiness during school
(p-.01), have more difficulty getting what they want from life (p-.01)
and have poorer self-esteem (p-.Ol) than is true for non-single women.
Married women experienced more happiness during school (p=.Ol) and are
more able to get what they want out of life (p-.OOO)than non-married
women.

The most striking noticable relationships among these dummy
variables is that married women are consistently better adjusted and
less likely to be full-time students than non-married women and single
women are apparently somewhat less well-adjusted and more likely to be
full-time students than non-single women. Divorced women are rather
heterogeneous, particularly on psychological adjustment variables.

When tested using analysis of variance by marital status,
self-esteem is significant at the .002 level but locus of control is
not found to be significant. The cumulative GPA differs at p-.02, but

the last one-term GPA does not differ.
COMPARISONS OF DROPOUTS AND PERSISTERS

When using only the two criterion groups of persisters to
graduation and withdrawers and ignoring those in the other categories

of student status, somewhat different patterns emerge.

139

SOCIODEMOGRAPHIC VARIABLES

The graduates and dropouts are not significantly related (or
significantly different) on any of the sociodemographic variables
except income (p-.025) and hours of employment per week (p-.OOO).
Somewhat surprisingly, family income is higher for dropouts than for
graduates, and withdrawers also report more hours per week of
employment.

Although income is not significant across all students,
perceptions about it are. A test on the adequacy of income for
dropouts and persisters is significant (p=.Ol4) but turns out to be in
the opposite direction from what was expected. The graduates report
their incomes as being'lg§§_adequate than do the dropouts.
Furthermore, the question of whether finances had been a problem
during college is significant (p-.OO3) with the graduates reporting
more of a problem with finances than those who withdrew. These
comparisons are highly intriguing. There is no relationship between
adequacy of income and a woman's number of years of education prior to
the present schooling.

EDUCATIONAL VARIABLES

On the educational dimensions, the rate of recent college
attendance is significant (p-.OOOO) with a majority of the withdrawers
having attended part-time while graduates were much more likely to
have attended full-time. Withdrawers had more education prior to
this entry to school than did graduates (p-.O47). Withdrawers are

also more likely than graduates to have achieved a previous degree

140

(p-.OO97). Neither of these last two differences are significant
across all student categories.

Primary Hypothesis 2 of this study states that graduates would be
more likely to have planned on completing a degree without breaks than
would dropouts. This is supported very strongly at the p-.OOO level.

No differences between dropouts and graduates emerge by the level
of spouses' education, contrary to what the literature would suggest
and different from the earlier analysis on all categories of student
status. A comparison is made of those whose spouses have only high
school education with those whose spouses have doctorates, since this
variable might determine the amount and adequacy of income. The
students with husbands in these two extreme educational categories do
not differ on their perceptions of the adequacy of income or on the
actual income itself, nor do these wives differ in the hours per week
of employment, although those whose spouses have a high school
education did find finances to be more of a problem than those whose
spouses had a doctorate (p-.O38).

CAUSES FOR PREVIOUS NON-COMPLETION OF EDUCATION

Graduates and dropouts do not differ on the rate they indicate
any specific causes for earlier non-completion of college nor do they
differ on the total number of reasons claimed for earlier
non-completion.

CRISES/CHANGES

Graduates mark a higher total of crises or changes contributing

to their return (p-.O35) and more recent crises since returning to

school (p-.026) than do withdrawers.

141

These criterion groups are significantly related on the same two
changes or crises preceding the return to education as are found
across all categories of student status: Changes or crises in the
family (n-31, withdrawers (w)-3, p-.02) and those called "personal
problems" (n-26, w-3, p-.Ol9) are marked more often by graduates than
by withdrawers. Unlike findings on all students, no significant
relationships are found between the criterion groups on changes or
crises resulting from the return to school.

REASONS FOR REENTRY

The fact that graduates have a longer list of reasons for reentry
than do dropouts is very highly significant (p-.OOO). These graduates
particularly have a longer set of career related reasons for return
than do dropouts (p-.05).

As when making comparisons by all student status groups, only one
of the specific reasons for entering college now, to prepare for a
job, shows a significant relationship to the criterion groups (n-118,
wr9, p-.0005). But on this one variable, the significance level is
higher when the comparisons are made only between these two groups
than when the comparisons were made across all student status groups.
RECENT REASONS FOR WITHDRAWAL

Again, several significant relationships on reasons for recently
anticipated or completed withdrawal emerge, although not as many as
when compared across all levels of student status. Family
responsibilities, job responsibilities, and lack of time are all
significant at the p-.0000 level. Other significant relationships to
the criterion groups are found for finances (p-.OOO3), too many roles

(p-.OOO6), personal problems, (p-.OO66), personal health (p-.0124),

142

scheduling problems (p=.00170) and other unspecified problems
(p=.OOO7). Withdrawers indicate problems more frequently in all these
areas than do graduates.

The withdrawers cite more reasons per person than the graduates
do for having withdrawn (p-.OOO), although this may be because they
have given more thought to why they left as well as for other reasons
cited above.

ACADEMIC ACHIEVEMENT

Primary Hypotheses 3a and 3b predicted that graduates would have
higher cumulative and recent grade point averages than withdrawers.
Recent academic achievement proves to be a more important
discriminator than historic academic achievement. The GPA's achieved
prior to this last entry to college are not significantly different.
However, now that they have accumulated more recent grades (which are
combined with their former records if they previously attended
Michigan State University and are calculated independently of previous
GPA's if they went to college elsewhere previously), cumulative GPA's
differ. One-tailed tests show that graduates have higher cumulative
GPA's than do withdrawers (pa.010). Graduates also have higher recent
one—term GPA's than withdrawers have (p=.O30).

PSYCHOLOGICAL ADJUSTMENT

In general, primary Hypothesis 1 on psychological adjustment
cannot be supported by itself. No significant differences are found
on life satisfaction as a whole while they were in school nor on life
satisfaction as a whole after they left school, contrary to
expectation. Although happiness while in school shows a significant

difference (p=.05), the direction is other than what was predicted:

143

Withdrawers report having been happier while in school than do
graduates. As predicted, graduates feel they can better cope with
and/or handle crises (p-.05) than do dropouts. No significant
differences are found on getting what they want out of life nor on a
single item self-assessment of locus of control. Neither the Locus of
Control scale nor the Self Esteem scale differentiate between dropouts
and graduates. All of these tests are one-tailed.

After using factor analysis with varimax rotation to analyze all
items of psychological adjustment including the individual items
within the locus of control and self esteem batteries as well as the
single items of adjustment, four factors are found. Primarily but not
exclusively, these factors and their coefficient alpha estimates of
reliability are:

Factor 1: Twenty-one of the 28 items in the Duttweiler's Locus
of Control scale (.83). Weightings on Factor 1 range from 56 to 24.
All except two items, items #1 and #17 of question #24 on the
questionnaire, clearly weight much more heavily on Factor 1 than on
any other factor. Both Item #1 and Item #17 of question #24 items are
nearly as heavily weighted on the second Locus of Control factor,
Factor 2. Item #17 is not very intercorrelated with other items in
either Factor 1 or Factor 2 and weights lightly on the other factors.
This item is not included in the four factors. Item #1 is more
uniformly correlated with Factor 1 and is retained in it. Factor 1
could be considered a self-reliance or belief in self factor.

Factor 2: Five items in the Locus of Control scale (.82). These
items have a common thread of self—assessment of one's role in a

group. Three items weight very heavily on Factor 2 (72 to 51) and

144

very lightly elsewhere. Item #23 weights moderately heavily on Factor
2 (41) and almost as heavily on Factor 1 (38). Item #3 weights
proportionately the same as item #23 (29 on Factor 2 and 26 on Factor
1). Both items are retained in Factor 2.

These items in Factors 1 and 2 are more clearly interrelated than
related to other items in the other factors.

Factor 3: All ten items on Rosenberg's Self-Esteem scale plus
one item, item #12 of question #24, from the Locus of Control scale
(.69). Weightings range from 74 to 33 with the highest factor loading
on any other factor being 25. This can be considered a self-esteem
factor.

Factor 4: The six single items of adjustment (.73). These are
the items assessing life satisfaction during and since college,
happiness during college, how well persons are getting the things they
want in life, ability to cope with or handle crises, and a
self-assessment of locus of control (questions #17, 18, 20 to 23 on
the questionnaire) . Factor loadings range from 70 to 22. The
weakest loadings on this factor are for the question on coping with
crises. This item is, in fact, the only item which is not placed on
the factor where it loaded most heavily (Factor 3). In all other
varimax rotations it belonged with items from Factor 4, and an
examination of the total correlation matrix reveals its most
consistent correlations are with Factor 4 items.

Factor 1 accounts for 102 of the total variance, Factor 2 for 5%,
Factor 3 for 92 and Factor 4 for 52 of the total variance in the

population.

145

Of these, only the first factor on the locus of control scale,
which is here called a self-belief or self-reliance factor, shows a
more internal score for graduates (p=.OS). No other factors show

significant differences.

Table 37 Psychological Adjustment by Graduates and Withdrawers*

Adjustment Variable One-tailed t: p=

Life Satisfaction During School .438 (ns)
Life Satisfaction Since School .234 (ns)
Happiness During School -.O41
Getting What Is Wanted in Life .094 (ns)
Gaping With Crises .044
Control of Own Life .305 (ns)

Analysis of Variance: pa
Locus of Control .075 (ns)
Self-Esteem .057 (ns)
Factor 1 (Locus of Control 1, Self-Reliance) .034
Factor 2 (Locus of Control 2, Role in Group) .355 (ns)
Factor 3 (Self— Esteem) .486 (ns)
Factor 4 (Adjustment) .096 (ns)

* Positive probability means graduates score more positively

COMPARISONS OF FINDINGS ON ALL SUBJECTS TO THOSE ON CRITERION GROUPS
It is instructive to compare findings on all subjects with those
of only dropouts and persisters. Some findings are significant across
all groups of subjects by student status but are not significant when
comparing only graduates to dropouts and vice versa. When findings
are significant in both analyses, the strength of the significance
also often varies. Furthermore, among all subjects, dropouts appear
to stand out from the rest more than do graduates. These comparisons

are most easily seen in tabular form in Table 38.

146

Table 38 Findings on All Subjects and on Criterion Groups*

All Criterion
Subjects** Members***

Variable (n=361) (n=232)
Age ns ns
Marital Status ns ns
Having Children ns ns
Number of Children .036 (c) [d] ns
Number of School-age Children .02 (c,g)[d] ns
Level of Spouse's Education .015 ns
Years of Previous Education ns .047 (d)
Income ns .025 (d)
Adequacy of Income us .014 (d)
Is income a problem? us .003 (g)
Hours of Employment Weekly .029 (d,s) .000 (d)
Enrollment Rate .000 (g)[d,s,t] .000 (g)
Reason for Previous Dropout

Family us .020 (g)

Personal Problems ns .019 (g)

Money .017 (s)[d,t] ns
Total Number of Reasons for Previous Dropout .03 ns
Change/Crisis Contributing to Return

Family .003 .02 (g)

Personal Problems .026 .019 (g)
Total Number of Contributing Changes/Crises us .035 (g)
Change/Crisis Resulting from Return

Health .0002 ns

Family Member Status .019 ns
Total Number of Resulting Changes/Crises us .026 (g)
Return to Prepare for a Job .003 .0005 (g)
Career Reasons for Return ns .05 (g)
Total Number of Reasons for Return us .000 (g)
Expect Degree with No Further Breaks .0003 [d] .000 (g)
Happiness During School ns .05 (d)
Getting What Is Wanted in Life .006 (g) ns
Ability to Cope with/Handle Crises ns .05 (g)
Possess Previous Degree ns .0097 (d)
Cumulative GPA ns .010 (g)
Recent GPA ns .030 (g)
Difference of Recent and Cumulative GPA ns .0001 (g)

* g - Graduate, d - Dropout, t - Transfer, 3 - Stopout, c - Current
Student
** ( ) - Higher and [ ] - Lower than Group Average
*** ( ) - Higher than Comparison Group

147

COMPARISONS BY MARITAL STATUS

The previously described dummy variables on marital status are
each used to compare those who have graduated or withdrawn. Singles
include 30% of the graduates and 16% of the dropouts, a difference
which is significant (p=.044). Married women include 472 of the
graduates and 63% of the dropouts. Nineteen percent of the graduates
and 212 of the dropouts are divorced.

Three separate and parallel sets of analyses using the three
categories of marital status highlight a number of significant
differences. Singles are younger than non—singles (p=.OOOO) and
marrieds are older than non-marrieds (p=.0006). Household income
during school was higher for married students than for those
non-married while in school (p=.OOOO),lower for singles than for
non-singles (p=.OOO) and lower for divorcees than for non-divorcees
(p=.0028). No comparisons show differences on years of previous
education nor expectations of completing a degree without further
breaks. Married women worked less than non-married women (p=.0025)
and single women worked more than non-single women (p=.0024).
ANALYSES BY MARITAL STATUS AND PERSISTER/WITHDRAWER STATUS

A very useful examination of dropouts and persisters is to look
at single women who graduate versus single women who withdraw, married
graduates versus married withdrawers and divorced graduates versus
divorced dropouts. This analysis sharpens awareness of what has
already been found: the marital status of women in this study is an
important variable.

No significant differences are found on any of the income

variables when comparing married graduates with married dropouts and

148

single graduates with single dropouts. More single (p-.OO4) and
married (p-.OO7) but not divorced women graduates expected to get a
degree than comparable women among the withdrawers (f: p-.Ol). The
divorced withdrawers claim significantly more income (p=.037) and more
adequacy of income (p=.040) than do divorced graduates. Income was
less a problem during schooling for divorced withdrawers than for
divorced graduates (p=.020).

Married dropouts report more hours per week of employment during
school than do married dropouts (p=.028), but no patterns of
difference are found among singles nor among divorcees.

Across all categories of single, married, and divorced status,
graduates were much more likely to attend full—time than were those
who withdrew. (Single and divorced: p-.OO3, married: p=.039).

Single graduates report more life satisfaction during school than
the single withdrawers (p-.O36), but the categories are reversed when
the single withdrawers report more life satisfaction since leaving
school than do single graduates (p=.050). Married (p-.017) and
divorced (p=.029) graduates both are more able to get what they want
out of life than married and divorced women who dropped out, and
graduated divorcees report being more able to cope with crises than do
withdrawn divorcees (p=.030). Both married (p-.026) and divorced
(p=.024) graduates have higher self-esteem than their withdrawer
counterparts.

The married graduates score higher (or more internally) than do
married dropouts on Factor 1 of the Locus of Control Scale (p-.O46).
Divorced graduates rank higher on Factor 3 or the self-esteem factor

(p-.O34) than do their withdrawer counterparts. Analysis of variance

149

shows the married/nonmarried difference between criterion groups on
the six individual items of psychological adjustment taken together as

Factor 4 is highly significant (p-.OOl).

Table 39 Dropouts Compared to Persisters Within Marital Status
Variable Single Married Divorced
Graduates/ Graduates/ Graduates/
Dropouts Dropouts Dropouts
1:: p- p= p=
Number of Children .225 (ns) .355 (ns)
Income .551 (ns) .799 (ns) .037
Income Adequacy .549 (ns) .374 (ns) .040
Financial Problem? .409 (ns) .439 (ns) .020
Hours Employed .065 (ns) .028 .173 (ns)
Enrollment Status .003 .039 .003
Years of Pre-Entry Education .769 (ns) .684 (ns) .241 (ns)
Cumulative CPA .222 (ns) .286 (ns) .082 (ns)
Last One-Term GPA .144 (ns) .140 (ns) .385 (ns)
Expect Degree Without Breaks .004 .007 .057 (ns)
Life Satisfaction During School .036 .486 (ns) .250 (ns)
Life Satisfaction Since School .050 .036 .143 (ns)
Happiness During School .280 (us) .090 (ns) .142 (ns)
Getting What Is Wanted In Life .070 (ns) .017 .029
Cope with Crises .148 (ns) .367 (ns) .030
Self-Esteem .128 (ns) .026 .024
Locus of Control .317 (ns) .059 (ns) .159 (ns)

TWO—WAY ANALYSES OF VARIANCE BY MARITAL STATUS AND CRITERION GROUPS

Because of the previously noted differences by marital status, a

set of comparisons are made between the dropouts and the graduates

using marital status as a second independent variable.

differences not previously visible emerge.

At this point

Among sociodemographic

variables, age shows a main effect, effects by marital status and an

overall significance.

effect for marital status and an interaction effect.

differences are highly significant.

Annual income has a main effect as well as an

Overall income

The hours of employment per woman

150

have a main effect, effects for both criterion group membership and
marital status and an overall significance.

Of the educational variables, the expectation of a degree
evidences an overall effect, an effect by criterion groups and an
overall strong significance. Enrollment status shows a main effect,
an effect for criterion groups, and an overall effect.

Several results are of interest on the psychological adjustment
variables. Significant main effects and a significant effect for
marital status are found for life satisfaction during and since
school, happiness during school, and getting what is wanted out of
life. Two of these have a significant interaction: life satisfaction
since school and getting what is wanted out of life. Most of these
findings are significant at the p-.Ol level overall. Coping with
crises shows a significant effect for the dropout/graduate category.
The self esteem scale shows no main or individual effects but is
significant overall. The locus of control scale shows no significant
effects at all, but the adjustment scale (Factor 4) which is primarily
a combination of the individual psychological items outside the
batteries (satisfaction, happiness, coping with crises, getting what
is wanted out of life, and self-assessment of locus of control) shows
a main effect, significant effects for marital status and overall
significance.

Academically, the cumulative GPA is a significant variable
overall and has both a main effect and effects for both criterion
groups and for marital status. The most recent one-term GPA does not
have a main effect but does show a significant effect by criterion

groups.

151

Table 40 ANOVA Using Graduates/Withdrawers and Marital Status

Criterion Marital

Main

Variable Effects
Age .001
Income .001
Hours Employed .001
Enrollment Status .001
Expectation of Degree.001
Cumulative CPA .002
Recent One-Term GPA .137 (ns)
Life Satisfaction

During School .013
Life Satisfaction

Since School .016
Happiness During

School .002
Getting What is

Wanted Out of Life .003
Coping with Crises .244 (ns)
Control of Own Life .713 (ns)
Locus of Control

Scale .143 (ns)
Self-Esteem Scale .108 (ns)

Factor 4 (Adjustment).OO4

Groups

Status

Inter-
action
Effects

Significance of f a

.443 (ns)
.150 (ns)
.002
.001
.001
.002
.023

.679 (ns)
.318 (ns)
.187 (ns)
.063 (ns)
.045

.574 (ns)
.127 (ns)

.055 (ns)
.083 (ns)

.001
.001
.021
.198 (ns)
.549 (ns)
.029
.818 (ns)

.005 (ns)
.008
.003
.003
.977 (ns)
.586 (ns)
.180 (ns)

.227 (ns)
.004

.700 (ns)
.018

.916 (ns)
.246 (ns)
.246 (ns)
.854 (ns)
.836 (ns)

.163 (ns)
.040

.420 (ns)
.021

.512 (ns)
.444 (ns)
.495 (ns)

.070 (ns)
.769 (ns)

Explained

.001
.357 (ns)
.700 (ns)

.233 (ns)
. 6

.018

152

All of these effects underline the importance of having analyzed
several variables using both marital status and the criterion groups
as independent variables, since marital status seems to have as much

or more effect as any other one criterion on other variables.

CORRELATIONS OF PSYCHOLOGICAL ADJUSTMENT VARIABLES

The secondary Hypothesis 1 stated that the measures of
psychological adjustment would be highly correlated with each other.
Every correlation except one (self-esteem scale with happiness during
school) is significant. Of those which show a significant
correlation, the strength of that correlation is examined. Several of
the psychological variables are moderately correlated with each other,
and all have stronger and weaker correlations with various other
psychological variables. The strongest relationships are between
happiness during school and satisfaction during school (.66),
satisfaction since school and the ability to get what is wanted in
life (.60), locus of control and self-esteem (.51), ability to get
what is wanted in life and self-esteem (.48), and the ability to
handle crises and self-esteem (.47). Rather surprisingly, although
the single-item locus of control question and the locus of control
scale are significantly related, the correlation is only mild (.21).

Those items which moderately correlate with several other items
(such as self-esteem) are found to be less useful in the better
discriminant functions because of that degree of overlap with other
useful variables. Some of those variables which are not significant

alone but which correlate less strongly with other variables (such as

153

locus of control) are more useful in the discriminant function because

they tap less overlapping areas for discrimination.

Table 41 Correlation Matrix of
Significantly Related Psychological Adjustment Variables

Psychological Variable 1 2 3 4 5 6 7 8
1. Satisfaction During

School .23 .66 .33 .14 .21 .17 .27
2. Satisfaction Since

School .23 .18 .60 .23 .38 .15 .43
3. Happiness .66 .18. .27 .12 .24 .18. ns.
4. Getting Things

in Life .33 .60 .27 .34 .42 .19 .48
5. Cope/Handle Crises .14 .23 .12 .34 .38 .32 .47
6. Self—Assessed Locus

of Control .21 .38. .24 .42 .38 .21 .39
7. Locus of Control .17 .15 .18 .19 .32 .21 .51

Scale
8. Self-Esteem Scale .27 .43 ns. .48 .47 .39 .51

The hypothesis that these adjustment items are correlated with
each other is supported by the near uniformity of significant
correlations, but they correlate mildly to moderately rather than as

highly as hypothesized.
DISCRIMINANT FUNCTION ANALYSIS

Several equations combining different sets of variables
discriminate quite well (near or above 80% accuracy) within the
randomly assigned groups (A and B) which each contain a split half of
the dropouts and of the persisters. Three final equations are formed
to test the predictive ability of the equation and to use for further

predictions.

154

The formula found to best discriminate the first group, group A,
can sort 83.622 of the 116 subjects in that group accurately into
categories of dropout or persister. When the same equation is used to
classify group B, it accurately identifies 84.482 of that group. It
also classifies accurately 81.472 of the total pool of dropouts and
persisters but is more accurate in identifying graduates in group B
and identifying dropouts in the total pool.

The formula found to best discriminate between dropouts and
persisters in group B can sort 87.072 of that group accurately. When
used to classify members of group A, this equation's classification
accuracy is 80.172. It places 80.172 of the total pool into the
proper groups.

The formulas found useful for discriminating the entire pool of
dropouts and persisters are generally somewhat longer and somewhat
less accurate than those found useful in the subgroups. However, one
ten-item formula correctly identifies 83.622 of the criterion group
membership of the total pool and then correctly classifies 86.212 of
group A and 83.622 of group B. Each of these items, when weighted by
its unstandardized canonical discriminant analysis coefficient
(generated by the computer program) is added to the other weighted
items to arrive at a total score. When these total scores are then
used for all members of the pool and the results are tested for
statistical significance by analysis of variance, this formula is
found to be significant at the p-.OOS level and the separate formulae
which best identify groups A and B are each found to be significant

when tested on the pooled criterion groups at p-.OOl.

155

This formula uses the following variables with the order in which
they are listed indicating the relative importance in the formula of
that item's predictive power:

Multiplier for weighting*
Expectation of completing the bachelor's

degree without further breaks .77
Total number of reasons given for last

entry into college -.35
Happiness during school -.70
Presence of a previous degree .99
Assessment of how well one is getting

the things wanted out of life .44
Hours per week of employment .17
Whether finances are a problem .23
Cumulative GPA -.39
Locus of control scale score .91
Single/nonsingle status -.43
(Constant) +2.02

*Positive weights mean graduates' scores are 1933; on the question as
worded on this questionnaire. They are also more often single than
non-single.

This equation predicts that of the 107 current students in the
study, 74 are likely to graduate and 33 are likely to withdraw. The
program has generated a list of case numbers showing exactly which
subjects are predicted to be in those two different groups.

This cross validation study thus shows it is possible to find a
combination of variables which, taken together, rather accurately
predict whether undergraduate women will persist to graduation or will
drop out without a degree. Each of the formulae which discriminate
one subgroup can be used on the other randomly selected subgroup with
a reasonably high degree of accuracy.

All of the better equations use a rather consistent minimum
pattern of types of information: 1) Some combination of psychological

variables, 2) Some form of assessment of academic achievement, 3) The

156

number of reasons cited for reentering college, 4) Whether or not a
previous degree is held, and 5) One or both of the items assessing use
of time by the rate of enrollment and/or hours of employment. Some of
these equations also use the perception of whether finances were a

problem and/or single or married status.

SUMMARY

This study has found a wide variety of variables which are useful
in understanding how to describe this population of undergraduate
women at Michigan State University. More importantly, a very useful
set of variables, used alone and in combinations, do significantly
identify differences and relationships between these women who
continue their education to a bachelor's degree and those who drop out

of the university without a degree.

CHAPTER 5
DISCUSSION AND RECOMMENDATIONS

Outcomes in this study weave fascinating patterns. First, the
response rate is rather high considering that the population sampled
was defined and addresses were frozen nearly two Years before contact
was initiated. For the approximately 165 responses received in the
first two weeks, addresses given now were compared to a those in a
student directory from September, 1984. Approximately one-fourth of
those giving addresses have moved, yet were reached and responded.
Some also indicate a name change during this period. Even if
undeliverable surveys are counted, a good-sized 572 of the entire
sample drawn replied. Although any dropoff from a 1002 return leaves
less explained in a study, it is useful to gain information on
five-eighths of the women who actually were reached.

The responses include a remarkably high number of graduates
considering the assumptions which were made by anyone consulted on
campus prior to the survey being conducted. In fact, this would have
been a relatively high number even if the entire sample had responded
to the survey. And it is exciting to have as large a group as more

than forty withdrawers respond.

LIMITATIONS OF THE STUDY
' RETROSPECTIVE DATA

It is important in reviewing these findings to recognize this
study's limitations. As cited in the opening chapter, the

retrospective nature of this survey may color the responses given. We

157

158

do not know whether the withdrawers choose to paint themselves in a
rosier light than is accurate. We do not know whether persisters,
simply because they know they have successfully completed their
degrees, may see things as being more positive as they look back than
they would have seen them while they were in the midst of the
educational process.

Even though it is impossible to know the effects of the
retrospective viewpoint from which these respondents are operating, it
is still significant that many differences exist between the two
groups, and that these differences are not all necessarily in the
direction predicted nor are they all in a similar direction. It is
particularly noteworthy that the dropouts expressed more happiness
during school while the graduates expressed more satisfaction. That
may be a rather accurate presentation of the difference between
persisters feeling satisfied with what they did and their
accomplishments while in school despite whatever else was going on in
their lives versus those who did not complete an education and
therefore had less with which to be satisfied still perhaps enjoying
life while they were in school.

Although retrospective studies have limitations, the limitations
will be the same in any autopsy (after-the-fact) study.

SETTING OF THE STUDY

The setting of this study in a large public university in a small
midwestern city limits generalizations. It may help us understand
more about older undergraduate women, but caution must be used in
extrapolating results from this study and applying them in another

setting, whether that setting differs geographically, in size, or in

159

type of educational institution (community or four-year college or
university, public or private).
UNKNOWN MAKEUP 0F POPULATION

A further limitation is the fact that we do not know who makes up
the entire pool from which this sample is drawn. Therefore we do not
know what portion of the dropouts or of the graduates or of the other
categories truly exist in the population. We can hypothesize that
there is a higher response rate from graduates and that withdrawers
are underrepresented, but we cannot test these ideas.

Nor are we able to test how much nonrespondents vary from
respondents because the university confidentiality procedures do not
allow us to know who were contacted but did not respond. If a random
sample of non-responders could have been contacted by telephone, this
variance could have been tested.

As it is, we must rely on generalizing from previous studies on

older women students in the literature where nonrespondents are
contacted. These studies commonly find nonrespondents to fit the same
model as respondents or to vary on only one or two demographic
dimensions such as age. However, because we do have a reasonably
large number of withdrawers and many graduates, we are able to assume
robustness of the sample relative to the assumption of normal
distributions.

A recent study addresses the argument in the literature about
whether nonresponse bias is a serious or a minor problem (Hogan,
1985). This study compares data gathered from graduate cohorts from
thirteen junior colleges during six consecutive years and draws

conclusions particularly about research in Student Affairs

160

Departments. The highest response rate, 672, is achieved from new
graduates in the year procedures were changed to conform with the
design and implementation procedures (including threedwave mailings)
used here. Analysis is done on the differences from the previous
years' data when the response rate was only 352 and on differences
from known information about the population. The lower response rate
results show few differences from what is expected. The higher
response rate results differ from known values only in that more women
replied than expected. And even with the lower response rate, no
substantive effects are found on correlational relationships. This
study on very recent graduates (satisfied consumers who were
locatable?) would seem to affirm the probability of having here
gathered rather accurate data.

One of the problems of attempting to study those who dropped out
of the university is that those who have left the university may have
less reason to want to remain in touch or to want to respond to
anything sent to them by the university. They may be very
disenchanted, they may be angry, they may just not care, or they may
be too caught up in whatever the reasons were which caused them to
leave the university to take the time and energy to respond to this
survey. Furthermore, according to some of the responses of the
sample, particularly of current students, some women who have not yet
graduated or who may yet have or plan on some connection with the
university have some anxiety about responding to a survey which comes
from the university because they fear there may be reprisals at some
point if they speak negatively, despite assurances of anonymity in the

cover letter. Therefore, a response rate of 41 women or 112 of the

161

sample who have withdrawn from the university is probably a rather
high response rate for dropouts.

It was assumed that many more would have dropped out but that it
would be difficult to get them to respond to the survey. As noted in
a previous chapter, many of these women who have left Michigan State
University responded only after having received all three waves of
mail. It thus seems most appropriate to have designed such a
comprehensive mailing series and/or to have colorfully personalized
envelopes and letters to reach this particular group, since they seem
to have responded best to the third contact.

USE OF UNDERGRADUATE WOMEN ONLY

The fact that only undergraduate women are studied makes this
population more homogeneous to study but limits our understanding of
the total older student population. Specifically, we learn nothing
about graduate students nor about undergraduate men at the university.
Therefore, generalizations here need to be limited specifically to
older undergraduate women. However, this limitation also allows us to
better and more accurately understand this one group of older students
than if we had sampled those other groups, and questions can be
specifically tailored to the older undergraduate female situation.
There is no intent to generalize beyond older undergraduate women. In
fact, a strength of the study is that it was intentionally limited to
this population (which should have less variability than studying
older students as a whole) specifically to allow us to be more

definitive in findings about one subgroup.

162

NON—OBJECTIVE NATURE OF SURVEYS

As noted in the opening chapter, surveys and self-report studies
are, by their nature, not objective studies. Therefore, the
subjective biases of the respondents must be taken into consideration.
Instruments particularly for psychological measures obviously may be
thought to suggest a more desirable and less desirable direction.
Although literature cited earlier suggests this influence is small in
most of the formats used here, and although this issue was carefully
considered in the construction of the survey, this possible bias
cannot be ignored. Probably the most significant response to this
limitation is that, of those questions which are the same or nearly
the same as those used in national large sample surveys, the
proportions of responses across the entire sample are not different
from those in the literature.

Bradburn (1969) declares the norms for happiness in several
national samples to have much stability. He finds they average about
one-third very happy and 5 to 152 not very happy. In the present
study almost exactly one-third are very happy and twelve percent are
not very happy.

Lowenthal, et. a1. (1975) say older people are less happy than
the above parameters but are rather stable in happiness unless they
are under particular stress. Campbell, et. a1. (1976) believe that
use of the very happy category is declining. They find happiness
correlates .50 with satisfaction and that happiness is generally most
strongly correlated with marital status.

Robinson and Shaver (1973) find one-fourth completely satisfied,

two-thirds satisfied and about ten percent negative. Campbell, et.

163

a1. (1976) suggest singles and divorced women are less satisfied.
They find seven percent dissatisfied and 22 percent completely
satisfied. These compare to findings in this study of one-fourth
exactly completely satisfied, sixty percent satisfied, and about
sixteen percent, heavily weighted by the graduates, to be negative.

Robinson and Shaver (1973) discover that satisfaction is stable
over time and that satisfaction and happiness correlate moderately
(.47). They also learn satisfaction correlates very little (.06) with
social desirability. The present study finds happiness during school
correlates at a .66 level with life satisfaction during school.

Comparing findings on major national surveys with findings in
this study suggests that the presently found satisfaction and
happiness ratios are fairly typical of the population in general
except for an increase in dissatisfaction during school for graduates.
Interestingly, graduates become somewhat more satisfied after leaving
school and dropouts become less so.
INSTRUMENTATION

Finally, the study is limited by the questionnaire itself. It is
defined by and limited by the questions which are asked, the format in
which they are asked, and the questions which are.ggt_asked. All
studies share this limitation. We do not know all we might have known
if we had asked different questions. For example, women are not asked
what their educational goal was when they enrolled and whether, at the
time of enrollment, they intended to get a degree. Some may have had
a non-degree goal or their aim may have changed in the course of their
education. An aim which changed may have appeared in one of the less

than sure responses to the question on degree expectation. And an

164

achieved non-degree goal should have appeared in the question on
achieving all that was wanted as a reason to withdraw. A question of
intent could have been more clear.

The structure of some questions could be improved, particularly
when questioning about changes or crises. Although the question noted
such changes or crises could have been positive or negative, the
questions would have been strengthened if they had asked for this
direction to be indicated. We now simply do not know how the crises
or changes were perceived nor do we know whether possibly negative
responses could have triggered positive growth, although some marginal
comments suggest such a possibility.

It is possible that, to some degree, the number of significant
relationships pertaining to withdrawal could be inflated because only
withdrawers and current students were specifically asked to respond in
this section and graduates may have therefore responded in quite
reduced numbers. Because of this possibility, findings about reasons
for withdrawal and frequencies or reasons given by withdrawers should
especially be noted. It is likely, though, that statistical
comparisons on these items may be less valid and should be used with
caution. Another time, all subjects could be asked to respond to such
a question. Even then, the difference between the possibility of
withdrawal (which may seem like a light-weight question to some) and
the fact of withdrawal could make comparisons on this section less
than fully accurate.

However, when only dropouts and persisters are compared on
reasons for withdrawal, fewer significant differences are found than

when comparisons are made across all categories of student status.

165

This dropoff is a bit of an enigma if the lack of response here from
graduates is truly problematic, since one might expect the lack of
graduate responses to have emphasized the differences when analyzing
only the two groups.

Nevertheless, one question remains unanswered: If everyone had
responded to this question, would these differences have remained
significant? Alternatively the question could be phrased, are the
relationships significant only because a higher percentage of
withdrawers answer this section than did graduates? Such questions
are unanswerable.

The format could be simplified although it would mean lengthening
the survey instrument itself. Although the pilot study did not
encounter this problem, some women did mark several answers to the
three multiple response questions but then did not go back (as per
instructions) to indicate a primary response. Therefore, a second set
of all possible responses perhaps should be offered to elicit the one
main response. Unfortunately the response rate on the "main reason"
is lower than the response rate to the request to check all that apply
on three questions. However, since this dropoff was not large and the
numbers of responses remain high, the data still is useful.

Despite these instrumentation limitations, we have learned much
that is of value. Hindsight could improve nearly every questionnaire.
This one has served rather well.

Given the limitations on generalizability from this study, it
remains important to not underplay what we hg!g_learned. Each piece
of knowledge which is contributed to the literature is useful now and

is one more springboard for learning more in the future. Therefore,

166

we must value that a new piece of significant information is added

here, despite its limitations.

DISCUSSION OF MAJOR HYPOTHESES
H1: PSYCHOLOGICAL ADJUSTMENT

This study has three primary hypotheses, one of which has two
parts. Three secondary hypotheses are also advanced. The first
hypothesis is that among women undergraduates recently in college,
those who persist to graduation are more likely to show better
psychological adjustments as measured by having higher self-esteem,
internal locus of control, reporting more happiness and life
satisfaction, coping better with crises, and feeling they are getting
more of what they want out of life than those who dropped out. Of
these psychological adjustment variables, graduates do see themselves
as being better able to cope with and handle crises than do dropouts.
Although locus of control as a complete measure does not differentiate
these groups, the first factor in the locus of control scale (which
includes the large majority of the items in the scale as found by
factor analysis) does differentiate the dropouts from the graduates,
with the graduates being in the direction of a higher internal locus
of control.

Happiness during school also is significant differentiater.
However, it is significant in the opposite direction to that
predicted. Withdrawers claim more happiness for their school period
than do graduates. The satisfaction measures show no significant
difference between dropouts and persisters. An interesting variation

noted while examining the frequencies is that graduates report

167

themselves as having been more happy than satisfied during school and
dropouts report themselves as having been more satisfied than they
were happy.

The other specified measures do not differentiate the groups,
although locus of control turns out to be a necessary part of a
discriminating equation.

Clearly, psychological adjustment, at least as measured by the
collection of six single items and two batteries used in this study,
is not a highly homogeneous differentiation variable. At least
partial explanations of this finding may be found elsewhere within the
survey. For example, neither those who drop out nor those who stay
nor current students list that self-confidence is much of a problem
for them. Only twenty out of the entire 361 women and only three of
the actual withdrawers list self-confidence as one possible reason for
withdrawal. And although personal problems are cited as reasons for
previous withdrawal and/or as causes or results of the return to
school, more personal problems are listed by graduates than by those
who left.

POSSIBLE EXPLANATIONS OF FINDINGS ON HYPOTHESIS 1

Four different explanations of this phenomenon are possible:

1. The first possibility is that the respondents simply either
do not wish to paint themselves in a poor light or that the
respondents who do answer the survey are a better adjusted portion of
those who have been at Michigan State University and are particularly
a better adjusted portion of those who withdrew. As noted in the
limitations section, if this were fully true, no significant

differences would emerge. Yet many did. And when marital status is

168

used as a second independent variable with criterion groups, the
pcyshological variables gg_become significant discriminators,
primarily in the directions hypothesized.

Although we do not know about nonresponders, previous studies
suggest that the need to present socially desirable responses on a
survey is relatively low as noted in the review of the literature.
And previously cited data on mailed surveys seems to establish their
relative freedom from response bias. Therefore, although this
possibility may have some validity, other explanations appear to be
more useful.

2. A second possibility is that psychological adjustment is not
adequately defined and/or measured in this study. Because of the
complexity of the construct of psychological adjustment and of the
subconstructs used here (such as life satisfaction or locus of
control), our ability to accurately measure these intrapersonal
variables is as yet evolving. The instruments may not be able yet to
pick up variance which may truly exist in the population. For
example, since the modal score on the self-esteem instrument is the
highest possible score, it appears the instrument is not sophisticated
enough to pick up variability in self-esteem. Furthermore, in several
discriminant function analysis equations and when comparing married
graduates to married dropouts or divorced graduates to divorced
dropouts, self—esteem does differentiate in the predicted direction.
Other constructs may similarly yet need improved methods of
measurement. Only by further studies which define psychological

adjustment in the same way as in this study, in studies using other

169

definitions and/or measures or in the future use of more sophisticated
instruments can we learn more.

It is useful to note that in the discriminant analysis, various
psychological adjustment factors are necessary variables in almost
every attempted equation which classifies accurately at least
three-fourths of the persisters and withdrawers. Several of these
equations include three to six of the variables which are considered
here to be part of psychological adjustment. The final equation used
here includes three such variables: locus of control, happiness during
school, and getting what is wanted out of life. Curiously, because of
intercorrelations of the variables, the adjustment item which is
significant alone in differentiating persisters and withdrawers is not
chosen in the discriminating equation: coping with crises.

It does seem that although this hypothesis does not hold well in
isolation, it is an important ingredient in a broader understanding of
dropouts and graduates, particularly because of its effects when using
marital status as a covariable and when finding a combination of
variables to best predict who will drop out and who will persist.

Psychological adjustment continues to be a difficult construct to
define and operationalize well, but we do better to use it as a
relatively imprecise tool now than to ignore it because of lack of
precise operationalizing.

3. The third possibility is that people who do come back to
school are quite well adjusted or they would not have had the courage
to undertake this venture. This does seem to be part of the
explanation. Further education for some purpose is clearly important

enough to older students for them to uproot or to change their

170

established patterns of life enough to attend college at a stage of
life when they are at least twenty-five years old, although we do not
know fully what meaning this change has for these students.

The literature is quite clear that many people consider returning
to college who do not actually do so. In the Adult Services office of
this university and the Department of Lifelong Education, many inquire
about a possible return to school who do not come. Issues such as not
believing they can do it, perceived concrete self—deficiencies in
academic skills, the inability to manage time, or the fear of culture
shock when being an older student in a youth environment, or the lack
of family support may well have kept out women with poorer
psychological adjustment.

It is remarkable that, even among those who withdrew, the
responses of lack of support, lack of interest or other psychological
issues of the twenty-one options offered for reasons for withdrawal
are each checked only by three or fewer dropouts. Can it be that
those who approach a major (massive?) university such as this one are
the more self-confident people and that a similar study done in a less
threatening environment such as a community college or a commuter
college may find people who are less sure of themselves?

Can it be that, as the literature would suggest, women may be
rather anxious about their ability to cope and lacking in
self-confidence during the first several weeks or months that they are
students but that they do gain self-confidence as they begin to find
they actually can make it in the university? It is possible that some
of these people may have shown poorer adjustment only during their

initial period as students but made the necessary adaptations to cope

171

with their student roles. If they had been tested during the first
month of school, the adjustment variables could have been different.

If this idea is true, only those whom this study caught very
early in this beginning segment of their educational career would be
at the floundering, poorly adjusted stage. Others would more likely
show the self-assurance that often accompanies success. Under this
possibility, few of the current students are likely to report that
early stage of their education due to their reporting what their
experience is now, two years at least into school. The withdrawers
either remained longer than their first term or two or dropped out
early but are now looking back from a year and a half later and
attempting to recapitulate their experience.

Undergraduate returning women may simply be primarily
well-adjusted, more so than stereotypes would say. They may still
feel stresses. They may say or find they have difficulty coping. But
they are handling their roles relatively well, as evidenced by the
preponderance of graduates in the sample. They may have had a strong
sense of self in order to have returned at all. They may be a
self-selected, well-adjusted group of people. Good adjustment does
not mean things will not go wrong for them. Rather, it means they
handle things relatively well and do not let life get them down too
often. They tell of crises that have occurred. They may have had
situational times when their adjustment may have been poorer
temporarily. But they may simply be survivors, copers, and on the
whole, rather well-adjusted people as Wertheimer and Nelson (1977)

suggest.

172

4. A fourth possibility is that moderator variables or more
complex models need to be considered which incorporate rather than
ignore psychological adjustment variables. This seems very
reasonable, given the findings of this study. As noted in the
analyses of variance using marital status and criterion groups as
independent variables, satisfaction during school, satisfaction since
school, getting things that are wanted out of life and the self-esteem
battery all significantly differ, although they did not differ when
analyzed with the variable of marital status ng£_included. When using
bg£h_of these independent variables, only the self-assessment of life
control and the locus of control battery still are not significant.
These findings raise the important question of why marital status has
so much effect on these adjustment factors.

In the individual tests comparing married with not married,
single with not single, and divorced with not divorced women, the
divorced women in the study are clearly a highly heterogeneous (or a
highly unpredictable homogeneous) group: Only a few results are
significant for them. This fits with the flow of ideas which
originally triggered this study:

At first it was thought that those women who entered the
university under much stress, particularly in cases of divorce when
women were recovering from one chapter of life and trying to prepare
for a new chapter while perhaps also coping with the weight of
finances and the lack of child care and a move, may find coping with
student life quite difficult. After observing such women for a while,
it became clear that some women are overwhelmed by the crises or

changes of various types in their lives. But many more who are

173

struggling mightily nevertheless cope very well. It becomes clear
that many of these stressed women do get good grades and feel rather
proud of themselves. They manage to cope financially and usually
without a lot of debt. Many cut their standard of living severely and
often relocate into inexpensive housing. They see it as critically
important to get the degree in order to have a desirable future. Many
endure stresses which others may consider overwhelming.

Therefore, rather than saying, as originally intended, that women
under much stress would drop out, the question of how well they could
manage despite the stress and how well they were adjusted seem to be
more likely to determine whether they stayed. This premise still may
be true for some women. However, alternatively it may be true that
those who were not able to cope did not try to come here.

Furthermore, the primary reason for returning to school which
differentiates between graduates and dropouts is that graduates are
more likely to indicate at a highly significant level that they wanted
to prepare for a job. Perhaps the clarity of their concrete pragmatic
concern kept them in school no matter what else happened.

A reasonable explanation includes acknowledgement that the third
possibility is important: Those women who attempt an education at a
major university such as this one are rather well adjusted and,
although perhaps stressed, primarily do have the psychological health
to persist. This theory needs to be combined with the effect of
moderating variables such as suggested in the last possibility above.
The issue is apparently more complex than the original hypothesis
indicated, but the hypothesis is not necessarily therefore totally

wrong. Instead, we need to continue to look for the combinations

174

which best predict the educational future of these women, as was done
here by discriminant analysis, in order to understand better the
interactive nature of their behavior. We should apparently neither
abandon the search for nor the belief in the effect of psychological
variables nor can we confine our study to such variables to the
exclusion of other more circumstantial parts of the lives of older
undergraduate women.

Some evidence that there may be stress in lives but that the
students can cope with it comes from the fact that both the total
number of previous crises and crises resulting from the return to
college were cited by more who nevertheless graduated than withdrew,
although those graduates were less happy than withdrawers. Apparently
the graduates experience a fair amount of stress but manage to stay in
school anyway. A very interesting question would be to pursue exactly
what those crises or stressors were which propelled students into
school to see if they were more life-changing elements or more severe
stressors which simply had to be lived with or overcome for the
graduates than for the withdrawers.

It is also important that the graduates saw themselves as more
capable of dealing with crises. Perhaps that alone is a significant
and sufficient explanation of why they were able to stay when
withdrawers did not. Perhaps is less important to ask if they are
happy or satisfied and more important to ask if they have the goal of
graduating and whether or not they are able to deal with the crises
which come their way in styles which will keep them from dropping out.

A striking finding of the study is that, among married women, the

predictions on psychological adjustment almost entirely all hold and,

175

with the exception of the happiness question, are in the predicted
direction. For the single women, the results are somewhat less
clear-cut although not as diverse as with the divorced women. It is
the married women graduates who are primarily better adjusted and the

singles who are more clearly in jeopardy schologically.

H2: EXPECTATION OF A DEGREE

The second hypothesis is that women who persist to graduation
will be more likely to have believed that they would graduate than
those who dropped out. This hypothesis is strongly supported in this
study (p<.OOl). As was thought when this project was designed, it is
possible that the simple question of expectation of completing a
degree is the most critical single question that can be asked to
differentiate possible future dropouts from future persisters early in
their educational careers. This variable was usually first and often
otherwise second in strength of differentiating ability in nearly all
of the more useful discriminant analysis equations. If this item
continues to hold its power in other studies, it could be a most
useful question for use in a school because it may not be seen as
threatening to be asked by many different people in several settings
in a university. It could also be asked on an admissions
questionnaire, although at that time people could possibly assume an
honest negative response could prevent admission or cause future

discrimination by the institution.

176

H3: TOTAL GRADE POINT AVERAGE

The third hypothesis has two parts. The first part is that those
who graduate will have a higher grade point average than those who
drop out. This hypothesis is supported by the study.

The second part is that persisters will have a higher recent
(last one-term) grade point average than withdrawers. This, too, is
supported although less strongly than the first part of this
hypothesis.

Since the cumulative GPA for many women had to balance off the
total previously accumulated grades if they had an earlier record at
Michigan State University, the gain in cumulative GPA over the
previous GPA is very important. Given that such averages haze.
improved, one wonders why the last term's grades do not differentiate
these two groups as well as or better than cumulative GPA's. Can it
be that the focus moved away from grades to future plans and
expectations for those about to graduate?

It is noteworthy that although a cumulative grade point average
does distinguish students, the grade point average held before this
entrance to college does not significantly differentiate the criterion
groups. This means that, contrary to formulas found very useful on
younger undergraduates, previous educational achievement may have less
to do with persistence than the students' purpose or other variables
in their lives. It seems most important that these people did not
leave college previously because of academic difficulties with any
great frequency, nor are they expecting to leave nor have they left

because of academic difficulties now.

177

Whether they can handle their academic work does not seem to be a
critical problem for these students. Rather, other segments of their
lives seem to effect more what choices they make. For example,
although those who have not dropped out suggest that finances would be
an important reason for leaving now, for those who actually did leave,
other issues such as time pressures, role overload and family
responsibilities are more critical reasons for departure.

This does not say withdrawers are not able to cope academically.
It does say they see themselves as caught in the squeeze of trying to
balance several roles and decided to lessen that pressure at least for
a while. It is also important to notice that withdrawers rarely said
they were satisfied with their choice to leave. Instead they see this
as an interruption which will lead to future education at a later
time. This seems to confirm that although dropouts may get lower
grades, these grades are not so low that these women are in academic
jeopardy. The lower grades may reflect, therefore, an overload rather
than lack of ability and may also reflect a lack of belief in or
intent to compete a degree now.

One procedural question in the design of this study was whether
to ask subjects for their GPA's or whether to get their informed
consent to obtain grade records from the university registrar. At
least in this population, the decision to ask the subjects to supply
grades appears appropriate and perhaps even preferable. All except
two provide cumulative GPA's, which may well be more than the number
who would have signed a consent form. Nearly as many provide recent
one-term GPA's. Even when asked to provide GPA's from before this

last entrance to college, 902 or those with previous college credits

178

do so, and this information would not have been available from the
registrar. Furthermore, the most common response is to give a GPA to
one decimal point. Many give it to the second decimal point. These
women appear likely to be good sources of quite complete and rather

accurate information.

DISCUSSION OF SECONDARY HYPOTHESES

H1. CORRELATION OF ADJUSTMENT MEASURES

Three secondary hypotheses are tested. The first is that
measures of psychological adjustment will be significantly correlated
with each other. This is tested in two ways, first using correlation
coefficients which assesses both significance and strength of
association. Many of the items show a moderate correlation with each
other. Few are highly correlated, and some whose correlations are
significant are only weakly related. One surprisingly weak
significant correlation (.21) is between the single—item
self-assessment of locus of control and the battery assessment of
locus of control. Perhaps a single-item question needs to be more
carefully constructed, or they may correlate better in another
population, or the battery or the single question may have received a
more skewed response than expected.

Secondly, the factor analysis highlights three interesting
results:

1) As Duttweiler (1984) proposes, the locus of control
instrument has two primary factors. However, in the current
population, the two factors which are found primarily in her

instrument use somewhat different subsets of items than she used in

179

her factors and are therefore somewhat different factors than what she
found. This may be because her study uses general college populations
of both sexes and of all ages and this study focuses on a narrower
population. Nevertheless, the two factors are relatively cohesive and
internally relate much more strongly to each other than to most other
items.

2) The self-esteem measure is also quite cohesive and the items
are also better related to each other than to items outside the
battery.

3) The single-item measures of psychological adjustment
correlate as well with each other as do the items in the other intact
scales. Together the single items have a standard score alpha level
of .73, a relatively good reliability estimate.

Therefore, these scales seem to have both face validity and
statistical reliability.

Since the correlation coefficients are moderate at best, it
cannot be assumed that one of these measures or items can fully stand
for another, but instead each measures some overlapping space as well
as some independent variation.

H2: SOCIODEMOGRAPHIC VARIABLES

This secondary hypothesis is that women who persist to graduation
and those who withdraw do not differ on the sociodemographic variables
of age, marital or parental status, income or employment.

In the cases of age, marital status and parental status, the
hypothesis is upheld. The two groups do differ in income and hours
worked per week. Interestingly, the withdrawers have higher income, a

variable which is highly correlated with being married, but are also

180

more likely to be working themselves than are the graduates. If
differences had been predicted, one might expect withdrawers to be
more likely to be working. But they could also have been expected to
have more financial problems and less income than graduates.

From these two variables alone, it would appear as though
withdrawers have an adequate life elsewhere. They seem to already
have jobs and it is possible they may not need the degree as much as
do the graduates in order to create an economically acceptable life
for themselves. This would be explained in part by the adequacy of
income and the problem of income questions. Although income itself is
not useful in discriminant analysis equations, the perception of
whether or not finances were seen as a problem is a useful ingredient.
Perhaps the most usefulness of this variable results from the fact
that t-tests show the item of the problem of finances to be much more
highly significant than the also significant differences in income or
its perceived adequacy.

Contrary to the literature, spouses' level of education is not
correlated with income in the family nor with whether or not the woman
finishes a degree. Also contrary to suggestions in the literature,
whether or not one is a parent is not a critical variable. It appears
that those who graduate and those who withdraw are both equally likely
to be parents and that the degree of family demands may not have
critically affected whether or not they remained as students. In
fact, withdrawers who are parents have fewer children than persisters
have, although the dropout experience before age twenty-five took
place for family reasons at a significantly higher rate for graduates

than for the withdrawers.

 

181

As a whole, the women in this study are less likely to be married
or to be parents than nearly all those groups studied in the
literature with more than expected here in both the always single
category and the divorced category.

H3: DISCRIMINANT FUNCTION EQUATION

The final secondary hypothesis suggests that an equation can be
formed which would be quite helpful in identifying those who would
persist to graduation or drop out without graduating. Ultimately, the
fact that such an equation is formed here which is both highly
significant and is rather good at prediction may be the most important
finding of the whole study. Apparently right questions are asked and
useful information is learned here, since the purpose of the study was
to identify dropouts and persisters and that has been done above the
802 level. Further work is yet needed to refine the identifying
variables and to broaden the generalizability of such a set of
predictors.

Discriminant function analysis does not permit inclusion within
one equation of variables which are strongly correlated with each
other. Therefore, some variables which may help discriminate between
the criterion groups do not appear in this equation although they
frequently appear in other equations. For example, happiness and life
satisfactions during school each often would appear in an equation
when used in a set of variable without the presence of the other. But
both do not appear together because of their higher intercorrelation
(.66) since they then do not have enough residual independent
variance. However, the variation of life satisfaction which does not

overlap the variation of happiness cannot be included at all.

182

Another variable which, by itself, is a relatively strong
differentiator is the reason of returning to school to prepare for a
job, a response marked more often by graduates. However, this cannot
be used in the same equation as the total number of reasons for recent
reentry, since a variable which is part of a larger variable cannot be
used in the same equation with the broader variable.

These limitations of discriminant analysis can exclude useful
information.

Discriminant function analysis is a very useful technique,
despite these limitations, for combining a variety of variables in a
regression equation to identify which elements add useful information.
The discrimination probability of chance alone is that half of the
total criterion pool would be dropouts and half would be persisters.
In stepwise analysis, at each step a new variable is added to improve
the discriminant probability and the previous collection of variables
is examined to discover whether any provide information which is no
longer effective at discriminating between criterion groups and should
be removed. When no variables which still can assist in
discriminating at a significant level remain to be used in an
equation, analysis ceases. At this point, the relative strength of
the discriminators is given and the subject pool is classified into
groups.

Some equations make more errors of predicting women to be
graduates when they actually are withdrawers. Others make the
opposite error. Since it seems most critical to identify potential

dropouts, equations correctly identifying the largest proportion of

183

dropouts are chosen over nearly equally good equations making the
opposite error.

The discriminant function equation does correctly predict group
membership for 84.02 of the graduates and for 81.62 of the
withdrawers. However, the base rate of graduates and dropouts in the
population appears to be quite skewed toward graduates. The uneven
sample sizes of graduates and dropouts increases the prior probability
of getting a larger'ggmbg£_of errors in miscategorizing actual
graduates (Meehl and Rosen, 1955). Therefore, of the 170 women who
are predicted to graduate, 962 are actually graduates and 42 are
actually dropouts. Of the 62 women who are predicted to be dropouts,
502 actually are dropouts, since a miss rate of only 162 of the
dropouts is 31 women and a hit rate of 81.62 of the dropouts is also

31 women.

Table 42 Classification Results of Discriminant Function Analysis

Number of Predicted Group Membership
Actual Group Cases Graduates Dropouts
Graduates 194 123 84¥O 31 16.0
Dropouts 38 7 18.4 31 81.6
Current Students 107 74 69.2 33 30.8

However, within the 62 women who are predicted to be dropouts are
81.62 of all the dropouts. The 62 women is a much smaller group of
women to target for possible program planning than to have targeted
all 232 women in the criterion groups, and this group of 62 women

misses only seven of the actual dropouts. Therefore, the errors are

184

in the preferred direction. In the future if the ratio of dropouts to
graduates continues to be in the range found in this study, the
assumption of unequal sample size could be factored into the
discriminate analysis a priori and could perhaps produce greater
accuracy.

The fact that one equation can correctly identify such a high
rate of two randomly constructed subgroups of the dropout/persister
population is exciting. This means we can break down a very complex
issue into manageable dimensions and can predict most of those who
will persist to graduation and can target reasonable pools of those
who may drop out if we are willing to look at these older women
undergraduates as complex wholes whose actions are predicted by a
collection of sociodemographic, academic and psychological equations
using varying information from all of these domains.

Rather surprisingly, the final equation does not include
part-time versus full—time enrollment because of the high correlation
of that item with the hours of employment weekly. Nor does life
satisfaction make a difference because of its correlation with
happiness. Self-esteem is a factor which seems important until the
less likely (farther from significant) locus of control scale is added
and surprisingly is a significant factor in the analysis. But this
selection causes self-esteem to become no longer useful.

Interestingly, the perception of whether income is a problem is a
better discriminator than adequacy of income, although adequacy comes
somewhat closer to paralleling actual income. Additionally, both
adequacy and the problem status of income are better differentiators

than income itself.

185

CONCLUSIONS

What have we learned?

We have found a population of fewer married women and fewer
mothers than expected who dropped out at a younger age more often than
expected for lack of interest or goals and less often for
family/marriage reasons than the literature suggests. A surprisingly

large number of changes or crises in their lives both contribute to

. -

the decision to reenter and result from the return to college. They :
return for reasons of personal satisfaction more often than the
literature leads us to expect.
We have learned that we can differentiate between dropouts and
graduates with a rather high degree of efficiency and that we can
predict with above eighty percent accuracy those who will persist to
graduation and those who will not.
Furthermore, we can predict such persistence with a rather brief

questionnaire. The entire questionnaire used in this survey takes

approximately ten to fifteen minutes to complete. A questionnaire
using primarily only the items identified by discriminate analysis to
classify women as persisters or dropouts would take no more than five
minutes at the most to complete.

It can be very useful to attempt to pinpoint who is more likely
to drop out and then to become more personal in trying to find out
what might help those potential withdrawers to stay and complete their
education if that is what they desire. It is possible that some
interventions, pragmatic or supportive, in which the university could

engage to could reduce the dropout rate of older women undergraduates

186

and perhaps others as well. On the other hand, we must recognize that
often situations external to the university (job or family, for
example) require time and energy now, and we may need simply to accept
that some of these women will come in and out of school several times.

We have learned that the issue of who graduates and who does not
is more complex than simple psychological adjustment, but that such
adjustment does appear to be an integral part of this issue when taken
in conjunction with other variables.
THEORY

These findings may fit well with a very recent theoretical model
by Bean and Metzner (1985) which attempts to define a model of
attition for older students. The premise is that the attrition
process differs for what they call the "nontraditional" students, in
that they are more effected by the environment external to school
while the social integration process or lack thereof for younger
students within the educational setting is critical. They do not
differentiate older students by sex.

In this model, attrition is affected by four general variables:
1. Grade point average, which they see as highly influenced by high
school grade point averages but which the current study's findings
lead one to suspect is not highly important.
2. Intent to leave, which they see is influenced by two factors. One
influencer is psychological outcomes which they see as the person's
conception of the utility of education, their satisfaction with it,
their goal commitment, and the stress they experience. The other

influence on the intent to leave consists of academic variables which

187

include study habits, academic advising, absenteeism, uncertainty
about majors, and course availability.

3. Background and defining variables, which they describe as
consisting primarily of high school performance and educational goals
but also including age, enrollment status, residence (on or off
campus), ethnicity, and gender.

4. Environmental variables, defined as including finances, hours of
employment, outside encouragement, family responsibilities, and the
opportunity to transfer.

In their conceptual path model, statistical effects, both
directly and indirectly, of a given variable on dropping out can be
calculated. Compensatory interactions are suggested. For example,
environmental variables are expected to be more important for
nontraditional students than academic variables. Therefore, when
environmental and academic variables are both favorable for
persistence, the student will remain. When both are poor, the student
will leave. When academic variables are good but environmental
variables are poor, the student will leave and the positive effects of
academic variables would not be seen. But if environmental support is
good and academic variables are poor, the student will remain since
the environmental support would be more important than the poor
academic environment. Bean and Metzner see a similar compensatory
effect between academic outcome (GPA) and psychological outcome with
the later being more important. Thus good students who are stressed
or getting little satisfaction, have low goal commitment, or are

seeing little practical value in their education may leave.

$1.1?

188

These authors suggest that age per se is irrelevant except
insofar as it indirectly suggests the stage of family responsibility
or level of employment being experienced. They see hours of
employment weekly effecting withdrawal and amount of classes taken.
Intent of achieving degrees or taking courses for pleasure should be
ascertained and degree oriented people should be separated from the
others in attrition models, they say. High school GPA's are seen
primarily as an indicator of ability rather than of predicted actual
achievement.

Environmental variables are seen as primarily outside the ability
of the college to control but may have a major effect on retention.
Interestingly, Benz and Metzner suggest what this current study
suggests elsewhere, that finances are primarily a socially acceptable
excuse which can be used to cover more personal reasons for withdrawal
such as lack of commitment or "internal personal liabilities". They
do find evidence in the literature that older students were more
likely to work than not but did not find evidence that working older
students were more likely to drop out than were non-working students.
They wisely say that perhaps college GPA is less predictive of
persistence for part-time older students than for more traditional
students. Educational aspirations, defined as the highest level of
college education desired, and goal commitment are assumed to be
related in this model.

It does seem important in this present study, particularly when
taken in light of the many, many comments written, that the student's
perception of how practical the education will be becomes more

significant when it becomes more difficult to attend classes. If

189

scheduling, parking and family or job responsibilities, finances,
etc., become problematic, it may be that the personal satisfaction
which many people cite as a reason for education, could decline. The
presence or lack of utility of a degree could then become more
important in keeping students rather than just in attracting them in
the first place. Benz and Metzner (1985) find few studies assessing
the relationship of satisfaction and attrition, but suspect this
variable could be important.

The review of the literature chapter in this study indicates
there has previously been no theoretical basis which is particularly
helpful in predicting attrition. The present study attempts to
broaden theory by suggesting that more emphasis needs to be placed on
the psychological variables and personal expectations. Perhaps less
emphasis should be placed on the concrete, external and situational
variables.

The model offered by Bean and Metzner appears to be a good
amalgamation of previous inadequate theories while answering many of
the previous theoretical weaknesses. At this point, their model seems
to accept many of the significant findings from this study. Their
model needs to incorporate more psychological variables in direct and
interactive ways, should attend yet to both specific reasons for
reentry as well as the sheer number of such reasons, and needs to
attend to marital status as an interactive variable at least for women
and to instrumental ability to gain what is wanted out of life and to
cope with stresses as well as assessing what stresses occur. It
further should look at crises and stressors in students' lives more

carefully.

190

IMPLICATIONS

This study has a variety of implications. Some are rather clear.
Others are questions which are raised by the study but may or may not
yet have answers.

Many variables, especially those of psychological adjustment,
stand out more strikingly when marital status is controlled. Many of
the stresses of multiple role balancing are family related. Yet it is
the very women who are combining marriage, often including children,
with other commitments and then graduate who are appearing better
adjusted psychologically than the married dropouts or single or
divorced graduates. Can it be that some married women have a stronger
support system at home which both emotionally encourages them and
provides for them economically and perhaps takes care of some home
chores also? Can it be that married women who Eggg_more income go
directly to work at whatever level they are already qualified, and
those married women who come to college are under less pressure of
necessity for completing a degree quickly and for paying their way in
the meantime which leaves them better able to cope with school at
whatever pace is necessary? And what does this say about the married
dropouts?

For married women at least, it does seem to be more accurate to
say, as does the study's first hypothesis, that either they are better
adjusted psychologically and graduate or are less well adjusted and
withdraw. A kernel of a larger idea appears to be embedded here which
we lack adequate information to assess. Are those married older women

undergraduates who drop out less well adjusted in general? Or would

191

they also say they are having difficulties at home which temporarily
cause problems?

Withdrawers in general indicate that time pressures and family
responsibilities as well as job responsibilities are causes of their
withdrawal. But we do not know whether married dropouts are less
competent and less well adjusted people or are just under more stress
than the married students who persist. It is worth noting here that
graduates report more changes or crises contributed to the decision to
return and also more changes or crises resulted from the return to
schooling. This may lend credence to the more general adjustment
explanation than to the current stress explanation.

While most of the concern for older students expressed in the
literature is for married women or for divorced mothers, both of whom
have important visible relational ties, perhaps always single women
who are in school at a later age have been neglected and need more
assistance. They may lack active support systems, have smaller (often
quite small) incomes than others, and may well be totally responsible
for their own regular financial support and college expenses. Some
write of the humiliation of having to move back home at this stage.
They seem to gggg_to get the degree for future economic support or for
doing work they really want and yet often need to work at the same
time to afford this. Two of the most frequent problems voiced by the
always single women are the difficulty of obtaining financial aid when
they have previously had adequate income and the problem of
maintaining a job for income while coping with inadequate or

inflexible scheduling policies at the university.

. -“fl'w-a
fa" n. S. '9'

HF

192

Older women students are an exciting population, a growing
population, and a committed population as evidenced by the number of
graduates in this two year period and as represented by the response
rate from admittedly busy women. Many of them indicated their
pleasure at being subjects of interest to their university, even while
many took the opportunity to direct criticism at their alma mater. If
we are to understand the comments made by these consumers as being the
signposts of where the university could better accomodate these
students, perhaps the attrition rate could be lessened. The
university needs to become more sensitive to the problems of the
growing population of older students within its gates which are within
the university's power to change.

The frustration many older students have with this university,
particularly including such issues as parking and scheduling, could be
ameliorated at least in part by deliberate intent by the school.

Then, perhaps the students' sense of satisfaction and utility would be
increased and the attrition rate could be decreased.

On the other hand, since most dropouts say they plan to continue
their education at a future time, we can ask whether this means the
attrition of older students should be of less concern to us than it
is. Alternatively, since the withdrawers primarily have‘ggg achieved
all they wanted in education, the university may need to become.gggg.
concerned in addressing their needs. Some situations cannot be
changed, such as those of the women who write that their children need
them at this stage but that they will return later. Other dropouts

and a majority of the transfers say that scheduling problems are

ass-rm“- 9]

E‘t. ..

193

important barriers for them. Perhaps the university truly gagDattempt
to be more flexible for them.

As indicated in the introduction, the older students are
projected to be the wave of the future in at least partially stemming
the expected decreasing enrollment in higher education. It is
precisely the type of information found in this study, both on the
questionnaire and in the open-ended responses, which is rarely
gathered and which can give a university good insights into how it is
being viewed by its customers. This university and other institutions
of higher education may need to undergo some serious self-examination
of their own attractiveness to older students and then design ways to
woo and keep such students while other more traditional enrollments
decline.

Although university controlled issues are not stated in the set
of predictive variables directly, many of the ten items of the
predictive equation can be affected by the university such as
arranging better financial aid possibilities so students can work
fewer hours, scheduling classes more creatively for those who do work
more hours, and assisting students in achieving their goals at the
university by finding out what changes within university control could
facilitate older students to better get the things they want here and
making those changes.

From another perspective, we must acknowledge that several items
in the predictive equation are intrapersonal. Withdrawers were enough
happier while in school to make that item of third importance in the
equation. What shall we make of this happiness among withdrawers?

Does unhappiness drive graduates to finish? Are graduates too driven

194

to enjoy the journey? Is this item a retrospective glow over a part
of their lives which withdrawers specifically say they wish could have
gone on? Perhaps older women students can get counseling and/or
advising which helps them more effectively get the things they want in
life ot that helps them move to a more internal locus of control. If
so, would this assistance increase their retention rate? We don't
know the answers to these questions yet.

We may legitimately conclude that if dropouts are more likely to
already possess a degree, to have higher incomes, to be married and to
be working, many of them could be considered to be doing relatively
well already. Perhaps more attention should be paid instead to the

concerns expressed by all students in the open-ended question.
RECOMMENDATIONS

Now that we have this information about older undergraduate women
in this university, what shall be done with it?

Several uses can be made of it. If a university department such
as the Division of Women's Programs, Student Affairs, Adult Services
or the cross-departmental Adult Student Network wishes to attempt to
identify potential dropouts, a brief questionnaire asking only the
questions used in the discriminant function equation or those plus
others on which dropouts and persisters differ significantly could be
sent to entering and/or current older students. Alternatively, a
question could be added to the admission application assessing the
expectation of completing a degree. However, such a question used in

such a place could raise fears of whether a negative answer would

195

result in denial of admission and could lead to falsification by
potential dropouts.

Once such identifying has been done, advising, counseling,
support groups or services, and revamping of university policies and
procedures (scheduling, parking, hours of office opening, etc.) could

be specifically offered to older students in general and/or

...A
-‘,‘ 9‘

specifically to those at more risk for dropping out. Additional

I
.

information could be provided to older students by mail before or
after arrival at the university which could let them know of common
pitfalls or experiences of older students and which suggests to them
where within or outside of the university they can get support or
pragmatic assistance.

The written responses from 245 women comprise information from
nearly 252 of all older undergraduate women from fall term, 1984.
Their views, both as categorized in this document and as originally
stated, can be disseminated in the university. Perhaps their
collective wisdom and experiences can bring about some changes in
scheduling, in attitudes toward them, in sources of support made
available, or in some other ways which could change the view shared by
many respondents that Michigan State University is not "user friendly"
to older students.

Recommendations for future study include that most of what is
done here be now tested in other settings to define how generalizable
the findings are. Particularly it would be important to ask the
questions which significantly or nearly significantly differentiate

dropouts and persisters in this study to see whether the same or a

196

similar discriminant function would differentiate dropouts from
graduates in other settings.

Secondly, because current students also responded, this study
designs two ways in which a followup can be carried out after some
future time lapse. As of now, we have a current address for a
majority of current student respondents and we know who they are.

They can be followed in one of two ways. Information can be gathered
from public university records for those who give their names to learn
whether those students graduate or drop out in the future. We can
then see whether we properly predicted their outcome. It would also
be useful to survey these people in two or three years to see in what
ways they differ in how they describe themselves then from how they
see themselves now. It would also be of interest to see how they then
retrospectively describe themselves and whether their retrospective
memories match current descriptions. The groundwork has been laid for
a good longitudinal study. The current analysis suggests that of the
current students assessed, 74 are likely to graduate and 33 are likely
to withdraw. Specific case numbers matching these predictions have
been generated. It would be important to see how well the present
discriminant function predicts their future student status and what
changes they might require in that formula.

As noted earlier, further studies need to incorporate more
clarity on the changes or crises and whether they are positive or
negative, and how much meaning they have. A question should be added

on what their specific educational goal was when they entered college.

197

In general, this study has remained exciting. Those people who
are its focus deserve much credit for the effort they have undertaken

and their persistence and success at achieving their goals.

 

’ILF.LI. .. 4 t .4. EFL

BIBLIOGRAPHY

BIBLIOGRAPHY

Aanstad, Judy. "A Study of Mature Women at Sante Fe Junior College:
A Topical Paper." Florida Community Junior College
Inter-institutional Research Council, 1972. ERIC Document
Reproduction Service: ED 057 797.

Adelstein, Diane, Sedlacek, William, and Martinez, Alyce. "Dimensions
Underlying the Characteristics and Needs of Returning WOmen
Students." College Park, MD, University of Maryland Counseling
Center, 1979. ERIC Document Reproduction Service: ED 216 597.

American Women: Report p£.the President's Commission on the Status of

Women. Washington, D. 0.: Government Printing Office, 1963.

Ancheta, Bernard. "Counseling Needs of Traditional and Nontraditional
Community College Students." Journal p§_College Student
Personnel. 21:6 (1980): 564-567.

Anderson, Richand and Darkenwald Gordon. Participation and
Persistence ip American Adult Education. New York: College
Entrance Examination Board, 1979.

Andrews, Frank M. "Construct Validity and Error Components of Survey
Measures: A Structural Modeling Approach." Public Opinion
Qparterly. 48:2 (1984): 409—442.

Andrews, Frank M. and Crandall, Rick. "The Validity of Measures of
Self-Reported Well-Being." Social Indicators Resegrch. 3:1
(1976): 1-19.

Apps, Jerold W. The Adult Learner pp.Campus: ‘A_Guide‘fpp.lnstructors
gpg.Administrators. Chicago: Follett Publishing Company, 1981.

Aslanian, Carol B. and Brickell, Henry M. Americans ip_Transition:
Life Changes pp Eggsons for Adult Learning. New York: College
Entrance Examination Board, 1980.

Astin, Alexander W. "College Dropouts: A National Profile." ACE
Research Reports. 1 (1972): 1-71.

Astin, Alexander W. "Personal and Environmental Factors Associated
with College Dropouts Among High Aptitude Students." Journal pf
Educational Psychology. 55:4 (1964): 219-227.

Astin, Alexander W. Preventing_Students From Dropping Opp, San
Francisco: Jossey-Bass, 1975.

198

199

Astin, Alexander W. In Foster, Donald E., Astin, Alexander W. and
Sherer, James. "Student Persistence: Some Stay, Some
Don't-Why?" College gpg University. 48:4 (1973): 298-306.

Astin, Helen S. "Adult Development and Education." In Astin, Helen
S. (Ed.). Some Action prHer Own: The Adult Woman and Higher
Education. Lexington, Mass.: Lexington Books, 1976a.

Astin, Helen S. "A Profile of the Women in Continuing Education." In
Astin, Helen S. (Ed.). Some Action pf_Her Own. Lexington,
Mass.: Lexington Books, 1976b.

Astin, Helen Stavridou. "Continuing Education and the Development of
Adult Women." The Counseling Psychologist. 6:1 (1976c): 55-60.

Badenhoop, M. Suzanne and Johanson, M. Kelly. "Do Re-entry WOmen Have

Special Needs?" Psychology pf Women Qparterly. 4:4 (1980):
591-595.

Baier, John Leonard. "An Analysis of Undergraduate Student Attrition
at Southern Illinois University at Carbondale, 1970-1973."
Dissertation Abstracts, 1974.

Bailey, Kenneth D. Methods pf Social Research. New York: The Free
Press, 1978.

Bandura, Albert, "Self-Efficacy: Toward a Unifying Theory of
Behavioral Change." Psychological Review. 84:2 (1977):

Bandura, Albert. "Self-Efficacy Mechanism in Human Agency." American
Psychologist. 37:2 (1982): 122-147.

Barnett, Rosalind C. and Baruch, Grace K. "Wamen in the Middle Years:
A Critique of Research and Theory." Psychology pf Women
Qparterly. 3:2 (1978): 187-197.

Batdorf, L. L. "Counseling Services for Adults." The Personnel and
Guidgnce Journgl, 55: (1976): 160—164.

Bean, John P. and Metzner, Barbara S. "A Conceptual Model of
Nontraditional Undergraduate Student Attrition." Review pf
Educational Research. 55:4 (1985): 485-540.

Bestor, Dorothy K. "Will a Different Degree Help?" Working Woman. 8
(March, 1983): 68, 70, 73.

Bishop, John and Van Dyk, Jane. "Can Adults Be Hooked on College?
Some Determinants of Adult College Attendance." The Journal‘pf
Higher Education. 48:1 (1977): 39-62.

Black, Sionag M. and Hill, Clara E. "The Psychological Well-Being of
Women in Their Middle Years." Psychology pf Women Qpprterly.
8:3 (1984): 282-292.

200

Blocker, D. H. and Rapoza R. S. "Professional and Vocational
Preparation." In Arthur W. Chickering (Ed.). The Modern
American Colleg_, San Francisco: Jossey-Bass, 1981.

Boshier, Roger. "Educational Participation and Dropout: A
Theoretical Model." Adult Education. 23:4 (1973): 255-282.

Bradburn, Norman M. The Structure p£_Psychological Well- i g.
Chicago: Aldine, 1969.

Brandenberg, Judith Berman. "The Needs of WOmen Returning to School."
Personnel and Guidance Journal. 53:1 (1974): 11-18.

-ver‘-v

Brewin, Chris R. and Shapiro, David A. "Beyond Locus of Control:
Attribution of Responsibility for Positive and Negative
Outcomes." British Journal p§_Psychology. 75 (1984): 43-49.

 

1'1““

Bross, Dorothy R. "Night College Courses for the Older Women." Adult
Leadership. 15:7 (1967): 233-234.

Brown, Ric. "Locus of Control and Sex Role Orientation of Women

Graduate Students." College Student Journal. 17:1 (1983):
10-12.

Brown, Steven D. and Heath, Linda. "Coping with Critical Life Events:
An Integrative Cognitive-Behavioral Model for Research and
Practice." In Lent, Robert (Ed.). Handbook.pf.Counseling
Psychology. New York: Wiley, 1984.

Bryne, Michael M. "Characteristics of Dropout and Dropin Liberal Arts
Students at Lansing Community College and Identificatiol:n of
Institiutionally Controllable Variables Affecting Student Holding
Power." Doctoral Dissertation, Michigan State University. 1960.

Bugaighis, Margaret A., Schumm, Walter R., Bollman, Stephen R., and
Jurich, Anthony P. "Locus of Control and Marital Satisfaction."
The Journal pf_Psychology. 14 (1983): 275-279.

Burgess, Paul. "Reasons for Adult Participation in Group Educational
Activities." Adult Education. 22:1 (1971): 3-29.

Cagiano, Annette, Geisler, Margaret, and Wilcox, Lee. "The Academic
Performance of Returning Adult Students." The College Board
Review. No. 106 (Winter, 1977-1978): 13-16.

Campbell, Angus, Converse, P. E. and Rodgers, W. L. The Qu_ality g
American Life: Perceptions, Evaluations, and Satisfactions. New
York: Russell Sage Foundation, 1976.

Carlsen, Mary Louisa Beird. "A Four-Year Retrospective view of the
Educational Experience of a Group of Mature Women Undergraduates
Students." Doctoral Dissertation, University of Washington,
1973. Dissertation Abstracts International. 35:1 (1974): 153A.

201

Carlson, Elliot. "Adults Are Rediscovering the Campus." Dynamic
Maturity. 12 (1977): 4-7.

Carnegie Commission on Higher Education. Opportunities for Women ip
Higher Education. New York: McGraw-Hill, 1973.

Carnegie Commission. Priorities for Action. New York: McGraw-Hill,
1973.

Carnegie Council on Policy Studies in Higher Education. Three
Thousand Futures: The Next Twenterears for Higher Educgtion.
San Francisco: Jossey-Bass, 1980.

Chandler, Joan M. and Gaelerstein, Carolyn. "Students Over Thirty:

Need Their Classes Be Different?" Alternative Higher Education.
7:1 (1982): 44-51.

Clarke, Marnie A. "Transitional Women: Implications for Adult
Educators." Adult Leadership, 24:4 (1975): 123-127.

Clements, Kathleen. "Emotional Characteristics of Mature Women
Students in Education." Eastfield College, Dallas County
Community College System, 1974. ERIC Document Reproduction
Service: ED 087 980.

Coopersmith, Stanley. The Antecedents prSelf-Esteem. San Francisco:
W. H. Freeman and Co., 1967.

Cope, Robert G. "Are Students More Likely to Dropout of Large
Colleges?"
College Student Journal. 6:2 (1972): 92—97.

Cottle, Thomas J. "Overcoming an Invisible Handicap." Psychology
Today. 13:8 (1980): 89.

Cross, K. Patricia. Adults p§_Learners: Increasing Pgrticipation and
Facilitgtipg_Learning. San Francisco: Jossey-Bass, 1981.

Cross, K. Patricia. "Adult Learners: Characteristics, Needs, and
Interests." In Peterson, Richard E. and Associates. Lifelong
Learning 3p.America. San Francisco: Jossey-Bass, 1982.

Daniels, Arlene Kaplan. Welcome and Neglect: The Ambiguous Reception
pf_Re-entry Womenp_t_:__a__PrivateI Elite University. Northwestern
University: Program of Women, 1979.

Datan, N. and Hughes, F. "Burning Books and Briefcases: Agency,
Communion, and the Social Context of Learning in Adulthood."
Academic Psychology Bulletin. 7 (1985): 175-186.

DelDin, Betty Zelestine. "A Descriptive and Analytical Study on Adult
Students Who Voluntarily Withdrew from Michigan State University

202

During Fall Term, 1979." Doctoral Dissertation, Michigan State
University, 1980.

Demo, David H. "The Measurement of Self-Esteem: Refining Our
Methods." Journal p£_Personalityflppd.Socigl_Psychology. 48:6
(1985): 1490-1502.

DeWolf, Virginia and Lunneborg, Patricia W. Descriptive Information
pp Over-35 Undergraduate Students. Seattle: Washington
University Bureau of Testing, 1972.

Diagnostic and Statistical Manual pf_Mental Disordepgp 3rd Ed. (DSM
111). Washington, D.C.: American Psychiatric Association, 1980.

l
Dillman, Don A. Mail and Telephone Surveys: The Total Desigp_Method.
New York: John Wiley and Sons, 1978.

Dollar, Robert J. "Attrition of College Education Students: A
Discriminant Analysis." College Student Journal. 19:2 (1985):
152-156.

Doty, Barbara A. "Why Do mature Women Return to College?". Journal

of pp; National Association for Women Deans, Administrators and
Counselors. 29: 4 (1966): 171-174.

Douvan, E. "Capacity for Intimacy." in Chickering, Arthur W. The
Modern American Eollege. San Francisco: Jossey-Bass, 1981.

Durcholz, Pat and O'Connor, Janet. "Why Women Go Back to College".
Change Magazine. 5:8 (1973): 52-53.

Duttweiler, Patricia A. "The Internal Control Index: A Newly
Developed Measure of Locus of Control." Educational and
Psychological Measurement. 44 (1984): 209-221.

Duttweiler, Patricia A. Personal Communication. April, 1986.

Dwinell, Patricia. "A Needs Assessment of Nontraditional Women
Students at the University of Georgia." Athens, GA: Georgia
University College of Education, 1980. ERIC Document
Reproduction Service: ED 195 219.

Edwards, John N. and Klemmack, David L. "Correlates of Life

Satisfaction: A Re-Examination." Journal prGerontology. 28:4
(1973): 497-502.

Ekstrom, Ruth B. Barriers §p_Women's Participation ip_Post-Secondary
Education: A Review fo the Literature. Princeton: Educational

Testing Service, 1972.

 

203

Elsner, David M. "Mrs. Suzy Coed: More Older Women Return to
College; Most Do Very Well." The Wall Street Journal. 180:50
(September 12, 1972): 1.

Erdwins, Carol J., Tyer, Zita E., and Mellinger, Jeanne C.
"Personality Traits of Mature Women in Student Versus Homemaker
Roles." The Journal 2§_Psychology. 105 (1980): 189-195.

Erickson, Mildred Brinkmeier. "An Analysis of Selected
Characteristics and Needs of Adult Undergraduate Students
Attending Michigan State University Fall Term 1966." Michigan
State University. Doctoral Dissertation, 1968.

Erickson, Mildred B. "Counseling Needs of Adult Students." Michigan
State University Paper presented at American Personnel and
Guidance Association Convention, New Orleans, March 23-26, 1970.

Eriksen, Erik H. "Identity and Life-Cycle." Psychological Issues.
(1959, No. 1).

Espersson, Margaretta Anna. "The Mature Woman Student Returning to
Higher Education in a Liberal Arts College for Adults." Doctoral
Dissertation, Columbia University, 1975. Dissertation Abstracts
International. 36 (1975) 1242A.

Farmer, Helen S. and Fyans, Leslie J., Jr. "Married Women's
Achievement and Career Motivation: The Influence of Some
Environmental and Psychological Variables." Psychology g§_Women

anrterly. 7:4 (1983): 359-372.

Feldman, Sandra Marilyn. "The Competent Woman: A Multivariate
Investigation of Sex Roles, Achievement Motivation, Self-Esteem,
and Locus of Control." Doctoral Dissertation, California School
of Professional Psychology, 1980.

Fisher-Thompson, Jeanne. ‘Ag_0verview‘g§_Re-Entry Women: Meeting_the
Enrollment Challenge. .5 Field Evaluation Draft. Washington, D.
C.: Association of American Colleges, Project on the Status and
Education of Women, 1981.

Fisher-Thompson, Jeanne and Kuhn, Julie A. Re—entrz Women: Relevant
Statistics, §_Field Evaluation Draft. Washington, D. C.:
Association of American Colleges. Project on the Status on the
Status of Education of Women, 1981.

Fitts, William H. Tennessee Self-Concept Scale. Nashville, Counselor
Recordings and Tests, 1964.

Folland, Laura, Pickett, Eunice and Hoeflin, Ruth. "Adult Women in
College: How Do They Fare?" Journal g£_Home Economics. 65:5
(1977): 28-31.

204

Foster, James and Gade, Eldon. "Locus of Control, Consistancy of
Vocational Interest Patterns, and Academic Achievement." Journal
gf Counseling Psychology. 20:3 (1973): 290-292.

Geisler, Margaret P. and Thrush, Randolph 5. "Counseling Experiences
and Needs of Older Women Students." Journal gf_the National

Association for Women Deans, Administrators and Counselors. 39:1
(1975): 3-9.

Gillies, Jean. "Should Mom Go Back to School?" Farm Journal. 86:9
(1962): 59, 83.

Gitter, A. George and Mostofsky, David I., "The Social Indicator: An
Index of the Quality of Life." Social Biology. 20:3 (1973):
289-297.

Glass, J. Conrad, Jr., and Harshberger, Richard F. "The Full-Time,
Middle-Aged Adult Student in Higher Education." The Journal‘gf
Higher Education. 45:3 (1974): 211-218.

Goodsell, Willystine. The Education g§_Women: I£§_Socia1 Background
and Its Problems. New York: The Macmillan Company, 1924.

Graulich, David J. "Graying of Campus." The Wall Street Journal.
189:16 (January 24, 1977): 1

Gray, Eileen. Everyyoman's Guide tg_College. Millbrae, California:
Les Femmes, 1975.

Greenfeig, Beverly R. and Goldberg, Barbara J. "Orienting Returning
Adult Students." In M. L. Upcraft (Ed.). Orienth Students to
College. New Directions for Student Services, No. 25. San
Francisco: Jossey-Bass, March, 1984.

"Grownups on Campus." Newsweek. 98:3 (December 21, 1981): 72-74.

Halfter, Irma T. "The Comparative Academic Achievement of Women."
Adult Education. 12:2 (1962): 106-115.

Hansen, Gary R. and Lenning, Oscar T. "Differences in Characteristics
and Outlooks of Men and Women College Students at Various Age
Levels." Paper presented at the North Central Special Interest
Group for Community College Research, Ann Arbor, July, 1973.

ERIC Document Reproduction Service: ED 078 840.

Harrington, Fred Harvey. The Future 2;.Adult Education. San
Francisco: Jossey-Bass, 1977.

Hechinger, Fred M. "Education's New Majority." Saturday Review. 2
(September 20, 1975): 14-16, 18.

Hildebrand, Verna. "Academic Orientation of College Students' Wives."
The Personnel and Guidance Journal. 45:6 (1967): 596-599.

205

Hildreth, Gladys J., Dilworth-Anderson, Peggye and Rabe, Sabrina M.
"Family and School Life of Women over Age Fifty Who are in
College." Educational Gerontology. 9 (1983): 339-350.

Hiltunen, Wandalyn Axthelm. "Adults as College Freshmen." .Ihg_
Journal g§_College Student Personnel.‘ 6:4 (1965): 208-211.

Hogan, Roseann R. "Response Bias in Student Follow-Up: A Comparison
of Low and High Return Surveys." College ggg_University. 61:1
(1985): 17-25.

Honzik, M. P. and MacFarlane, J. W. "Personality Development
Intellectual Functioning from 21 months to 40 Years." In L. F.
Jarvik, C. Eisdorfer, and J. E. Blum (Eds.). Intellectual
Functioning in Adults. New York: Springer, 1973.

 

Hooper, Judith Oakey. "Returning Women Students and Their Families' 1
Support and Conflict." Journal of College Student Personnel. 20
(1979): 145-152.

Hooper, Judith O. and March, G. B. "A Study of Older Students
Attending University Classes." Educational Gerontology. 3
(1978): 321-330.

Hooper, Judith O. and Traupmann, Jane. "Women Students over 50: Why
Do They Do It?" Journal 2£_College Student Personnel. 25:2
(1984): 171-172.

Hopson, Barrie. "Response to the Papers by Scholssberg, Brammer and
Abrego." The Counseling Psychologist. 9:2 (1981): 36-39.

Houle, Cyril O. The Inguiring Mind. Madison: The University of
Wisconsin Press, 1961.

Hudesman, John; Avramides, Basil; Loveday, Charles; and Wendell,
Anne-Sojourner. "Locus of Control and Allocation Phase: Impact
on the Academic Achievement of Special Program Students Over
Three Semesters." College Student Journal. 19:1 (1985): 23-29.

Hull, Dale. Maturity 2§“§_VariableIiEDPredicting Academic Success.
Columbia, Missouri, Missouri University College of Education,
1970. ERIC Document Reproduction Service: In O45 039.

Ice, Orva Lee. "An Analysis of the Academic Achievement of Adult
Undergraduate Students in the Liberal Arts Program at Macomb
County Community College During the Academic Year, 1968-1969."
Doctoral Dissertation, Michigan State University, 1971.

Johnson, Deborah Hazel, Wallace, Karen Weiss, and Sedlacek, William E.
"A Comparison of the Needs of Returning and Traditional Students
by Sex." Journal of the National Association for Women Deans,

Administrators ggg_Counselors. 42:3 (1979): 14-18.

206

Kahnweiler, Jennifer Boretz and Johnson, Patricia L. "A Midlife
Developmental Profile of the Returning Women Student." Journal
‘g£_College Student Personnel. 21:5 (1980): 414-419.

Kaplan, Alexandra. Speech on "Self-in-Relation." Michigan State
University, East Lansing, April 26, 1986.

Kaplan, Susan Romer. "A Feminist Cinderella Tale: Women over Thirty
in Graduate and Professional School." Journal gf_the National

Association for Women Deans, Administrators 22g Counselors. 45:2
(1982): 9-15.

Kasworm, Carol. "Lifespan Differences Between Student Groupings."
Journal of College Student Personnel. 23: (1982): 425-428.

Kelman, Eugenia and Staley, Bonnie. "The Returning Woman Student:
Needs of an Important Minority Group on College Campuses.
Student Development Report, Volume 12, No. 2." Ft. Collins,
Colorado: Colorado State University Counseling Center, 1974.
ERIC Document Reproduction Service: ED 103 747.

Kernaleguen, Anne and Conrad, Sister G. "Analysis of Five Measures of
Self-Concept." Perceptual 29g Motor Skills. 51 (1980):
855-861.

Kimmel, Earnest W. In Erickson, Lance; Kimmel, Ernest W.; Murphy,
Michael T. and Newcomer, Wilma J. "Back to School: The

Older-Than-Average Student." College Egg_University. 51:4
(1976): 679-692.

Kirk, Cyhthia Formanek and Dorfman, Lorraine T. "Satisfaction and
Role Strain Among Middle-Age and Older Re-entry Women Students."
Educational Gerontology, 9 (1983): 15-29.

Kjerulff, Kristen and Wiggins, Nancy Hirschberg. "Graduate Student
Styles for Coping with Stressful Situations." Journal gf_
Educational Psychology. 68:3 (1976): 247-254.

Klecka, William R. Discriminant Analysis. Beverly Hills: Sage
Publications, 1984.

Knowles, Malcolm S. Higher Adult Education ig_the United States: The
Current Picture, Trends and Issues. Washington D.C. Committee on
Higher Adult Education, American Council on Education, 1969.

Knox, Alan B. and Sjogren, Douglas. "Research on Adult Learning."
Adult Education. 15:3 (1965): 133-137.

Kowalski, Cash. The Impact gg_Persisting,ggg_Nonpersisting Students.
New York: Philosophical Library, Inc., 1977.

Krings, Donna. "Meeting the Counseling Needs." Adult Leadership.
24:8 (1976): 311-313.

207

Lance, Larry, Lourie, Joan and Mayo, Cynthia. "Difficulties of
Re-Entry Students by Sex and Length of School Interruption."
Journal of the National Association of Women Deans,
Administrators and Counselors. 42: 3 (1979): 39-42.

Lavin, David E. The Prediction of Academic Performance. New York:
Russell Sage— Foundation, 1965.

Lefcourt, Herbert M. Locus gf_Control: Current Trends ig_Theory,ggg
Research, 2nd Ed. Hillsdale, N. J.: Lawrence Erlbaum
Associates, 1976.

Lenz, Elinor and Hanson, Marjorie. So You Want to Go Back to School:

[Facing the Realities g; Re-entry. New York: McGraw-Hill—1977.

. :‘I-ym‘JlfixL‘ an."

‘- ..L -m-

Letchworth, George E. "Women who Return to College: An

Identity-Integrity Approach." 'Ihg Journal gf_College Student
Personnel. 11:2 (1970): 103-106.

fil‘

Levenson, Hanna. "Activism and Powerful Others: Distinctions Within
the Concept of Internal-External Control." Journa1,g§
Personality Assessment. 38:4 (1974): 377-383.

Levinson, Daniel, Darrow, Charlotte M., Klein, Edward B, Levinson,
Maria H., and McKee, Braxton. The Seasons‘g§“§.Man' 3 Life. New
York: Alfred Knopf, 1978.

Lipman-Blumen, Jean. Editorial in The Personnel and Guidance Journal.
53:9 (1975): 680.

Lowenthal, Marjorie F., Thurnher, Majda and Chiriboga, David. Four
Stages g§_Life. San Francisco: Jossey-Bass, 1975.

Lunneborg, Patricia W., Olch, Doris R. and DeWolf, Virginia.
"Prediction of College Performance in Older Students." Journal
,2£_Counseling Psychology. 21:3 (1974): 215-221.

McCrea, Joan M. "The New Student Body: Women Returning to College."

Journal g£_the National Association for Women Deans,
Administrators and Counselors. 43:1 (1979): 13-19.

McDonald, Kathryn. "Women in Higher Education: A New Renaissance?"
College Board Review. #111 (1979): 10-11.

McGraw, Linda K. "A Selective Review of Programs and Counseling
Interventions for the Reentry Woman." The Personnel Guidance
Journal. 60:8 (1982): 469-472.

McGuire, Edward C. "Psycho-Social Characteristics of Students Who
Voluntarilyfl Withdrew from Rutgers, The State University, Newark,
New Jersey." Doctoral Dissertation: Wayne State University,
1969. Dissertation Abstracts International. 30:2 (1969): 552A.

208

McKennell, Aubrey C. and Andrews, Frank M. "Components of Perceived

Life Quality." Journal'g£_Community Psychology. 11:2 (1983):
98-110.

Maeroff, Gene. "Schools Feel Impact of 'Graying Population'". 391
York Times. (January 8, 1978) 13: 1,10.

Magarrell, Jack. "The Enrollment Boom Among Older Americans: One in
Three College Students Is Now Over 25 Years Old." The Chronicle
‘g£_Higher Education. (May 4, 1981): 3.

Magill, Kathleen and Cirksena, Kathryn. Problems and Information
Needs 2f.Women Re-entering Higher Education. Stanford: Stanford
University Center for Research on Women, 1978.

Mardoyan, Jennifer L., Alleman, Elizabeth and Cochran, John R.
"Adapting University Couseling Centers to Meet the Needs of an
Older Student Body." Journal g§.College Student Personnel. 24:2
(1983): 138-143.

Marks, E. "Student Perceptions of College Persistence, and Their
Intellective, Personality and Performance Correlates." Journal
‘g§_Educational Psychology. 58:4 (1967): 210—221.

Markus, Hazel. "Continuing Education for Women: Factors Influencing
a Return to School and the School EXperience." 1973. ERIC
Document Reproduction Service: ED 078 296.

Marple, Betty Lou N. "Adult Women Students Compared with Younger
Students on Selected Personality Variables." Journal'ggughg
National Association for Women Eggns, Administrators and
Counselors. 40:1 (1976): 11-15.

Martin, John D. and Coley, Leslie A. "Intercorrelations of Some
Measures of Self-Concept." Educational agg_Psychological
Measurement. 44:2 (1984): 517-521.

Maslow, Abraham H. Toward g_Psychology 9;.Bei g. New York: Harper
and Row, 1970.

Maudel, Gail R., Butcher, James N. and Mauger, Paul A. "A
Multivatiate Study of Personality and Academic Factors in College
Attrition." Journal g£.Counseling Psychology. 21:6 (1974):
560—567.

Meehl, P. E. and Rosen, A. "Antecedent Probability and the Efficiency
of Psychometric Signs, Patterns, or Cutting Scores."
Psychological Bulletin. 52 (1955)): 194-216.

Metz, Betty Ann. "Comparative Study of High School Records and
College Achievement of a Select Group of Adult Undergraduates."
Doctoral Dissertation, New York University, 1966. Dissertation
Abstracts International. 27:4 (1966): 925A.

209

Mirels, Herbert L. "Dimensions of Internal Versus External Control."

Journal g£_Consulting and Clinical Psychology. 34:2 (1970):
226-228.

Mishler, Carol, Frederick, Dennis, Hogan, Thomas P., and Woody,
Saundra. "Adult Students' Pace Toward Graduation." College and.
University.

58:1 (1982): 31-41.

Mitchell, DeWayne W. "The Over-35 Student." 'Improving College gnd_
University Teaching, 18 (1970): 21-22.

Monsour, Karem J. "Education and a Woman's Life." In Dennis,
Lawrence E. (Ed.). Education and g Woman's Life: Proceedinggngg
‘§h§_Itasca Conference gg_the Continuing_Edugation‘g£.Women,
Itasca State Park, Minnesota. Washington, D.C.: American Council
of Education, 1963. 9-21.

Morstain, Barry R. and Smart, John C. "A Motivational Typology of

Adult Learners." ‘Ihg.Journalig§_Higher Education. 48:6 (1977):
665—679.

Morstain, Barry R. and Smart, John C. "Reasons for Participation in
Adult Education Courses: A Multivariate Analysis of Group
Differences." Adult Education. 24:2 (1974): 83-98.

Muskat, Hermine Susan. "The Relationship Between Locus of Control,
Scholastic Aptitude, Academic Performance and Voluntary College
Interruption of Commuter College Freshmen." Doctoral
Dissertation, Boston University, 1978a.

Muskat, Hermine S. "Women Reentering College: Some Basic Ingredients

for Curriculum Development." The Personnel ang.Guidance Journal.
57:3 (1978b): 153-156.

Myers, Cora H. "Special Problems Encountered by Mature Women
Undergraduates." Journal ginghg_National Association £g£_Women
Deans, Administrators ggQ.Counselors. 27:3 (1964): 137-139.

Neugarten, Bernice L. "Adult Personality: Toward a Psychology of the
Life Cycle." In Neugarten, Bernice L. (Ed.). Middle Ag; and
Aging. Chicago: University of Chicago Press, 1968.

Newcomb, Theodore Meade. The Acguaintance Process. New York: Holt,
Rinehart and Winston, 1961.

Nowicki, Stephen. "Predicting Academic Achievement of Females from a
Locus of Control Orientation: Some Problems and Some Solutions."
Paper presented at the Annual Convention of the American
Psychological Association, Montreal, 1973. ERIC Document
Reproduction Service: ED 087 542.

Osborn, Ruth Helm. "Characteristics, Motivation, and Problems of
Mature Married Women College Students: A Status Study of
Selected

 

210

Students at the George Washington University." Doctoral
Dissertation, The George Washington University, 1963. American
Doctoral Dissertations (1963): 69.

Ostar, Allan W. "Part-Time Students: The New Majority for the
1980's." The Chronicle pf.Higher Education. 23:6 (October 7,
1981): 56.

Otten, Mark W. "Inventory and Expressive Measures of Locus Control
and Academic Performance: A 5-Year Outcome Study." Journal‘pg.
Personality Assessment. 41:6 (1977): 644-649.

Owens, W. A. "Age and Mental Ablilities: A Second Adult Follow-up."
Journal p£_Educational Psychology. 57:6 (1966): 311-325.

Pantages, Timothy J. and Creedon, Carol F. "Studies of College
Attrition: 1950-1975." Review.p§ Educational Research. 48:1
(1978): 49-101.

Perkins, Stanley A. "Success of Mature Age Students in University
Studies." The Alberta Journal p£_Educationa1 Research. 17:2
(1971): 105-112.

Pervin, Lawrence A. and Rubin, Donald B. "Student Dissatisfaction
with College and the College Dropout: A Transactional Approach."
The Journal prSocial Psychology. 72 (1967): 285-295.

Pirnot, Karen A. and Dunn, Wendy L. "Value Priorities of Adult
Students." Journal of the National Association for Women Deans,

Administrators and Counselors. 46:4 (1983): 22-42.

Porter, Lee. "Adults Have Special Counseling Needs." Adult
Leadership. 18:9 (1970): 275-276.

Powell, Arnold. "Alternative Measures of Locus of Control and
Prediction of Academic Performance." Psychological Reports. 29
(1971): 47-50.

Powell, Arnold and Vega, Manuel. "Correlates of Adult Locus of
Control." Psychological Reports. 30 (1972): 455-460.

Preston, William G. "Adults as Regular Community College Students: A
Comparative Analysis of Some of Their Characteristics and
Perceptions and Those of College-Age Students, Summary of
Findings and Conclusions." Unpublished paper from Diablo Valley
College, Pleasant Hill, California. n.d. ERIC Document
Reproduction Service: ED 121 368.

Prociuk, Terry J. and Breen, Lawrence J. "Internal-External Control,
Test Anxiety and Academic Achievement: Additional Data."
Journal‘QEDPersonality Assessment. 41:6 (1973): 644-649.

211

Prociuk, Terry J and Breen, Lawrence J. "Internal-External Locus of
Control and Information Seeking in a College Academic Situation."
Journal p£_Socia1 Psychology. 101 (1977): 309-310.

Rawlins, Melanie. "Life Made Easier for the Over-Thirty Undergrads."
The Personnel 92g Guidance Journal. 58:2 (1979): 139-143.

Rawlins, Melanie and Davies, Kathy. "Today's Challenge: Adults in
College." ‘Ldfelong_Learning;_ Th; Adult Years. 4 (May 1981):
12-13, 27.

Reed, Jeffrey G. and Murphy, Michael T. "Academic Performance of

Mature Adults and Veterans." College'gpd.Univer§ity. 50:2
(1975): 129-144.

Reehling, Jean E. "They Are Returning: But, Are They Staying?"
Journal p§_College Student Personnel. 21 (1980): 491-497.

Reimal, M. William. "A Study of Factors Affecting Attrition Among
Women Re-entering Formal Education." Unpublished Doctoral
Dissertation. University of Northern Colorado, 1976. ERIC
Document Reproduction Service: ED 160 165.

Richter, Donna L. and Witten, Charles H. "Barriers to Adult Learning:
Does Anticipation Match Reality?" Journalip§_College Student
Personnel. 25:4 (1984): 465-467.

Roach, Rose Marie. "Honey, Won't You Please Stay Home?" The
Personnel gpg_Guid§nce Journal. 55:1 (1976): 86-89.

Robinson, John P. and Shaver, Phillip R. "Measures of Social
Psychological Attitudes." Measures pf Social Psychological
Attitudes. Ann Arbor: Survey Research Center for Social
Research, 1973.

Roehl, Janet E. and Okun, Morris a. "Depression Symptoms Among Women
Re-entering College: The Role of Negative Life Exents and Family
Social Support." Journalip§,College Student Personnel. 25:3
(1984): 251-254.

Roelfs, Pamela J. "Teaching and Counseling Older Students."
Findings. 2:1 (1975): 5-8.

Rosenberg, Morris. Society and phg_Adolescent Self-Image. Princeton:
Princeton University Press, 1965.

Rossman, Jack E. and Kirk, Barbara A. "Factors Related to Persistence
and Withdrawal Among University Students." Journal‘pf_Counseling
Psychology. 17:1 (1970): 56-62.

Rotter, Julian D. "Generalized Expectancies for Internal Versus
External Control of Reinforcement." Psychological Monographs:
General gpg.Applied. 30:1, Whole No. 609, 1966.

212

Ryan, Melanie. "The Mature Woman in Higher Education: What's a Nice
Old Girl Like You Doing in a Place Like This?." 1979. ERIC
Document Reproduction Service: ED 174 295

Ryden, Muriel B. "An Adult Version of the Coopersmith Self-Esteem
Inventory: Test-Retest Reliability and Social Desirability."
Psychological Reports. 43 (1978): 1189-1190.

Sands, Roberta and Richardson, Virginia. "Educational and Mental
Health Factors Associated with the Return of Mid-Life Women to
School. "Educational Gerontology. 10 (1984): 155-170.

Sears, Pauline S. and Barbee, Ann H. "Career and Life Satisfactions
Among Terman's Gifted Women." In Stanley, Julian C., George,
William C., and Solano, Cecelia H. (Eds.). The Gifted and
Creative: lA_Fifty-Year Perspective. Baltimore: The Johns
Hopkins University Press, 1977, 28-65.

"Relationships Existing Among Race, Student Financial

Selby, James E.
Journal p£_College Student

Aid, and Persistence in College."

Personnel. 14:1 (1973): 38-40.
A Study gf_ the Macture women Students Attending Day

Sensor, Phyllis.
Classes §E_Riverside City College During the Spring Semester,
1964. Riverside City College, California, 1964.
"A Study of Adult Undergraduates: What Causes

Sewall, Timothy J.
Them to Seek a Degree?" Journal p§_College Student Personnel.

25:4 (1984): 309-314.
Sharp, Lawrence F. and Chason, L. Ralph.
Variables in Predicting College Student Attrition."
19:5 (1978): 388-393.

College Student Personnel.
"The ESSENCE Woman's Guide to Going Back to
13:7 (1982): 40-42, 45.

"The Use of Moderator
Me:

Smalley, Barbara.
School." Essence.

Smallwood, Kathie Beckman. "What Do Adult Women College Students
Really Need?" Journal prCollege Student Personnel. 21:1

(1980): 65-73.
"Differences in College

Starr, Ann, Betz, Ellen L. and Menne, John.
Academic Dropouts, Nonacademic Dropouts,

Student Satisfaction:
Journal p§_Counseling Psychology. 19:4

and Nondropouts."
(1972): 318-322.

Steele, Gertrude M. "The Mature Woman Student at Skyline College:
Research Report Number 1." San Bruno, California: Skyline
ERIC Document Reproduction Service: ED 129 367.

College, 1974.
"Financial Aid for Lifelong

Steiger, Joann M. and Kimball, Barbara.
Learning: The Special Case of Women." School Review. 86:3

(1978): 395-409.

213

Stephan, William F. and Wheeler, Joseph C. "Facts and Figures." Adult
Leadership. 18:6 (1969): 171-172, 197-200.

Suchinsky, Richard T. "The Older Female College Student: An
Educational Challenge." National Association p§_Student
Personnel Administrator's Journal. 9:4 (1981): 29-41.

Tinto, Vincent. "Dropout from Higher Education: A Theoretical
Synthesis of Recent Research." Review‘pf Educational Research.
45:1 (1975): 69-125.

Tittle, Carol Kehr and Denker, Elenor Rubin. "Re-entry Women: A
Selective Review of the Educational Process, Career Choice, and

Interest Measurement." Review pf_Educ§tional Research. 47:4
(1977): 531-584.

Tomlinson-Keasey, C. "Role Variables: Their Influence on Female

Motivational Constructs." Journal p£_Counse1ing Psychology.
21:3 (1974): 232-237.

Traub, Gary S. "Relationship Between Locus of Control and Grade Point

Average in Freshman College Students." Psychological Reports.
50 (1982): 1294.

"Trends in Higher Education." The Chronicle prHigher Education.
24:17 (June 23, 1982): 10.

Truax, Ann. "Special Admissions for Adult Women." College Egg.
University. 50:4 (1975): 579-580.

U.S. Bureau of the Census. Current Population Reports. Educational
Attainment ithhe United States: March 1981 and 1980. Series
P-20, No. 390, August, 1984.

U.S. Bureau of the Census. Current Population Reports. School
Enrollment: October 1981 (Advance Report}. Series P-20, No.
373, February, 19833.

U.S. Bureau of the Census. Statistical Abstract p£_the United States
1984. (104th edition). Washington, D.C. 1983b.

 

Verner, Coolie and Davis, George S. "Completions and Drop Outs: A
Review of Research." Adult Education. 14:3 (1964): 156-176.

Warchal, Peter and Southern, Stephen. "Perceived Importance of
Counseling Needs among Adult Students." Journal p§_College
Student Personnel. 27:1 (1986): 43-48.

Watkins, Beverly T. "More Women Coming Back to Campus." .'_I'_h_e_
Chronicle p§_Higher Education. 9:14 (December 23, 1974): 6.

Weinstock, Ruth. The Graying pfnppg_Campus. New York: Educational
Facilities Laboratories, 1978.

214

Wells, D. Collin. "Some Questions Concerning the Higher Education of
Females." American Journa1.p§_Sociology. 14 (1909): 731-739.

Wertheimer, Barbara and Nelson, Anne. "Into the Mainstream: Equal
Educational Opportunity for Working Women." Journal 2; Research
and Development ip Education. 10:4 (1977): 60-76.

Westin, Jeanne. "Back to School Again." Weight Watchers. (May
1975): 51-57.

White, Carla J. "Adult Student Support Groups on the Campus."
National Association pf Student Personnel Administrator's
Journal. 21:2 (1984): 55-58.

Wolfgang, Mary E. and Dowling, William D. "Differences in Motivation

of Adult and Younger Undergraduates." Journal‘p£_Higher
Education. 52:6 (1981): 640-648.

"Women: Second Wind." Time. 82:21 (Nov. 22, 1963): 56-57.

Women's Bureau Employment Standards Administration, U.S. Department of
Labor, Continuing Education Prggrams and Services §p£_Women.
Washington D.C. 1971.

Young, Barbara H. "Welfare Mothers in College." Journal prCollege
Student Personnel. 18:1 (1977): 38-44.

APPENDICES

APPENDIX A

COVER LETTER

215

MICHIGAN STATE UNIVERSITY

 

”WMWIANIEATIONS mmo-mm-mnu
W's IIOGIAIS

11 July 19%

Dear Older Present or Former Woman Student:

You are special! You are part of a growing group of women who attend college at a
'non-treditional age. You are receiving this letter as an older (25 and over) woman undergraduate who
was enrolled at lichigan State Ihiversity in Fall term of 1984. You've probably balanced several roles
and lived through a busy life within and outside the university. We would like to know more about your
experiences and about your life as a student. lie think it has been difficult and may even have not
worked out for you. You've pioneered as an older student, and your help in understanding that
experience now may better help others along that path.

lie are assuming that some of you have by now graduated, sue of you have left the university (with
or without an intent to return to college) and some of you are arrest students. 'l'he enclosed
questionnaire asks for information about who you are and how you've done academically. It also covers
a wide range of experiences, attitudes and feelings. Although these very widely in most groups, we
went to learn what these have been like for you. We are interested in whether those who stay in school
until graduation differ in any ways fra those who do not, whether those who are wrrently enrolled
show the same differences, and whether we might be able to predict those who want to min in school
but may leave in time to offer them sose assistance. Your participation is inortant to increase
understanding about possible differences among these groups. Please tell us how a life has worked,
including any positive or negative motions to your experience here. Our informal channels tell us
too little wont older students, less about you who graduate, and least of you who left without
graduating. We'd like to understand you all better. Yhis survey took only about fifteen minutes to
cowlete in our pretest.

The more of you who respond, the more sure we can be that our picture is complete. The registrar
has provided the mailing labels, so we do not know who is receiving this letter. We ask for your name,
address and m to know duo not to recontact now and also so it could be possible to recontact sue
of you in the future for a possible longitudinal follow-up. If you do not wish to provide your name,
address, etc., we would still like you to complete the survey anonymously. You are free to participate
at the level you desire. No one has to respond, and we will never know who did not.

Replies will be treated with strict confidence. lanes will be i-ediately separated from
responses and will be kept securely. The code number you create would allow us to correlate any future
information with your present responses while saintaining caplete anonymity. All results will be
reported anonymously. Within these limits, a sunny of results will be available to you upon your
request. Please contact Dorothy llercer at (517) 355-8270 with any questions.

We're looking forward to learning what you have to teach us: Please complete and return the
enclosed survey in the enclosed st-ped envelope as soon as possible. lhank you so mach for your time
and your assistance.

Sincerely, Sincerely,
A/[f'r 4&&*;/.'/1’/,.
‘ \n/
' Diana Algra, Di or Dorothy lie r
Division of Women's Programs Consultant

Depart-eat of Huan Relations

MSL' 's a Min-sane Anna/Eyed W3 Inuit-nos

APPENDIX B

SURVEY QUESTIONNAIRE

216

COFIDBITIK HESTIMIRE
roe PRESS" ~40 FMR W194 INIVERSITY SIMS

ELEASE CREATEACMEMER CMSISTIMSNYM DATE (I: GIRTHPLIBTHENMHNDDAYU:
YOUR NOTHER’S BIRTH. (Ex. Mine is I2 I2 42 7 ID)
Code nmber:

 

: PLEASE PLACE A MC! max IN TIE m BEFME EACH WIATE INNER
MESS A SPECIFIC MSTIOI GIVES OTIER INSTRUCTIMS.

I. “T IS voun CURRENT STIDENT STATUS?
I) Graduated (Bachelor’s level) Indicate nonth and rear:
(Include a June, "86, graduation here.)
_2) Currently enrolled at Michigan State University
_3) Transferred to another school in (month and rear):
_4) Have not been enrolled since (month and rear):

 

 

 

 

 

 

 

2. AGE: 3. WITH. STATUS: 4. CHILDREN:

I) 25-29 I) Never Narried (RITE TIE MR

_2) 30-34 2) Married IN EACH DUNK)
3) 35-39 3) Separated Ages:

____4) 40-44 ___4) Divorced I) None
5) 45-49 S) Hidoeed 2) 0-5

__6) 50-54 3) 6-IG
7) 55-59 4) 199
G) 60*

S. W, "MT IS Tl-E LEVEL or YOUR l-IJSGMD’S EDUCATICN?
_l) Sane or all high school __3) Bachelor's __5) Doctorate
2) Sue college 4) Master’s

6. “FIBER (F DIFFERENT TIMES YOU WE WERE!) COLLEGE:
__3 or core tines

7. YEARS OF EDUCATIIN CWLETED BEFORE THIS LAST TIIIE YOU ENTERED COLLEGE:
High school/GED
__College: I 2 3 4 or nore years

G. NILE Yw KRE IN SCIIIDL TliIS- LAST TIME, illl GOULD YOU DESCRIBE YM EMILY "CUE?
__I) Very adequate 3) Inadequate
__2) Adequate 4) Very Inadequate

9. TOTAL WIN. museum) INCOIE MING TIE LAST "REE TENS (OR LESS) TMT YOU WE
BEEN BROLLED:

 

__2> ss,ooo-9,m __7) s30,ooo-34,999
3) $10,000'I4,m 3) $35,000-39,999
___4> “5,000-19,999 9) «0,000 or nore

217

N

IO. aw m0 YW DESCRIBE FIMCES MILE IN COLLEGE?

I) A major problem 3) A minor problem
2) A moderate problee 4) Not a problem

II. DURING TIE LAST THREE TENS (M LESS) TMTYWIEREMLED, MTIENEMBEKRE

YOU EMPLOYED:
I) Not at all 4) 2I-3O hours per week
2) HO hours per week S) 3I-40 hours per week
3) II-20 hours per week 6) over 40 hours per week

12. UHAT HERE YOIR REASCNS FOR NOT COIPLETIW CilLEGE DEFINE AGE 25? (PLEASE CHECK ALL
THAT APPLY. TIEN NAKE A m CIECK GEFIRE TIE mm REASIN.)

I) Narriage 8) Academic difficulties

2) Naving children 9) Lack of interest

3) Personal problems IO) Lack of purpose or goal

4) Family responsibilities II) Lack of money

 

 

5) Lack of support fru family I2) wanted to work
6) Lack of support frm friends I3) Conflict of Job and studies
7) Lack of time I4) Other
(Specify)
[3 TE FQLLQING Tug sgmmg. PLEASE CIECK ALL TINT APPLY. TIEN IhDICATE IN THE GIMK

m YOUR CHOICES YOUR ASSESSIBIT N TIE SEVERITY N TIE NE on DRISIS. USE THE
FOLIIJING KEY:
KEY: 4-I'ligh Moderate Mild I‘Little

13. HEN YW C(NSIDER WT CAUSED YOU TO OUTER COLLEGE THIS LAST TIIE, DID MY PERSWL

 

 

 

CHANGE OR CRISIS. POSITIVE 0R NEGATIVE. W WEN
YOU DID? IF SO, IN UMT AREMS) MS TIE WE IR CRISIS MD NW SEVERE “AS IT?
I) None 6) Job
2) Narriage 7) Nove—
3) Family 3) Change of status of family
4) Nealth Dub"—
5) Personal problems 9) Other
(Specify)
I4. DID MY PERSINAL CWGE Ill CRISIS, POSITIVE on NEGATIVE, T Y

W
m mo causes? IF so, IN wiAT AREAIS) uAs TIE ewes on 031513 mo um
ssvsns was 11? (AREAS uses HAY nanny nor as “his sons ASAaovs.)

 

 

 

 

 

 

I) None 6) Job
2) Narriage 7) Have
3) Family G) Change of status of family
4) Health member
5) Personal problems 9) Other
(Specify)

 

I5. mm uses your REASINS FOR ansams CII.LEGE nus LAST me? (PLEASE CHECK ALL THAT
APPLY. new ms A m cascx serous TNE mm ensue.)

 

I) Prepare for a job 5) Personal satisfaction
__2) keep or upgrade present Job 6) Get a degree

3) Prepare for a career change 7) To be seen as a success

4) Pressure frm others 3) Other

(Specify)

 

I6.

I7.

I9.

21.

218

W. no you MTICIPATE C(I‘IPLETING Your moisture ossnss
amour amuse BREAKS? W» uses YOU suns was nosr or
ms TIME THAT You IERE IN snout m1 YOU mo STAY m sum no FINISH voun
scenes?

_I) Yes, quite sure 3) Unsure much of the time
2) Sure much of the time 4) No
Y , WOOYOUFEELAGIIITYOIRLIFEASAMLEIIIILEYWMVE

E84 IN SCHOOL? W, III-IDIDYIIIFEELAGOUI’YMLIFEASAIIIILE
HIILE YOU HERE IN SCHOOL?

_I) Very satisfied 3) Mostly dissatisfied

_2) Mostly satisfied __4) Very dissatisfied

W8 WWYWFEELWYMUFEOSGWLE
nun

_;_i> Very «(mm 3) Mostly dissatisfied
2) Mostly satisfied 4) Very dissatisfied

W Peon TIE uuvsnsmr, Poe vim REASONS) om you LEAVE
momma STATE mwsnsrrv? w, IF you room NOT muss
swoon. no). um um: as TIE REASONS)? cuscx As mm AS my APPLY. THEN so neon
”MAmmwmrmmnmm:

 

_I) Achieved all I wanted frtm school I3) Needed a break or change

2) Family responsibilities I4) Finances

3) Personal health IS) Academic difficulties
__4) Personal problems Id) Problems with scheduling
_S) Illness/death in the fuily I7) Problems with faculty
_6) Job responsibilities I8) Problems with university

7) Too many roles bureaucracy
_3) Not enough time :9) Lack of interest in my classes
_9) Lack of self-confidence 20) Moved out of the area

IO) Lack of family support 2)) Other (Specify

 

 

II) Lack of support from others (Specify:)
I2) MSU did not have the classes or the program I wanted

 

TAKING ALL THINGS TOGETHER, aw MPPY MOLLD YOU SAY TIAT YOU ME GEBf Olltlhh's THE
TIME YIIJ MVE SEEN IN SCIDOL?

_I) Very happy 3) Not very happy

_2) Pretty happy 4) Very unhappy

IN GETTIM TIE TMIMS you WT WT OF LIFE, NW WELL WULD YOU SAY YIIJ ME DOIlIi?
I) Very well 3) Poorly
2) Adequately 4) Very poorly

IN ASSESSING IIIJ YOU CWE NITN CRISES, III! (MILD YOU SAY you ME CRISES?
_I) Very well 3) Not very well
_2) Adequately 4) I am overdielmed

WWCMROLDOYOUIWEWERYMLIFE?
_I) Complete control 3) Little control
_2) Much control 4) No control

24.

219

PLEASE READ EACH S'I’ATB‘ENT. FOR EACH ns. oscnos am you: w, on
A‘m'ruos, scams or ssmvion wouw as.

USE TIE FELWIW KEY TO IMICATE YM

RESPINSE:

3

MRELY
(Less than ")7.
of the time)

I)
2)

 

 

3)
4)
5)
6)

 

 

 

 

7)
O)

 

 

9)
IO)
II)
I2)
I3)
I4)

 

IS)
I6)
I7)
IO)

I?)
20)
21)

23)
24)
25)
26)

 

27)

20)

25.

22)“

A G
OCCASI MLY
(About 307.

of the time)

C
SOETIIES
(About half
the time)

D
FREDIENTLY
‘ (About 707.
of the time)

E
USII'ILLY
(More than 90%
of the time)

then faced with a problem I try to forget it.

I need frequent encouragement fru others for me to keep working at a
difficult task.

I like Jobs where I can make decisions and be responsible for my own work.

I change my opinion when sueone I shire disagrees with me.

If I want something I work hard to get it.

I prefer to learn the facts about suething fro mane else rather than
have to dig them out for myself.

I will accept jobs that require me to supervise others. ,

I have a hard time saying 'No' when sueone tries to sell me something I
do not want.

I like to have a say in any decisions made by any group I am in.

I consider the different sides of an issue before making any decision.

that other people think has a great influence on my behavior.

Uhenever suething good happens to me I feel it is because I have earned it.

I enioy being in a position of leadership.

I need smeone else to praise my work before I am satisfied with what
I have done.

I am sure enough of my opinions to try to influence others.

when something is going to affect me I learn as much about it as I can.

I decide to do things on the spur of the moment.

For me, knowing I have done something well is more important than
being praised by saneone else. ‘

I let other peoples’ demands keep me frmn doing things I want to do.

I stick to my opinions when sueone disagrees with me.

I do what I feel like doing, not what other people think I ought to do.

I get discouraged when doing something that takes a long time to achieve
results.

when part of a group I prefer to let other people make all the decisions.

when I have a problem I follow the advice of friends and relatives.

I enjoy trying to do difficult tasks more than I enjoy trying to do easy tasks.

I prefer situations where I can depend on swune else's ability rather
than Just my own.

Having someone important tell me I did a good Job is more important to me
than feeling I have done a good .iob.

when I am involved in suething, I try to find out all I can about what
is going on even when saeone else Is in charge.

MING TIE LAST TREE TEMS (on LESS) TMT you IERE WED, WICM “RE Yul

PRIMRILY?

26.
27.

WT IS YIIIR TOTAL OMDE POINT “MOE? (AS CLOSE AS POSSIBLE)
WT “AS YOUR LAST DIE-TERI OMDE POINT “MOE? (AS CLOSE AS POSSIBLE)

I) A full-time student
2) A part-time student

_3) Goth airing the last
three terms

 

 

220

23. MAT (AS YOIR TOTAL GUIDE POINT m EFME you BfTEREO CilLEfi TMIS LAST TIrE?
(PLEASE GIVE TIE HIST ACQMTE MR POSSIBLE.)
I) (Number as close as possible) 3) Cannot remember
2) None (No previous college)

29. IF YOU MD COLLEGE CREDITS DEERE YOU ENTERED TIIIS LAST TIME, HETIER you CM OR
WOT MR YM GUIDE POINT AVEMGE, m YOU ESTIMATE YOIR MES SINCE
RETURNING ARE:

I) NI’IOP than before 3) Linear than before
2) Approximately the sue as now

30. PLEASE READ EACH STATBfENT. DECIDE “(TIER YOU AGREE OR DISAGREE UITII EACH
STATBIENT MD TO MT DEGREE. ItOICATE YM RESPINSE ACCIRDING TO TIE FOLLGIING
KEY:

REY: I 2 3 4
Strongly agree Agree Disagree Strongly disagree

I) I feel that I a a person of worth, at least on an equal basis
with others.
2) I feel that I have a nimber of good qualities.
3) All In all, I . inclined to feel that I am a failure.
4) I am able to do things as well as most other people.
5) I feel I do not have much to be proud of.
6) I take a positive attitude toward myself.
7) On the ediole, I - satisfied with myself.
8) I wish I could have more respect for myself.
9) I certainly feel useless at times
IO) At times I think I a no good at all.

 

 

 

 

 

 

 

 

 

TIE NDCT TIREE OLESTINS ARE MY FOR TMSE Hill) umnnaa ERIN MSU:

 

3i. DO YOU INTRO TO C(NTINUE Yillt EOUCATIIN?
_I) Yes, at MSU by next fall 4) Elsewhere by next fall
_2) Yes, at MSU suetime S) Uncertain
3) Yes, suewhere, suetime 6) No

 

W

2. DO YOU FEEL GOG) ABOUT voun OECISIIII TO LEINE?
I) Always
_2) Usually

Occasionally
Rarely or never

3)

_4)

33. DO YOU HISM YOU HAD STAYED IN CELEGE?
_I) Frequently 3) Occasionally
_2) Snetimes _4) Rarely or" never
34. W: HAT ELSE DO YOU TMINR IS IIOORTMT TO iom IN A STIDY OF

OLDERWHIODOMDOWTCIII’LETEMINDEWTEDEOREE?
(USEMCK IFIEEOED.)

 

 

 

 

 

 

 

221
Name . Student 0
Street and Number Phone I t )
City State Zip Code

 

 

Again, thank you so much for your help and cooperation. Nay you fulfill the
hopes and dreams that brought you to NSU as an older student!

APPENDIX C

POST CARD

222

July 17, I986

Dear Present or Former Uoman Student,

Thank you so much for completing and returning the
questionnaire we sent you last week. He are excited to be
learning more of the experience of older women who return
to school. It has taken a special kind of determination
for you to have come back to school, however long you have
been able to attend.

Lie know this may be a very crowded time for you, so we
are especially grateful that you have taken approximately
I5 minutes of your time to participate in this survey.

Diana Algra, Dir ctor Dorothy M rcer, Consultant
Division of meen’s Programs

APPENDIX D

FOLLOW-UP LETTER

223

 

MICHIGAN STATE UNIVERSITY alma/w ‘/

wmmoruuwmmmus mum- “0.2440“

wownrs recounts w! m I
W W W
46

24??”

Dear Older Present or Former Woman Student:

Two weeks ago we sent a questionnaire to all of you who were women
undergraduate students at least twenty-five years old at Michigan State
University during Fall term. 1984. So far we have not received a response
from you, although we realise that you may have returned the survey
anonymously or since this letter was written. This letter is being mailed
anonymously to all on the original mailing list except those who responded
giving their names.

We need your information if you're willing to participate to have as
complete a picture of older women undergraduates as is possible. Results
so far are beginning to draw an interesting picture. but it is possible that
our understanding may change as we hear from more of you. We realize
this may be a hectic time of year but hope you will find about fifteen
minutes to comgilete this questionnaire.

If you have already responded, we thank you. In case you misplaced
the survey. a second copy is enclosed. If you have questions or would
like to talk about your experience. please contact Dorothy Mercer at
355-8270 or 371-4312. Thank you for taking the time in your busy lives to
help us learn more about all of you.

Sincerely ,

gum/6%..

Diana v.3. Algra
Division of Women's Programs

magma

Dorothy Mercer
Consultant

"SL1: am Allie-nation: Aden- IEqul Opportunity [maul-mam

Wills!IIHIIIIHIIHWIIIIIIIIIIHIIHIIHIIH!NI