“‘4 u. ".1“ ..Aw. . '-

-..v n;-

WAI-

 

f._r‘- .

. ‘..
c.-\\;‘~,}.,
.‘~"-$‘3‘“.33.

4“

O"
“I'M?

. ivy" " fin
'11
. 9‘

u

‘iff'l'lf-

IA"

41' ‘I’ '
f" '35.); (9’)},

0.2-.

I-J'hk

o

w.“ ?’
:Q‘L'M fig?" .
#4“th 3' .. «4-,... Etc? ngé};

ky.

I
u

”UVF: 5

U.
0.",

-;
:l

‘ I
‘ '31.... ‘ 77‘ £1733

rm: ‘ h Aim??? 5%“ w ..

‘4‘. ‘1: .

it

*1. "F3. ”
haiyw‘k 'fi ' ‘1 ‘ “ff?“

1“
9

\‘;-

; 1M
:d‘t‘i

I

, J} {m3

‘4 P u 1

'21,.“ I": (41':
' nu”

IV.
l.a’f'- ,

. .r
' f
“tad/ALL

’24:

V'It'v.\

‘ I
».:',':

$2.

.‘I
a.
. a

p.
3* .

u
. hm
it” (”7'th ;' I

L‘-

 

w

- . , . ,r 3,

"Egg: ‘: I , J'Uwiv‘

”Ii, ‘ . 4.1._ .h _ ”mu-4:? Mfg”

:12“: ‘2‘; ,‘f t 7’1' I’d," 2
"J‘J*

"if“ m“. 1;:

.14}. 1’33““
’J-Jl-IJ‘. . piggy, ,7"..J?h ',.
'.r.' 6:2.ng ,.- 3
112745;!!!

'5..ng

 

 

 

 

37"?”«7 .Wj’fv ‘J/ I)»
4:! EM)?

I
2'

 

l

 

 

LIBRARY 1 i
Michigan State
University

 

 

This is to certify that the

dissertation entitled
MID-LIFE MARRIED WOMEN'S
PERCEPTIONS OF THE RETIREMENT LIFE
STAGE: AN EXPLORATORY STUDY WITH
IMPLICATIONS FOR PRE-RETIREMENT EDUCATION
AND HOME ECONOMICS/HUMAN ECOLOGY

presented by

Betty Johnson McMahon

has been accepted towards fulfillment
of the requirements for

Doctoral degree in mflLEsmlogy

 

Major professor

Date April 18, 1988

 

MS U is an Aflirmau've Action/Equal Opportunity Institution 0-12771

 

 

 

 

 

 

 

" MSU

LIBRARIES

m

 

RETURNING MATERIALS:
Place in book drop—to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

MID-LIFE MARRIED WOMEN'S
PERCEPTIONS OF THE RETIREMENT LIFE
STAGE: AN EXPLORATORY STUDY HITH

IMPLICATIONS FOR PRE-RETIREMENT EDUCATION
AND HOME ECONOMICS/HUMAN ECOLOGY

By
Betty Johnson McMahon

A DISSERTATION

Submitted to .
Michigan State Univer51ty .
in partial fulfillment of the requ1rements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Family and Child Ecology

1988

 

 

 

—_._..___
#q‘fi

« ~ 3// it"

ABSTRACT
MID-LIFE MARRIED WOMEN'S PERCEPTIONS OF THE RETIREMENT

LIFE STAGE: AN EXPLORATORY STUDY HITH IMPLICATIONS FOR
PRE-RETIREMENT EDUCATION AND HOME ECONOMICS/HUMAN ECOLOGY

By
Betty Johnson McMahon

This study explored perceptions of retirement among
married mid-life women. Women's economic status, family
and marriage satisfaction, work status, work type, health,
life events, community type. and age were the independent
variables. Changes in perceptions of retirement over a
two year period were also explored. The dimensions of
retirement perceptions. are: control/anxiety, attitudes
toward time use, residence selection. adequacy of wife's
retirement income, and life continuityu Relationships
between personal, environmental, and interaction
characteristics and perceptions of retirement were
studied.

This study examines a Michigan subsample of the NC-
164 project, "Stress, Caping, and Adaptation in the Middle
Years of the Family“. Data were collected in 1983 and

1985 by mailed self-administered questionnaires.

Betty Johnson McMahon
The study sample consisted of 154 women who responded in
both years. The sample was primarily Caucasian, middle-
income, and high school educated and was evenly divided
among full-time employed, part-time employed, and non-
employed. Half lived in the open country; the other half
was divided between small town and urban.

This study provided hypotheses for further
exploration of perceptions of retirement, suggested
hypotheses for exploration of differences between employed
and nonemployed women, supported the ecological framework
for studying retirement, provided a better understanding
of perceptions of retirement among mid-life married women,
and presents implications for Home Economics/Human
Ecology and pre-retirement educators.

Family satisfaction was the strongest predictor for a
sense of control. When employed women were tested
independently marriage happiness was the strongest
predictor of sense of control. Nomen did not perceive a
strong sense of control in retirement. Women had a
positive attitude toward time use in retirement, with
full-time employed women having the most positive
attitude. Full-time employed women perceived their
retirement income as more adequate to meet retirement
needs than part-time employed or nonemployed women.
Perceived income adequacy predicted a change in residence,

and the lower the perceived income adequacy the greater

Betty Johnson McMahon
the perception of residence change. The fewer the number
of major life events, the higher the sense of life
continuity. Perceptions of retirement became less
positive as age increased and became more negative over

the two-year period.

DEDICATION

To
my husband
for his love. support, and
faith in my ability.

ACKNOWLEDGEMENTS

Completion of this doctoral dissertation was made
possible through the cooperation and contributions of a
number of individuals to whom I am grateful. My
dissertation committee deserves special recognition:

Dr. Margaret J. Bubolz. Dr. Barbara Ames, Dr. Dennis
Keefe, and Dr. James Marshall.

It has been a special privilege to work with my
chairperson, Dr. Margaret J. Bubolz. She has given
expert guidance. freely of her time, and inspiration. I
am indebted to her. .

Dr. Barbara Ames has made a difference. Throughout
the several years of this academic journey she has been a
source of academic wisdom, support, and caring.

I appreciate the contribution of Dr. Dennis Keefe.
His advice at a critical point in the process helped me to
find the direction I needed.

I thank Dr. James Marshall for his enthusiastic
support and guidance in my cognate area. His contribution
is appreciated.

Bruce Haas, statistician, was an indispensable
contributor to this dissertation. His contribution of

numerous hours of consultation is greatly appreciated.

Special thanks is given to my family. They provided

understanding,

needed . My doctoral

without them.

encouragement,

and a sense of priority as

program could not have been done

TABLE OF CONTENTS

LIST OF TABLES ...................
LIST OF FIGURES
Chapter

I. INTRODUCTION ................
The Challenge ..............
Purpose of the Study ............
Relationship of This Study to NC-164. . .
Theoretical Approach ...........

Human Ecological Framework .......

Modified Person-Environmental

. Interaction Model ..........
Significance of the Research .......
Limitations and Assumptions .......

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

Women and Work
Married Women .............
Retired Employee/Homemaker .......
Employment Type ............
Women/Men ...............
Summary ............. . .

Retired omen and Economics .......
The Last Fifty Years in Brief .....
Comparisons of Segments of

t e Retired Population ........
Economic Resources ...........
Summary ................

Rural Women ...............
Rural Family Well-being ........
Stress - Rural Females ........
Work Patterns and Marriage

Satisfaction ..........

II.

Page
viii

X

Chapter

III.

IV.

Support Systems . . . .........
Retirement Satisfaction ........

Summary

Women and Stress .............
Interpersonal Networks .........
Family Related Stress .........
Stress Buffers .............
Stress and Age .............

Summary

Locus of Control .............
Summary ................

METHODOLOGY ...............

Research Questions ............
Research Design .............
Research Variables ............

Procedure Used

Selection of Sample ...........

Description of Core Sample

Description of Study Sample .......

Research Instrumentation

Retirement Perceptions Assessment
Instrument ............

Health Difficulty Scale. . .. ......
Kansas Family Life Satisfaction Scale..
Marriage Hap iness Scale ........
Environmenta Marker Index .......

_ Analysis of Data .............
RESULTS ..................
Dependent and Independent Means .....
Research uestion l ...........
Research uestion 2 ...........
Research uestion 3 ...........
Research uestion 4 ...........
Research uestion 5 ...........
Research uestion 6 ...........
Research uestion 7 ...........
Research uestion 8 ...........
Research uestion 9 ...........
Research uestion 10 ...........
Research uestion 11. . .' ........

V. DISCUSSION, CONCLUSIONS, IMPLICATIONS

AND SUMMARY ................
Discussion ................

vi

101

102
102

127
127

Chapter Page
Perceptions of Retirement ......... 128
Control/Anxiety ............ 128
Residence Selection .......... 130
Time Use ................ 131
Adequacy of Wife's Retirement Income .. 132
Life Continuity ............ 133
Research Questions ............ 134
Research uestion 1 .......... 134
Research uestion 2 .......... 137
Research uestion 3 .......... 138
Research uestion 4 .......... 145
Research uestion 5 .......... 145
Research uestion 6 .......... 146
Research uestion 7 .......... 146
Research uestion 8 .......... 147
Research uestion 9 .......... 148
Research uestion 10 .......... 149
Research uestion 11 .......... 150
Conclusions ............... 152
Implications for Research ........ 156
Imfllications for Home Economics/
uman Ecology ProfeSSionals ...... 162
Implications or Pre-retirement Education 163
Summary ................. 165
APPENDICES
APPENDIX
A. Retirement Perception Assessment
Instrument ................. 167
B. Retirement Perception Assessment
Instrument (Modified) ........... 168
C. Health Difficulty Scale ........... 169
D. Kansas Family Life Satisfaction Scale. . . . 170
E. Marriage Happiness Scale .......... 171
F. Environmental Marker Index ......... 172
G. Correlation for Perceptions of .
Retirement by Independent Variables
for the Regression Subsample ....... 173
LIST OF REFERENCES ................... 174

vii

Table

0301th

10.

11.

12.

13.

14.

LIST OF TABLES

Age of Women - 1985 .............
Family Size - 1985 ..............
Family Income by Community Size - 1985. . . .
Employment Status of Women - 1985 ......
Type of Employment - 1985 ..........

Em loyed Women's Personal Income by Community
15’.
ize - 1985 ................

Summary of Analysis Techniques ........
Meaggsof Dependent and Independent Variables,

Correlations of Environmental and Personal
Variables with Perceptions of Retirement. .

Step-wise Regression of Perceptions of
Retirement by Continuous EnVironmental and
Personal Variables ............

Analysis of Variance for Community Size with
Perceptions of Retirement .........

T-test Comparison of Rural and Small Town
Women on Perceptions of Retirement

T-test Comparison of Rural and Urban Women
on Retirement Perceptions .........

T-test Comparison of Small Town and Urban
Women on Perceptions of Retirement

viii

Page
85
86
86
87
88

89
100

103

104

105
107
107
108

108

Table Page

15. Analysis of Variance for Wife's Employment
Status with Perceptions of Retirement . . . 110

16. T-test Comparison of Nonemployed and .
Part-time Employed Homen on Perceptions of
Retirement ................. 110

17. T-test Comparison of Nonemployed and _
Full-time Employed Women on Perceptions of
Retirement ................. 111

18. T-test Comparison of Part-time and
Full-time Employed Women on Perceptions of
Retirement ................. 112

19. Step-wise Regression of Perceptions of
Retirement for Employed Women by All
Continuous Environmental and
Personal Variables ............ 114

20. Step-wise Regression of Perce tions of
Retirement for Nonemplo ed omen by All
Continuous Environmenta and
Personal Variables ............ 116

21. Frequency Distribution of Environmental
Markers Among Women Between 1983 - 1985 . . 119

22. Mean Stress for Environmental Marker Events . 120
23. Analysis of Variance for Husband's

Occupation Type with Perceptions

of Retirement ............... 122

24. T-test of Mean Differences Between
1983 - 1985 for Perceptions of Retirement . 124

25. Correlation of Variance for Wife's Employment

by Independent Variables for the Regression
Sub-Sample ................. 173

ix

Figure

LIST OF FIGURES

Modified Person-Environmental Interaction
Model ...................

Modified Person-Environmental Interaction
Model (Design A) .............

Modified Person-Environmental Interaction
Model (Design 8) .............

CHAPTER I

INTRODUCTION

In spite of the increased emphasis on the aging in
our society, the issue of women and retirement has been
neglected. At the present time, little is known about
mid-life women's expectations or perceptions about
retirement“ Many of the perceptions and expectations
about retirement are formed in the mid-years. How one
perceives the future has an impact on the decisions which
shape the future. Expectations for the future operate in
different ways. Expectations can become self-fulfilling
pr0phecies, or be a cause of discontent when unfulfilled.
Faulty perceptions can lead to decisions which would not
be made if more accurate information were available.

Perceptions are derived from environmental and
personal factors. Parr (1980) noted that researchers may
have fallen into the trap of confusing the environment
with perceptions of the environment. How people perceive
their environment, as well as dimensions of the
environment, need to be investigated. Understanding the
perceptions can make a meaningful contribution to the
understanding of person-environment interactions, and

outcome decisions. Further study of person-environmental

2
characteristics and perceptions of'iwomen preparing for
retirement can help to identify concerns, needs, and
trends, thereby suggesting ways to facilitate adaption and
life satisfaction during the retirement years.

Retirement is a major life transition for both men
and women; however, research has focused almost entirely
on the male retiree (Schmall & Staton, 1980; Johnson &
Price-Bonham, 1980; and Kasworm & Wetzel, 1981). When
women have been included, it usually has been to ascertain
their perception of and reactions to the retirement of
their husbands. Schmall & Staton (1980) cited three
primary reasons for focusing on women: (1) the present and
projected labor force participation of women (2) the
increasing proportion of female retirees, and (3) ‘the
longer life expectancy of women.

Labor force participation among women has been
steadily increasing. In 1980 the average labor force
participation of women between age 60 and 65 was 45%. By
1990 the percentage is expected to be 47%, and 71% by
2020 (Sp;ial_§gcurity_fiullgtin, 1985). The group of women
reaching retirement age in 2020 is projected to have a
more consistent labor force participation history, for
benefit computation purposes, than women reaching
retirement age in 1990 or even 2000. Fowles (1983) noted
that during the last few decades there have been sharp
declines in the proportion of men in the labor force at

3
older ages; however there has been a steadily increasing
proportion of women entering, re-entering, or remaining in
the labor force. This trend has implications for the
future economy.

Women are living longer than men. Only 50 years ago,
just as many males as females were reported at ages 65 and
over, but there has been a steady decline in the
proportion of men, and, hence, an increasing excess of
women since that time. In 1978, life expectancy at birth
for females exceeded that for males by nearly 8 years.
Expectation of life at birth in the United States in 1978
was 69.5 years for males and 77.2 years for females. The
divergence of male and female mortality has occurred in
spite of the fact that some important differences in
lifestyles and roles of men and women have been
diminishing (Cugren1_£gpulatign_3epgrts, 1984).

Related to longevity of women is the risk of poverty.
Women are more affected than men by changes in family
composition such as death of a spouse, divorce, or
separation (Duncan, 1984 and Warlick, 1985). Warlick
(1985) noted that more often than not, private pension
benefits fall sharply with the death of the husband, even
when he was fully vested. Workers who reach retirement
age can, and frequently do, elect to drop the survivor's
benefits in exchange for higher income while both are

alive.

4

Death of the husband can have serious implications

for the adequacy of Social Security benefits as well.
Social Security provides a spouse's benefit equal to 50%
of the retired worker's benefit. Thus a retired worker
and spouse (who is not entitled to a larger benefit on the
basis of covered employment) together receive a benefit
equal to 150% of the amount the worker would have received
if single. Upon the death of the husband, the wife's
benefit is reduced by one third. The reduction may be
greater if both spouses receive worker's benefits in their

own right.

M11339:

The challenge for educators, policy makers, and
families is to develop an understanding of women's
perceptions and expectations for this vital portion of
their life in order to facilitate a realistic and
satisfying retirement for women. Women have unique work
and socialization patterns. Johnson and Price-Bonham
(1980) and Schmall and Staton (1980) indicate that one
research challenge is to increase awareness of the
heterogeneity of working women. Diversity of economic
status, career path, and marital status needs to be better
understood. Johnson and Price-Bonham (1980) point out that
women need help, particularly in financial planning for

retirement. Many women have been socialized to expect

5
that men will take care of them throughout their entire
life. In addition, women are in the lower paying
occupations; therefore, they receive lower retirement
benefits leaving them with less actual retirement income.
In order to increase the effectiveness and appropriateness
of pre-retirement education, counseling, and retirement
policies affecting women, there needs to be a greater
understanding of women's unique expectations, perceptions,

and concerns.

W

The purpose of this study is to develop a better
understanding of perceptions of retirement among married
mid-life (aged 35 - 65) women.with at least one child.
The dimensions of perceptions of retirement to be explored
are control/anxiety, time use, residence selection,
adequacy of wife's retirement income, and life continuity.
In addition, changes in perceptions over a two-year period
will be examined. Antecedents (independent variables) for
the idimensions. of’ perceptions of’ retirement (dependent
variables) have been selected to reflect the heterogeneity
of’ married women. The study' will examine employment
status and type, stress from life events, perceived
health, marriage happiness, family life satisfaction,
rural/urban residence, perceived income adequacy, and
husband's employment status and type, as well as the

6
demographic characteristics of age, education, income,

years married, and family size.

B 1 l' l' E II' S! l ! NC-lfi4
This study examines a segment of the data collected
for the NC-164 regional project, "Stress, C0ping, and
Adaptation in the Middle Years of the Family,” supported
by the Cooperative Research Service of the United States
Department of Agriculture, the Michigan State University
Agriculture Experiment Station, and the Experiment
Stations of the other states involved. Nine states are
involved in the project: Indiana, Iowa, Kansas, Kentucky,
Louisiana, Michigan, Minnesota, Missouri, and Nebraska.
This study will utilize data collected only in Michigan.
The primary focus of the core study (NC-164) was
stress among rural families during mid-life. The core
study included both rural and urban residents, varied
employment status and types for men and women, and
multiple life events. Measures of life stress, health,
family happiness, and life satisfaction were
incorporated. The scope and depth of the core data
facilitate the examination of individual differences among

mid-life women's perceptions of retirement.

7
WW
HumanJcQJmiQLELamemrk

This study was based on a human ecological framework.

A human ecological framework examines the
interdependencies between humans and their environments.
A strength of the human ecological framework is the
holistic approach to influences on human behavior' and
potential. It takes into consideration relationships
between family members, between family members and their
work and other environments, and between family and the
socio-economic system, as well as characteristics of the
individual members of the family. Bubolz, Eicher, and
Sontag (1979) suggest that phenomena must be examined in
the wholeness of interaction and interdependence. It is
through interaction with the environment that humans form
perceptions, make decisions, and satisfy needs.

Bubolz et al. (1979) define the environment as the
sum total of the physical, biological, social, economic,
political, aesthetic, and structural surrounding of the
individual unit. A human ecological framework
conceptualizes the individual/family unit as nested in
three inter-related environments: the natural environment,
human constructed environment, and human behavioral
environment. Examples in this study of the human
behavioral environment include family members, the human

constructed environment contains community and work place:

8

the place of residence, rural, small town or urban
represents the natural environment modified by human
action.
W

This study uses a Modified Person-Environmental
Interaction Model to structure the research. Parr (1980)
incorporated the concept of perceptions into an
ecological framework. The Person-Environment Interaction
model consists of four components; personal
characteristics, environmental characteristics, mediators,
and behavior. The model was modified to add interaction
between person and environmental characteristics, which in

turn impact on mediators.

Figure l. EEBSQN2ENIIBQNU£NIAL_INIEBAQIIQN_MQQEL

person environment
characteristics characteristics
mediators
behavior

Source: J. Parr, 1980, MM, edited by
L. W.Poon, pp.

The Parr model begins with the assumption that it is
the behavior in environments that is of primary interest
to both theoreticians and practitioners. We really want
to know how persons might function in a given setting.

It is suggested that the best way to study a complex
system without destroying its functional integrity is to

9
choose one kind of behavior and to look at possible
interactive influences on the quality of that behavior.
For this study the concern is how mid-life married women
perceive their retirement life stage in order to help
understand how they make decisions regarding retirement.

Four components are seen as essential parts of an
interaction system at any one point in time: environment,
person, mediators, and behavior. The environment for
this study includes community, family size, and economic
factors. Person characteristics incorporate health,
education, employment status and type, satisfaction and
age. Interactions between environment and person in this
study are life events and income perceptions.

The mediating components are the persons' perception
of the environment, the importance to them of
environmental characteristics, what they expect to receive
from the environment, how they expect to be able to change
it, and the experiences that match or have discrepancy
between persons and the environment. These perceptions
are influenced by experiences, priorities, and
expectations. By mid-life a person has incorporated
varied perceptions, experiences, and expectations. It is
through the mediating variable that the inputs of person
and environmental variables are synthesized and
transformed to become behavioral outcomes. In this study,

mediating variables of perceptions of control, continuity,

10
and change for the retirement years are of primary
interest.

The behavior outcomes include the decision about what
resources to use in pre-retirement planning, the value
placed on planning, economic programs, and family
interactions, all of which impact on quality of life in
retirement. The behavioral outcomes are beyond the scope
of this study.

The arrows drawn between the components are not
considered exhaustive or exclusive. Other relationships
could be conceptualized. Conceptualization could differ
from one situation to another. The purpose of the model
is to sensitize the investigator to the likelihood that at
least four components are operating in a person-

environment interaction.

5' 'Ei E II B I

Those who work with pre-retirement planning can
utilize the data of this study with women who are
preparing for the retirement life stage. A strength of
this research study is that it takes into consideration
the diversity of women. As more of the diversity is
considered in a single study, the understanding that is
gained becomes more sensitive.

Three major classifications of work status are used,
full-time paid work, part-time paid work, and

11
nonemployment. The category of nonemployment can be
further' divided into homemaker, retired, disabled, and
unemployed. Work can be further classified by the type of
position held. This allows consideration of diverse work
histories.

The age span of the sample is 35 to 65, with age 65
being reached between 1983 and 2013. The age span
reflects the changing role of paid work and family in the
lives of women.

Women who live in rural comunities have been less
studied than their urban counterparts. Approximately half
of the sample live in the open country or in a community
of less than 2,500 persons. The sample allows for
comparison of rural and urban women.

Stress from major life events is examined. Williams
and Bubolz (1986) found that women perceived a larger
number of potentially stressful events had occurred and
reported higher overall degree of stress.

Pospisil (1979) observed that longitudinal research
opens new areas of information to the researcher. This
study reports on data on the sample population at two
points in time, 1983 and 1985. Two years can be
considered a relatively short longitudinal study.
However, the depressed economic conditions in Michigan in
1983 made it a stressful time for workers on the farm and

in the automotive industry. In 1983 the unemployment

12
rate for the United States was 9.9 and 14.2 for Michigan.
The per capita personal income in the United States was
$13,608; and $11,830 for Michigan. (Mi;higan_§1ati§tigal
Abstnagts, 1986-87). By 1985, Michigan had begun an
economic recovery. In 1985 the unemployment rate for the
United States was 7.2 and the Michigan rate was 9.9. The
per capita personal income for the United was $13,867, and
for Michigan it was $12,093. (Michigan__§tatistical
Abstracts, 1986-87). Due to the difference in economic
conditions between 1983 and 1985, there may be important
differences in stress from life events and perceptions of

control over the future.

l' 'l l' l E l'

Characteristics of the subjects and the manner in
which the sample was gathered determine how extensively
the findings can be generalized. The primary limitation
of this study is the imbalance of race and income. The
sample is predominantly Caucasian. Persons with higher
education and income levels returned the questionnaire in
greater numbers than persons with lesser formal education
and lower income.

In the original study, the intent was to explore
stress of intact families. The sample was purposely
limited to married persons.

The sample was a random sample; data were gathered by

13
mailed questionnaire. The questionnaire dealt with
sensitive areas of finances and family interactions.
Those who agreed to return the questionnaire may feel more
positive about their ability to cope and their
relationships. The sample may under-represent persons who
are less satisfied with their lives.

The following (assumptions underlying 'this research
have been identified:

1) There will be age cohort differences among the
women respondents. Each age cohort will be exposed to
somewhat different socialization influences.

2) Perceptions are formed out of the complex fabric
of life and are influenced by multiple factors. Only a
small portion of the influences can be explored in any
given study.

3) Women have historical and Socialization patterns
which differ from those of men.

4) The respondents will have sufficient knowledge of
self and trust in the institution gathering the
information to respond in an honest and reasonably

accurate manner.

CHAPTER II

REVIEH 0F LITERATURE

The review of literature is divided into five
sections. The first section examines the relationship
between women and their labor force participation. The
second section deals with economic issues of women in
retirement. The third section examines rural women.
Section four explores women and stress. The fifth section

discusses locus of control issues.

W

The relationship between women and work will be
examined from four perspectives: work patterns of married
women, comparisons of women who work in the market place
and full-time homemakers, work patterns of different types
of employment, and a comparison between women's and men's
work histories.

Mattied_flgmen

Keating and Jeffrey (1983) suggested that women are
heterogeneous in their propensity for labor force
participation. In Keating and Jeffrey's exploration of

married and unmarried retired women, four work patterns

14

15

for married women were identified; an interrupted pattern
where women leave the labor force for a period and return,
an interrupted pattern where the women leave and re-enter
the labor force twice, an unstable pattern which is
characterized by several gaps in employment, and the
delayed pattern in which the women entered the labor force
for the first time in their mid-thirties. These patterns
are similar to those found by Sorensen (1983) and Lopata
and Norr (1980). In contrast, Keating and Jeffrey (1983)
found that the majority of the single women followed a
continuous uninterrupted work career from early twenties
to early sixties, and when they had gaps, they were fewer
and shorter. Married women had a mean of 2.1 gaps, while
the never married had a mean of 0.68 gaps. As would be
expected, married women gave family reasons for work gaps
more frequently than those who were unmarried. Eighty-
three percent of the married women gave family reasons for
career gaps. They included (in descending order)
marriage, pregnancy, family illness, and husband's
transfer. The most common reasons given by singles were;
lack of available jobs, job training, and poor health, and
care of aging parents. Both groups retired at an average
age of 62 years.

Type of occupation impacts on work patterns.
Sorensen (1983) suggested that attractive jobs at the time

of marriage provide one reason for postponing the decision

16

to leave the labor force and that household work for women
with low-status jobs may provide a more attractive
alternative. The majority of women Lopata and Norr (1980)
surveyed were located in the secondary labor force, which
is characterized by easy entry and reentry, and also by
lower pay and shorter career lines than the male-dominated
primary labor force. In addition, delays in marriage and
motherhood were related to occupational achievement,
mainly through increased education.

The research suggests that younger women are more
likely than older women to be employed. Rexroat's study
(1985) found that those who had children after 1968 were
more likely to be employed while rearing children than
those who entered the child rearing stage before that
time. Having children in 1968 and bearing children over
the twelve year period, substantially reduced the number
of years employed, about 1.3 years with each child. Lopata
and Norr (1980) concurred that despite their heavier
childcare responsibilities and other factors that inhibit
their employment, younger women are as likely to be
employed as older women. Younger women are tending to be
committed to both family and career.

Sorensen (1983) noted that age at the time of
marriage had a significant effect on labor force
participation. Women who married at an early age were
more likely to have followed a pattern that included work

17
and family. Women who married late were less likely to
leave the labor force at any time, but if they did leave
at the time of marriage or first birth, were less likely
to return to the labor force.

Rexroat (1985) indicated that the husband's attitudes
toward the wife's employment had a considerable impact: as
the husbands's approval of his wife's employment
increased, so did her market experience. Lopata and Norr
(1980) reported that nearly 50 percent more women who
rated their husband's attitudes toward their work as
highly favorable were employed than those whose husbands
scored low on the same dimension.

The husband's occupational status had varying
effects. Philliber and Hiller (1979) studied the
perceptions of job status among working wives. It was
found that the extent to which a wife identified with the
middle class depends to a large extent on the occupation
of her husband. The probability that a working woman
married to a husband with a middle-class job will identify
with the middle class increases 0.99 percent for each
increase of one point in the prestige of the occupation
she holds. However, women married to husbands with
working-class jobs do not change their’ perceptions of
their social status on the basis of their job attainments.
Philliber and Hiller (1979) suggested that women whose

18
husbands held middle-class positions lived in an
environment where achievement was valued.

Job mobility is affected by husband's career path.
Ferber and Huber (1979) studied families where both
husband and wife had a Ph.D. The women were far more
likely than men to permit their location to be determined
by spouse's career interest. Thus, even highly educated
women tend to subordinate their own to their husband's
career. The less favorable mobility of women reduced
their annual earnings.

Sorensen (1983) noted that women married to men with
high educational attainment choose one of two very
different patterns of employment, staying in home or
pursing a career while raising a family.’

In general, married women have multiple career
patterns from continuous work to full-time homemaker.

A frequent pattern is an interrupted job history, with
family being the primary reason for ‘the interruption.
Several influencing factors have been identified:
occupational status of the woman, age cohort, age when
married, husband's attitude towards the wife's
employment, and husband's occupational status.
Betited_£mnlgxee£flgmemaker

Fox (1977) and Depner and Ingersoll (1982) concluded
that women's retirement was not a trouble-free or stress-4
free transition. Difficulties in retirement, which had to

19
do with morale were; level of social contact, reduction
of income, and perceived health.

The findings on morale are mixed. Fox (1977)
reported that retirees had significantly lower
psychological well-being than housewives or those still in
the labor force. Retirement had real consequences for
women in terms of reduction of income, general level of
social contact, and possible perception of’ health. In
contrast, Jaslow (1976) found that women who never worked
had lower morale than women who were retired. Jaslow's
study found women who had worked in the market place
tended to be the younger, the healthiest, and best off
financially, while the women who had never worked tended
to be the oldest, poorest, and in worst health.

Social contacts and friends were found to be an '
antecedent to a positive evaluation of life in old age,
particularly for the retirees (Fox, 1977; Keith, 1982; and
Depner, 1982). Keith (1982) noted that even though
importance of friends was important for all older women,
the sources of the social contacts were somewhat different
for retired women and homemakers. For example,
involvement in formal organizations (both civic and
religious) was especially important to the well-being of
retired women. Depner (1982) found retired women belonged

to an average of 1.87 organizations and housewives 1.45.

20

Fox (1977) indicated that women who worked much of
their lives had no fewer social resources, in fact were
more socially involved than lifelong homemakers and had
more involvement in informal interactions with friends and
neighbors than those still employed. However, the
retirees had a lower perceived level of social contact
even though they talked to more people on a given day than
housewives. Housewives were more likely to include
spouses and children in their social networks, while
retirees were more likely to name friends (Depner, 1982).
Keith (1982) found that fewer than 10 percent of both
groups of women had no intimate friends, but more retired
women (67%) than homemakers (54%) had two or more close
friends. Twenty percent of both groups saw a best friend
daily.

Fox (1977) found that even when there were few
differences in the education level of retirees,
homemakers, and still employed, the retirees had
significantly lower family incomes than the iother' two
groups. In contrast, Jaslow'(1976) found those who had
never worked to be by far the most economically
disadvantaged. The difference may be explained in part by
the difference in sample ages; the mean age for Fox was
60.7 and the age range for the Jaslow was 65 and older.
The older sample might include more widows who had never

worked.

21

Fox (1977) and Keith (1982) compared the health of
retirees and homemakers. Fox (1977) found health to be
worse among the retirees than among housewives. Keith
(1982) notes that valuation of health was more strongly
associated with loneliness among the retired than among
homemakers.

In sumary, social contact, health, and income are
related to level of morale among retired women and
homemakers. However, there is some indication that
retired women perceive fewer social contacts. The
findings on health and family income are mixed, with age
and marital status possible intervening factors.
W

There have been a limited number of studies examining
difference among women in different types of employment.
Kilty and Behling (1985) examined retirement expectations
of four professions: lawyers, social workers, high school
teachers, and college professors. The study found that
intentions to retire early related to being a high school
teacher, having venturesome investments, and planning for
a new career. High school teachers had the most positive
attitude toward retirement, and lawyers had the least
positive attitude.

Women suggested older ages for retirement than the
men in the Kilty and Behling (1985) sample. The age at

22
which the respondents planned to retire, as well as their
consideration of early retirement, were both related to
planning a new career. Positive attitudes toward
retirement were related to expectations about more
involvements, such as new friends, interacting with old
friends, and planning volunteer activities.

Block, Davidson, and Grambs (1982) studied
professional women and found a significant relationship
between work patterns and retirement satisfaction, with
continuous workers expressing more satisfaction than
intermittent workers. Continuous workers differed in
several demographic characteristics from intermittent
workers. Specifically, continuous workers were employed
longer, had done more retirement planning, retired most
often to pursue leisure activities, were more likely to be
single or widowed, and were less likely to have very high
or low incomes. Only three factors were significant
predictors of satisfaction with retirement; health, pre-
retirement planning, and income after retirement. Social
resources were not a significant predictor of
satisfaction.

Prentis (1980) investigated two groups of female
white-collar workers: professionals (lawyers, certified
public accountants, physicians, and college faculty) and
general employees (employees of banks, utilities,
hospitals, and colleges). The majority of both groups of

23

women anticipated retirement, appreciated some of the
changes that will occur, and were confident about making a
satisfactory adjustment. Their optimism toward the
future years may be understood in part by their median age
of 42 and present good health. The younger persons tended
to be more favorable in their attitudes toward retirement
than the older persons. Those in general employment had a
greater interest in retirement. Sixty-five percent of
the professionals and 35% of the general employees
intended to work until age 65. Fifty-five percent said
they would not retire based on their spouse's actions.
This was especially true for the professionals.

The overall concerns of both groups of women were as
follows: financial, missing work, time use, health,
loneliness, missing friends at work, getting along with
her spouse, and a feeling life was over. The younger
women did not perceive they would miss their work, but
women over 50 did. For the total sample, more married
women indicated they would miss their work than single
women. Professionals expressed the feeling that.‘work
would be missed.

Enjoyable aspects of retirement were as follows:
travel, free time, time for friends and family, hobbies,
opportunities to follow other pursuits, and no schedule.
Forty-two percent indicated they had done some pre-
retirement planning, with those earning $25,000 and over

24

doing the most. Non-planners indicated things would take
care of themselves.

In summary, older married professional women
indicated the strongest work attachment. Positive
retirement attitudes were related to being younger, having
a continuous work history, adequate income and health, a
social network, and pre-retirement planning. Being older,
married, and professional were associated with more
negative retirement attitudes. Early retirement was a
more likely consideration for high school teachers and
those who had higher incomes and new career plans.
Womenlflen

Men have been the primary focus of retirement
research. Only recently have women been included in the
studies. Atchley (1976) compared male and female
attitudes toward work and retirement. The study
controlled for age, marital status, and education. It was
found that older men in general were more likely than
older women to become accustomed to retirement in three
months or less. Women were less likely to make quick
adjustments to giving up their jobs and were more likely
to report feelings of loneliness. and depression after
retirement. Women perceived themselves as having less

contact with friends and less participation in voluntary

25
association. Atchley (1976) concluded that there is an
absence of sex difference in the importance of work, and
women indeed think that work is important. Retirement can
be a problem for women and they reported more difficulty
getting used to it than men.

Levy (1981) focused on men's and women's perception
of their future and adjustment to retirement. The
findings indicated that women who expressed an
unwillingness to retire, even after a year's time, lacked
a sense of the future and were poorly adjusted. They felt
their life was over. Health was not a factor.
Conversely, women who wanted to retire, whether they were
healthy or ill, sensed some definite length of time ahead
of them and a continuous sense of themselves. The
findings for men were different; for males willingness to
retire or not became relatively insignificant after' a
year's time.

Seccombe and Lee (1986) also found that women had a
somewhat lower level of satisfaction in retirement than
men. The antecedents of health, income, and marital
status were similar for both genders. No net effect from
occupational status was found for either gender.

Atchley (1982) compared men and women in the process
of retirement to determine what was most essential for
high satisfaction in retirement. Among women, the
economic aspect of retirement seemed to dominate the

26

decision to retire later. Women were much more likely to
retire before age 65 than men. These women tended to have
high social status, to be married, and to be in good
health. On the other hand, women were more likely to plan
retirement at age 70 or later than men. These women
tended to have low social status and be unmarried. More
women had negative attitudes toward retirement before they
retired than after. The only control variable that
changed the attitude after retirement was living
arrangement. Living with a spouse' decreased positive
attitudes toward retirement.

Inadequate income and poor health depressed attitudes
toward retirement for both men and women in retirement,
but adverse effects were greater for women than men. On
the other hand being in very good health was a tremendous
advantage for women but made no difference for men.
Retired women's attitudes were influenced more by both
adverse and advantageous circumstances than retired men.
Women were more likely to retire if married but that did
not necessarily lead to a positive attitude.

In general, both men and women are affected in
retirement by income and health factors, but women's

attitudes are more sensitive to shifts in circumstances.

27
59mm

Women are heterogeneous in their propensity for labor
force participation. Patterns of labor force
participation range from full time homemaker to full labor
force participation. Professional careers and being from
a younger age cohort were positively associated with
continuity in the labor force, conversely non-professional
employment and older ages were associated with interrupted
patterns of work. Single women were more likely to have
continuous labor force participation than married women.
Family factors such as number of children, husband's
attitude towards wife's employment, husband's career path,
and age of marriage were identified as factors in the
wife's pattern of labor force participation.

It is unclear whether employed wives or homemakers
have the highest level of satisfaction during retirement.
For both groups income, health, and social contacts are
important. The findings on marital status were mixed.
This may indicate that the quality of the marriage is a
factor as well as the marital status. Women were more
responsive to both positive and negative factors than men.

Women were inclined to retire both earlier and later
than men. Marital status, need for income, health,
continuity in the labor force, and job type were found to
be factors in the decision to retire. Married women with
high social status and good health tended to retire early,

28
while those who were single with lower incomes and social

status retired later.

WW

Retirement income of women reflects both events which
affect retirement income for the population as a whole and
patterns that are unique to women. This section will
begin with a brief overview of retirement trends in the
last 50 years. Second, differences in the economic well-
being of segments of the population will be compared.
Finally, sources of income will be discussed.
1! I I E'El M . B . I

There is little doubt that the economic well-being
of the elderly in relationship to the population as a
whole has improved over the last fifty years. Uhlenberg
and Salmon (1986) noted that during the Depression of the
19305 attention was drawn to the desperate plight of the
elderly, when dependent elderly persons probably reached
an excess of 50%. One response was enactment of the
Social Security Program as an income floor. Over the next
twenty years little further attention was given to the
elderly. A new concern emerged in the early 19605. In
1959, 35.2% of the population aged 65 and over fell below
the official poverty line. As a result there was the
enactment of programs such as Medicare, low cost housing,

transportation services, legal services, home-delivered

29

meals, and homemaker services (Uhlenberg & Salmon, 1986;
Cook & Kramek, 1986). Every year between 1965 and 1971,
benefits were increased, raising payments by almost 50%.
During the same time period Congress tripled the level of
earnings covered by the Social Security tax (Schultz,
1985). Since 1975 Social Security benefits have been
automatically increased each year whenever the Consumer
Price Index increased by 3% or more. In 1982 the
percentage of those over age 65 below the poverty line was
down to 14.6%. For the first time since the concept of
the poverty line was introduced, the proportion of old
people in poverty was smaller than the proportion of the
total population (Uhlenberg and Salmon, 1986). By 1984
the rate declined even further to 12.4%, while the general
population rate was 14.4%, and 22.2% for children
(Crystal, 1986). .

Although the general well-being of older persons has
risen, the policy changes of the 1980-84 period have not
had an even impact. Among those changes were reductions
of matching funds to states and subsidies for a wide
variety of welfare programs. For the first time there
was a ceiling placed on federal matching monies for
medical assistance, a tightening of food stamp eligibility
requirements and reduced food stamp benefits, restricted
eligibility' for> cash and in-kind assistance (available
mainly to the working poor), and reduced funding to states

30

for low-income energy assistance and social service
programs (Uehara, Goron, & Beeman, 1986). As of 1984,
elderly families with incomes above $32,000 started paying
income tax on some of their Social Security benefits. In
addition, higher labor’ and asset incomes also led to
increased tax liabilities. Those increases were not fully
offset by income tax cuts that were instituted over the
period (Moon, 1986).

As a group, elderly Americans fared better than
others since lower rates of inflation and income tax
reductions more than offset the negative effects. Females
living alone experienced a growth of 15.5% in disposable
income. These increases should be tempered, however, by
the knowledge that even by 1984, half’ of’ all elderly
individuals had disposable incomes of less than $5,406 and
the bottom 20% averaged less than $3000. What created
hardship for those living outside the nursing homes was
the 6% cut in Medicare and Medicaid outlays. The cut put
a severe strain on budgets of low-income elderly who had
to spend substantial shares of their incomes on health
services. Health care cuts coupled with other cuts left
the poor and near poor less well off. Twenty-two percent
of the elderly were near poor, those whose earnings fall
below 125% of the poverty level. Conversely, those
families in the highest 40% were better off (Moon, 1986;
Uehara et al., 1986; Cook & Kramek, 1986).

31

Measuring well-being among the elderly is a difficult
task because money income is not the only indicator
(Anderson, 1985; Riley, Hess, & Bond, 1983; Crystal, 1986:
and Cook & Kramek, 1986). Riley et al.(1983) noted that
as individuals age their economic holdings are
increasingly important in relationship to human capital.
Among the resources that are difficult to measure are
family input of work and gifts, home ownership (2/3 of the
elderly own their own home), buying and management skills,
and net worth. In-kind transfers such as Medicare, food
stamps, and rent subsidies, unlike cash, do not substitute
for other goods, but do add to well-being. The overall
effect may be an under reporting of the elderly well-
being. Uehara et al. (1986) suggested that those who are
older and poor perceive themselves as having fewer
informal resources and less likely to utilize formal or
informal sources of support.

In conclusion, in recent history those over 65 have
income distributions similar to the general population,
with some segments being much better off than others.
However it is difficult to get a totally accurate
understanding due to the complexity of in-kind transfers
and under-reporting of varied resources.

.1... ' u . -.“-. . l- ;- -. 'u, . '..

Economic well-being can be measured by dollar income
as percent of the poverty level or viewed in relationship

32

to previous earnings and the earnings of other segments of
the population. Riley et al, (1983) noted that the
feelings of relative deprivation were more important to
financial satisfaction than actual income. If persons are
unable to maintain a previous standard of living they may
be experiencing relative poverty. Uhlenberg and Salmon
(1986) compared the relative income of older women between
1960 and 1980. The first comparison was between current
elderly and those of previous generations. As previously
noted, changes in the recent years have improved the
position of more recent cohorts entering old age relative
to those who preceded them. Between 1960 and 1980 the
percentage increases were greater for those at the lower
end of the income distribution than for those at the upper
end.

Uhlenberg and Salmon (1986) next compared current
income to earlier income. There was a substantial drop in
real income for married women as they moved from late
middle age into old age, so that old aged income averaged
only about two-thirds of the previous income. The drop
for those entering old age in the 1970's was much greater
than for those who entered old age in the 1960's. This
probably reflects the increased labor force participation
of women and the trend toward earlier retirement of
husbands. The third comparison was made between the
elderly and contemporary younger adults. Older women have

33

much lower incomes, averaging in most cases about half of
younger women. The fourth comparison was inequality
within the older population. Overall income inequity was
found to be somewhat less among more recent cohorts
entering old age than among preceding ones. The relative
advantage of high income women to others declined
significantly. The disadvantage of the 75-79 age group
compared to the 65-69 declined. The disadvantage of older
unmarried women compared to married ones was large and
persisted over the time period studied. Unmarried women
had earnings that were 40% of their married counterpart.

O'Rand and Landerman (1984) explored the difference
in income for men and women. For both men and women
education is a major determinant of late life economic
status. O'Rand and Lanerman (1984) noted, however, that
men and women tend to follow notably different pathways
through work and early family roles, their status and
opportunities in later life vary accordingly.

The O'Rand and Landerman (1984) study focused on sex
difference in the impact on retirement income of family
and childbearing events and work history characteristics.
The findings indicate that, in general, both women and men
approach retirement with comparable work and family
histories, but only women's retirement income levels come
to be significantly determined by roles in both domains.
There is a dollar cost for each year of delayed entry into

34

the labor force. For women, each child costs $82. per
year in private or government pension income and about
$48. in social security income. For men, the number of
children and age of first full time job have small though
significant effects on social security income only.
Significant family life cycle costs are present for women
and not for men.

Comparisons were made between women and men in
favorable pension industries. A position in a favorable
pension industry increased women's retirement by $275 and
men's by $395, and each dollar increment in hourly wage in
1969 returns nearly $300 in retirement income to men,
while women receive $184 per year more for each wage
dollar.

Warlick (1985) compared the composition of’ family
income for families with aged male and female heads to
determine why poverty during old age is so heavily
concentrated among the families of aged women. The study
focused on the roles of work histories and marital status
in conjunction with rules of the current pension system.

Warlick (1985) reported that the economic status of
elderly people rose dramatically to equal, if not surpass,
that of the non elderly population. However, the rise in
per capita income had not been shared by all aged families
and persons. The problem of poverty is particularly acute
for the very old, minorities, and elderly women who head

35

their own household. Two of every three poor, elderly
persons are women, and two-thirds of these women live
alone. Women account for 70% of all aged persons eligible
for Supplemental Security Income. One in every four
families with a female head had income below the poverty
threshold, and another 15% were near poverty. Three
factors were identified as major determinants of poverty
among older women: marital status, work history, and
fertility experience.

Social Security accounted for almost 50% of the
total income of families headed by aged women but only 40%
of that for families with male heads. In addition, Social
Security benefits alone were sufficient to satisfy the
need of families with male heads, on average, but fell
short by about 10% for families with female heads. The
Social Security rules assume that a single individual
needs no more than two-thirds as much income as a couple
to attain the same level of economic well-being. In
contrast, the poverty lines used by government in their
official account of number of poor imply that a single
individual needs 80% of the income of a couple to achieve
equal economic well-being. This has serious implications
for women in death of a spouse. The problem is compounded
because the work careers of the vast majority of today's

elderly women fall into two catagories; no participation

36

or interrupted participation making pension contributions
small or nonexistent.

lick and Smith (1986) explored men and women who had
experienced widowhood. The study was restricted to those
persons who were non-poor prior to loss of spouse. lick
and Smith (1986) noted that widowhood is an inevitable
later-life event. In 1960 approximately 12% of all
households in the United States were headed by widows or
widowers. By 1970 it had increased to 14%, and it is
projected to be 17% by 1990. In 1973 widows outnumbered
widowers five to one. The study findings indicated that
nearly 50% experienced at least one year of poverty within
five years of loss of spouse. Men did not experience as
significant a drop in real total family income due to the
death of a spouse. After the first year of widowhood 11%
of the widowers, and 17% of the widows had fallen into
poverty, if the near poverty group were included it became
29% and 25% respectively. Widowers were less likely than
widows to experience changes that lead to extreme poverty.

Morgan (1986) explored the relationship between money
management. experience, discussion with spouse prior to
death about financial issues, seeking of financial advice,
and poverty. Fifty-four percent of the widows in the
sample were below the 1974 poverty level after all sources
of income were counted. It was found that having been

employed, discussion of financial plans prior to spouses

37

death, and receiving financial counseling were associated
with lower reports of poverty. Simply having experience
in managing money did not preclude a decline in subjective
economic well-being or poverty. Poverty was much more
common among new widows than those who had been widowed
longer. Possibly the longer widowed were eligible for more
benefits.

In general, women are more at risk to be poor or
marginally poor during retirement than men. ' A
combination of brief or no work history and widowhood
increased the risk of being less well off economically.
Conversely, employment, marriage, and financial planning
increased economic well-being.

We;

Retirement is an economic decision for both men and
women. Shaw (1986) suggested that women 'do respond to
economic incentives in their work decisions. They were
more likely to remain in the work force when their own
earning potential was high or their husbands' income was
low. Shaw's findings indicated that women who are
eligible for a pension or Social Security are more likely
to stop working than those without resources. The strong
influence of women's pension eligibility on their
retirement plans was evident. At each age women who would
become eligible for a pension at that age were much more

likely to plan to retire than were ‘women who had no

38
pension. Women who would be eligible for partial pension
but receive a full pension later were deterred. Women
who had no pension at all were more likely to plan to work
until age 65 or beyond.

Retirement income is based on a three-tier system,
the base being social security, then pensions, and
followed by individual resources. Social Security
benefits are based on earnings. Benefits are a calculated
average of lifetime earnings, with years out of the labor
force counting as zero. The earned amount is based on the
average earnings since 1950 (or age 21) and the lowest
five years are dropped. So the proportion of wage
replacement by social security is based on earnings
history, age of retirement, and family size. Benefits are
weighted to provide a greater adequacy to lower earners.
Social Security benefits by themselves are modest and
would not support a high standard of living. This is
particularly relevant for women since Social Security is
their major source of economic security. (Block et al.,
1981; Kahne, 1981; and Schultz, 1981). Program
improvements in the last 20 years have boosted benefit
levels for both women and men, yet the rate of increase
has been greater for men (Tracy & Ward, 1986). In 1980,
the average female wage earner who spent her career
employed in manufacturing in the United States received an
old-age pension under Social Security that was less than

39

two-thirds of the benefit level of her male counterpart.
Lower wages do not explain why women's benefits in
proportion were smaller in 1980 than 1960, 61% and 67%
respectively. This is particularly perplexing given that
women's average wages rose at a faster rate than men's
over the period. The benefit formula itself contributes
to the lack of parity (Tracy & Ward, 1986).

Wives who engage in paid work receive a primary
benefit, plus the difference in amount, if higher, that
would have been paid to them as a dependent wife. Their
tax contribution may not create any benefits if the
husband's benefit is the highest. Thus working wives may
be disadvantaged by lower pay, intermittent work, and tax
contributions that do not create additional benefits
(Kahne, 1981 and Holden, 1980).

If a woman is widowed, she is entitled to equivalent
of 100 percent of the husbands benefit, but loses her 50
percent. In the case of divorce, the marriage had to have
lasted ten years to make the wife eligible for any benefit
on her former husbands' earnings. In each case, the wife
must be age 62 to collect. This leaves the widowed or
divorced woman with even fewer benefits than the married
female. Widows account for the majority of the aged in
poverty (Kahne, 1981).

Private pensions have added to the well-being of the
retired. Approximately half of the private labor force is

40

covered by a private pension system. Those who are
covered are predominately men. .Historically, only one in
five women retiring from jobs in private industry could
expect to receive a pension, and those may be half the
size of pensions received by men (Block et al., 1981).
This picture has changed with increased labor force
participation rates among married women. Labor force
participation rates have increased since 1960 and continue
to rise. Of the women who retired in 1982, 31% were
entitled to pensions.

Many private pensions, while still needing
improvement, provide a 60 - 80% wage replacement rate when
combined with Social Security. Schultz (1985, 1981)
suggested issues that threaten the viability of private
pension plans: a) inequity and inadequacy of vested
benefits for highly mobile workers and women who have
intermittent work histories, b) most pensions are not
adjusted for inflation, c) survivors benefits may not be
available, and d) pension benefits may be integrated with
Social Security and thereby reduced.

Delayed mandatory retirement may be beneficial to
women who desire to continue in the labor force in later
years (Campbell, 1979 and Schultz, 1980). Women have
already been working to age 65. Women were opposed to
mandatory retirement on the grounds that having started

work at ages forty or forty-five they were not ready to

41

retire as long as they felt themselves capable of working.
For women over 65 labor force participation has remained
around 10% for the last two decades. For both men and
women, part-time work is the most conInon among the very
old. Iams (1986) found that women are two to three times
more likely to be working if they were not receiving
pension benefits than if they' were. Unmarried women
generally were more likely to be working than were married
women. Among married women, employment was more frequently
noted if their husbands were employed. Women with no
health problems were slightly more likely than other women
to be re-employed after retirement.

Morgan (1984) examined labor force participation of
widows. The most important reason that widows give for
work is the same as that reported by working women in
general: economic need. The sample had a median age of
62.5 and nearly one-third reported working at the time of
their husbands death. Of those not working, one fourth
considered looking for work. Age of widowhood related to
seeking employment, the younger women being more likely to
seek employment. Women reporting reduction in standard of
living were more likely to consider employment.
5311mm

In summary, the economic well-being of persons over
age 65 has steadily increased over the last five decades.
Currently, the percentage of older persons below the

42
poverty line is comparable to the 'general population.
This speaks to the success. of’ improvements in Social
Security benefits, Medicare and Medicaid, and other social
programs. However, not all is positive. Within the aged
population are pockets of poor and near poor, with non-
married and/or older females being the most likely to
occupy those positions. Two of every three poor persons
are elderly women, and two-thirds of these persons live
alone. Women account for 70% of all aged persons eligible
for Supplemental Security Income. Cuts in social programs
during the early 1980's negatively impacted on the well-
being of those who were the most at risk. Higher
inflation or further program cuts could change the

favorable position of many elderly persons.

BumLWemen

One in every four persons age 65 and over lives in a
rural area, yet this population has been understudied
compared to their urban counterparts. This section will
highlight the plight of the rural family in recent years,
discuss family stress, and the relationship between work
patterns and marriage satisfaction. The formal and
informal support systems of rural retirees will be
addressed, and finally, retirement satisfaction will be

examined.

43

Rural FamilvjeJLBeiM
Little, Proulx, Marlowe, & Knaub (1987) reviewed the

 

impact of farm legislation, tax policy, foreign trade, and
weather conditions on the well-being of farm families. In
brief, changes in farm policies made it possible for
farmers to borrow during the 1970s as never before. Many
borrowed against their land values to meet the then
growing global needs. When land prices dropped and the
market decreased they were left in financial trouble. By
1977, the total farm income had fallen 42% below the 1973
high. By 1978-79 farm family stress was at it peak, with
many losing their farms and their way of life. From 1980
to 1985, the overall economic situation declined
significantly for many farmers. Farm assets fell $53.8
billion. From 1982 to 1986, land values fell more than
27%. Export income reached an all-time high in 1981; it
then sharply declined and resulted in deceased farm
income. In 1982 more farm families sought off-farm
employment, with 60% of the income from non-farm sources.
Each of these factors had a profound impact on farm
families. In January 1985, over 370,000 American farms
were reported to be in serious to severe financial stress.

While 80% of American farm families remained
functioning within an acceptable debt limit, others
experienced serious financial difficulty. The difficulty
-stems from borrowing too heavily, high interest rates,

44

severe drought, decreased farm prices, and a shrinking
world market. Predictions for the near future are
decreased number of farms, a gradual decline in farm
family well-being, and more dependency on non-farm income.
Each of these predictions have implications for the
retirement well-being of rural women.
W

Berkowitz and Perkins (1984) studied stress among
farm women and found, like their urban counterparts, that
farm women experience stress associated with role
conflicts between home responsibilities and work outside
the home. Role conflict is negatively associated with
husband support and is unrelated to farm or home task
loads. Both husband's support and role conflict are
related to stress experience. Finally, husbands play an
' important role in mediating stress experienced by their
wives. Thus, the degree of involvement in different roles
and the potential conflicts between them may not be as
important as the interpersonal dynamics when performing
the roles.

Walker, LS. and Walker, J.L. (1987) examined
stressors and their symptoms and found farm females showed
a statistically higher mean for stress symptom items than
males. Stressful events most predictive of symptoms for
farm women included conflict with spouse, time pressure,
political issues, government regulations, financial

45
worries, and geographic or personal isolation. Over half
of these stressors were events which were specific to
farming. The most frequently occurring symptoms of
stress were trouble concentrating, sleep disruption,
change in health and increase in arguments.

Tight money was the most frequently occurring
stressor among two-generation farm families (Weigel, R. &
Weigel, D., 1987). A frequent stressor for mothers and
daughters-in-law was not being part of the operation.
Four factors accounted for the 54.2% of the variation in
stress among the farm families: feeling like labor rather
than management, difficulties in conmunications and team
work, value and generational conflict, and competition
between farm and family. Among the most useful coping
mechanisms was having faith in God, fun and physical
activities, and talking with others. Women reported more
stressors than men, which could be related to their
peacekeeping role within the family.

H I E l! I H . S I. E l'

Researchers recently have become interested in work
patterns of rural families, and their relationship to
marriage satisfaction. The underlying assumption was that
farm families spend more time working together and may
experience more stress. However a review of the work
patterns indicate that much work is non-fanm. Women who
live on the farms are increasing their participation in

46

paid off-farm work. While employment levels of non-farm
women have doubled in the last 50 years, the labor force
participation rates for farm women have quadrupled.

In 1983, 46% of farm women were part of the labor force,
while the rate for non farm women was 52%. Women's off-
farm jobs tend to be in nonagricultural occupations, while
the opposite is true for farm men (Meiners & Olson, 1987).

Scanzoni and Arnett (1987) noted that, more rural
wives than urban wives were employed, and they also work
more hours per week. However, rural employed wives
receive relatively less for their efforts than do urban
employed wives. The high annual income was $30,732 for
rural wives and $41,652 for urban wives.

Bokemeier and Maurer (1987) reported that, in
general, women in dual-earner households remain
responsible for domestic tasks and child care and are more
likely to accomodate their husbands' career than visa
versa. Beach (1987) noted that home-working wives, those
who work in cottage industry or farm tasks, lack a
standard work day, and the work is punctuated by
interruptions for personal and family chores. Meiners and
Olson (1987) suggested that farm women reduce their
household work time in response to paid work (as do non
farm women), however, they do not reduce housework if they
participate in the running of the farm. It seems they add
the farm work to their household work.

47

Bokermeier and Maurer (1987) compared marital quality
of farm couples who are dual-earners, where at least one
moonlights beyond regular employment, single earners, and
the unemployed. Little or no difference in marital
quality of couples with different labor involvement was
found when controlling for other variables. It appeared
that employment status or number of jobs that a couple had
was not significantly associated with tension,
communication, and sociability levels. The couples'
marriages were not adversely affected by increased labor
force participation. Other factors such as sex role
orientation and, particularly, similarity of husband's and
wife's sex role orientation scores, were more strongly
related to marital quality. Lower levels of communication
were present among couples with greater differences in sex
role orientation scores and with a more egalitarian sex
role orientation of the wife. -

Scanzoni and Arnett (1987) explored marital differences
between rural and urban couples. The findings revealed
that rural respondents had significantly fewer years of
education than urban respondents. Urban respondents
expressed a more liberal viewpoint of father and husband
gender roles than farm respondents. The men emerged more
traditional than women, conversely, the women more
strongly preferred egalitarianism. The rural respondents

48
perceived conflict resolution more positively and were
more devout than urban respondents.
W

.Scott and Roberto (1985) examined the inter-
relationship of the use of informal and formal sources of
support by a group of older rural adults who lived in
areas of less than 2,500 persons. The sample was divided
between poor (below the poverty line of $5,498) and those
above the poverty line. Respondents below the poverty
line were in poorer health, were older, and were over
represented by females, minorities, and widows. Of those
respondents who had children, 65.4% and 51.5%, of the poor
and higher income groups respectively, had at least one
child living within one hour's travel time.

In this study Scott and Roberto (1985) found three
patterns of support network use for each group. For both
the poor and higher income groups, proximity to children
was crucial to involvement with children; both groups
relied upon filial support. A pattern of reliance upon
friend support was associated with widowhood for both
groups. The poor relied upon their children for both help
and social interaction, while higher income persons did
not rely as greatly on children for help. The higher
income group sought organized social outlets, volunteer
opportunities, and formal services. Formal service use
was found to be slight for poverty level elderly persons.

49
The study provides evidence that formal services, when
used, are being used in conjunction with support from
children and friends rather than taking their place.

Scott and Roberto (1987) compared elderly rural and

urban men and women's patterns of informal sources of
support. The finding indicated women are more invested in
kin relations than men. Likewise, friendships are more
extensive and meaningful for older women. There were few
urban-rural differences in the type of aid received from
children, and nearly half saw a child within the week.
However, a greater proportion of rural females received
assistance when ill in comparison to urban females.
Rural women also reported a greater percentage of giving
and receiving help from friends. Friendships appeared to
be a more salient feature of rural support networks than
urban networks.

The findings of Brown (1985) speak to the value of
family’ and friends for rural elderly. Brown studied
formal services in small rural communities. When asked to
give an account of how service programs for the elderly in
local communities had been initiated, none of those
interviewed mentioned formal planning processes or efforts
by official advocates as important. What was seen as
vital were local leaders who identified closely with
community and group values and norms and who would work on
behalf of the elderly.

50

Not to seek outside help was seen as a source of
pride. Rural elderly persons were apt to know less about
available social services, be more indifferent or hostile
to government-supported programs and more difficult to
mobilize for participation in social programs (Brown, 1985
and Harbert & Wilkinson, 1979)

Krout (1987) explored rural-urban differences in
services and activities of senior centers. Rural centers
have fewer services and activities. Increased number of
activities related to higher incomes and a greater portion
of the population being over age 75. It would appear that
the smaller budgets and staffs of rural senior centers add
to rural disadvantage. However, rural centers offered
more in-home services. It may be that rural elders found
participation in the‘service network more acceptable if
the services are delivered in the home. Transportation
also may have an impact.

B I. l 5 II E I.

Dorfman and Moffett (1987) compared retirement
satisfaction for married and widowed rural women. The
sample was stratified by town versus rural and included
women who reported they had retired from a paying job
within the past ten years.

Self-perceived health was the strongest predictor of
satisfaction for both married and widowed women. Social

participation and performance of volunteer roles related

51

positively for both groups. Perceived financial adequacy
was the strongest predictor of satisfaction for married
women. Age was negatively related to satisfaction for
married women. The older the respondent the less
retirement satisfaction. The proportion of close friends
retained from the pre-retirement period was the strongest
indicator of satisfaction for widowed women, but
negatively related for married women. Frequency of
contact with close friends and neighbors was positively
related to satisfaction for widowed women only. The more
frequently a widowed woman received aid from her
neighbors, the more negatively she described her
retirement.

These findings are in agreement with an earlier
examination of retirement satisfaction among rural females
by Dorfman, Krout, and Heckert (1985). Perceived
financial adequacy was the strongest indicator of
retirement satisfaction. Other indicators were quality of
relationship with confidant, frequency of aid from
relatives in the past two years, a positive change in
number of voluntary memberships, and occupational
prestige. Length of retirement was negatively related to
the satisfaction with relationship with other persons.
Summatx

Farm families in the 19805 have experienced economic
difficulties due to a complex interplay of government and

52

banking policies, world markets, and natural
environments. As a result work patterns have altered to
include more non-farm work for both men and women. Stress
and marriage satisfaction for farm wives is more closely
related to shared beliefs about appropriate sex roles than
the increased diversity of labor. The closer the belief
systems of husband and wife, the less the stress and
higher the marital satisfaction.

Family roles and relationships were especially
important to women. Farm women report a higher level of
stress than men and many of the stressor events (family
relationships, shared business decision making) relate to
family interactions. The women and the poori (mostly
women) rely more on family and friends than formal
systems.

A strong value of independence was indicated. Life
satisfaction decreases with age, conversely dependence on

friends increased. Being able to reciprocate was valued.

Ransom

Women perceive that a larger' number' of’ potential
stressful events occur in the family, and report more
personal interaction events and a higher overall degree of
stress than men (Williams S Bubolz, 1986; Seigler &
George, 1982; and Kessler & McLeod, 1984). Topics

53
discussed in this section includes interpersonal networks,
family related stress, stress buffers, and stress and age.
Welded:

Kessler and McLeod (1984) argued that the greater
vulnerability to stress is due to the greater emotional
involvement of women in the lives of those around them.
Women reported a number of events that men failed to
mention. The latter consisted of life crises that had
occurred to members of the respondents' social networks:
family, friends, and neighbors. Women reported that such
network events were very distressing to them, while men
seldom mentioned events of this type. Homemakers reported
the highest level of exposure to network events, followed
by women in the labor force, and lastly men in the labor
force.

Seigler and George (1982) concur that women were more
likely to mention interpersonal and community events while
men were more likely to mention personal events as
stressful. It was noted that when the events described
are those that happen to others, the ability to control
and actively cope may be limited.

W

Domains of family life are particularly salient to
the degree of stress and how life as a whole is perceived
by women (Sontag, Bubolz, & Slocum, 1979; Williams &
Bubolz, 1986; and Seigler & George, 1982). Sontag, et al

54

(1979) found the most important factors influencing stress
level were family life, self, family income, and housing;
with family life making the greater contribution. Seigler
and George (1982) reported family issues fell into five
major categories: marriage and courtship, establishing a
home, parenting, relationships with parents, and
alcoholism in the family. These are similar to the issues
reported by Bubolz and Keefe (1980). They found that
Michigan women identified family life, children, love and
affection, and personal health as major life concerns.

Williams and Bubolz (1986) focused on family life
events which were more frequently reported and perceived
as stressful. Mid-life women reported more frequently than
men that a relative or a pet had died, and more health
events. Women reported menopause, but it was not
identified as a health issue among men. Seven individual
events were perceived as more stressful for women than for
men. They were: 1) death of family member, 2) family pet
dies, 3) member demands new privilege, 4) household chores
pile-up, 5) death of a close friend or confidant, 6)
member experiences serious emotional problems, and 7)
family member involved in the courts. Kandel, Davies, and
Raveis (1985) examined stress level in relationship to
roles. The findings indicated that stress has a stronger
effect on level of depression when the stress originates

from the household roles. In descending order, the

55
household role, marital role, and occupational role create
depression.

Seigler and George (1982) found parenting with one's
own children (including emptying and refilling the nest)
and relationships with one's own parents across the life
cycle as major family life issues for women. Aizenberg
and Harris (1982) noted that the middle generation is
increasingly feeling the squeeze between the older and
younger generations as they attempt to cope with both
simultaneously. In 1980, when there were 2.1 people aged
65 to 85 to every person 45 to 49, the ratio was twice as
large as in 1940. Given the current mortality rate, this
trend will continue.

Boss, Pearce-McCall, and Greenberg (1987) examined
the effects of boundary ambiguity on parents when children
leave home. It was found that mothers report a higher
level of boundary ambiguity. Mothers tended to be
slightly more preoccupied with the adolescent who left
home. In particular, mothers reported that they thought
about the adolescent who had left and felt lonely more
often than fathers did. For men, boundary ambiguity was
associated with somatic symptoms; for women, it was
associated with their general attitude toward present
life. The difference between parents decreased as the
time increased since the adolescent left home.

56

There has been a steady increase in young adults
'reroosting". Wise and Murry (1987) reported that the
percentage of adults ages 18-34 who live with their
parents has gone from 22.7 percent in 1960 to 30.3 percent
in 1983. The increase of 18-24 year-olds was even more
sharp; 45.8 in 1960 to 58.9 in 1983. There was a 98
percent increase in married couples with two 18-24 year
olds residing in the parental households. There is an
indication in the research that middle-aged parents do not
welcome this trend. Clemens and Axelson (1985) noted that
80% had not planned to have a child at home at this time.
In the Clemens and Axelson study 47.6% of the respondents
complained of negative effects. Adult children in the
home can have significant negative effects upon a
marriage, particularly as the children get older.
Respondents commonly referred to the relationship as
strained, causing conflict and adding stress.

Mancini and Bleiszner (1985) noted that the trend is
not limited to young adults, but includes the middle-aged
child (40 to 60 years old). Both generations place a high
value on independence; neither is especially interested in
living with the other. These mature people indicate that
their relationships often have many costs and benefits.
The relationship requires an investment of time and energy
as they try to maintain a warm, productive, and mutually
satisfying association.

57

The middle generation finds itself caught between
trying to meet their own needs and meeting the needs of
their parents and children. Wise and Murry (1987) point
out that a very conservative estimate of the number of
people who are involved in parent-care at any given point
in time is well over five million. In addition, modern
children provide more care and more difficult care to
parents over larger periods of time than they did in times
past.

It is agreed that women offer significantly higher
levels of overall assistance to aging family members than
men, (Stone, Cafferata, & Sangl 1987; Horowitz, 1985; and
Brody, Kleban, Johnsen, Hoffman, & Schoonover, 1987).
Stone et al. (1987) investigated a national sample of
caregivers to non-institutionalized frail elderly. This
study substantiated the often reported observation that
informal caregivers are predominately female, with wives
and daughters providing the bulk of the care. Daughters
were twice as likely as sons to assume the primary
responsibility with no assistance. Less than 10% reported
the use of paid services, and those who did rely on formal
care were assisting the more severely impaired elders.
Of the overall caregiver population, 80% provided unpaid
assistance 7 days a week. Horowitz (1985) suggested many
adult children did perceive caregiving as stressful, with
women perceiving a higher degree of stress. Daughters

58
were more likely to report their leisure time had been
affected, their emotional state had changed for the worse,
and their plans for the future had been negatively
affected. Caregiving also was reported to impact on their
social, emotional, and familial lives.

The type of assistance offered by daughters differs
from sons. Daughters were more likely than sons to
provide assistance with personal care hygiene tasks;
whereas, sons were more likely to provide assistance in
decision-making and provision of financial assistance
(Stone et al., 1987; Horowitz, 1985; and Montgomery,
Gonyea, and Hooyman, 1985). Montgomery, et al. (1985)
examined the subjective and objective burden of
caregiving. Objective burden was defined as the extent of
disruptions or change in various aspects of the
caregivers' life and household. The tasks that related to
objective burden 'were nursing care, bathing, dressing,
walking, transportation, and errands. These are tasks
primarily associated with female caregivers. It was
hypothesized that freedom or lack of freedom could be a
critical factor in explaining the relationship between
objective burden and personal care tasks. The reported
tasks are confining in both terms of geographical
proximity and scheduling of time. Subjective burden was
described as the respondents' attitudes toward or
emotional relationship with the caregiving experience.

59
The younger and employed caregiver had a more negative
attitude and emotional relationship.

Stone et al. (1987) found that 44% of the daughter
caregivers were in the labor force. It also was noted
that the majority of the respondents expected working
daughters rather than sons to quit their .jobs or to
rearrange their work schedules for parent care. Brody et
al. (1987) explored the relationship between care of frail
mothers and working women. The findings indicate that
daughters who work continue to meet their responsibilities
to their families, jobs, and elderly parents who need
help. When confronted with competing demands on their
time and energy, they give up their own free time and
opportunities for socializing. Mothers of working
daughters did not receive less overall care than those of
nonworking daughters. Substantial proportions of working
women reported that parent care had made them miss work,
caused work interruptions, made them lose pay, robbed them
of the energy to do their work well, and limited their job
choices. Of those who quit work, their mothers were the
oldest and most disabled, and the daughters were more
likely to be the sole helper. The daughters' occupational
level, deterioration of daughters' health, and daughters'
feeling restless were associated with physical and mental
impairment of the mothers.

60
matters

Family life is related to life satisfaction as well
as increased stress. Sontag et al. (1979) noted women who
feel very satisfied with their family life are most likely
to feel very satisfied with their total life. Herzog,
Rodgers, and Woodworth (1982) found marital status (being
married) had the strongest effect on subjective well-
being, displaying a similarly strong effect across all age
groups. Likewise all age groups viewed their children as
an important source of happiness.

Kandel et al. (1985) found the lowest level of
depression associated with stress was among those women
who held combinations of roles involving the most complex
set of obligations: marriage, employment, and parenthood.
The highest level of depressive symptoms was experienced
by women who were single, not working and not a parent.
Married women showed statistically lower levels of
depression than single women. Those who were parents
experienced less depression than those who were not
parents.

Lin, Woelfel, and Light (1985) found that depression
associated with undesirable life events was reduced when
help came from strong rather than weak ties. For each
level of undesirability, help from a spouse/lover was
related to lower levels of depression than help from a

close friend, an acquaintance, or helping professional.

61

The married experienced lower levels of depression from
stressful events than the unmarried. Eaton (1978) found
that when the married experienced a stressful event they
maintained better mental health. For the unmarried, the
addition of the support of a single individual had more
positive impact on mental health than more widespread
groups or voluntary associations.

Wethington and Kessler (1986) examined the
differences between actual support and perceived support
on the buffering of stressful life events. It was noted
that it was only when the event was evaluated as
threatening and the coping capacities evaluated as
inadequate that the event produced stress. It was found
that the stress-buffering effect of social support was
more strongly linked to the perception that support was
available than to the effects of actual supportive
behaviors. It may be that personal coping competencies
are bolstered by mentally cataloging one's reserve of
ancillary coping resources, including available
supporters. Self-reliant copers were those individuals
who maintained strong ties to support networks, which
provided a sense of assurance that help would be available
if needed.

W

Sands and Parker (1980) examined perceived

stressfulness across three age cohorts: ages eighteen to

62

twenty-four, thirty to forty, and sixty-five to eighty-
six. One pattern that stood out was that the elderly
persons rated death - related events (death of spouse,
family member, or close friend) as requiring significantly
less readjustment than younger groups did. Elderly
persons also rated "changing to a different line of work"
as less stressful. In contrast, the elderly persons rated
“change in social activities", “feelings of slowing down",
“vacations" and "holidays“ as more' stressful than the
younger group.

Herzog et al. (1982) noted that observed declines in
numbers of life events, and presumably related stress,
resulted in older respondents being more likely to rate
their lives as free and as easy than younger respondents.
However, the decline in stress was less if the respondent
had poor health, low family income or lower level of
education.

Samar!

Women reported higher levels of stress and more
stressful events than men. The women's extensive network,
as well as family, were sources of stress. Family stress
stemmed from family life, children's demands for new
freedoms, children leaving and returning home, death,
household chores, the marital relationship, relationship
with parents, and caregiving to frail parents.
Conversely, families played a stress buffering role. When

63
women found their family life satisfying they reported
life as a whole satisfying. When stressful events
occurred assistance, or perceived availability of
assistance from significant others, increased coping and
diminished negative results of the event. The literature
suggests that the strong relationship between family and
stress level for women is based on their emotional
involvement and awareness. It also may be related to

their willingness to share personal feelings.

LgcuLQLContLol

Most studies show that persons who perceive that they
control their rewards through their own actions tend to
have higher life satisfaction and self esteem, fewer
mental disorders, and stronger ego function (Kivett,
Watson, & Busch, 1977). Kivett et al. (1977): examined
the relative importance of several physical, psychological
and social variables to the type of control that middle-
aged adults perceive over their environment. The study
found that 58% of the variance was accounted for by social
variables; occupation, religious motivation and education.
Occupation was the most important. Individuals who
perceived having more control over their environmental
rewards were more likely than others to be in job types
that provided positive reinforcement through control of

persons or powerful objects. For example, persons in

64
administrative jobs which require supervision of other
people and operators of large powerful machinery perceived
more control over their environments than laborers or
service workers.

Religious motivation was an important predictor of
perceived control regardless of the relative influence of
other variable. Thosei who perceived less control over
their environment were more extrinsically motivated.
Education contributed no increments.

The psychological variables of actual self-concept
and ideal self-concept accounted for 25% of the variance.
Physical variables (age, sex, self-rated heath) explained
16% of the variance. The findings of the Kivett et al.
(1977) study do not support the findings of other studies
with respect to the importance of self-reported health to
perceived control of middle adulthood. One reason may be
the relative small amount of variance of health ratings.
Health may be a more salient factor in an older
population.

For women between the ages of 21 and 63, a
statistically significant relationship was found between
internal locus of control and depression (Walford-Kraemer
& Light, 1984). Low control over life was related to high
depression. Married women had the lowest level of
depression. Women who earned incomes of $30,000 or better
had lower depression scores; while those who earned under

65
$10,000 had the lowest depression scores. Women with a
college education had a higher sense of internal control;
while those with a high school education had a lower sense
of internal control. Residence made a difference, with
rural women having high internal control and small town
women low internal control.

McKenna and Nickols (1986) noted that the level of
internal control relates to the way women prepare for
retirement. Women who have an internal locus of control
were inclined to be more active investors and financial
planners. Kasworm and Wetzel (1981) noted that women who
focus predominantly on inner family activities or who have
been protected by their spouse from the realities of life,
financial planning, and decision making may experience a
lesser sense of control and greater stress. Sense of
control tends to be a gender issue, women were found to
be more inclined to an external locus of control than men.

Steitz (1979) found age differences among women in
locus of control. Teenagers and retired women tended to
respond with very low or very high internal control
scores. Mid-life females perceived a greater degree of
influence on powerful others and were more like males in
their level of internal control.

In general, higher levels of internal control were
associated with self-esteem, financial aggressiveness, and

leadership. External control was associated with

66
depression, a higher propensity for mental illness, and

lethargy.

Swan
The literature highlights the complexity of the

environment in which married women formulate their
perceptions of retirement and carry out their life during
the later years. Characteristics that relate to well-
being in retirement are varied, interrelated, personal,
and environmental. An important factor is work history.
Married women have multiple work patterns; continuous,
interrupted, and nonemployed. Those who work continuously
tend to be either working from economic need and/or
involved in a profession. They are more likely to enjoy
retirement benefits, however, may feel the greatest income
loss at time of retirement. The women with the
interrupted work patterns generally give family reasons
such as marriage, children, husbands's attitude, or
husband's career move. The interrupted workers tend to
be in low entry level jobs and have few benefits. The
nonemployed women focus on the family and are dependent on
husband's earnings. There are inconclusive findings
concerning the comparative well-being of those who retire
from employment and the nonemployed, both groups were
found to have the highest level of well-being.

67

Economics plays a major role in retirement. As a
whole the economic well-being of the elderly population
has risen to the point where the poverty level is on par
with that of the general population. However, there are
pockets of poor within the elderly population, and they
tend to be comprised of older, single/widowed, and less
healthy females. Two of every three poor are women.
Approximately 50 percent of the women are dependent on
Social Security. This leaves these women at economic risk
since Social Security is based on the generally lower
earnings of the women or' a portion of the husbands'
earnings. There is a higher level of economic well-being
for those who have vested pensions and who earned enough
to draw on their own Social Security. Although younger
women are tending toward continuous work and higher level
positions than the women who preceded them, women will be
at economic risk for years to come.

One group that is experiencing particular difficulty
is farm wives. They have a higher level of employment
participation due to the economic vulnerability of
contemporary farmers. The farm wife is more likely to
experience stress, role conflict, and have a lower paying
position than her urban counterpart.

Psychological (as well as economic factors), are
associated with perceptions of retirement well-being.

Women report more stress and more stressful events than

68

men. Much of the stress is related to family issues such
as launching children, reroosting children, events in the
family, spouse relations, parent care and relationships,
role strain, and family death. Conversely, families play
a stress-buffering role. Women who are~ married with
children enjoy the highest level of coping and life
satisfaction. A second psychological factor is the degree
of internal control. The greater the internal control,
the higher the level of self-esteem and positive coping
behavior.

Several concerns related to retirement well-being
have been identified: financial security, perceived
economic adequacy, retirement policies that allow options
for the nonemployed and underemployed, health, time use,
social networks, marital relationships, parent
relationships, family relationships, and locus of control.
Each concern is related to how realistically a woman
perceives her options in retirement and the satisfaction
level experienced in retirement. The literature dealt
primarily with findings among the retired rather than with
influences on the perceptions of retirement of the pre-
retiree. Perceptions are of concern because of their
influence on the decision process.

CHAPTER III

METHODOLOGY

This chapter; presents the research questions, the
research design, the research variables, the procedure
used, selection of sample, description of the core sample,
description of the sub-sample used in this study, the
interview schedules, and analysis of data. This study
explores the perceptions of retirement of mid-life married

women .

W
The research was guided by a number of basic
exploratory questions. The questions fall in two
categories: the relationship between selected

characteristics of women and their perceptions of
retirement and change between 1983 and 1985. Perceptions
of retirement, (the dependent variables), include a sense
of control/anxiety, residence selection, time use,
adequacy of wife's retirement income, and life continuity.

Question one is the basic relationship question.
Questions two through nine focus on the relationship

69

70

between specific characteristics and retirement
perceptions. Question two concerns community type; three
employment status; four and five income; six age; seven
and eight environmental markers and question nine
husband's employment type. Question ten seeks information
on type of change that is experienced. Question eleven
links selected characteristics with changes in retirement
perceptions.

(1) What is the relationship between characteristics
of mid-life married women and their perceptions
of the retirement life stage?

(2) What is the relationship between community type
and perceptions of the retirement life stage?

(3) What is the relationship between employment
status and perceptions of the retirement life
stage?

(4) What is the relationship between the portion of
family income earned by women, personal income
or family income and their retirement
perceptions?

(5) What is the relationship between perceived
income adequacy and perceptions of the
retirement life stage?

(6) What is the relationship between age and

retirement perceptions?

71

(7) Which environmental markers do women report as
most stressful?

(8) What is the relationship between perceptions of
retirement and degree of stress pile-up from
environmental markers and number of
environmental markers?

(9) What is the relationship between perceptions of
retirement and husband's employment type?

(10) What changes in retirement perceptions did the
women experience between 1983 and 1985?

(11) What is the relationship between the
characteristics and change in perceptions of the

retirement life stage?

BMW

This study uses the survey' method. It examines
families at two points in time and collects information on
family change during the intervening time between data
collections. The primary focus of this study is two-fold:
Model A explores the relationship between the person-
environment and interaction characteristics of women and
perceptions of retirement, the independent and dependent
variables are measured at time two data collection. Model
B explores the relationship between characteristics and
changes in retirement perceptions between time one and

time two. In the model, person-characteristics are those

72
which relate mainly to the women themselves. Environment-
characteristics refer to some aspect of the environment,
while interaction-characteristics emphasize inter-
relationship with the environment. Since all components
are related, it is recognized that placement of the
characteristic is somewhat arbitrary.

Figure 2 “I! I

IND

 

 

Environment Characteristics

 

 

IDESIGMI
Independent Varjebles (l-Z)
Person Characteristics
Education
Age

Employment status
Employment type

Years married

 

Self perceived health
Marriage happiness

 

 

 

\L

Family income

Wife's income

Family size

Community type

Husbands employment
type

 

 

J,

 

Interaction Characteristics

 

 

 

Perceived family income adequacy
Environmental markers

Family life satisfaction

 

 

 

W J,

 

.----—--)

Mediators

 

 

Perceptions of Retirement:

Control/anxiety
Residence selection
Time use
Work income

Life continuity

 

 

l

 

 

 

 

“'"" "'°"’°" BEHBVIOT‘ "“"'""""

.. .. -_..--..-..------..>

Figure 3 IN! I

1-10' ' ' '- 01

WM

11:21

9‘. ‘1. TA.

 

Person Characteristics

 

Environment Characteristics

 

 

Education
Age
Employment status
Employment type

Marriage happiness

Years married

 

Self perceived health

 

 

Family income

Wife's income

Family size

Community type

Husband's employment
type

 

I}

‘ Interaction Characteristics

 

 

 

Perceived family income adequacy
Environmental markers

Family life satisfaction

 

 

 

. ib
W W

 

Mediators

 

 

Perceptions of Retirement
Control/anxiety
Residence selection 4‘

Time use

Work income

Life Continuity

 

 

 

-------------- (BehaviorF---«~—~-~~-~-----‘

 

 

75
B I M . I]
DEPENDENT VARIABLES
B l' l E I.

Ceneeptuel_fle£initien; The intentions for, or mental
image of life after withdrawal from one's position,
occupation, or active working life.

Qpe:etiene1_nefinitien; In the retirement section of
the questionnaire respondents were asked a series -of
questions pertaining to their perceptions of retirement.
Each question had five possible responses which ranged
from strongly disagree to strongly agree. The instrument
measured perceptions along five dimensions;
control/anxiety, residence selection, time use, adequacy
of wife's retirement income, and life continuity.
CQDILQILAnzielx

Ceneeptuel__nefinitien; The degree of mastery and
positive attitude toward one's retirement life stage, with
low levels of' mastery and negative attitudes creating
anxiety.

Qpe:etienel_flejinitien; Control/anxiety was measured
by four items: a) I am comfortable thinking about
retirement, b) I have a sense of control over my future,
c) I am looking forward to retirement, and d) My
retirement needs will be covered by my spouse. The
higher the score, the greater perceived sense of control.

76
B 'I 5 1 l'

Ceneeptuel_nejinitien; Whether the individual plans
to change their current residence during the retirement
lifestage.

Onezetiene1__ne11nitien; Residence selection was
measured by two items: a) I will move to a different home
and b) I will move to a different conInunity. The higher
the score the more likely there will be a major change in
residence.

Iim£_flie

W The degree of positive or
negative attitudes toward time use and leisure time
activities during the retirement life stage.

W Four items measured time
use: a) I will have more time for recreation, b) I will
start new hobbies, c) I will enjoy having more free time,
and d) I will have plenty to do with my time. Higher
scores represent a more positive attitude toward time use
in retirement.

!l E H'E . B l' I I

Ceneeptual Oefjnjtjen; The perception of whether the
retirement earnings from the wife's employment are
sufficient to cover her financial needs in retirement.

Operetienel__nejinitieg; Adequacy of wife's
retirement income was measured on a single item: The

retirement plan through my work will cover my financial

77
needs. The higher the score the greater the perceived

adequacy of the wife's retirement income.

HE C l' “I
Consensual—Definition; Life is perceived as

carrying the same theme and general activities throughout
the adult life span.

Qnenetienel_flejinitien; Life continuity is measured
on a single item: My life will remain basically the same
as it has been. The higher score presents little change
in life's pattern.

INDEPENDENT VARIABLES
mm

W: The classification of
community of residence is based on the population density.

W Community type is
categorized into three groups: a) rural - residence is
open country or village of less than 2,500 persons,

b) small town- more than 2,500 persons but less than
50,000 persons, and c) urban - 50,000 persons or more.
Ecrceixed.1ncgme_Adeguasx

Ceneentua] Definjtjen; The cognitive perception of
how well the income meets family financial needs and
wants, not actual income.

Wen; Perceived economic adequacy
is measured on a five point scale: a) won't be able to buy

some necessities, b) will be able to meet necessities

78
only, c) will be able to afford some of things we want but
not all we want, d) will be able to afford about
everything we want, and e) will be able to afford about

everything we want and have some left over.

Emflmenmm

Ceneentnel_ne11nitien; The level of participation in
paid work. ‘

Qpeeatiene]_peflnj_tjen; This concept is defined by
categorizing paid work participation as full-time in the
labor force (35+ hours/week), part-time in the labor force
(less than 35 hours/week), and nonemployed (no
participation in the labor force). The nonemployed
designation includes the full-time homemaker, retired

persons, the disabled, and those who are laid-off.

We:

Wen; The field of work in which a
person is engaged.

QeemjjeneJ—Deflejtjen; Eight categories of
employment are used to classify the work in which a
person is engaged for pay. The categories were those used
in the 1980 U.S. Census. The categories range from
professional through blue collar and service work.
Seli.£erseixed_flealtb

Ceneeptuel__ne11nitien; The degree of physical

difficulties as perceived by the individual.

79

Onenetienel_ne£1nitien; A measure of twelve items is
used to assess perceived health difficulties. The
respondents indicate the frequency with which each item
occurs on a five point scale from never to almost always.
The higher the cumulative score the more health
difficulties.
Enmilx_LiI£_§iIi§InsliQn

CeneemjLDefinitjen; The level of gratification
one receives from the relationships of the nuclear family
unit; husband, wife, and children.

Qnenetienel_fle£1nitien; The four item Kansas Family
Life Satisfaction measure is utilized. This scale
provides a life satisfaction score for each respondent.
The higher the score, the greater the satisfaction.
Marriage_flanniness

Ceneeetnel_ne11nitien; The degree of contentment the
respondent reports with the husband/wife relationship.

Qnenetienel__flejin1tien; Marriage happiness is
measured by a seven point scale from extremely unhappy to
extremely happy. The respondents report their own
perception.
Eamilx_Size

Ceneentuel__nejinitien; Number of members in the
nuclear family - husband, wife, and number of children

born to wife and/or husband.

80

Onenetienel_nejinitien; All children who were born to
the wife and/or husband are considered part of the family
whether or not they are members of the household. Only
the nuclear family is considered even if extended family
members are part of the household.
Enn’mmentaLflanker

Ceneentnel_nefinitien; An environmental marker is a
signal of change in the fabric of one's life that is
received from the environment, or from personal factors,
or from interaction between environments and the person.
There is a changed perception of self and self in
relationship to the environment in some meaningful way.

l f' i ° The Environmental Marker

Index is used to indicate the number of life-changing
events that has occurred in the past two years. The
events include death of a significant other, change in
family structure, menopause, and retirement. The greater
the number of events the higher the score.
WWW

Ceneepeue1_D_e_f_j_nj_tjen; The degree of stress
generated by multiple life-changing events.

QpenafleneLQeflnjjjen; Stress pile-up is measured
by a cumulative score. The degree of the disturbance is
reported on a five point scale from not at all to
extremely. The sum of the degree of disturbance reported

81
for each life-changing event yields the score. The higher

the score the greater the stress pile-up.

W

Project NC-164 utilizes the survey method. Data were
collected from families at two points in time by a mailed
questionnaire. The first data collection took place in
February, 1983; the second, in March of 1985. The
questionnaire was developed by the research team with a
primary focus on the study of stress among rural families.
Two questionnaires, one identified as husband and one as
wife, two postage paid return envelopes to encourage
independent completion, and a letter of introduction were
mailed to each sample family. Follow-up post cards and a
second set of questionnaires were mailed to encourage
participation. All procedures and instruments were
approved by the University Committee on Research Involving

Human Subjects prior to the undertaking of the research.

W
The sample was selected randomly from a list,
provided by the Donelly Corporation (a company which
collects and distributes list of names and addresses for
businesses and organizations across the nation) of
families in the middle years of the family life cycle in
four rural counties and one city in the Standard

 

82

Metropolitan Statistical Area in the state of Michigan
(Williams & Bubolz, 1986). Middle stage families were
defined for Project NC-164 as families in which the wife
was between the ages of 35 and 65; the family must have
had at least one child. The rural country and the urban
area were selected to be representative of the state in
terms of median family income and educational attainment
of persons age 18 and older. Since the primary interest
was stress among rural families, they were over-
represented in the sample. The rural and small town
samples came from Hillsdale, Charlevoix, Alpena, and
Dickenson counties. The urban sample came from Flint. It
is recognized that Flint and the selected counties were
experiencing varying levels of economic stress at the time
of initial data collection. The final sample for the
first data collection was 546 (263 women; 243 men),
representing a 34% response rate. There are multiple
reasons for the one-third response rate; mailed
questionnaires are easy to ignore, the questionnaire was
lengthy and complex, and no payment was made to the
respondents. In addition, the researchers believe that
some respondents self-selected themselves out because they
did not fit the criteria. Over thirty persons either
telephoned the researchers or returned blank
questionnaires, stating they were elderly or young and not
middle-aged or did not have any children.

83

The second data collection was in March of 1985. The
instrument was mailed to husband and wives in 270 repeat
families. The response rate was 58.5%. This study
comprises the women who responded at both data collection
points.

Smith (1985) compared the 1983 sample characteristics
with those of the rural and urban population of the state.
Respondents had somewhat higher education and income
levels than the state average. Racial composition of the
urban sample was not proportionally representative of the
composition of the city sampled. (The rural counties had
very few racial minorities). The under-representation of
minorities, low-income, and the less formally' educated

needs to be kept in mind when making inferences.

Descriptign_gt_sore_5amnle

The core sample, all persons from the 1983 data
collection, had a mean age for women of 45.1 and 48.1 for
men (Williams & Bubolz, 1986). The mean number of years
married to the present spouse was 24. Sixty one percent
of the families had two or three children; 34% four or
more, and 5% one child.

Seventy-two percent of the sample was rural and 28%
was urban. The sample was 95% white. Religious
preferences of the respondents were: 67% women and 63% men
Protestant, 23% women and 30% men Catholic, and the

84
remainder other. Most men were employed full-time; women
were evenly distributed over full-time employed, part-time
employed, and full-time homemaker. The mean educational
level for women was 12.9 years and 13.1 years for men.
The mean family income reported by women was $32,384 and
$35,087 by men.

0 . l' E S! I S ]

The sample for this study includes the women who
responded at both data collection points, 1983 and 1985.
One hundred and fifty four women returned both
questionnaires. The study sample was not intended to be
proportionally representative of race, education, or
urban/rural classification. One of the primary interests
of this study was to explore changes in perceptions from
time one (1983) to time two (1985). Based on the 1985
data collection, 94.9% of the sample was Caucasian, 1.9%
black. Three and nine-tenths percent could not be
classified because of missing data. The mean age was 49.1
years. The mean education level was 13 years. The mean
number of children per family was 3.15. The mean years
married was 27.3. Forty nine percent (76) of the sample
lived in the open country or in a community of less than
2,500, 27.3% (42) in a community of 2,500 to 50,000, and
22% (34) in an urban area of more than 50,000. Tables 1
through 3 present frequencies for age, number of children

85
per family, and family income. There was very little
difference in the demographic for study and core samples.
Table 1 indicates the percentage of persons in five age
catagories; over 50% of the women are between the ages of
35 and 49. Table 2 displays the breakdown for number of
children in the family with the majority having two or
three children. Table 3 presents family income for the

total sample and by community size.

Table 1
Age of Women - 1985 (n = 154)

 

 

Age No. of Women Percent of Women
35 - 49 84 59.6
50 - 54 20 14.2
55 - 59 17 11.0
60 - 65 14 9.1
Over age 65 6 3.9

Missing data 13 8.4

 

86
Table 2
Family Size - 1985 (n = 141)

 

I. ' [I ' ‘I I. ' -ll 0 .' ‘ -" 0 ill I

 

1 10 7.1
2 40 28.4
3 48 34.0
4 18 12.8
5 16 11.3
6 7 5.0
7 1 .7
8 -O- -O-
9 1 .7
Table 3

Family Income by Community Type - 1985

 

 

Total Rural Town Urban

(n = 1261* Ln = 61) (n = 37) (n = 28)
Under $20,000 18.8 23.0 10.8 14.2
$20,000 - $39,999 41.4 44.0 40.5 39.2
$40,000 - $59,999 28.1 25.0 30.0 35.7
$60,000 - $79,999 7.8 7.0 8.1 10.7
$89;QQQ_end_ever 3.9 1.6 19.8 .0

 

 

* Data for those who reported income

87

Tables 4 and 5 detail factors that relate to the
women's employment and income. Table 4 presents the
employment status of women. Approximately half of the
sample is employed: full-time employed 27.9%, part-time
employed 22.7%, and nonemployed 49.3%. The nonemployed
included homemakers, retired, disabled, and those seeking
work. Table 5 presents the employment types. Table 6
gives the income of women for the total sample and by
community type.

Table 4
Employment Status of Women - 1985 (n = 154)

 

 

Employment Status No. of Women Percent of Women
Employed full time 43 27.9
Employed part time 35 22.7

Nonemployed 76 49.3

 

88

Table 5
Type of Employment - 1985 (n = 81)

 

Employment type Percent of Women

 

Professional Specialty workers

Executive, Administrative, Managers

Technicians and Related Support Workers

Sales Workers

Administrative Support and Clerical Workers
Precision Prod, Craft, Repair

Machine Operators, Assemblers, Inspectors
Handlers, Equipment Cleaners, Helpers, Laborers
Farming, Forestry, Fishing

Service Workers

22.

21.

W i-I
ONONNomNi-IN

 

89
Table 6
Employed Women's Personal Income by Community Type - 1985

 

 

Total Rural Town Urban
(n - 74*) (n - 36) (n - 23) (n - 15)
Under $20,000 70.2 75.0 69.6 60.0
$20,000 - $39,999 21.6 16.7 21.7 33.3
$40,000 - $59,999 6.7 8.3 4.3 6.7
$60,000 - $79,999
$80,000 and over 1.4 4.3

 

* Data for those who reported income

W

The research instrument utilized for this study was
developed by committee members of the Agricultural
Experiment Station regional research project NC-164, North
Central Region. The instrument was a self administered
mail questionnaire. It included a variety of components
representing the varied interests of the research team,
including a section devoted to the special interests of
the researchers from Michigan State University.

The research instrument for the 1983 data collection
was thirty-two pages in length and took approximately one
hour to complete. The instrument was pre-tested twice in
the process of development. Each state pre-tested the

90
instrument and then pooled the information to identify
problems within the instrument.

The research instrument for the 1985 data collection
was similar to that for 1983. However, some changes were
made. Questions were added concerning changes in family
events and economic status between the two data
collections. Some questions were rewarded to make the
instrument clearer for the respondent. Questions which
were confusing or not as central to the overall purpose of
the project were eliminated. For example, open-ended
questions were reduced to one (omitted questions 9, 32).
A family health scale (question 20), a coping measure
(question 31), and a conInunity service measure (Question
37) were omitted. This enabled the respondent to complete
the questionnaire in a shorter length of time.

Data from both collection points are included in this
study. However, only a portion of the data gathered will
be utilized. Only information gathered from Michigan
women who responded at both collection points will be
used. The sections of the instrument which are relevant
for this study will be presented next. I
B l' l E l' E I I l I

The Retirement Perceptions Assessment Instrument
(Appendix A) was developed by this researcher for data
collection with the Michigan sample. This instrument is
based on the 1985 questionnaire, the second data

91

collection point. Perceptions of the retirement life
stage are measured along five dimensions: control/anxiety,
residence selection, time use, adequacy of wife's
retirement income, and life continuity. Factor analysis
was used to help select the five dimensions and the
salient items for each dimension. Each of the dimensions
are measured on a five point scale with the lower score
representing strong agreement and the higher score strong
disagreement. The scales have been reversed for analysis
so that five represents strongly agree and one strongly
disagree.

Three dimensions were identified through factor
analysis. Each included two or more items. The
control/anxiety dimension contains four items and has a
reliability of .69. The reliabilities were computed using
Cronbachs' Alpha Model. The factor loadings range from
.60 to .84. The higher scores represent a sense of
control while lower scores represent greater anxiety. The
possible range of scores is from 4 - 20.

The residence selection dimension has two items with
a .87 correlation. The factor loadings were .93. Higher
scores indicate a plan to change residence. The possible
range of scores is from 2 - 10.

The time use dimension contains four items and has a
.74 reliability. The factor loadings range from .54 to
.85. Higher scores suggest a positive attitude toward

92
time use and the lower score a less positive attitude.
The possible range of scores is from 4 - 20.

Adequacy of wife's retirement income and life
continuity are single items which did not factor with
other items. On the adequacy of wife's retirement income
dimension, the higher the score, the greater the
perception of providing for retirement income from their
own work. On the life continuity dimension, the higher
score represents less change. It is believed the single
items may stand independent of the other dimensions
because of the diversity of the sample.

Not all questions were used in both time one and time
two therefore, a modification of the Retirement Perception
Assessment Instrument (Appendix B) was developed to assess
change between time one and time two data collections.
Only questions which appeared in both instruments were
used for comparison between time one and time two. The
control/anxiety dimension has three repeat items with a
reliability of .42 at time one and .56 at time two. The
residence selection dimension included two repeat items,
the correlation was .59 at time one and .87 at time two.
Two repeat items are included in the time use dimension.
The correlation was .49 at time one and . 31 at time two.
The two single items are the same at time one and at time
two.

93

H Ill 0 [E' 1! S 1

The Health Difficulty Scale (Appendix C) is a twelve
item measure. The twelve items are general difficulties
which can be a health difficulty' by itself’ or' be an
indicator of another illness. Each respondent indicates
the frequency with which each item occurs on a five point
scale from never (1 point) to almost always (5 points).
The points are added for the twelve items, the higher the
score the more health difficulties. The range is five to
sixty points. This scale, and all succeeding scales,
were the same at time one and time two.
N E '] I'E S l' E I. 5 I

This is a four item scale which measures satisfaction
with four types of relationships: (1) mother and father
with each other, (2) parents and children, (3) siblings'
relationship with each other, and (4) family relationships
in general. The scale was an outgrowth of research to
evaluate differences in perceived quality of life between
rural and urban areas. Schumm, McCullum, Bugaigis,
Jurich, and Bollman (1986) reported on the Kansas Family
Life Satisfaction Scale (Appendix D). The scale was
administered to 620 families in twelve states. Family life
satisfaction was substantially correlated with
satisfaction with quality of life (r = .46 for mothers,
p < .001) as well as with locus of control ( r = .36 for
mothers). In the reported study the scale proved to have

94
acceptable internal consistency reliability, near or above
the .80 criterion.-

Each of the four items are measured on a seven point
scale that ranges from completely dissatisfied to
completely satisfied. The points range from 4 to 28; the
higher the score the higher the satisfaction.

H . H . S ]

The marriage happiness scale (Appendix E) is a one-
item, seven-point, self report measure. The participants
describe how happy they are with their* marriage from
extremely unhappy to extremely happy. Shectman, Bergen,
Schumm, and Bugaighis (1985), in a study of 61 women in a
childbirth class, found an internal consistency
reliability of .90.

WW2;

The Environmental Marker Index (Appendix F) was
constructed from eleven items of the Family Events Scale
of the questionnaire» which was a modification of the
Family Inventory of Life Events and Changes (McCubbin &
Patterson, 1980). The Family Inventory of Life Events and
Changes (FILE) was developed as an index of family stress
in an effort to assess the pile-up of life events
experienced by the family. McCubbin and Patterson (1980)
describes FILE as a 71 item self-report instrument which
is designed to record the normative and non-normative life

events and changes experienced by a family. Some

95
infrequently occurring items are included that were
considered major stressors (e.g. death of a parent). The
overall internal reliability for FILE is .72.

McCubbin and Patterson (1980) suggested families
experience a pile-up of stress from demands and changes,
particularly in the aftermath of major role changes or
death. In this study the focus will be on those major
changes which are reflected in the losses and work-family
transition dimensions of the measure. Eleven items were
selected from the FILE measure that signal a change in the
fabric of life and may lead to a change in one's
perceptions of self and self in relationship to the
environment. The loss and work-family dimensions
correspond to those dimensions found critical during mid-
life by developmental theorists. Gould (1979) views mid-
life as a time of transitions, where one changes one's
view of self in relationship to the environment. A major
challenge is the acceptance of one's death. The reality
of the death issue is brought to consciousness by death of
significant others. Levinson (1978) emphasizes the
challenge of changing roles between generations and career
evolution. During this time of change the individual
experiences emotional turmoil and confusion.

The Environmental Marker Index can be scored two
ways. First, by adding the number of events which the
respondent indicated had happened within the last two

96

years. The greater the number of yes responses the more
events that occurred. Eleven yes responses are possible.
The second way is by creating a stress pile-up score. The
pile-up score is based on the degree of disturbance caused
by the event on a five point scale. The greater the
disturbance, the higher the stress pile-up score. The
range is 0 - 55.

W

A tape was constructed which combined the data from
women who participated in both the 1983 and 1985 data
collections. For the original data each question was
assigned a variable number, a label, a value, and column
number or numbers for the computer card. The values
correspond to the response options. The data were key
punched at The Michigan State University Computer Center.
All data were checked with the original coding forms to
insure that they had been entered accurately into the
computer.

Missing data were coded as 9; inappropriate data as
8. The study will report on all 154 participants, but
each analysis will be done on actual cases. Some persons
responded on one item and not on another which created
missing data rather than missing cases.

The variables under investigation are nominal,

ordinal and interval. The interval level data use a five

97
or seven point scale. For the purpose of statistical
analysis, the ordinal data were treated as interval data
even though the difference between scale points may not
always be the same.

For question one a Pearson Product Moment Correlation
was performed to determine the general direction and
strength of the relationships between the dependent
variables and the independent variables for all 154 sample
participants. The independent variable of employment type
is a categorical variable and will be tested in question
nine. The assumptions of a linear relationship, homo-
scedasticity, no truncation, and bivariate normality are
met. A step-wise regression analysis was performed on all
variables, only those found significant were reported. A
Pearson Correlation was performed to determine the
relationship between the independent variables and to
ascertain if there was a problem of multicollinearity.
Age and years married were found to be highly correlated
(r a .87), due to the higher coefficient for age (r 8.22)
it was entered first" Marriage happiness and family
satisfaction were highly correlated (r =.69), with family
satisfaction having the highest coefficient (r = .33).
Step-wise regression enters the strongest correlation into
the regression first, thus taking the cannon variance.
The residual, that variance which is distinct for the

second independent variable, is what is accredited to the

98
second independent variable. In each analysis the
residuals were not significant. All other correlations
were low (below r - .3).

Questions 2, 3, and 9 were analyzed using an ANOVA.
If appropriate, multiple comparison tests followed. Step-
wise regression analysis was performed for questions 4, 5,
and 8. Each dimension of the dependent variable will be
treated as a single dependent variable for regression
analysis because the dimensions of the retirement
perception instrument do not have interval properties.
Question 7 was descriptive, and frequency distributions
were created. Question 10 was analyzed using a T-test and
Question 11 using a correlation. The data analysis were
done utilizing the SPSS program. Results with p 5 .05
were considered significant.

Table 7 presents a summary of the questions,
independent variables, and analysis techniques. The
regression analysis selected out only those cases which
had complete data (n = 105). A Pearson Correlation was
conducted on the regression sub-sample to determine if the
sub-sample differed from the whole sample. Few
differences were found (see Appendix E). There were 54%
fewer nonemployed in the regression sub-sample than in the
total population. The number of employed women was 78 for
the full sample and 54 for the sub-sample. The number of
nonemployed women was 76 far the full sample and 35 for

99
the sub-sample. The total and sub-sample» means were
similar on family size, 3.15 and 3.0 respectively. The
pattern for community type was very similar' for both
samples. The difference in employment status was the
primary difference and should be kept in mind when

interpreting the results.

100
Table 7
Summary of Analysis Techniques

 

 

009in on Independentfiflnfle—Amlxsjueshnm
1 All Correlation &
Regression
2 Community type ANOVA, T-tests
3 Employment status ANOVA, T-tests,
regression
4 Income Regression
5 Perceived income Regression
adequacy
6 Age Regression
7 Environmental Frequency, means
markers
8 Environmental Regression

stress pile-up

9 Husband's employ- ANOVA
ment type

10 T-test

11 All Correlation

CHAPTER IV

RESULTS

The objective of this study was to develop a better
understanding of the perceptions of retirement among
married mid-life women (aged 35 - 65) with at least one
child. The diversity of women's economic status, family
and marriage satisfaction, work status, work type, health,
life eVents, community size, and age were incorporated
into the research design. The second objective was to
explore the impact of change on perceptions of retirement.
This is an exploratory study of the relationship between
selected characteristics of married women and five
dimensions of retirement perception; control/anxiety, time
use, residence selection, wife's retirement income, and
life continuity.

The results are presented in twelve sections. The
first section presents the means for dependent and
independent variables. The next eleven sections present
the results for the eleven research questions. The
questions are arranged from general to specific. The
first question provides an overview, questions two through
nine provide more in-depth analysis, and questions ten and

101

102

eleven deal with change between 1983 and 1985. The
statistical procedures and test statistics are detailed in
Chapter 111.
W

In Table 8, the means for dependent and independent
variables are presented to guide the reader in
interpreting the statistical findings. Means were
calculated for all dependent and all independent variables
except employment status, family income, wife's income,
community size, and husband's employment type. Dummy
variables were created for employment status, family
income, wife's income, and community size. The higher the
response number, the higher the level of the variable
under consideration. The dummy' variable 'consisted of
three to five responses. For summary data see Tables 3,
4, and 6. Employment type was a categorical variable.

Beseeecnjyeemu What is the relationship between

characteristics of mid-life married
women and their perceptions of the
retirement life stage.

A Pearson Correlation was conducted on all 154
respondents in order to take advantage of as much data as
possible. Table 9 reports the correlations which were
statistically significant. A step-wise regression

analysis of the 105 persons with complete data follows.

103
Table 8
Means of Dependent and Independent Variables, 1985

 

 

Variable Mean. N Egssible
DEPENDENT
Control/anxiety 10.33 126 4 - 20
Residence selection 5.19 126 2 - 10
Time use 15.64 129 4 - 20
Adequacy wife's
retirement income 2.35 132 1 - 5
Life Continuity 2.49 154 1 - 5
INDEPENDENT
Education 13.10 132 O - 22
Age 49.10 134 37 - 68
Perceived health 27.08 127 5 - 60
Marriage happiness 5.33 123 ' 1 - 7
Years married 27.30 144 7 - 47
Family size 3.15 141 0 - 9
Perceived income
adequacy 3.08 132 1 - 5
No. environmental
markers 1.25 149 O - 11
Stress Pile-up 4.34 154 O - 55

Family satisfaction 22.39 118 4 - 28

 

104
Table 9

Correlation of Environmental and Personal Variables
with Perceptions of Retirement (TOTAL N = 154)

Retirement Life
Control Residence T' ' ' '

Education .179 .212 -.147

(n=139
Age ge M) -.168 -.220 -.249
"81
Employment .224 .302 .411 .204 .165
status
(n=154)
Health
(n=130)
Marriage .262 .177
(Hapginess

 

Years married -.131
(n-14
Famil income .145
ns 28
Wi e's income
(n-103
Famil{5512e .297 .226 .288 .300

Communit) size -.202
(n=152
Ineome36 adequacy -.201
"=1
Environmental .207 .205
markers
(n=154)
Environ. pile-up .140
(n=154
Famil sat. .215
(n- 32)

 

r U

; o ' 2 0. i ‘I 1. ‘ .v i 00.0 < i or

All independent variables, with the exception of
employment type, were entered into a step-wise regression.
Table 10 reports those! which *were found statistically

significant.

105
Table 10

Ste -wise Regression of Perceptions of Retirement
by ontinuous Environmental and Personal Variables (n=105)

Variable Beta E E . TE Sig- T
WILL!
Family life
Satisfaction .281 .079 2.98 0.003
WON
Perception of income
adequacy -.312 .085 -3.36 0.00l
No. of environmental
markers -.185 .034 -1.99 0.048
IIME USE
Wife's employment
status .250 .062 2.63 0.009
W
None
LLEEJQNIINILLLY
No. of environmental
markers -.205 .042 -2.13 0.035

 

 

 

Family life satisfaction is the strongest predictor
of a sense of control in retirement. Perceptions of
income adequacy and the number of environmental marker
events were significant predictors for residence
selection. The greater the perceived income adequacy,
the less likely the women perceive they will move during

106
retirement. The larger the number of environmental
markers in the last year, the less likely they perceive
they will move during retirement. Wife's employment
status was a significant predictor for time use in
retirement. The full-time working wife had a more
favorable attitude toward time use. The dependent
variable of adequacy of wife's retirement income had no
significant predictors. The number of environmental
marker events experienced was a significant predictor for
life continuity. The fewer number of environmental
markers experienced the more likely life is viewed as

continuous.

Reseenen_0ne§tien_z; What is the reIationship

between community type and
eerceptions of t e retirement
ife stage7
Table 11 reports the results of the ANOVA for this
question/ No statistically significant differences were

found for community type.

107
Table 11

Analysis of Variance for Community UNFe
with Perceptions of Retirement (n-l )

 

 

WW5 F 5191—“
Control/anxiety 2 1.82 .91 .184 0.832
Residence selection 2 12.23 6.11 1.175 0.312
Time use 2 30.20 15.10 2.391 0.096
Adequacy wife's

retirement income 2 5.60 2.80 2.650 0.075
Life continuity 2 1.08 .54 .51 0.602

 

However, since some difference approached
significance, a series of t-tests were performed on
community type to determine if there were any differences
between rural, small town and urban residents.

Table 12

T-test Comparison of Rural and Small Town Women
on Perceptions of Retirement

Retirement (n=76) n-42)
E l' ] IS I n T 5.

 

 

Control 8.70 10.10 -2.13 0.035
Residence 4.26 4.55 -0.62 0.534
Time use 13.66 14.38 -0.79 0.433
Adequacy wife's

ret. income 2.03 2.02 0.01 0.992

W 1.23 -1.91 0.019

 

 

108
Persons in small towns perceived they will have a
higher sense of control and greater life continuity during

retirement than persons living in rural communities.

Table 13

T-test Comparison of Rural and Urban
Women on Retirement Perceptions

 

 

TRetirement (n=76) (n-34)
Ween T 51°91
Control 8.70 8.47 0.24 0.811
Residence 4.26 4.71 -0.68 0.502
Time use 13.44 12.44 0.90 0.371
Adequacy wife's

ret. income 2.03 1.26 2.89 0.005
Continuity 2.43 2.17 0.85 0.40l

 

The rural wife was more likely than the urban wife to
look towards her retirement plan through her work to cover
financial needs in retirement. It should be noted that
47% of the rural, 54% of small town, and 44% of urban

women were employed.

109
Table 14

T-test Comparison of Small Town and Urban
Women on Perceptions of Retirement

 

Retirement (n=42) (0'34)
We“ T 51°91.

 

Control 10.09 8.47 1.72 0.091
Residence 4.55 4.71 -0.24 0.812
Time use 14.38 12.44 1.49 0.143
Adequacy wife's

ret. income 2.02 1.26 2.73 0.008

Centjnuity 2.92 2.18 2,35 0,922

Those living in small towns perceived greater life
continuity and were more likely to perceive that their
work retirement plan was adequate than urban residents.

In summary, small town wives were unique in that they
were more likely to feel a sense of control and view life
as continuous than either rural or urban residents. Both
rural and small town women looked toward their work
retirement plan to cover their financial needs during
retirement to a greater extent than those living in an

urban residence.

Beseenen_0nestien_3 What is the relationship

between employment status
and perceptiens of the
retirement life stage?

110
Table 15

Analysis of Variance for Wife's Emplo ent Status
with Perceptions of Retirement n=114)

 

 

Bet. Eereeptjens gt :5 nsn FF Sig.
Control/anxiety 2 28.43 14.21 3.04 0.052
Residence selection 2 55.90 27.99 5.64 0.005
Time use 2 16.79 8.40 1.31 0.273
Adequacy wife's ret.

income 2 13.97 6.99 7.19 0.001
Life continuity 2 4.20 2.10 2.05 0.134

 

Significant differences were found for residence

selection and adequacy of wife's income.

control/anxiety approaches significance.

The dimension of

A series of trtests were performed to determine if

there were differences between full-time employed, part-

' time employed, and nonemployed women.

Table 16

T-test Comparison of Nonemployed and Part-time Employed
Women on Perceptions of Retirement

 

 

 

Retirement (n=76) (n=35)

r ' n n m l ine_mean .I Sig.
Control 7.84 10.57 -3.97 0.000
Residence 3.31 6.22 -5.54 0.000
Time use 11.03 15.31 -4.53 0.000
Adequacy wife's

ret. income 1.62 1.80 -0.73 0.466
Continuitv 2.18 2.97 -3.00 0.003

 

111
Women who are employed part-time perceived a higher
sense of control during retirement, were more likely to
change residence, have a more positive attitude toward
time use, and feel their life will be more continuous
than women who are nonemployed.
Table 17

T-test Com arison of Nonemployed and Full-time
Employe Women on Perceptions of Retirement

 

 

Retirement (n=76) (n-43) .
WW 1 - m .
Control 7.84 9.72 -2.79 0.006
Residence 3.31 5.00 -3.72 0.000
Time Use 11.04 16.39 -6.03 0.000
Adequacy wife's

ret. income 1.61 2.27 -2.73 0.007
Continuity 2.18 2.65 -1.92 0.057

 

Full-time employed women are more likely to perceive
a higher sense of control, a change in residence, a more
favorable attitude toward time use, and a greater sense of
adequacy of their retirement income than nonemployed
women. They also were more likely to perceive life as
continuous, but the difference was not significant at p <
.05.

112

Table 18

T-test Comparison of Part-time and Full-time
Employed Women on Perceptions of Retirement

 

 

 

 

Retirement ne35) (n-43)
WW1 . i 519...
Control 10.57 9.72 1.66 0.100
Residence 6.22 5.00 2.29 0.025
Time use 15.31 16.39 -1.70 0.093
Adequacy wife's

ret. income 1.80 2.27 -2.01 0.048
Continuity 2.97 2.65 1.35 0.18

 

Part-time employed women are more likely to perceive
a change in residence than full-time employed women. On
the other hand, full-time employed women are more likely
to look toward their employment for their retirement
financial needs than part-time employed women.

In sumary, both part-time and full-time employed
women perceived a greater sense of control than non-
employed women. Employed women, full-time and part-time,
perceive that they are more likely to move in retirement
than nonemployed women, with part-time employed women
indicating highest expectation of change in residence.
Both full-time and part-time employed women indicated a
more positive attitude toward time use than nonemployed

women. It should be noted that full-time employed women

113

were more likely than part-time employed women to perceive
a favorable attitude toward time use, but the probability
level was not significant (p <.093). Full-time employed
women are more likely to perceive their retirement income
as adequate for financial needs than both part-time and
non-employed women. Part-time employed women are more
likely to perceive life as continuous than both full-time
and nonemployed women, but the difference was significant
only between part-time and nonemployed women.

A regression was conducted for employed women, both
full-time and part-time, to examine the contribution of
working to the perceptions of retirement. Table 19
reports the results which were found to be statistically
significant.

114
Table 19
Step-wise Regression of Perceptions of Retirement for

Employed Women by All Continuous
Environmental and Personal Variables (n=80)

 

 

WM r 50. T Sig;
CQNIBQLLANXIEII

Marriage happiness .361 .131 3.51 0.000

Age -.217 .047 -2.12 0.037
’BESIDENC£_S£L£CIIQN

Perceived income

adequacy -.275 .069 -2.61 0.011

Community size .247 .061 2.34 0.022
IIME_US£

Age -.252 .104 -2.42 0.018

Family satisfaction .286 .082 2.84 0.005

Employment status .208 .040 2.00 0.049
ADEQUACI_HIEELS_BEIIB£M£NI_INCQM£

Community size -.247 .051 -2.30 0.024

Family size -.232 .054 -2.16 0.033
LIE£_£QNIINUIII

Personal income -.250 .062 -2.30 0.024

 

Marriage happiness and age were significant
predictors of sense of control for employed women.
Employed women perceive more control as marriage happiness

increases. However, when all women are included, family

115
satisfaction is a better predictor than marriage happiness
(see Table 9). In addition, younger employed women
perceived more control at retirement than older employed
women.

Income adequacy and community size were significant
predictors of residence selection for employed women.
Employed women who perceive their income as inadequate are
more likely to expect to change their residence. Employed
urban women perceive a greater likelihood of moving than
small town or rural women.

Age, family satisfaction, and employment status are
significant predictors for positive attitudes toward time
use for employed women. Younger employed women perceived
they would have more time in retirement than older
employed women. Employed women who perceived higher
family satisfaction indicated a more favorable attitude
toward time use than those with lower family satisfaction.
Full-time employed women had a more positive time use
attitude than part-time employed women.

Community type and family size were significant
predictors of adequacy of wife's retirement income for
employed women. Employed women who live in smaller
communities perceive their retirement income as more
adequate than urban women. Employed women with smaller
families perceive their retirement income as more adequate

than women with larger families.

116

Wive's employment income was a significant predictor
of life continuity for employed women. Employed women
with lower personal income perceive life as more
continuous than women with higher incomes.

A regression was performed for; nonemployed women.
The dimension of adequacy of wife's retirement income was
omitted. There were no significant predictors of
control/anxiety or time use for nonemployed women. Table
20 reports the predictors that were statistically _
significant.

Table 20
Step-wise Re ression of 1Retirement Perceptions

for Nonempleyed omen by Al Continuous Personal and
nvironmental Variables (n=43

 

 

W L Sig...

CQNIBQLLANXIEIX

None

BESIDEN£E_SEL£CIIQN
Income adequacy -.401 .161 -2.85 0.006

IIME_U§E

None

LIEE_£QNIINUIII

No. environmental
markers -.369 .136 -2.58 0.013

 

117

Income adequacy is a significant predictor of
residence selection for nonemployed women. Women who
perceive a lower level of income adequacy are more likely
to change residence.

The number of environmental marker events experienced
was a significant predictor of life continuity for
nonemployed women. The fewer environmental marker events
that occurred, the more likely nonemployed women perceived
life continuity in retirement.

The reported results for questions 4, 5, and 6 are
derived from the Pearson Correlations and regressions
which are reported in Tables 9 and 10 (pages 104 S 105).
The questions relate to income and age.

Beseenen_0ne§tien_1 What is the relationship
between the portion of amily

income earned by women,
personal income, or family
income and their retirement
perceptions?

Family income was significantly correlated with time
use (r s .145, p <.05). The greater the income, the more
positive the attitude toward time use. Regression
analysis did not find family income, nor; proportional
income, to be significant predictors for any of the

dimensions of retirement perceptions .

 

118
Beseaneh_0uestien_§_, What is the relationship

etween perceived ineome
iflfiqiéiiréigniiiiifitéiQZeif
Perceived income adequacy was significantly
correlated with residence selection (r- .20l, p <.009).
The lower the perceived adequacy, the more likely the
women perceived they would move from their home or
community during retirement. Perception of income
adequacy accounted for .085 of the variance for residence

selection. (r. sq. .085, T = -3.36, p <.001).

Reseenen_0ne§tien_§ What is the relationship

between.age and retirement
perceptions?

Age had a significant correlation with three
dimensions: control (r - -.168, p <.023), residence
selection (r - -.220, p <.004), and time use (r - -.249,
p <.001). As age increased, women perceived a decreased
sense of control, a decreased likelihood of changing
residence, and less positive attitude toward time use
during retirement. The regression analysis did not show
age as a significant predictor for any of the dependent

variables.

119
Reseenen_0ne§tien_z Which environmental markers

do women report as most
stressful?

A frequency distribution of the environmental
marker events which the women experienced in the previous
year was generated.
Table 21
Fre ency Distribution of

u
Environmentlii Markers Mong Women Between
983 - 1985 (N - 154 )

 

EnxjmmnentaLHaLkeLmnt—___mguencv

Marital separation occurs 3

 

(Divorce - separation)

Widowed . 2
Marriage of a family member 24
Family member moves out of home 41

(for independence, school, or marriage)

Major wage earner retires from work 5
Death of husband's of wife's parents 13
Death of brother or sister 4
Death of close friend or confidant 12

Aged parent(s) become seriously ill or

disabled requiring direct care 18
Member experience menopause 32
Aged parent committed to institution or

or placed in nursing home 4

120

The most frequently cited event was a member moving
out of the home, the least cited event was widowhood.

Mean stress for each of the environmental marker
events was calculated. The calculations were done for the
participants who indicated the event had occurred. There
was a possible range of 1 - 5 for each event.

Table 22
Mean Stress for Environmental Marker Events

 

 

Enyinennental Markers N. Mean St;Qe1;__
Death of sibling 3 4.33 .577
Death of family 15 4.20 .862

member
Death of parent 12 4.16 .937
Ill parent 18 4.05 .938
Marital Separation 2 4.00 .000
Death of close friend 10 4.00 1.054
Widowed 2 4.00 .000
Parent

institutionalized 3 3.25 1.700
Wage earner retires 5 2.60 1.140
Member moves 39 2.59 1.200
Menopause 30 2.57 .858
Marriage family 24 1.67 .868

member

 

121
Death of a significant other was the most stressful
event; and marriage of a family member was the least

stressful.

Resenren_0ue§tien_8 What is the relationship between

325$§2t3?"§tr§2§"?ieffiiiiiii“t ...
environmental mar ers?

Stress pile-up was calculated by measuring the
reported stress level for each event. A range of 0 - 60
was possible. Time use had a significant correlation with
the degree of stress pile-up from environmental markers
(r - .137, p <.042). Women with higher stress pile-up
indicated a more favorable attitude toward time use in
retirement than women with lower stress pile-up.
Regression analysis did not find stress pile-up to be a
significant predictor of perceptions of retirement.

The number of events experienced was a significant
predictor for change of residence and life continuity.
Women who reported fewer of environmental marker events
were more likely to indicate that they would change
residence during retirement and were more likely to
perceive life as continuous.

Beseeren_0uestien_2_ What is the relationship

between perceptions toward
retirement and husband's
employment type?

122
An Analysis of Variance was performed to examine the
relationship between husband's occupation type and

perceptions of retirement.

Table 23

Analysis of Variance for Husband's Occupation
Type with Perceptions of Retirement (n=106)

 

WHOM df 55 ms F 519..

 

Control/anxiety 5 72.36 14.47 0.87 0.499
Residence selection . 5 106.32 21.26 3.25 0.009
Time use 5 232.84 46.56 1.35 0.251
Adequacy wife's

retirement income 5 14.38 2.87 1.58 0.171
L11e_eentinuitv 5 5-36 1.07 0.52 0.759 '

 

Using Analysis of Variance, it was found that only
residence selection varied by husband's occupation. An
examination of the means found that women with husbands in
a sales occupation, followed by those in administration,
indicated they were more likely to change residence during
retirement. When the husbands occupation was professional
specialists, or in production, they were less likely to
change residence.

To add another dimension, women's occupational
patterns were also examined. Occupational type approached
significance for residence selection (p <.059). An
examination of the means indicated that women in sales and

123
management were more likely to move in retirement. Women
in service work were least likely to more in retirement.
Occupational type was significant for life continuity
(p <.005). Women in clerical or administrative assistant
positions were found to perceive more life continuity
during retirement than women who were professional

specialists or in sales.

Beeeenen_0neetien_10 What changes in retirement
perception did the women
experience between 1983
and 1985?

Table 24 reports the results of the T-tests of mean
differences between 1983 and 1985 for perceptions of
retirement. The mean differences indicate change from
1983 to 1985 for the perceptions of retirement. The time
use dimension was measured on the change of perception of
time for recreational activities. The number of
respondents for the other items in time one were too few

to be used in the statistical analysis.

124
Table 24

T-test of Mean Differences Between 1983 and 1985
for Perceptions of Retirement

 

i“ ‘ll‘l " ., .-.| ': l‘.[ ': I .

Coetrol{39§iety 8.72 10.36 +1.65 0.000

n.

Reeidenge selection 5.58 5.15 -0.43 0.213
"8

Time use 4.56 4.08 -0.48 0.000
(n - 115)

Adequacy wife's
retiregent income 3.60 2.35 -1.25 0.000
"3

Li e continuity 2.92 2.34 -0.57 0.000
(n = 154)

 

From 1983 to 1985 there was an increase in sense of
control. However, women were less likely to perceive a
change in residence, had a less positive attitude toward
time use, perceived the wife's retirement income as less

adequate and life as less continuous.

Beseenen_0ne§tien_11 What is the relationship

between the characteristics and
changes in perceptions of the
retirement life stage?

A change in the perceptions of retirement was
calculated by subtracting 1983 from 1985. Change in
perceptions of sense of control in retirement was
correlated with education (r - -.19, p < .049),
employment status (r - -.127, p < .01), health (r = -.19,
p <.045), marriage happiness (r = .34, p < .001) and

family satisfaction (r = .34, p < .002). An examination

125
of the means indicated that the lower-educated,
nonemployed, healthiest, most happily married, and those
with higher family satisfaction had the greatest increase
in their sense of control between 1983 and 1985.

No significant correlations were found between the

independent variable and change in residence selection.

Employment status correlated with change in
recreational activities (r - .212, p < .037). The
employed women had the most change, and expected less time
for recreational activities.

Age correlated with change in adequacy' of’ wife's
retirement income (r -.246, p < .028). As age increased,
perceptions of wife's retirement income adequacy
decreased. 'The relationship between stress pile-up from
environmental markers to change in adequacy of wife's
retirement income was close to significant
(r = .206, p < .055). The greater the stress pile-up, the
less likely the respondents perceived the wife's income as
adequate.

Change in life continuity correlated with health
(r - -.224, p < .13) and family satisfaction (r = .178, p.
< .042). Those who perceived a greater number of health
difficulties had the greatest decrease in sense of life
continuity. Respondents who increased the most in family
satisfaction had the greatest increase in a sense of life

continuity.

126
The discussion of the findings is contained in the
final chapter. Conclusions, implications, and a summary

follow the discussion of findings.

CHAPTER V

DISCUSSION, CONCLUSIONS, IMPLICATIONS, AND SUMMARY

Based on a modified conceptual model of person-
environmental interaction developed by Parr (1980), this
research sought to determine the contribution of personal,
environmental, and interaction characteristics to the

perceptions of retirement for mid-life married women.

0' .

The discussion is presented in two sections. The
first section will review the five dimensions of
perceptions of retirement and summarize the results. The
second section will focUs on the findings for each of the
eleven research questions.

The discussion of the dependent variables and
question one will be based on the findings of the Pearson
Correlation and step-wise regression analysis. Question
one will focus on the independent variables for which
further analysis was not performed. It should be noted
that there was a discrepancy between the findings of the
of Pearson Correlation and the step-wise regression. The

Pearson Correlation took advantage of as much data as

127

128
possible and included all 154 respondents. The regression
analysis selected out only those cases which had complete
data (n = 105). The difference in employment status was
the primary difference and should be kept in mind when
interpreting the results.

Questions two through nine will reflect the findings
of more aggressive analysis of selected independent
variables. Questions ten and eleven will address the
findings for change between 1983 and 1985.

Beneeptjens 9: Retirement

The Retirement Perceptions Assessment Instrument is
comprised of five dimensions: control/anxiety, residence
selection, time use, wife's retirement income, and life
continuity. The first dimension to be addressed is
control/anxiety.

Centnellenxiety: Factor analysis found four salient
items measured the control/anxiety dimension. Three of
the items related to the degree of locus of control: 1) I
am comfortable thinking about retirement. 2) I have a
sense of control over my future. 3) I am looking forward
to retirement. The fourth item related to the spouse: 4)
My retirement needs will be covered by my spouse. It is
not surprising that the husband's ability to provide
should factor with the wife's sense of control. All women
in the sample were married, and 72% were either
nonemployed or part-time employed.

129

The Pearson Correlation indicated that family size
had the highest correlation with control, the larger the
family the greater the perceived control. However,
regression analysis found family life satisfaction to be
the only significant predictor, with family satisfaction
explaining eight percent of the variance. Since women
have a managerial role within the family, those with
larger families may play a larger managerial role.
Control is a component of management. Kevitts, Watson,
and Busch (1977) noted that those who perceive an
opportunity to control others feel a greater sense of
internal control. The family role may be primary in the
sense of control for women.

Other variables found to correlate positively with
sense of control were marriage happiness, employment
status, family satisfaction, and the number of
environmental markers experienced. A negative
relationship was found between age and control, the lower
the age, the greater the perceived control in retirement.
All but age and employment status are family variables.
Environmental markers are primarily family events.

This study did not find that women perceived a strong
sense of control during retirement, the overall mean was
10.33 out of a possible 20. Kevitt, Watson, and Busch
(1977) linked a positive sense of control with greater
overall life satisfaction and self-esteem. A low sense of

130

control was related to depression. Steitz (1979) found
mid-life females had a greater sense of control than
either teenagers or elderly. The data suggest that married
women experience a relatively low level of self control
across the life span. This finding has implications for
the emotional and economic well-being of women. The issue
of control tends to be a gender issue. Seccombe and Lee
(1986) noted that antecedents for well-being are similar
for both men and women, however, women have a lower level
of well-being. The socialization of women is reflected in
their work patterns, work type, decision making ability,
and the control they feel they have over their
environment.

Besidenee__5e1eetien; The residence selection
dimension contains two items; 1) I will move to a
different home. 2) I will move to a different community.
The Pearson Correlation found changing residence to be
most positively related to employment status, with those
working full-time the most likely to move. Perceived
income adequacy was negatively correlated with
anticipation of moving. Regression analysis supported the
findings that the less adequate the women perceived the
family income, the more likely they thought they would
move in retirement. Perceived income adequacy explained
nine percent of the variance in residence selection. The
data suggest that changing of residence in retirement is,

131

at least in part, an economic decision. Possibly, women
working full time are doing so from economic need and feel
they will not be able to stay in their current resident
during retirement. Or, conversely, their economic
condition may allow them to move. A curvilinear
relationship between economic adequacy and change of
residence may exist.

The number of environmental marker events experienced
explained an additional three percent of the variance.
The fewer the number of events, the more likely the women
perceived themselves as moving during retirement.
Possibly those persons who had experienced few' major
changes in their current life were looking forward to
change in retirement. It also could be related to the
negative correlation between age and residence change, as
age decreased the likelihood of changing residence
increased. Because of the nature of the environmental
events they are more likely to be experienced by the older
persons. The older person who is experiencing major life
events may be seeking stability, while the younger person
seeks change.

IjniLuse; The time use dimension was a cluster of
four items: 1) I will have more time for recreation. 2) I
will start new hobbies. 3) I will enjoy having more free
time. 4) I will have plenty to do with my time. Six
percent of the variance was explained by employment

132

status. The full-time employed had the most positive
attitude toward time use; the nonemployed had the least
positive attitude toward time use. Those who work full
time perceive they will have more time for extracurricular
activities in retirement, while the nonemployed feel
things will remain basically unchanged. It should be
noted, however, that the overall mean (see Table 8) was
15.64 out of a possible 20, indicating a generally
positive view toward time use in retirement. Fox (1977),
Keith (1982), and Depner and Ingersoll (1982) found that
both retired women, and those who were nonemployed, were
active in organizations and social activities in
retirement. The perception and the reality seem to be in
agreement. .

Other contributing variables to a positive attitude
toward time use were family size, education, number of
environmental marker events, marriage happiness, family
satisfaction, and environmental stress pile-up. Age had a
negative relationship, the older were less positive.

W The fourth
dimension, adequacy of wife's retirement income, was a
single item: The retirement plan through my work will
cover my financial needs. Employment status was
positively related, the full-time employed were more
likely to feel their needs would be covered by their own

work than the part-time employed. The regression found no

133

significant variables. The mean for this variable, 2.35
out of a possible 5, suggests that the majority of women
did not perceive that retirement income from their own
earnings would be sufficient to meet retirement needs.
Married women's work patterns tend toward interrupted,
while single women have a more continuous pattern
(Keating & Jeffrey, 1983; Sorenson, 1983; and Lopata &
Norr, 1980). Married women leave the labor market for
marriage, pregnancy, and husband's transfer. O'Rand and
Landerman (1984) noted that interrupted work patterns had
economic consequence for retirement income. This may
change with younger cohorts of women.

Ljfe Centjnujty; The life continuity dimension was a
single item: My life will remain basically the same as it
has been. Family size and employment status positively
correlated with life remaining the same. The respondents
with larger families indicated life would be continuous.
Those with larger families would presumably be more
involved in family activities, and those would continue
over time. To a lesser degree those who were full-time
employed perceived life as continuous. Possibly those who
are employed perceive they will be able to maintain the
same standard of living in retirement.

As education increased, perceived life continuity
decreased. The educated person may be better able to

perceive a wide range of options, and more willing to

134
attempt something new.

The regression analysis found the fewer number of
environmental markers the respondent experienced to be a
predictor of life continuity, explaining four percent of
the variance. The fewer the environmental markers, the
greater the expectation of continuity. Dohrenwend (1973)
noted that the occurrence of' each life event usually
evokes, or is associated with some adaptive or coping
behavior. The greater the number of events, the greater
the amount of change. A negative cycle can be created:
change leads to stress, stress to risk of illness, and
illness to change. Death of a significant other or a
change in the family structure requires adaption to a new
pattern of behavior. Current change may be projected into

the future to some degree.

Remunjuesttons

Research Question l: What is the relationshi between
characteristics of mid- ife married
women and their perceptions of the
retirement life stage?

Fifteen independent variables are included in the
study, they are items which the literature review
suggested as salient in the formation of perceptions of
retirement. Not all fifteen items were of equal interest
to the researcher. This discussion will address those

variables which were not the focus of later research

135
questions: education, health, marriage happiness, family
size, and family satisfaction.

muen; Education was correlated with attitudes

toward time use, residence selection, and life continuity.
Education had the strongest correlation with perceptions
of positive time use (r = .2115, p < .006). The educated
may have a greater interest in hobbies and recreational
activities. There was a positive relationship between
education and change of residence (r -.179. p < .018).
A negative relationship was found between education and
life continuity (r = -.147, p < .042). Change was a
common factor in changing of residence and lower perceived
life continuity. The educated may be more open to
different options. All the correlations for education are
weak; and this is reflected in the lack of findings for
education in the regression analysis.

1131110.; Health was not correlated with any of the
perceptions of retirement, nor did the regression analysis
find it to be a significant predictor. The health
difficulty instrument was limited to general symptoms, and
did not include questions about major or chronic illness.
The lack of relationship may be related to the limitations
of the instrument. The mean age for the sample was 49.1
years; the lack of findings for health may also reflect a

healthy sample. The mean perceived health difficulty

136
score was 27.08, with 60 indicating a greater number of
health difficulties.

Ingeniege__fleee1ne§§; Marriage happiness was
positively correlated with control (rs .262, p < .00l) and
a positive attitude toward time use (r - .177, p < .019).
Regression analysis did not find marriage happiness to be
a significant predictor for perceptions of retirement. As
noted earlier, there is a strong relationship between
marriage happiness and family satisfaction, with family
satisfaction accounting for the majority of the variance.
Marriage happiness accounted for the remaining variance.

Eamjly Satjsfaetjen; Family satisfaction was
positively correlated with sense of control (r - .215,

p <.007), and accounted for eight percent of the variance.
Although amount of variance explained was not large, the
finding remained Stable across the total sample and the
regression sub-sample. Women with higher levels of family
satisfaction had a greater sense of control. These data
highlight the centrality of family for' women. Women
report more family events and a broader family network
than men (Williams S Bubolz, 1986; Seigler 8 George, 1982;
and Kessler & McLeod, 1984). Family factors, such as
relationships, family structure, family income, and family
roles are related to women's stress level and family
satisfaction. Walker, LS. and Walker, J.L. (1987)

identified several control issues for women: division of

137

labor, isolation from decision making, and whose values
predominate.

Eamjly Size; For the total sample, family size was
positively correlated with sense of control (r = .297,
p < .000), change in residence (r - .226, p < .002),
attitudes toward time use (r - .288, p < .000), and life
continuity (r . .300, p < .000). The regression analysis
did not find family size to be a significant predictor for
perceptions of retirement. It should be noted, however,
that the regression sub-sample had fewer nonemployed women
than the total sample, and family size may have been a
factor for nonemployed. For the sample population family

size was related to retirement perceptions.

Research Question 2: What is the relationship between
community type and erceptions of
the retirement life 5 age?

Although analysis of variance did not find community
type to be significant for perceptions of retirement,
t-tests were performed to determine if there were
differences between rural, small town, and urban.
Differences were found, the small town wives were more
likely to feel a sense of control and view life as
continuous than either the rural or urban residents. Both
rural and small town women looked toward their work
retirement plan to cover financial needs during retirement

to a greater extent than those living in an urban

138
residence. Some element of the small town is unique;
small town women perceive a higher sense of control, view
life as more continuous, and feel their retirement income
will more adequately meet their needs. One underlying
factor may be the economic well-being of those who live in
small town. The income may be sufficient to allow a sense
control, maintain a similar life style, and cover needs
from the wife's income. In general, small town women had

higher incomes.

Research Question 3: What is the relationship between
employment status and perceptions of
the retirement life stage?

The wife's employment status was significantly
related to sense of control, residence selection, and
adequacy of wife's retirement income. The results of the
t-tests highlight the differences between full-time, part-
time and nonemployed women. Nonemployed women perceive
less control, were less likely to change residence, were
less likely to feel their own retirement income as
adequate to meet retirement needs, viewed life as less
continuous, and have a less positive attitude toward time
use than either full or part-time employed women. Full-
time employed women are more likely to find their
retirement income adequate than part-time employed women.
Part-time are more likely to move than full-time employed

women. There were no significant differences between full

139
and part-time employed women on sense of control, attitude
toward time use, or life continuity. It is interesting
to note that the employed woman perceives she will be more
likely to change residence, but views life as more
continuous than the nonemployed woman. This would suggest
there is another factor other than residence influencing
life continuity in retirement.

A regression analysis was performed for both employed
and nonemployed women to examine the contributions made to
retirement perceptions by the independent variables.
These contributions will be discussed for each of the
dimensions of retirement.

genteel; Marriage happiness was a significant
predictor’ of sense of control for' employed women; it
explained thirteen percent of the variance. Respondents
perceived a greater sense of control as marriage happiness
increased. Regression analysis did not find a significant
predictor of sense of control for the nonemployed women.
Given that the regression sample has fewer nonemployed
women than the full sample used for the correlations, the
correlations may be a more reliable finding for the
nonemployed. The correlation found that sense of control
correlated with family satisfaction.

The concept of locus of control is related to
overall life satisfaction. Jackson (1979) found family is
the domain which yields the greatest amount of

140

satisfaction and is the strongest indicator of perceived
overall well-being. Family satisfaction is a stronger
predictor of life satisfaCtion for women than men. Family
satisfaction ‘was the' strongest predictor regardless of
employment status; however, a measure of marriage
happiness was not included in the Jackson study. Kivett,
Watson, and Busch (1977) concurred that women with higher
sense of control have higher overall life satisfaction.

Fox (1977) found that housewives had a higher level
of life satisfaction in retirement than retired women;
conversely Jaslow (1976) found retired women had higher
life satisfaction. The ambiguity of the findings may be
related to different influences on sense of control for
employed and nonemployed women. These findings suggest
life satisfaction is related to quality of marriage for
the working women and family satisfaction for nonemployed
women. The literature provides insights to this
difference. Fox (1977) found retired women had as many
friends as those who were nonemployed, but perceived
themselves as more socially isolated. Dorfman and Moffett
(1987) noted that having close friends from work was
strongly related to satisfaction for widows, but
negatively related for married women. A possible
interpretation is that during the working years women were
free to have social contacts within the work setting.
However, during retirement they may feel they should spend

141
time with their husbands and perceive social interaction
restricted.

Atchley (1982) noted another possible autonomy
issue: women were more likely to retire if married, but
retirement did not lead to a positive attitude. Levy
(1981) found that retiring unwillingly, even after years
passed, was significantly associated with a lack of sense
of the future and with poor adjustment. For working
women, a sense of control may relate to egalitarian
decision making within the marriage, and be a greater
influence than overall family satisfaction.

Family satisfaction is important for both employed
and unemployed women; however, family solidarity may be a
primary issue for nonemployed women. Walker, L.S. and
Walker J.L. (1987) in a study of farm wives, found
conflict with spouse, family communications, lack of team
work, and intergenerational conflict related to stressful
family life. Control may be related to harmony in the
home work place for nonemployed women.

Age was a significant predictor of sense of control
for employed women. The younger women perceived a higher
sense of control. The younger women may be more likely to
have a more modern marriage with more equal decision-
making powers and more equal division of labor. Age

explained five percent of the variance.

142

fleeinenee_§e1eetien;_ Perceived income adequacy was
a significant predictor of change of residence for both
the employed and nonemployed women. The lower the
perceived income adequacy, the more likely they expected
to change residence. Perceived income adequacy explained
seven percent of variance for the employed women, and
sixteen percent for the nonemployed women. Lower incomes
may not have been perceived as adequate to maintain
current living arrangements.

Community type explained another six percent of the
variance for the employed; those who live in urban
conmunities perceived a greater likelihood of changing
residence than those in small towns or rural communities.
This suggests that employed urban women may perceive their
income as least adequate to maintain their current
residence, or may have a greater desire to move.

Iine_!.Lse; No significant predictors of attitudes
toward time use were found for nonemployed women. Age,
family satisfaction, and employment status were
significant predictors of time use for employed women.
The younger employed women were more likely to perceive a
positive attitude toward time use in retirement than older
employed women. Age explained ten percent of the
variance. Employed women with higher family satisfaction
had a more positive attitude toward time use. Family

satisfaction accounted for eight percent of the variance.

143
Another four percent of the variance was explained by
employment status. Full-time employed women perceived a
more positive attitude toward time use in retirement than

part-time employed women.

Admacuflfiefijetinemenunme; Only employed
women were included in the regression for adequacy of
wife's retirement income. Community size was a
significant predictor for adequacy of their retirement
income for the employed. Women in smaller comunities
perceived a more adequate retirement income than those
living in urban comunities. Urban women may expect a
greater disparity between income and needs. Conliiunity
size explained five percent of the variance.

Family size was a significant predictor for adequacy
of wife's retirement income. Women with smaller families
perceived their income as more adequate than those with
larger families. Larger families require more resources
for support than smaller families. Family size explained
five percent of the variance.

Life_§entinnity: The predictors were different for
employed and nonemployed women. The wife's employment
income was a significant predictor of life continuity for
employed women. Employed women with lower incomes
perceived life as more continuous than women with higher

incomes. It is presumed the part-time worker would earn

144
less, may be less dependent on their earnings and
therefore may feel there would be less change in life-
style. Personal income explained six percent of the
variance.

The number of environmental marker events experienced
was a significant predictor of life continuity for
nonemployed women. The fewer environmental marker events
that occurred, the more likely the nonemployed women
perceived life continuity in retirement. Environmental
markers are indicative of major life changes. The changes
can impact on future family structure and functioning.
The number of environmental marker events explained
fourteen percent of the variance. In general, a greater
number of independent variables explained a portion of the
variance in the perceptions of retirement for the employed
women than for the nonemployed. Personal and
environmental characteristics tended to be salient
predictors of retirement perceptions of employed women;
the interaction characteristics tended to be significant
for the nonemployed women. The personal and environmental
variables focus on income, community and personal
demographics, whereas the interaction variables were
family characteristics. Perceptions were drawn from

different contexts.

145
Research Question 4: What is the relationship between the
portion of family income earned by
women, personal income, or family
income and their retirement
perceptions?

The income variables were weakly correlated with
retirement perceptions. There was a significant
correlation between a positive attitude toward time use
and family income (r - .145, p < .05). The greater the
family income, the more positive the attitude toward time
use. There is the implication that increased income
allows for more activities. Regression analysis did not
find family income, proportion of the family income earned
by the wife, or the wife's personal income significant for
any of the dimensions of retirement perception.
Perceptions of retirement are related. to something other

than actual income.

Research Question 5: What is the relationship between
perceived income adequacy and
percegtions of the retirement life
stage.

Perceived income adequacy was a significant predictor
of residence selection (r sq. .085, T . -3.36, p < .001).
The lower the perceived income adequacy, the more likely
the women perceived they would move from their home or
community during retirement. Perceived income adequacy
accounted for .085 of the variance for residence
selection. These findings suggest that those who perceive

their income as inadequate anticipate having to change

146
residence. It could be surmised that their current

residence is too costly to maintain as a retirement home.

Research Question 6: What islthe relationship between age
and retirement perceptions?

Age had a significant negative correlation with three
dimensions; control (r - -.168, p < .023), residence
selection (r - -.220, p < .004), and time use (r - -.249,
p <.001). The strongest correlation was between age and
time use; as age increased the women had a decreasingly
negative attitudes toward time use. As age increased the
women perceived they would be less likely to change
residence and have less control. The correlations were
weak and regression analysis did not find any significant
predictors for age. lit can be inferred that perceptions

of retirement are not strongly associated with age.

Research Question 7: Which environmental markers do women
report as most stressful?

Those environmental marker events which were reported
the most frequently (member moves out (39), menopause
(30), and marriage of a family member (24)) had the lowest
reported stress levels. The reported stress level was
between "not at all" and "slightly“. Marriage of a family
was reported as the least stressful environmental marker
event (mean 1.67). Boss et al (1987) found mothers felt
some loneliness when children first left home, but then

147
decreased with time. Normal family life events are not
generally perceived as stressful.

Those environmental marker events which were reported
least frequently (death of significant others) were
reported as 'quite' to “extremely“ stressful (mean
4.33 - 4.00). Death of a sibling was the most stressful
event, and death of a parent most frequently reported
event (13). Death of a spouse was the least frequently
reported event (2). Kessler and McLeod (1984) concurred
that loss of a loved one is the most stressful event.
However, women seemed to be able to adjust better than men
to the loss of a spouse. The stress level for loss of
spouse was 4.00 in this study. (5.0 was the highest
possible score.)

Research Question 8: What is the relationship between
perceptions of retirement and the
degree of stress pile-up and the
the number of environmental markers?

A positive correlation was found between the degree
of stress pile-up and a positive attitude toward time use

(r = .137, p < .042). Women with higher stress pile-up

indicated a more positive attitude toward time use in

retirement than women with lower stress pile-up.

However, the correlation was weak and regression analysis

did not find stress pile-up to be a significant predictor.

The overall level of stress did not appear to predict

perceptions of retirement.

148

The number of environmental marker events experienced
was a significant predictor of change in residence
(T - -1.99, p < .048) and life continuity (T = -2.13,
p < .035). Women who reported a fewer number of
environmental marker events were more likely to indicate
they would change residence and perceived life as more
continuous. The number of environmental events
experienced had a greater influence on perceptions of
retirement than the amount of’ accumulated stress; 3.4
percent of the variance for residence selection and 4.2
percent of the variance for life continuity could be

explained by the number of environmental marker events.

Research Question 9: What is the relationship between
£33235?2"§m3io§i§i{i'€§3§f"d

The wives' perceptions of retirement was tested based
on the husband's employment type. Husband's employment
type was significant for residence selection. Husbands
who were employed in sales, followed by those in
management, were more likely to change residence in
retirement. Husbands whose occupations were professional
specialist, or in production, were least likely to change
residence. A similar pattern was found for women's
employment type. Employment type approached significance
(p < .059) for residence selection. Women employed in

sales and management were most likely to perceive a change

149
in residence during retirement. Women employed in service
work were least likely to perceive a change in residence.
Lower perceived income has consistently correlated with
change of residence, possibly those in sales and
management perceive a lower level of income.

Employment type was found significant for life
continuity of women. Women in clerical or administrative
assistant positions perceived life as more continuous;
while women in professional specialist or sales positions
perceived life as less continuous. Travel and/or overtime
may be associated with sales and professional specialist
positions. The end of employment could make noticeable

changes in life patterns.

Research Question 10: Wgelzrgll‘iggeeefiijn thpeercIeci’ineeilenet):f
between 1983 and 1985? perience
There were significant changes in retirement
perceptions between time one (1983) and time two (1985)
even though it was only a two-year period. The
respondents had a greater sense of control, a less
positive attitude toward time use, perceived their
retirement income as less adequate, and perceived life as
less continuous in 1985 than in 1983. In 1983, Michigan
was in an economic recession, the unemployment rate in the
five counties of the sample population ranged from 11.9 to

18.2. The national unemployment rate was 9.9. In 1985,

150
the counties had an unemployment rate that ranged from
ll.3 to 15.5. The national unemployment rate was 7.2
(W, 1986-87). The economic
conditions may have changed enough to allow a greater
sense of control, a more stable economy tends to support a
more stable family life. Walker, Borland, Bubolz,
Cameron, and Dulmes (1986) examined the data of the NC-164
project, and concurred that women's family life
satisfaction and perceived income increased between 1983
and 1985. However, the changes may not have been great
enough to support a sense of continuity, or perception of
adequacy of the wife's retirement income, or of the
ability to use retirement time for recreational
activities. In 1985, the unemployment rate was still
considerably higher than the national average» ‘There
could have been a sense of vulnerability concerning the

future.

Research Question 11: What is the. relationship between
the characteristics and changes in
perceetions of the retirement life
stage.

CentnellAnniety: ‘The strongest correlations with
change in sense of control were family satisfaction,
marriage happiness, and employment status. Those who were
least satisfied with their marriage and family increased
the most in sense of control. The nonemployed had the
greatest increase in sense of control. This speaks to the

151
hypothesis that family life is more satisfactory when the
economy is more stable. Possibly the nonemployed wives
feel the greatest anxiety when time are uncertain.

Iime_flee; The change for perceptions of time use was
measured on one item; more time for recreational
activities. The employed women had the most change,
perceiving less time for recreation. There is the
possibility they feel they may need to continue to work
longer.

AW Aoe correlated
with change in perceptions of the wife's retirement
income, all respondents decreased in sense of adequacy.
The recession may have raised their awareness of needs and
what is required to meet those needs.

ije_Centinnitx; Those who perceived their health as
less satisfactory changed the most in their perceptions of
life continuity, life was seen as less continuous.
Herzog, Rodgers, and Woodworth (1982) noted that health
difficulties were related to a less free and easy life. A
decline in health can restrict both activities and
relationships. Conversely, those who experienced the
greatest increase in family satisfaction had the greatest
increase in sense of continuity. Family satisfaction
seemed to have a stabling effect.

152

Conclnsjons
The Modified Person-Environmental Interaction Model
is useful in exploring perceptions of retirement. A
diverse mix of personal, environmental, and interaction
characteristics influence retirement perceptions
(control/anxiety, attitudes toward time use, life
continuity, residence selection, and adequacy of wife's
retirement income) of women. While the total percentage
of variance in retirement perceptions explained by the
independent variable in this study was not large, several
significant relationships were found. The conclusions
that can be inferred from this research are:
Education
0 Education level is a weak predictor for
perceptions of retirement.
Age
0 The older respondents had a more negative
perception of retirement. The sense of control,
positive attitudes toward time use, and
anticipation of changing residence decreased as age
increased.
Employment status
0 Perceptions of retirement and their antecedents

differ by employment status.

153

0 Personal and environmental characteristics are the
most frequent predictors for perceptions of
retirement for employed women and interaction
characteristics for nonemployed women.

0 Nonemployed women have the most negative
perceptions of retirement. The perceptions of
retirement are more negative for all dimensions
than for full or part-time employed women.

0 Full-time employed women perceive their retirement
income as more adequate to meet retirement needs
than part-time or nonemployed women.

0 Part-time employed women feel they are most likely
to change residence in retirement.

0 When women, employed and nonemployed, are

« considered, many women have a low sense of control.

Employment type

0 Women employed in clerical or administrative
assistant positions perceived life as more
continuous.

Self perceived health

0 Perceptions of health are not strong predictors of

perceptions of retirement for mid-life married

women .

154

Marriage happiness

0

Marriage happiness is a predictor of sense of
control for employed women. Those who perceive
their marriages as most happy perceived the

greatest sense of control.

Years married

Years married is too strongly correlated with age

to be an independent predictor of perceptions of

. retirement.

E . l ] C! l . I. ,

Family

Wife's

Family

income

Family income was not a strong predictor for
retirement perceptions.

income

Wife's actual income is not a predictor for
retirement perceptions.

size

Larger family size relates to a more positive
attitude toward time use, a higher sense of
control, a view of life as continuous, and a

- change in residence. The relationship is stronger

for nonemployed women.

Community type

Rural, small town, and urban residents perceive
retirement differently. Small town residents have

155
a greater sense of control and perceive life as
most continuous. Small town and rural residents

- perceive the wife's retirement income as more
adequate than urban residents.

Husband's employment type

0 There is a greater perception of changing residence
during retirement when the husbands are employed in

. sales or management positions.

I ! I. Ch I l l' ,
Perceived family income adequacy

o Perceived family income adequacy is a better
predictor of retirement perceptions than actual
income. Those who perceive an inadequate income
anticipate changing residence.

Environmental markers

0 There is a relationship between the number of
environmental marker events experienced and life
continuity and residence selection. Those who
experience fewer events perceive life more
continuous and are more likely to anticipate a
change in residence.

0 The greater the stress pile-up from environmental
marker events, the more positive the attitude
toward time use in retirement.

0 There is an inverse relationship between reported

frequency of an event and the perceived stress

156
created by the event. Normal events (i.e., family
member moves out, family member gets married, or
menopause) are not considered stressful. Infrequent
events (death of a significant other) which
represent loss, were most stressful.
Family life satisfaction

0 Family life satisfaction is positively related to

sense of control for nonemployed women.
Change;

0 There were changes in retirement perceptions

~ between 1983 and 1985. As the women became older,
they perceived their retirement income as less
adequate. Those who experienced the most increase
in marriage happiness and family satisfaction
increased most in sense of control. The

. nonemployed had the greatest increase in sense of
control. The attitude toward time use among
the employed was more negative in 1985. Those who
perceived the most health difficulties decreased in

sense of life continuity.

ImnljsatjonsJoLBeseamn
As has been previously indicated, the intent of this
research was not that of hypothesis testing but rather the
intent was to develop initial conjectures which can serve
as the foundation for further study of the relationship

157

between aspects of the environment, personal, and
interaction characteristics and perceptions of retirement
of women. The ideas generated for further inquiry
include: hypotheses to guide further exploration of the
dimensions of retirement perceptions; antecedents of
family and marriage satisfaction, and sense of control for
employed and nonemployed women; further exploration
incorporating the entire Modified Person-Environmental
Interaction Model, and refinement and validating of the
Retirement Perceptions Assessment Instrument.
Wham

The results of this exploratory research suggest

several hypotheses for further testing:

1) Predictors of sense of control in retirement
differ with employment status.

2) Perceived control will vary with employment
status.

3) The more time spent in employment, the more
positive the attitude toward time use in
retirement.

4) The younger the employed woman, the more likely
she is to perceive her retirement income as

adequate.

158
5) Those with a low or high perceived income are
more likely to anticipate a change of residence
in retirement.

6) The fewer environmental marker events in mid-life

the more continuous life is perceived.

7) Perceptions of retirement becomes more negative

as age increases.

The testing of these hypotheses will determine to
what extent the findings of this study can be generalized
to other populations.

.o- .l. -u .’ . 'cp -. - . o. o

The findings of this study indicate that marriage
happiness predicts sense of control for employed women and
family satisfaction predicts sense of control for
nonemployed women. Although there is overlap between
marriage happiness and family satisfaction, there appear
to be differences as well. These conclusions point to a
need for further study of the differing needs of the
nonemployed and employed women that underlie the different
predictors for sense of control. The literature offers
some guidance to possible hypotheses.

1) The more freedom an employed woman perceives she

will have in selecting and interacting with a
social network, the greater her perceived sense
of control in retirement.

159
2) Employed women who feel they must retire at the
same time as their spouse will perceive less
control.

3) Nonemployed women who report family conflict will

perceive less control.

4) Nonemployed women who perceive role ambiguity

will perceive less control.
H |'[' I E 'E . l ] I l l' H I ]

The Modified Person-Environmental Interaction Model
was only partially tested by this research. The research
focused on the model components of person characteristics,
environment characteristics, interaction characteristics,
and mediators (perceptions). The behavior component was
not studied. The model suggests that perceptions
influence behavior which in turn feeds back to person and
environmental characteristics. The findings that person
characteristics, environmental characteristics, and
interaction characteristics influence perception is the
first step. Study of the relationship of retirement
behavior to mid-life perceptions of retirement would
further understanding of the relationship between
perceptions and action. Following a sample across time
would give insights to the importance of retirement
perceptions to the quality of life in retirement. The
data from this study suggest possible research questions.

160

1) 00 higher levels of perceived control lead to
more aggressive retirement planning?

2) Does lower perceived income adequacy lead to
delayed retirement?

3) Does higher stress pile-up from environmental
markers impact on health or overall quality of
life in retirement?

4) How do women with positive and negative attitudes
toward time use differ in their' selection of
activities in retirement?

5) What pre-retirement perceptions have the greatest
influence on retirement life satisfaction?

8 l' l E I. ! l I l I

The Retirement Perceptions Assessment Instrument was
designed and piloted in this study; and provided a good
’beginning. Currently there is not another instrument that
focuses on the perceptions of retirement investigated in
this study. In order to further research in this area the
instrument needs to be refined and validated. The
dimensions of retirement perceptions need to be enlarged
to make the instrument more robust and sensitive.

The control/anxiety dimension contains four items.

It does not directly address a sense of control in social
interactions and economic decision making. Both social
interaction and economic adequacy are important components
for retirement well-being and should be incorporated.

161

The time use dimension contains four items that
relate primarily to activities, but does not address
interactions with family and friends, nor is a value
placed on how time is used. The literature has
consistently stressed the importance of family and friends
to women. Questions that address the value placed on
activity, social interaction, family gatherings, and
family care-giving should be added.

The residence selection dimension consisted of two
questions related to possible change. There is no way to
determine if the change in residence is perceived as full
filling a dream or is an adaption to scarce economic
resources. Additional questions are needed on the reasons
for anticipated change.

The wife's perceived income adequacy was a single
item. This dimension should be enlarged to include
perceived income adequacy from a variety of sources. In
this study perception of income was more fruitful than
questions about actual income.

The continuity dimension was a single item. It did
not address the component parts of continuity or the
desirability of continuity. Items on life style, family,
health, and work would focus on possible components.
Attitude questions would enlighten the desirability of
life continuity.

162
mu . . 0‘ 0 .on- 01.11.111.| o 00 ' o ' 0‘.

In 1978 Brown and Paolucci stated the mission of home
economics as follows:

The mission of home economics is to enable
families, both as individual units and generally
as a social institution, to build and maintain
systems of action which lead (a) to maturing in
individual self-formation and (b) to enlightened,
cooperative participation in the .critique and
formulation of social goals and means of
accomplishing them.

The findings of this research suggest at least two
areas that need to be focused on in the training of family
professionals if the stated mission is to become a
reality. First, this study indicates that a major portion
of the women in this study have a low level of sense of
control. For 1985 the mean was 10.33 out of a possible
20, up from 8.72 in 1983. The mid-life women of this
study were socialized to be supportive of husband and
family, even if that means interrupted employment patterns
and acquiescence to husband's financial decisions. In
reality a majority of women are living the later part of
their life cycle as independent persons, and living with
the economic reality of decisions made in young and mid-
life. Women are expected to live eight years longer than
males and loss of spouse is related to risk of poverty.

163

Providing education and supportive programs that build a
sense of control is important for females of all ages. A
sense of control is linked to a sense of self-worth, life
satisfaction, and assertive financial planning (Kivett,
Watson, and Busch, 1977). It is vital to the well-being
of future men and women to educate and socialize all
children, not just males, to have independence of thought
and a sense of responsibility for self.

The second finding is the relationship between family
satisfaction and marriage happiness and sense of control.
This suggest that if family professionals desire to build
a sense of control in women they must be skilled in the
area of marriage and family relationship therapy. Morris
and Joanning (1986) noted that home economics is
particularly suited to carry out the mission of family
therapy. Home economics is not as narrowly bound as other
disciplines when looking at the family and is able to draw
upon other disciplines as well as their rich heritage.
Home economics can provide the family therapist with a
sound base and flexibility that is needed to address the

control issues of families.

I 1' I. E E “E II I El Ii
Although there appears to be growing national
interest in pre-retirement planning programs, particularly

on development of programs directed toward subgroups that

164

have been previously overlooked, pre-retirement planning
programs are doomed to ineffectiveness and even failure
unless they are structured to address the needs and
interests of subgroups of women (Szinovacz, 1982). This
study suggests that women's needs in retirement are
strongly associated with family, particularly in the area
of sense of control. Marriage happiness is a predictor
of sense of control for employed women, and family
satisfaction is related to sense of control for
nonemployed women. A review of connion pre-retirement
program formats indicate that programs include
presentations by lawyers, insurance persons, tax experts,
and social security experts (McMahon, 1981 and Prentis,
1980). There are exceedingly few programs that include
family professionals to address the concerns of role
conflict, autonomy, and communication. Programs that
offer skill building in conflict resolution and
interpersonal communication skills which can be applied to
emotional, economic, and time use issues should focus on
family needs.

The findings on age support the notion that the
earlier persons are involved in the pre-retirement
planning process the better. Older women had more
negative perceptions of retirement. Early intervention
could take advantage of more positive perceptions. In
addition the majority of women do not perceive their own

165
retirement earnings as adequate to meet their needs.
Early intervention could lead to more aggressive money
management.
Summon:

The study reported is an exploratory study of
perceptions of retirement of mid-life married women. The
research sample is a sub-sample of 154 participants from
the NC-164 regional project, "Stress, Coping, and Adaption
in the Middle Years of the Family".

The purpose of the study was to develop a better
understanding of perceptions of retirement among married
mid-life women. There is a need to understand the
heterogeneity of women's environment and perceptions in
order to better prepare them for retirement. The
ecological approach allowed for the study of a complex
system without destroying the functional integrity of the
influences on perceptions.

This study found that each of the dimensions of
perception of retirement was correlated with a different
pattern of personal, environmental, and interaction
characteristics. All characteristics were related to one
of the five dimensions to some degree except health. In
brief, family satisfaction was the strongest predictor for
a sense of control. When employed women were tested
independently, marriage happiness was the strongest
predictor of sense of control. Women did not perceive a

166

strong sense of control in retirements Women had a
positive attitude toward time use in retirement, with
full-time employed women having the most positive
attitude. Full-time employed women perceived their
retirement income as more adequate to meet retirement
needs than part-time or nonemployed women. Perceived
income adequacy predicted a change in residence, the lower
the perceived adequacy the greater the perception of
residence change. The fewer the number of environmental
marker events experienced the higher the sense of
continuity. The perceptions for retirement for all
dimensions became less positive as age increased.
Perception changed over a two year period, generally
becoming more negative.

Women are a heterogenous population who formulate
perceptions from a complex mix of environmental, per50nal,
and interaction characteristics. This study gives
important insights to the perceptions of retirement on
which mid-life women base their subsequent decisions about
this vital life stage. It provides several suggestions
for further research, for home economics/human ecology,
and for pre-retirement educators. Furthermore, it

supports an ecological model for the study of retirement.

APPENDICES

APPENDIX A

RETIREMENT PERCEPTIONS ASSESSMENT INSTRUMENT

167
RETIREMENT PERCEPTIONS ASSESSMENT INSTRUMENT

We hsve been discussing things that influence your present
feelings. Nowlthink about your feelings and expectations
abont the retirement stage of your life. Use following
sca e.

* 1. Strongly agree 4. Disagree
Agree 5. Strongly Disagree
3. No opinion

Control/anxiety

Time II
60-i I am comfortable thinkinn about retirement.
60- I have a sense of contro over my future.
60-o My retirement needs will be covered by my
spouse.
60-h I am looking forward to retirement.

Residence selection

60-e I will move to a different home.
60-f I will move to a different community.

Time use
60-d I will have more time for recreation.
60-5 1 will start new hobbies.
60-r I will enjoy having more free time.
60-c I will have plenty to do with my time.

Adequacy of wife's retirement income

60-p The retirement plan through my work will
cover my financial needs.

Life continuity

60-q My life will remain basically the same as it
has been.

* For the purpose of analysis the scale was reversed.

APPENDIX B

RETIREMENT PERCEPTIONS ASSESSMENT INSTRUMENT
(MODIFIED)

168
RETIREMENT PERCEPTION ASSESSMENT INSTRUMENT

We hsve been discussing things that influence your present
feelings. Now think about your feelings and expectations
abont the retirement stage of your life. Use following
sca e.

* 1. Strongly agree 4. Disagree
2 Agree _ 5. Strongly Disagree
3. No opinion
Control/anxiety
Time I Time II
60-a 60-j I am comfortable thinking about
retirement.
60-b 60-k I have a sense of control over my
future.
60-h 60-0 My retirement needs will be covered by
my spouse.

Residence selection
60-e I will move to a different home.
60-f I will move to a different community.
53-e I will move to a smaller home.
53-f I will move to a larger home.
Time use
53-d 60-d I will have more time for recreation.
Adequacy of wife's retirement income

60-f 60-p The retirement plan through my work will
cover my financial needs.

Life continuity

59 60-q My life will remain basically the same
as it has been.

* For the purpose of analysis the scale was reversed.

APPENDIX C

HEALTH DIFFICULTY SCALE

169
HEALTH DIFFICULTY SCALE

We are interested in your health. Please use the codes
given below to indicate how often the following items
apply to you.

1. Never

2. Seldom

3. Sometimes
4. Frequently

5. Almost Always
Q 27. How often have you:
a. had trouble sleeping ......

b. had accidents ..........

c. been irritable ......... _

d. been depressed ......... _

e. smoked cigarettes, cigars, or
a pipe ............ _

f. used prescription drugs .....
9. had a weight problem ...... ._
h. used alcohol .......... _
i. found it difficult to relax. . .

j. had headaches .......... _

k. had mdscle tension, nervous
indigestion or anXiety .

l. had colds or flu ........ _

APPENDIX D

KANSAS FAMILY LIFE SATISFACTION SCALE

170

KANSAS FAMILY LIFE SATISFACTION SCALE

Please circle the number which best describes how
satisfied you are with each of the following:

053.
Com letel Complete]
Disgatisfied Satisfiedy
b. Your family life 1 2 3 4 5 6 7
c. Your relationship with
your spouse 1 2 3 4 5 6 7

d. Your relationship with
your children 1 2 3' 4 5 6 7

e. The relationship your
children have with
each other 1 2 3 4 5 6 7

APPENDIX E

MARRIAGE HAPPINESS SCALE

171

MARRIAGE HAPPINESS SCALE

Please circle the number which best describes how happy
you are with your marriage. The middle point represents

happy'.

Q. 59. Extremely Extremely
Unhappy 1 2 3 4 5 6 7 Happy

APPENDIX F

ENVIRONMENTAL MARKER INDEX

172

ENVIRONMENTAL MARKER INDEX
(Time 2 Data Collection)

1 a Not at all 3 - Moderately 5 = Extremely
2 - Slightly 4 - Quite

Please read each of the events listed below and mark
whether it was experienced by any family member in the
last two years.. If yes, please Circle the number showing
how disturbing it was.

last 12 Mo. Yr. before the
last 12 mo. 1 2 3 4 5

Yes No Yes No
1. Marital separation
occurs (Divorce -
Separation)

2. Become widowed ldeath of
family member

3. Marriage of a member

4. Member moves out of home
(for independence, school
or marriage)

5. Major wage earner
retires from work

6. Death of husband's or
wife's parents

Death of brother or sister

8. Death of close friend and
confidant

9. Aged parent(5l become seriously
ill or disab ed requiring
direct care.

10. Member experiences menopause

11. Aged parent committed to
institution or placed in
nurSing home.

* From questions 10a, 10b, 60 a, b, c, f, t, 2, aa, bb,
ii, jJ, and kk in the time two instrument.

APPENDIX G

CORRELATION FOR PERCEPTIONS OF RETIREMENT BY
INDEPENDENT VARIABLES FOR THE REGRESSION SUBSAMPLE

173
Table 25

Correlation for Perceptions of Retirement by Independent
Variables for the Regression Sub-Sample

 

(n 8 110
Independent . . Retirement .Life
. «l ‘ 'l 0 i‘ "1' ll‘ ['111’ '1 I
Education .19 -.18
Age -.17 -.20 -.21
“23.382“ .17 .25
Health -.17
Marriage .
happiness .26 .15
Years married -.17 -.16

Family income
Wife's income
Family size

Community size

Perceived income

adequacy -.28
Environmental

markers -.15 -.17
Environmental

stress pile-up

Family
Satisfaction .28 .19

 

; o ' . 0| . ‘o l- ‘ . o 00.0 I < I o

LIST OF REFERENCES

174

List of References

Aizenberg, R. 8 Harris, R. J. (1982). Family Demographic
Changes. The Middle Generation Squeeze,
The Journal of the Western Gerontologica oc1ety.
Winter 6- 41.

Anderson, T.A. f1985%.7 The Best Years of Their Lives,
fleysyeek -

Atchley, R. C. (1982). The Process of Retirement: Comparing

Women and Men, .o 1 R u
11' -. 01 0 i‘ ‘1 M1, e 1te y .
z1novsacz, age 'ress, - -=.

Atchley, R..C (1976). Selected Social and Psychological
D1fferences Between Men and Women in Later Life,

,No. 2, 204- 211.
Beach B. A. (1987 . Time Use in Rural Home-Working
Famillies, ' ,October, 412-416.

Berkowitz, A. 81 Perkins, H. W. (1984). Stress Among Farm
Women. Work and Family Interacting Systems, Journal
Wily. February. 161- 166

Block M. R. £1982). Professional Women. Work Pattern as a

Correla e of Retirement .Satisfaction, .-1-1'
-1111 110 - 01 0 1 i' - 1
e 1te- -y' 21novacz, age ress, = -
Block M.R., Davidson J. & Grambs. J. (1981). ngEn_QyfiL_
° Springer Pub
or , , - .

Bokemeier, J. 8 Maurer. R. (1987. Marital ualit
Conjugal Labor Involvement 0 Rural Coup es,
Relatlgns. October, 417-423.

174

175

Boss, P. & Pearce-McCall, D. 1 Greenberg, J. (1987).
Normative Loss in Mid-life Famihes: Rural, Urban,
2317i 4(‘i‘e3nder Differences, W. October.

E.M., Kleban, M.H. Johnsen, P.T., Hoffman, C. &.
Schoonover, C.B. (1987 . Work Status and Parent Care:

A Com arison of our Groups of Women, Ihe
Mideast. 21. 201-208.

Brown A.S. (I985). Grassroots Advocacy for the Elderly in
Small Rural Communities, 1he_fiergniglggist, 25,

Brody

415-423.
Brown M. & Paolucci, B. 61978). Home sfgngqggs; A
geiinitign, Washington, .C., American ome conom1cs
ssoc1at10n.

Bubolz, M.M., Eicher, J.B. 81 Sontag. M.S. (19793. The
12431113111 Ecosystem A Model. LBomeJmnomiss. prino.

Bubolz. M.M. 81 Keefe, D. (1980). Wily
, edited by M. Andrews an . oger,
nstitute for Family and Child Study, Colle e~ of

Human Ecology, Michigan State University. 81-10 .

Campbell S. 1979). Delayed Mandatory Retirement and the
Working omen. W. 12. 257-262.

Clemens, A.W. 81 Axelson, L.J. (1985)..The Mot-so-empt -
nest: The Return of the Fledgl1ng Adult. family
' , April, 259-264.

Cook, F.L. & Kramek, L.M. (1986). Measuring Economic
grdsggpnAmong Older Americans, W,

Crystal, S. (1986). Measurin Income and Ineguality Among
the Elderly, ' , 26, 57- 9.

MW. (1984). Series P-23. No. 138

Depner, C. and Ingersoll, B. (1982). Employment Status and
Social Support: The Ex-erience of the Mature Woman”,
.011‘1 1' ‘11'1’ '0 11- - 01 0 i‘ 1
33335211 eu1te: -y '. z1novacz, age 'u- 1cation,

Dohrenwend, 8.5. (1973). Life Events as Stressors: A
Methodological Inquiry, '
fignayign 167-175.

I I

176

Dorfman, L.T., Kohout, F.J. 8 Heckert, D.A. (1985 .
Retirement Satisfaction in Rural Elderly, Reseancn_&n
Aging. 1. 577-599.

Dorfman, L.T. & Moffett, M.M. (1987). Retirement
Satisfaction in Married 2aond Widowed Rural Women, 111:

Duncan, G.J. (1984)
University of Mic gan, nn r or, 1c 1gan.

Eaton W.W. (1978;. Life Events, Social Supports. and
gsychiatric ymptoms: A Re-analysis of the New Haven
a a

230-23 . ' '
Ferber, M. 8 Huber, J. (1979). Husbands, Wives, and
Careers, WW May.

315-325.

Fowles, D. (1983). The Changing Older Population, Aging,
May-June. 6-11.

Fox, J.M. (1977). Effects of Retirement and Former Work
Life on Women's Adaptation in Old Age.
. 32, 196-202.

Gould R.L. 1979) ILansI95mgt%nn51.5rfiythvang_£hange_in
Adult_Liiei Simon an c uster, ew or .

Harbert, A.S. & Wilkinson, C.W. (1979). Growing Old in
Rural America, Aging. 222. 36-40.

'HerZOc, .A.R., Rodcers, W.L. & Woodworth. J. 1982).
- - 1— l 0 O

1 ' :- ‘1 .1101! I ‘ ' ‘1'

'esearc 'eport er1es, |n1vers ty 0 '1c 1gan.

I

Holden, K. (1980 . The Homemaker' and Social Security:
Assumpggons ehind Reform Proposals, Generaiigns,

Horowitz, A. (1985). Sons and Daughters as Caregivers to
lder Parents: Differences in Role Performance and
Consequences. Ihe_fienontologist. 25. 612-617.

Iams, M. (1986). Employment 53f 5Regired-Worker Women,
Social_§ecunity_flulletin. . - .

177

Jackson, $.B.1(1979) Efiflgfi Effiw
npu 15 e 1ssertat1on, ast an51ng, , 1c 1gan

State University.

Jaslow, (1976&. Employment, Retirement, and Morale
21ona Older omen Jou1111_11_fierontology. 31. 212-

 

 

Johnson, C. K. 6 Price-Bonham. S. (1980). Women and
Retirement: A 38$t3ud? and Implications, family,
Relations

I 0

Kahne H. (1981). Women and Social Security: Social Policy

Adjusts 5_t0850cial Change“ . Aging_&_fluman_neyelonment.

Kandel, D.B Davies, M. & Raveis, V.M. (1985). The
Stressfulness of Daily Social Roles for Women:
Marital, Occupational and Householg4 goales. Jaunnal
oI_flealih_and_Social_Behayion

I D

Kasworm, C. & Wetzel. £1981). Women and Retirement:
Evolving Issues for. uture Research and Education
Intervention , International
Quarterly, 299-

Keating, M. 81 Jeffrey, (1983) Work Careers of Ever

Married and Never8 Married Retired Women, Ina

0

Keith P. M. (1982). Working Women Versus Homemakers;
Retirement Resources and Correlates .of Well -Being,
10m n' i‘ ‘1 0 - -. 01 0 i 1
33331231, o1teu 1y '. z1novacz, age 'u. 1cat1ons,

Kessler & McLeod (1984). Sex Difference in
Vulnerability to Undesiraaggefiliife Events, Anerjsan
Sociological_8eyiew

O I

Kilty K. M. & Behling, J. M. (1985). Predicting the
Retirement Intent1ons and Attitudes of Professional
Workers. Journal_oi_fierontology. AQ.J 219-227

Kivett, V.R., Watson. J. A. a Busch. (1977). The
Relative Importance of Physical, Psychological, and
Orientation in

Social Variables to Locus of Contro
Middle Age. 1111111_11_11111111111.n

178

Krout J. A. (19871. Rural-Urban Differences in Senior
Entegz _A9$tivi ies and Services, W,

Levi nson (1978) WWW.
Bal llantine Books. New or

Levy, 5. M. 1(981. The ad ustment of the Older Women:
Effects (of hronic 1.1 Health and Attitudes Toward

Retirement,

Lin, Woelfel, M. W. 8 Li ht, S.C (1985). The
Bufferin Effect of Social upport Subsequent to an
MmLoLflealmJndJosjal

Im ortan Life Event,

Bafiayinn,n .247-263.
Little L.F., Proulx, F. P. Marlowe, J. 8 Knaub, P. K.
{1987). The History of Recent Farm Le islation:
moliggthogs for Farm Families. W. 36.

Lopata, M. Z. & Morr, K. F. 1980). Changing Commitments of
American Women to 03k l3and Family Roles.

Mancini, J. 8 Bliezner, (1985). Return of Middle-
A ed Children to the argentlal4 Home,

McCubbin, H.I. 8 Patterson,

   

(1980).

A
I

 

 

 

 

 

 

 

 

 

o,s1 .1 r envear§1 1 .
Inifinyanion. ~am ly tiress 18:1. 0 ege

o ome conom1cs, Un1ver51ty o 1nneso:a.
McKenna. J. 81 Nickols, (1986). Retirement Planning
Strategies For MisdlYife Women, Jannna1__ni__flnma

Egnnnmigs, Winter, 34-36.

McMahon, B. J. (1981). Feasibility Study for Fre-retirement
Planning Program, Mich1gan State University,
Unpublished manuscript.

WWW. (1986-87). 12th edition,
ureau o u51ness esearch.

Meiners, J. E. 8 Olson, G. (1987). Household Paid, and
23 alclilWork Time of IFarm Women, Eamjly Relatjnns.

179

Montgomery R. L V. Gonyea, L G. & Hooyman, M. R. (1985L
Careg1ving and t e Experience of 1Subgective and
Obj ecti ve Burden. W.l

Moon, (1986). Impact of the Rea an Years on the
Distr1bution zgf Income of te Elderly, Ihg
5919:119199151. . - .

Morgan, 8é1986L The Financial Experience of Widowed
Women. vidence from the LRHS. W51. 25,

Morgan, £198..” Continuity and Change in the Labor
ForcLe

ctivitfl!53 of o_Re35cently Widowed Women,1ne

Morris . 81 Joanning. . (1986 Marriage and Famil
Therapy: A GrowthsArea (for61107me Economics, Jon-mi
9.f_fl9m9_E99n9mj_cs. unlner, 6-3

0' Rand, A. 81 Landerman, R. (1984). Women's and Men's
Retirement Income Status, Wm, Q,

March, 25- 43.

Parr, L (1980). The Interaction of Persons and Living
Environments. ' edited by L. W.
Poonlm6 American syc o ogica ssoc., Washington DC,

Philliber W.W. 8 Miller, ($19.79) Occupational
Atta1nments and Perceptionsv of tatus Among:b Working
Hives, Eb ruary'

59- 62.

Pospisil, L. (1979). The Tirolean Peasants of Obernberg: A
stud in Long-Term Research, “WW
W991 edite y oster eorge,

ayer c u er Col lson, R. Kemper, Academic
Press, New York,125-130.

Prentis, R. f(1980). White- Collar Working Women' s
Sgrception 0 Retirement, W11. 29,

Rexroat, C. 1985). Women's Work Expectations and Labor-

Market xper1ence in Early and Middle Famil Life-
Ciel“ e Stages. W. 51.

180
Riley ".W., Hess, 8.8: 8 Bond, K. (1983 . :c'nc '1

n 1 x I x '1,
Zgwgfnce r aum ssoc1ates 'u- 15 ers, ew Jersey,

Sands, J.D. & Parker, J. 1980). A Cross-Sectional Study
of The Perce1ved S ressfulness of .Several L1fe
Events,

Scanzoni, J. & Arnett, C. (1987). Policy 1m lications
Derived from a Stud of Rural and Urban arriages,
EamJJLBelj119ns. 0c ober. 430-436.

Schmall V.L. & Staton, M. 1980 . Women: For otten
Subjects for Retirement( ResLarch, Generagigns
October, 4, 18-19.

Schultz, J.H. (1985). To Old Folks with Love: Aged Income
Ma1ntenance in America, Ihg_fiezgn191ggi§1. 25,

465-470.
Schultz, J.H. (1981). Pension Polic at a Crossroads: What
Shgugg be the Pension Mix?, Ifi;_§gronxglogi§1, 21,

Schultz, L11. (1980). Ihfimmjmm, Wadsworth
Publishing Co., Ca., - .

Schumm, W.R., McCollum, E.E.,Bugaigis, M.A.. Jurich A.P.
& Bollman, S.R. (1986). Characteristics of the kansas
Family Life Satisfacti0é18$c9ag7e 9151(1) a Regional Sample,
Wm. . - .

J.P. and Roberto. K.A. (1985). Use of Informal and
Formal Support Metwogéi. 61% Rural Elderly Poor, lug
9919919193191. - .

Scott

Scott . and Roberto, K. 91987). Informal Supports.of

J.P
Older Adults: A Rura -Urban Comparison,
° , October 444-449.

Seccombe, K. & Lee. G.R. $1986). Gender Differences in
Retirement Satisfact on and Its Antecedents,
9n_A9.i.n.91 426-440.

0

Shaw, L., (1984). Retirement Plans of Middle-Aged Married
Women. 1119.69191119199111. 2.4.. 154-159.

181

Shectman, K. L. Bergen, M. 8., Schumm, W.R & Bugai his,
M. A., (1985). Characteristics of the Kansas arital

Satisfaction Scale among Female Partici ants
Community Childbirth Education Classes. Psycfiolggigal
11999112. 59.5 538

Siegler, I. C. and Georo 1e, L. K. (1982). 1- L'lU

[0 01 ‘1‘. on [a

. c-m-“t— - 1 . presented a
C1ent1 1c 'eet1ng o t e toston Society of
Gerontologic Psychiatry, November.

Smith,1985 - i- . of famjlx
( ) ‘ mm s act1 on:

 

 

 

. . 0| 0
mm-mIr-mn - npub ished
I1ssertat1on, . an51ng ' ., 'ichigan State
University.

Social Security Bulletin, (February 1985), 48.

Sontag, M. S. Bubolz, M. M. & Slocum A. C. (1979).
erceived Quality of Life of Oakland County Families:
A Prel1mi nary Report.

Eamjly Ljngg, Michigan gtate University flgricuiturai
xper1ment ation.

Sorensen, A. (1983). Women's Employment Patterns After
W

Marr a e,
311- -32?.

Steitz, J. A. (1979a. An Interaction- Transaction
Perspective on t e Perception of Control within
iggect Life Situations, Human Dexelopment. 22, 390-

Stone, R. Cafferata, G.L., & Sangl, (1987;. Caregivers
of the FrailzzElde6r11.A626 National rofile,
59199191991111 . -

Tracy M. B. & Ward, R. 91986). Trends in Old-Age
Pensi ions for Women: Bene it Levels in Ten Nations.
1960-1980.119.9919919199191. 29. 287-291

Uehara, E. Geron S. 8 BeemanE S. (1986‘. The Elderlg
Poor n the Reagan Era, , 25, 4 -55

Uhlenberg, P. 8 Salmon, (1986. Change in Relative

Income5 of 17(1))lder Women, A1960-l9 0, MW,

29.1

I I

182

Nalfor-Kraemer, P. 8 Light H. (1984), Depression and
Mastery in Women: Differences According to Personal
Characteristics. MW. 55. 710.

Walker, L.S. 8 Walker, J.L. (1987). Stressors and Symptoms
Predictive Distress in Farmers. 'Eamily__fielations.
October,374-378.

Walker, R. Borland, D., Bubolz, M., Cameron, C. & Dulmes,
P. (1986). -- -. -- - - ' _

f. o . 0 . o 1
epartment 0 am .y an 1 . ogy,
Human Ecology, Mich1gan State Un1versity.

Warlick, J. (1985). Why is Poverty After 65 a Woman's
Problem. WM. .10. 751-757.

Weigel, R. & Weigel, D. (1987). Identifying Stressors and
Coping Strategies in Two-Generation Farm Families,
am. October. 379-84.

Wethington, E. & Kessler, R. (1986). Perceived Support,

Received Su port and Adjustment to Life Stress
MW 21.78- 9

Williams, T.R. & Bubolz, M.M. (1986 . Family Life Events
and Perceived Stress: An xamination of Sex
Differences, Unpublished paper.

Wise, 6.". & Murry V.M. (1987) The Empty Nest: The Silent
Invasion on Two Fronts, Wigs.
Spring, 49-52.

Zick, C.D. & Smith, K.R. (1986). Inlnediate and Delayed
Effects of Nidowhood on Povert : Patterns from the
1970'5. W. Zfi. 69-675.

  
 
 

   
 
 

   

 

 

   

M'll'l‘fljflILlelLilfllflfijl'ljfljfll’lfllfl'fl