:4. . . .!EE.. r
. ‘ E’ III! ' E!!! EIE l" E II I PIE-a 1!!!! ”LEI!“ .. we»: . ._
E‘”EW‘EI1~!EEW! 'II E EEIII :IEEWI! NEE“: E EWE‘SI . .12
3.1! .‘ . !!EI ElI! ! ! £!IEEIIII E5!!! 13'! EE!!E,E-!-!!!‘(!j!!!!:!h .;.E.IEIIE.II fij‘; 1:22..
I IE" I ‘ E E:¥:E:.25!'E . I ‘
‘.,; II [RE-1!! EEE ELIE :5
IE ' EEE E!!! E E!!! E?!
I E I . I I‘M” I .
TE ' ' 'II! I , II‘II E I!!! .E WEE}!!!
E I E ‘EEEEI’IE E “E!!! E!!':'.
*I. ‘Ef ! . EM . . . E I"!!! EEEEITIIEEEE, EEEI
I2!!!1 I“ E!!! .2 I ' EE' IEIE‘ E1!
3.: . ! !! EEEEI'! I!!! I! E! ! ‘EIE‘E! {:3 .'E
7' E! 1II! 1' EIE El E III! "
E . E!!! EEE !!!!l!.i!!2. -. E
..'I I!!! !I!!I "J! !!L L!!!!'!! '1!!!J ! "E- ! !E1 «E III I
i» ., EL!!! "IE«II‘E!!E EEEEE EIEEWEEE'E‘ E I!
I1! ‘ IE! EEEI!!! I!!! ‘EEE! ' E! E II EEEE I I! '
III E!!! ,E! I E! E!!! E I! E! . . . .
. .lEI‘EI! !!!!!!!!!!EI!! I 'E E k”!!! ‘ ~ .I ' . ‘ I; =-'
.2! E!‘ ’E‘!!! !!.‘E!3‘! E!!I‘IIE. !!l ! IEEEEEEIEEEI'!!! I E!!! I win!!! 1. EIE ..' E274!
IIIE ‘II I" 1! I‘Ian. El!!! 1!“). {'EEQ‘E"
ME!!! . E! E. E! 2"! III E EIEEE IEEIEEE‘!‘IEEEEI’EEII! WEE-E!!! E!!!" EE
.3 II "' ‘EEE E!!! E!!! E! I!!! EE! IE EEEIEE‘E‘EE '3!!! IEEE‘ELEEE‘E 4.: $9.
Q! ‘ E E‘!!! ;r «EN! ‘E EEEEEIEEE!!! EEEE! E!!! ‘ EIEEEE . E78. IE~
‘3»!!! !! E!!! !!!! EEEEI !! EEE!.!IE ‘IELE E!!!! !E!!I!!El!.!!!!E!!!!!!!EE!E !!E‘;III I IE ;! EEIE'EI‘J EEIJIEIEIIIEI'IEEEEEEEIEr-IEEI EWE‘I'IIIII“ Wig; I J
1!; EEEEIE' EE! I!!! !!'E‘! I!!!'E:"!EEI!I! EEEEIE‘!!EE!!!!E!!E'! I! 2!! EEE. !!'!!I!! EL‘IIIIIIIII 3E“!!! . 'EIIv E'IEEIII "I IMIEE ”:EI~ EII 51%
' I !!.!!I 1 I E -' I ; i ”1!..- 'g , | .;
I'E,“I .“E w! 2. ‘I' IEI I!!! ' E ! 11‘!!ng E!!! 'EL-i :‘ i-‘JI. .3? -* m ~‘-
I2 :E. E E. EII 5...? ‘E!“““"!!E‘ IEIEE! E I!!! In!!! IE“... EIE'E‘EEII‘EE‘ * "1‘
. HIE I! !I ‘E: El!!! EI'III !I -E' I EI,!!!!!! !!!!l! !! ;I: E!! EE'!!' NEE!!! EIEEI QEEEIE II!!! "IIIIE’!!!55!!i EH91!!! 5'!ng If}! ‘IEIIIIIE IIIEIEII in: . .., II
IE II" III! III! EI! I EEEIEI IEIII I! ”II EI:.I,II ‘,:E- ‘ I'EI‘n'I'“ ' rEEIEE H!!! E 'E 21' ’3“
EII . IIII E ‘ 'EEEE EEI E E l EEI EIEEEEI _!E‘i!!l' EEEEIJ EIEIE EEEI I!!! EIEE I E E 1 LEIIEIIIIE! IEIII I EIEI.‘ ‘IE E‘!!E III! I!!! I! IIEII II Ifi‘éIEE Em IIEB‘Im‘v’gfiE, ”If":
.' ‘I ‘1! ! IE. ‘EI I EIEIE ,EI'E‘ E‘s H IEEI~IE E-‘EEEI' I
E T IE‘EIEI! EEEE!!!E EE.! 1"“! 1! 1!!I!=‘E'I!!! I l I ‘ IEEI E!!! Mn! :1 II' E!!: E!‘!‘ I!!! Elm-E!E i‘ 1:". I';!g. rIH‘E'E. "w“; :f‘! 194
:,. ..!...E.III!"E E"!!! EEE'IE! ;; ,, I"‘E . I..IEE'I-EE‘E-EQEIEEEI‘IIEIEII'I' ‘I -' ' 'IzI" IEEIE-Ii‘rIE 2-III:."!‘IEE 1:133?!
! !E '!'E !I 7!: '7‘!” :EEIEEE‘::.!;!):, I‘!£’!!!l!.!!!“,!!1%!!!“EE) R, 2.1!. .,lI . )I ‘!.E1‘Ev‘ ,-' .‘IE‘JEI.’ 9 ‘ lg.” .3"5
I!!! EI!E‘3! !I E l!!“ I I I!!!|I:! III"II'EI!" 'E! .E'I!!. II! EEEI}. ME! 2' :I1;E.EI.I: 'I I: I- II:':E_ E-I‘ {LII
.~:- EI E . ... I — I I :1. w E- «I. s; E 1!!
”E "IE!" "If! ‘1!!!" "" ’I' L"! E‘I‘EI. Ile'EEEIIEEI IIIIEIIII‘IIEI EIII "E"

  
     

‘IEEIEE IIIII‘EEEE
Elf!!! I!!!E

!!.!\|!! '1‘!!! Elm! J!!!
! I(E! EE'I E!!!!'. ' .E! E
‘ EEI‘ 1!! MI!!! . !I

I «I! ‘ E! E E E
|'H! hum”!!! 3E!!! ! !! EEIEI .E! M . I!!!
EI‘I ‘EI E!‘ III I'EI! LEI! "'II E“ E "E II“ III !!IEIE
I ‘IEE

‘ JEEIIIE‘EI I E El! [1 2 LEE! E!‘ IE!‘ E!!! ,l!‘ E!!!" Ex!" 'I'IEE .I!.' .‘E' E!!! I!!! ‘I'E!!!E!

E!!!

      
 

Im—u.

I: I.II" IIIE‘" ‘I I!“ I" IEEI!!E!EIIII|!IIEIII EEEIEEEE! EYE!!!" !!!!! u!
!! E'EII, “"EEE' EII‘I'IIE" I"! ‘ ‘....EE EEI‘E!!"'.."EE'IEI’IEwI!! IIE'E I I!!! E!!!

‘ '= ‘2..I. 1!! 2.! .. .E.,' I 2.! !! I"
""‘I' II E "III! ElIIIEIE! EEE .- ' E!
E1 III!!! 'IEIII'I‘IIIIE“E I!!! I! E

‘ II I! .2!!! I! I

E!!!! ‘E!!!!!! I
!I I!! EVE} E!!.IE: !!!I!!!!!! E ! WEE!!!
! !I !II!IE!!|E!! !!| I Ell!!!“ l!! EEIEI

!!
""IIII‘EIEE'II “I! E! E!!! EEEEEEIEE !!!!!‘!I E!!!1 EEEEEEEI

.
!I‘ >‘
!.

II‘ E!

an
-

 

!!I
I!E

I" III. E! E!
IIIEEEI‘E EEEIIIIEE‘I 'E‘EEEEE 'EEEIEE EEEEIEEEIIEIE’E

. I?‘-Q"#"‘?"‘I

3 1293A {0139 JIM/fl” ’ L.

This is to certify that the

thesis entitled

A.Need Assessment of Old People Living
Alone: Applied and Theorhetical
Considerations

presented by

Denis Gray

has been accepted towards fulfillment
of the requirements for

M.A. Jegree in Psychologx

Matti QAM

gajor professc‘

March 1978
I)ate

 

0-7 639

 

A NEED ASSESSMENT OF OLD PEOPLE LIVING
ALONE: APPLIED AND THEORETICAL

CONSIDERATIONS

BY

Denis Gray

A THESIS

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

MASTER OF ARTS

Department of Psychology

. 1978

ABSTRACT

A NEED ASSESSMENT OF OLD PEOPLE LIVING
ALONE: APPLIED AND THEORETICAL
CONSIDERATIONS

BY

Denis Gray

A review of the literature showed that old people
in general and old peOple living alone in particular are
frequent victims of a large number of exacerbating problems
including income, health, interpersonal, leisure, and
familial losses. While documentation of the prevalence of
these and other problem areas was prolific, the probability
of unique needs and extant remedies existing locally argued
for undertaking a local, programmatically focused, need
assessment of elderly individuals living alone in Lansing,
Michigan.

Although the primary focus of the study was applied,
a Life Satisfaction scale was added to the study instrument
to permit the examination and replication of some theore-
tically interesting hypotheses including examination of the
disengagement and activity theories of aging. A Compre-
hensive structured interview was conducted with 70 old

people who lived alone. Potential respondents were

Denis Gray

identified for the study by a number of local agencies and
churches. Interview data described the health, inter-
personal contacts, leisure involvement, problem areas and
service preferences of the study population.

Descriptive statistics and correlative analysis,
including partialing and multiple regression procedures,
were used to achieve study objectives and test study hypo-
theses. In addition, a cluster analysis of the research
data was performed. Results revealed that while many
respondents reported chronic illnesses and poor health,
only a small percentage of respondents (7.1%) reported
health as a very important problem. Only transportation
was seen by a majority of respondents as an important
problem. Although a large number of possible services
were acceptable to respondents, home health care type
services were most frequently viewed by respondents as
services they would use.

In addition, several categories of variables were
found to be significant correlates of Life Satisfaction.
These included health, interpersonal involvement (visiting),
and leisure involvement measures. However when corre-
lations with Life Satisfaction were examined separately for
respondents under 75 and respondents over 75 years of age,
different patterns of correlations were revealed for the
two groups. Only two of 15 correlations were significant

at the .05 level for the over 75 group, whereas 10 of 15

Denis Gray

correlations were significant for the under 75 group.
These results were seen as supportive of disengagement
theory for respondents over 75 but supportive of activity
theory for respondents under 75.

Based upon these results and their interpretation,
several suggestions were made for program development for

the elderly residing in the Lansing community.

To the memory of my father

and to Mom, Maryann and
Jenifer who help me travel

the road to find out.

ii

ACKNOWLEDGMENTS

While a good many people played significant roles
in making this study possible, there can be no doubt that
Bob Calsyn, my advisor, was the prime mover and source of
motivation for the completion of this final product. Bob's
methodological acumen and systematic critiques of early
proposals and drafts helped forge a thesis I felt proud to
submit to my Master's Committee. Probably of more signi-
ficance, however, was Bob's unremitting personal accessibi-
lity and support, which were felt even as I labored on this
product 1200 miles from my nest, in Tampa, Florida. During
those periods when I questioned whether I would ever com-
plete this study, Bob's support and friendship provided one
a quieting refuge which helped me see the light at the end
of the tunnel and proceed toward it. I must also thank the
other members of my Committee, George W. Fairweather and
Louis Tornatzky, for their ongoing support and encouragement.

While Bob helped to keep me going in this effort,
Jenifer helped to keep me going. She gave me her emotional
support when I needed it, her patience when I was short
on my own, and knowingly vanished into thin air when I

needed solitude. As Jenifer approaches her own deadlines

iii

and goals in the coming year, I hope I can provide her with
the same gifts she openly provided me.

Many others played significant roles in the ini-
tiation and completion of this study: Roxanne O'Connor,
Dorothy Payne, and the staff of the Tri-County Office on
Aging, who believed in the value of applied research and
provided me with the opportunity to collect my data;

Esther Fergus, who provided encouragement and support when
this product was only a vague idea, Jack Wakely, who helped
snip some of the red tape which threatened to scuttle this
effort as the words of service and academia were brought
together, and Caroline Clark who diligently assisted in

the interviewing process. A final word of thanks is also
owed those Lansing residents who participated in this

study who graciously took the time to tell us about their

needs and satisfactions.

iv

I.

II.

III.

TABLE OF CONTENTS

INTRODUCTION 0 O O O O O O O I O O O O O O O 0

Problems of the Elderly . . . . . . . . . .
The Elderly Living Alone--High Risk

Group . . . . . . . . . . . . . . . . . .
Assessing Local Need . . . . . . . . . . . .
Life Satisfaction--A Theoretical Need

Construct . . . . . . . . . . . . . . . .
Activity Theory vs. Disengagement Theory . .
The Present Research . . . . . . . . . . . .

METHODS . . . . . . . . . . . . . . . . . . .

Sampling . . . . . . . . . . . . . . . . . .
Initial Contact . . . . . . . .
Interviewers . . . . . . . . . .
The Interview . . . . . .
Description of Variables . . . .
Data Analysis . . . . . . . . . . . . . . .

RESULTS 0 O O O O O O O O O O O O O O O O O 0

Objective 1 . . . . . . . . . . . . . . . .
Objective 2 . . . . . . . . . . . . . . . .
Objective 3 . . . . . . . . . . . . . . .
Hypothesis Testing . . . . . . . . . . . . .
Hypothesis 1: Life Satisfaction will be sig-
nificantly related to reports of recent
age related losses . . . . . . . . . . . .
Hypothesis 2: Life Satisfaction will be sig-
nificantly related to various measures of
Health . . . . . . . . . . . . . . . . . .
Hypothesis 3: Life Satisfaction will be sig—
nificantly related to various measures of
Interpersonal Involvement . . . . . . . .
Hypothesis 4: Life Satisfaction will be sig-
nificantly related to various measures of
Leisure Involvement . . . . . . . . . . .

Page

10
14
21
24
27
27
28
29

30
40

41

43
51
54
60

63

66

70

74

Page

Hypothesis 5: Activity theory predictions
will be supported for younger respon-
dents (under 75) and disengagement
theory predictions will be supported

for older respondents (75 or over) . . . . 77
Overall Prediction of Life
Satisfaction . . . . . . . . . . . . . . . 80

Cluster Analysis . . . . . . . . . . . . . . 85

IV. DISCUSSION . . . . . . . . . . . . . . . . . . 98

Multiple Problems . . . . . . . . . . . . 99
Income . . . . . . . . . . . . . . . . . . . 101
Transportation . . . . . . . . . . . . . . 103
Health . . . . . . . . . . . 104

Crime . . . . . . . . . . . . . . 105
Nutrition . . . . . . . . . . . 106
Spare Time . . . . . . .
Other Problems . . . . . . . . . . . . . . . 108
Alternatives to Institutionalization . . . . 109
Hypothesis Testing . . . . . . . . . . . 110
Health . . . . . . . . . . . . . . . 112
Interpersonal Involvement . . . 113
Leisure Involvement . . . . . . 114
Demographic Determinants of Life

Satisfaction . . . . . . . . . . . . . . . 116
Overall Prediction of Life Satisfaction . . 116

Disengagement and Activity Theory . . . . . 117

V. CONCLUSIONS . . . . . . . . . . . . . . . . . 120
Future Assessments of Need . . . . . . . . . 122

APPENDICES

APPENDIX

A. Senior Citizen Questionnaire Consent
Fom O O O O O O O O O O O O O O O O O O O 124

B. Study Questionnaire . . . . . . . . . . . . 125

C. Breakdown of Demographic Characteristics
of Respondents . . . . . . . . . . . . . . 135

D. Breakdown of Interpersonal Involvement
Items by Source of Involvement . . . . . . 138

vi

Page

E. Variables Excluded from Cluster Solution
Because of Communalities Below .200 . . . 141

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . 142

vii

Table

1.

10.
11.

12.

13.

14.

15.

LIST OF TABLES

Activity and Disengagement Predictions of
Association to Life Satisfaction . . . . . .

Variables Assessed on Questionnaire . . . . .
Breakdown of Health Variables . . . . . . . .

Breakdown of Interpersonal Involvement,
Leisure Involvement and Loss Measures . . .

Breakdown of Problem Area Responses . . . . .
Breakdown of Total Problems Reported . . . . .

Breakdown of Responses on Service
Utilization O O O O O O O O I O O O I O O 0

Correlation of Demographic Variables to Life
satiSfaCtion O O O O O O I O O O O O O O O 0

Zero Order and Partialed Correlations of Loss
Measures to Life Satisfaction . . . . . . .

Correlations Among Loss Measures . . . . . . .

Zero Order and Partialed Correlations of
Health Measures to Life Satisfaction . . . .

Correlations Among Health Measures . . . . . .
Zero and Partialed Correlations of Inter-
personal Involvement Measures to Life

satiSfaction O C O O O O O O O O O I O O I 0

Correlations Between Interpersonal Involvement
measures 0 O O O O O O O O O O O O O O O O 0

Zero Order and Partialed Correlations of

Leisure Involvement Measures to Life
Satisfaction . . . . . . . . . . . . . . . .

viii

Page

22
32
44

47
52

55

56

61

64
65

67
69

72

73

75

Table

16.

17.

18.

19.

20.

21.

22.

23.

Correlations Between Leisure Involvement
measures 0 O O O C O O O O O O O C O O O 0

Comparison of Correlations with Life Satis-
faction for Respondents Below 75 Years and
Respondents 75 Years of Older . . . . . .

Multiple Regression Solution of Life Satis-
faction: Income, Interpersonal Involvement,
Health and Leisure Involvement . . . . . .

Correlation Matrix of Self Health Evaluation,
Income, Visiting Activity and Leisure
Activity . . . . . . . . . . . . . . . . .

Multiple Regression Solution of Life Satis-
faction: All Predictors Including Total
PrOblems O O O O O O I O O I O O O O O O O

V-Analysis of Key Cluster Structure . . . .

Correlations of Variables with Oblique Cluster

Domains 0 O O O O I O O O O O O O O O O I

Intercluster Correlations . . . . . . . . .

ix

Page

76

78

83

84

86
89

92
96

INTRODUCTION

Z_ Today our population has 20 million individuals, or
10 percent of the total population, over 65 years of age
(Maddox, 1971). This figure represents a percentage gain
in population of 700 percent for those over 65 since 1900
while the rest of the population has increased only 200
percent (Maddox, 1971). In addition to their pure numerical
growth it is noteworthy that the elderly already constitute
17 percent of the total voting population (Butler, 1975).
The future also seems to promise more of the same. Pro-
jected figures for the next quarter century indicate that
25 to 30 percent of the papulation may be over 65 years of
age by the year 2000 (Scott, 1971).

Over the last decade, the bureaucracy's recognition
of the present and potential political clout of the elderly
has resulted in increased interest in the problems affec-
ting this group and has ultimately led to expanded social
programming for the elderly. An example of this phenomenon
is the 300 percent increase in the Office of Economic
Opportunities's (OEO) Senior Opportunities for Services

budget between 1968 and 1972 (Butler, 1975).

Although similar expansions in programming occurred
for blacks and other minorities during the sixties, many
viewed the results of these efforts as disappointing,
pointing in particular to the ill-fated efforts of the War
on Poverty. While these previous efforts occurred at a
time when many assumed fiscal effort could be equated with
social change, the dawning of major social programming for
the elderly would appear to be coincident with a new more
realistic perspective toward the change process.

To a much greater extent, today's social program-
ming efforts must justify the need for specific programs
and additionally institute evaluations aimed at demonstrating
the worth and effectiveness of these efforts. While such
expectations are not panaceas, meaningful execution of
these activities in the course of developing and insti-
tuting programming for the elderly would optimize the
potential impact of recently allocated program dollars
(USDHEW, 1975) .

The present study represents an attempt to meaning-
fully institute the first of these activities, an assess-
ment of need, for a high risk group of elderly individuals.
This course of inquiry will be pursued through two comple-
mentary but distinct assessment approaches. The first will
consist of a programmatically oriented assessment of needs
and service preferences for a specific population, namely

the elderly of the Lansing, Michigan area. The second will

consist of a theoretically oriented hypothesis testing
inquiry concerning factors related to adaptation and

satisfaction among the elderly.

Problems of the Elderly

 

Health.--It is difficult to initiate a discussion of
the problems faced by old people without beginning with one
of the most widely existing problems of all-—poor health.
Figures indicate that 8 out of 10 old people have one or
more chronic ailments, producing disability in 45.1 percent
(Scott, 1971). It has further been estimated that 460,000
old people are bedfast, over one million are confined to
home and, all told, 6.5 million Americans over 65 years
cannot carry on their major activities (Scott, 1971).
Although only 5 percent of all old people are seriously
enough impaired to be in nursing homes and other institu-
tions, the fact that 80 percent of these are projected to
have both health and mental health problems (Scott, 1971)
portrays a very exacerbated situation necessitating full
time supervision at escalating costs.

Mental Health.--A1though there is considerable

 

support for the hypothesis that mental health (or stress
related) problems can cause health problems in the general
population (Holmes and Rahe, 1967), the prevalence of
organically caused mental illness in the older population
would seem to argue for a reversal of this cause-effect

relationship among our elderly group.

Recent trends have indicated that the prevalence of
organic brain dysfunctions is rising (Riley & Foner, 1968).
Some feel this situation is due to the wonders of modern
medicine which permits large numbers of inherently less fit
people to survive into old age, many of whom eventually
succumb to organic brain dysfunctions (Sjogren, 1959).
Regardless of the etiology of this situation, its impact on
individuals is staggering. Locke et a1. (1960) has put the
percentage of all mental health admissions over 55 years of
age due to organic etiology at 60 percent, while other
estimates (USDHEW, 1968) run as high as 80 percent.

Noninstitutionally based studies also reveal high
estimates of mental health problems in old people.
Passamanick (1962) reported almost five times as many cases
of psychoses in the 65 and over group than in the 35-64
year old bracket, based on clinical examinations. It
should be reiterated, however, that the part health related
problems play in these figures is considerable. Srole,
author of the well known Mid-Town Manhattan study, touched
on this point when he stated, " . . . the decline in bio-
logical vitality and concomitant ability to cope with
stress, increasing physical disability and illness cannot
be disregarded as significant determinants of the increase
in mental symptomaltology observed in the respondents of
advancing age" (1962). Clearly, the synergetic effect of
mental health and health problems among the elderly por-

tends a real and severe need.

Income.--The ability of our older citizens to pay
for the medical and related expenses they must necessarily
incur brings this discussion to what may be the most exacer-
bating problem to face all old people--income.

Health care expenses of older people is 2% times
that of younger individuals (Maddox, 1971). This must be
viewed in the context of the fact that the median income of
old people is 48 percent that of younger individuals and
one out of every four old people (compared to 1 out of
every 9 for other age groups) lives below the poverty level.
Some groups within the aged p0pu1ation are even worse off,
with more than one out of every two old people living alone
having incomes almost two thousand dollars below the poverty
line (Maddox, 1971).

Although some individuals claim this situation will
be altered as the recently aged, who are better educated,
and have been better paid, enter the upper age brackets,
recent data fails to indicate such a trend has begun.
Between 1968 and 1969 the number of elderly below the
poverty level rose by 200,000 while this number declined by
1.2 million for all other age groups (Butler, 1975). When
one views this evidence in light of the fact that most
older people are discriminated against in the job market by
mandatory retirement rules, and are further discriminated
against by income ceilings tied to receiving their own
social security benefits, it paints a very dreary economic

picture for those individuals entering their later years.

Other Problems.--Although income and health related

 

problems are the most commonly cited problems of older
pe0ple, other related problems are also of great concern.
For instance, transportation becomes a crucial problem for
older individuals as increasing years, disability, and
decreasing income places the most common and expensive mode
of transit, the automobile, out of reach for most elderly.

Housing is another major expense for all old people
and although the common view has it that all old people
have a comfortable unencumbered home, six million aged are
estimated to be living in substandard housing (Congres-
sional Committee year estimate). Poor nutrition, crime
victimization, consumer victimization, the unavailability of
leisure opportunities, age discrimination and lack of edu-
cation are only some of the additional problems facing
older people.

Unlisted up to this point but of great significance
is the emotional and role losses visited upon great numbers
of old people. At age 45-54, 80 percent of all men and 70
percent of all women live with a spouse, while at age 75
only 40 percent of all men and 15 percent of all women have
residence with a spouse (Maddox, 1971). The jobless worker
after a lifetime of employment, the parent without a child
to care for after years of caring, and the spouseless
family man or woman after years of sharing are the all too

common stereotypes of old age.

The foregoing has painted a very dreary picture of
the average elderly individual in our country. Of course,
not all old people have been ravaged by these problems.

In fact, most everyone can recollect the older individual
they have met who is in fine health, lacks nothing material
or otherwise, resides with their spouse and in general is
getting along fine. Regrettably, this picture can usually
be balanced by another which portrays an individual inordi-
nantly stricken by the ills of old age. Although there
might seem to be nothing inherently problematical about
being old and living alone, most of the facts about this
group, constituting one third of all old people, point to
old people living alone as an extremely high risk group in

many problem areas.

The Elderly Living Alone--High Risk Group

 

Elderly community residents residing alone consti-

tute approximately 33 percent of all individuals over 65

 

years of age nationally, 35 percent of all individuals over

65 years in Michigan and 31 percent of all individuals over

 

65 years in Ingham County (U.S. Bureau of Statistics, 1970).
A statistic descriptive of this group's demographics makeup

is age; 43 percent of all individuals over 75 years of age

 

in Michigan live alone (Michigan Office of Services to the
Aging, 1975). Hence the majority of old people living alone
come from the oldest of the elderly. Further, it has been

predicted that in the future, old people living alone will

constitute an even larger proportion of all elderly than

they presently do (Waddell, 1976).

Health.--A recent survey of the Michigan Office of
Services on Aging (1975) revealed that as a group, elderly
individuals living alone had the highest percentage re-
porting personal injury and illness, were the least able to
walk around their house, walk outside, and go up and down
stairs, and more frequently reported health care as a
problem. In terms of medical care, Rosow (1967) found that
13-36 percent of the elderly living alone had no one to
care for them during long term illnesses compared to 6-15
percent of the elderly not living alone. Possibly more
representative of the day to day health care situation of
this group is the fact that 65-85 percent of the elderly
who lived alone compared to 21-35 percent of the elderly
not living alone reported having no one to care for them

during short term illnesses (Rosow, 1967).

Mental Health.--There is considerable evidence that

 

old pe0ple living alone are inordinantly more susceptible
to mental health problems than their peers. Lowenthal
(1965) found that although only 32 percent of the elderly
population of San Francisco lived alone, 50 percent of the
admissions of over 65 year olds to a San Francisco Psychia-
tric ward lived alone. It is difficult to accurately

appraise whether this situation is truly due to a higher

prevalence of mental health problems for this group since,
as was stated earlier, individuals who live alone generally
have fewer at home care options. Nonetheless, the trend

of higher utilization of mental health care for this group
seems to persist even in outpatient services. In the
Lansing, Michigan area 62 percent of the over 60 year old
outpatient admissions at St. Lawrence Community Mental
Health Center (CMHC) lived alone, as compared with a commu-
nity wide representation of only 32 percent (St. Lawrence

CMHC Outpatient Utilization Statistics, 1974).

Income.--The disadvantaged state of old people
living along is probably most evident when considering
income statistics. Scott reveals that 89 percent of the old
people living alone have incomes below $3,000 and 62 percent
have incomes below $1,500 (1971). When one considers this
meager financial state in the context of the previously
cited high rate of health problems, the day to day uncer-
tainty and vulnerability with which this group must live is

quite evident.

Social Isolation and Problem Solving.--While some

 

elderly living with spouse or others experience isolation,
some evidence suggests living alone and isolation are
almost synonymous in old age. Turnstall (1966), using a
complex scaling system, reports that his data suggests

that 68 percent of all old people living alone are socially

10

isolated. However, he goes on to report that this number
constitutes 92 percent of all socially isolated old people.
The resulting psychosocial problems incurred (i.e.,
loneliness, feelings of helplessness) when elderly isolates
fail to receive interpersonal support has been pointed out
by Bennett (1973) in her research with elderly isolates.
In some respects, however, the frequent social isolation of
old people living alone can have considerably more far
reaching effects than simple loneliness. As has been
pointed out by the Institute for Interdisciplinary Studies
(115), "other people can serve as a resource to.aSSist in
solving problems. This is particularly the case with the
elderly who experience a general decline in health and
mobility. Friends and relatives can, through their support,
make c0ping with a complex environment easier for aged
people" (IIS, 1972). Thus it would seem to follow that old
peOple living alone would be at a distinct disadvantage
compared to their peers in coping in a number of potential
problem areas like transportation, nutrition, socializing

and use of leisure time.:3

TEAssessing Local Need

 

As the previous pages show, there is no lack of
evidence documenting the plight and needs of the elderly
in general and the elderly who live alone specifically.
Although this kind of documentation points to a need, it

nonetheless falls far short of the specificity needed in an

11

applied setting to effectively plan programs capable of
dealing with the multiplicity of needs which might exist
for the elderly in a specific community. Despite the
expansion in funding for programs for the elderly, alluded
to earlier in this introduction, the amount of resources
presently available (and likely to be available in the
near future) are limited.

Thus every local, municipal and federal agency
targeting their services to the elderly must establish
some priorities for social programming or risk wasting
scarce resources on low priority needs. As a recent Admini-
stration on Aging publication put it, "When many problem
areas are identified, how should an agency proceed? It is
likely that funds will not be available to effectively
attack all problems, so which should be addressed?"
(USDHEW, 1975). The obvious answer to the dilemma of
prioritizing social programming for the elderly is to
develOp re3ponses to problems which constitute the greatest
needs. Although such a position is sound, it clearly
avoids the real dilemma of how to judge or measure needs.

With the current trend toward revenue sharing,
many local communities are having to grapple with local
autonomy in program planning and implementation for the
first time. Given this new autonomy, should states simply
follow the federal lead and cues on priority needs and

municipalities follow suit? If not, how should states and

12

municipalities make their programmatic decisions? The
argument against following the federal lead in programming
decisions is the same as the argument in favor of revenue
sharing: "If it worked and was needed in Peoria, it might
not work nor be needed here!" Local social programmers are
more and more becoming aware of the unique needs of their
own community. Acknowledging this situation, many states
and communities have taken steps toward instituting locally
based need assessments. Siegel et a1. (1974) has called
such need assessments "one of the informational inputs
necessary to select priority problem areas and leading to
the subsequent development of programmatic responses."

Most communities have readily adOpted the philoSOphy of
program planning based on assessment of local needs,
although sometimes in a perfunctory manner merely to comply
with the numerous funding laws like PL 89—749 (Health
Planning and Public Services Amendments) which now require
such assessments (Warhiet, 1976).

The persuasiveness of the arguments and rationales
behind using assessments of need to help make program
planning and implementation decisions rest in large part on
a clear and specific conceptualization of what constitutes
a need. [While some may equate the presence of a problem
with a "need," in point of fact they are not always iden-
tical. A.more meaningful description of the concept at

issue might be "unmet need." Herein lies the quintessence

13

of the need assessment process: what appears to be a "real
problem" to the bureaucrat may not be experienced as such
by the affected individual; in such a case no need ever
existed although the omnipotant bureaucrat might decide
otherwisef>>Further, sometimes real problems do not con-
stitute "unmet needs" as when there is already extant,
adequate and effective natural or socially constructed
supports to deal with the problem; in this case the problem
may be termed a "met need." Since the kinds of problems
faced and the countervening forces present to deal with
them are apt to differ from community to community, the
priority rankings given various problems by the indigenous
target group can be considered a good measure of real

need (USDHEW, 1975).

Quite often epidemiological studies are equated
with need assessments. While epidemiological studies are
generally focused solely on measuring the prevalence or
incidence of a particular malady, the need assessment
process goes beyond this and examines applied issues rele—
vant to instituting change in the condition under study. As
part of the "planning for change" process, one must go
beyond knowing that a need exists and is of considerable
impact. In addition, a thorough needs assessment seeks
information regarding current services available to meet
the need and factors affecting the utilization of these

services by the target population. A needs assessment may

14

also seek information concerning the target population's
willingness to use various kinds of remedies, and under
what condition (i.e., cost and transportation considera-
tions). Thus for any target population but particularly
for high risk groups (Siegel et al., 1974), including
elderly living alone, a thorough practical need assessment
and not simply an estimate of morbidity, is essential to
effective social programming.

The foregoing issues certainly do not exhaust the
possible range and breadth of the tOpics relevant to a
need assessment, and certainly do not preclude the pursuit
of more theoretically oriented questions. The fact that
the appropriateness of performing locally based needs
assessments rests in large part on the uniqueness and indi—
genousness of the situation studied and the applied utility
of the data collected, dictates that perceived needs (the
target population's identification of problems) and the
acceptability of possible interventions constitute a minimal
skeletal framework of any need assessment.

Life Satisfaction--A Theoretical
Need Construct

 

 

While the target population's reports of needs are
an essential component of need assessment and appear to be
the most persuasive of legislative opinions (Waddell, 1976),
exclusive reliance on the data provided in such measures

is not without pitfalls. Although one should not

15

precipitously ignore target group problem ratings, it should
be recognized that some areas of need may be viewed as

more or less acceptable, or on the other hand, stigmatizing
by survey respondents. For instance, someone who had trans-
portation needs, and also felt alone and isolated (inter-
personal needs) might readily report the former, but might
fail to report the latter because of embarrassment or a
feeling of futility about their interpersonal needs.

Since relying solely on a "straight forward"
approach to assessing needs may bias ratings for certain
socially suspect or personally embarrassing needs, a careful
and complete need assessment strategy should include a
meaningful need related dependent measure. Such a measure
could then become the dependent variable upon which the
researcher could calculate regression equations for one
or a number of variables. Within the context of socio-
logical and psychological research on the elderly, various
life satisfaction measures have been used extensively as a
global assessment of the quality of life of older individ-
uals and have been highly recommended for use as an overall
measure of need (USDHEW, 1975). Use of such a measure also
provides the opportunity to perform multivariate examina-
tions of the data focused on identifying "outcome" related
variables, and can provide the basis for evaluating the
effectiveness of interventions targeted at remedying the

problem(s) at hand. Obviously, the value of such "outcome"

16

evaluations would depend upon the meaningfulness of the
outcome measure chosen. Fairweather emphasizes this point
in his work on experimental social innovation (1968) indi-
cating that only measures of proven validity (by virtue of
their association with the "social change outcomecmiterion")
are likely to be accepted by or produce much practical value

for society and its members.

General Theory of Stress and Life Satisfaction.--It

 

may be helpful at this point to view the determination of
Life Satisfaction within the context of an overriding con-
ceptualization of adaptation to stress.

Troubles, defined by Ross (1946) as "situations
outside the normal pattern of life . . . situations which
block the usual patterns of activity and call for new
ones," are common to all individuals and occur in all age
groups. Although individuals of different sexes, socio-
economic positions, and age groups tend to experience some-
what different troubles and classes of troubles (Dohrenwend
and Dohrenwend, 1969), the existence of a singular process
of adaptation to stress (or troubles) by all individuals
has been commonly held by various psychological and socio-
logical theorists (Wohlhill, 1970, Dohrenwend and Dohren-
wend, 1969).

Within their paradigm, Dohrenwend and Dohrenwend,
acknowledge the typically disruptive effect experiencing a

major personal or social stressor has on the individual.

17

However, basing their formulations on the physiological
work of Seyle they further posit the potential for other
factors in the individual's life to mediate (negatively or
positively) the effects of the stressor. Their conceptuali-
zation of an individual's adaptation behavior contains
three basic elements: (1) antecedent stressors, (2) ante-
cedent mediating factors (either internal or external to
the individual), and (3) the consequent response--which may
be adaptive or maladaptive (Dohrenwend and Dohrenwend,
1969). Thus an older individual's consequent response (in
terms of personal life satisfaction) to the problems they
experience may be conceived of as depending on the internal

and external mediating factors Operating in their lives.

Stressor--Personal and Social Losses.--Rosow (1973)

 

has called the various status and personal losses exper-
ienced by the elderly as a chronic prolonged and cumulative
crisis, with no comparable experience in any other age
group. Loss of job, close relative or friends, or spouse
has frequently been shown to be associated with low or
reduced life satisfaction.

Comparisons of life satisfaction among employed
and retired old people has generally shown higher morale
for employed individuals (Riley and Foner, 1969). Both
Thompson (1973) and Kutner et a1. (1956) have reported

higher levels of life satisfaction in employed old people

18

while Strieb (1956) comparing individuals with similar
levels of health and SES reported similar findings.

Loss of close relative (Lowenthal and Boler, 1965)
or spouse (Kutner, 1956) has frequently been associated with
greater unhappiness, worry and lowered life satisfaction.
However, Kutner (1956) has pointed to evidence suggesting
that the impact of familial and friendship losses is quite
temporal and is generally diminished with the passage of
time. Another qualification about the influence of widow-
hood on life satisfaction is the work of Neugarten (1964)
and Leviton (1976) indicating that this loss is most demorr
alizing when it is developmentally inappropriate. Thus,
this would suggest that the elderly for whom widowhood is
an expected developmentally appropriate loss may be less
affected by the loss of a spouse or various significant

others than younger individuals.

Stressor--Hea1th.--The psychological and socio-

 

logical literature is replete with studies citing the
relationship between health and life satisfaction in old
age. Although a number of studies have demonstrated the
positive associations of objective medical evaluations and
life satisfaction (Maddox and Eisdorfer, 1962; Pallmore and
Luikart, 1972), reports of significant correlations between
self health evaluations and life satisfaction are much more
common. When both are examined in the same study, results

have generally shown self health evaluations to be the

19

better predictor of the two (Bultena and Oyler, 1971). This
does not imply the two measures are unrelated, in fact,
objective health measures have at times been reported as

the best predictor of a respondent's self health evaluation
(Pallmore and Leukart, 1972). The number of studies citing
self health evaluation as the largest correlate of life
satisfaction is considerable (Lowenthal, 1965; Fowler, 1969;

Thompson, 1973; Maddox, 1963; Kutner et al., 1956).

Mediator--Individual Personality.--Research has

 

shown that individuals with favorable self images tend to
report higher life satisfaction despite the absence of other
factors normally associated with high morale (Riley and
Foner, 1969). Others report that individuals who regarded
themselves as disadvantaged (Kutner et al., 1956; Morrison
and Kristjanson, 1958) or old (compared to middle aged or
young) exhibited lower levels of life satisfaction than
their peers. Additionally, Pallmore and Leukart (1973),
have reported that individuals scoring high on internal
locus of control tend to have higher life satisfaction
scores. These idiosyncratic trends do not seem unusual
given the acknowledged association of personality charac-

teristics with various measures of adjustment.

Mediator—-Demographics.--Not unexpectedly a number

 

of sources identify income or socioeconomic status as an

important determinant of life satisfaction in old people.

20

For instance, Fowler (1969) found that income had a posi-
tive association with life satisfaction, while Edwards and
Klemmock (1973) reported income had twice the correlation
to life satisfaction of the next largest predictor (self
health evaluation). Paradoxically, there is considerable
evidence that the correlation between amount of money and
happiness is lowest in the older age groups (Riley and
Foner, 1969). Education which is usually used in computing
SES has independently been shown to correlate with life
satisfaction, however, there is very little evidence to
suggest that other demographic characteristics signifi-
cantly affect life satisfaction (Lowenthal, 1969; Riley and

Foner, 1969).

Mediator--Activities.--There is evidence that low

 

levels of overall activity are associated with low life
satisfaction and high rates of interaction and contact with
the environment are associated with high life satisfaction.
Various studies have shown positive relationships

between life satisfaction and the following: overall scores
on activity (Maddox, 1963), scores of interaction or number
of roles played, (Tobin and Neugarten, 1961) and nonsocial
activity (Kutner et al., 1956). The inhibitory effect of
various losses on activities is well accepted and may be a
factor here. Lowenthal (1969) has reported that withdrawal

(from activity) preceded by some kind of external loss

21

(e.g., health, job or spouse) is associated with lowered
life satisfaction while simple withdrawal is not.

The possible impact of activity on life satisfaction
is not simply resolved. A reduction in activity in a
single area (i.e., interpersonal), need not be associated
with lowered life satisfaction if compensated for by acti-
vity and involvement in other areas (Anderson, 1968;

Maddox, 1963; Bultena and Oyler, 1971).

Activity Theory vs. Disengagement Theory
In the main, most of the studies just described
support a theory of aging adjustment called the activity
or role theory. The activity theory posits that there will
usually be a reduction in overall life satisfaction in old
age. This would be due to the unavoidable losses of later
life and the subsequent reduction in social participation
and specific roles which almost necessarily accompanies
these losses (IIS, 1972). George Maddox (1971) has stated
this position as follows:
The social self emerges and is sustained in a most
basic way through interaction with others. The per-
sonality processes of ego involvement and object
cathexis are thus inextricably related to the demands
and constraints set by social structure: consequently,
structural constraints which limit or deny contacts
with the environment tend to be demoralizing and
alienating and to be associated with withdrawl or
various forms of aggressive behavior.

Thus activity theorists predict that the generally unavoid-

able reduction in activity experienced by older people by

22

virtue of the onset of various losses leads to lower

life satisfaction.

Table l.--Activity and Disengagement Predictions of
Association to Life Satisfaction.

 

 

Extrinsic Activity Theory Disengagement
Losses negative r LS Zero r LS
Activity (hi) positive r LS Zero r LS
Socioeconomic (hi) positive r LS Zero r LS
Health (good) positive r LS Zero r LS

 

In contrast to the activity theory of aging, the

{disengagement theory (Cuming and Henry, 1961) posits a
developmentally acquired response to stress and losses,
whereby with increasing age the elderly initiate their own
withdrawal from societal roles. Furthermore, in contrast
to activity theory the disengagement theory postulates that
this withdrawal leads to undiminished life satisfaction in
the elderly. Although few people would be surprised that
research indicates that internal personality characteristics
can mediate an individual's response to stress (as measured
by life satisfaction), the possibility that older people in
general acquire a disposition which changes or eliminates
the impact of various stressors on life satisfaction is
seen by many as unprecedented and antithetical to the

notion of a universal response to stress.

23

Beyond a doubt, activity and disengagement theories
of aging are contradictory. That this is the case is made
quite clear by an examination of the hypotheses one would
expect from these theories about the effect of various
variables on life satisfaction presented in Table 1.

When one examines the wealth of research done on
this area (including most of the work cited in the previous
section), it predominantly supports the activity theory.
Support for the disengagement theory of aging has not been
as strong. Riley and Foner (1969) have termed the evidence
from Cumming and Henry's (1961) initial work "suggestive
but highly tenuous" and cite only two additional studies
supporting their position. The first by Maddox (1966)
identifies a minority of respondents who exhibited the ex-
pected combination of low activity high morale. The other
study (previously cited) by Lowenthal (1969) found only

involuntary withdrawal to be associated with low life

 

satisfaction, meaning voluntary withdrawal, as disengage-
ment would suggest, has no association with life satis-
faction.

Although one of the difficulties in validating the
disengagement theory is the fact that no one has opera-
tionalized a universally accepted definition of the state,
Tobin and Neugarten (1961) and to an extent Cumming and
Henry (1961) agree that disengagement should coincide with

the older rather than the younger segments of the elderly.

24

In a study designed to test this notion, Tobin and
Neugarten (1961) found that there was actually a stronger
association between activity and life satisfaction among
the oldest of the elderly than was true for a younger
elderly cohort.

Since the present study presents a unique opportu-
nity to replicate the results of the study just cited, an
examination of the relative merits of disengagement and
activity theory will be undertaken by calculating separate
correlations corresponding to Table l, for respondents over

75 and respondents under 75.

The Present Research

 

The present study has two purposes in mind: (1) a
local assessment of need for the elderly individual living
alone in Lansing, Michigan, and (2) the testing of several
hypotheses concerning the relationship of several variables
with life satisfaction. Since each calls for a different
tactic of inquiry, it would seem most appropriate to con-
sider these two aspects of the study separately.

Thus, the examination of local need will be achieved
by fulfilling the following objectives:

Objective 1. To ascertain to what degree oldypeople living

 

alone in Lansing, Michigan, appear to be the

 

victims of various problems associated with

 

old age.

Objective 2.

Objective 3.

Objective 4.

25

To ascertain the prioritized importance of

 

various problems to these same respondents.

 

To identify what services among a large group

of fpossible" services are viewed as accept-

 

able interventions by these same respondents.

 

To ascertain if there is any empirically

based grouping or clustering of variables

 

which would represent a simplification of the

complexity of the situation under study.

 

In addition, this study will attempt to confirm the

following hypotheses:

 

Hypothesis 1.

Hypothesis 2.

Hypothesis 3.

Hypothesis 4.

Life Satisfaction will be significantly

 

related (negatively) to reports of recent

 

(previous two years) age related losses.

 

Life Satisfaction will be significantly
related to various measures of Health. .
(Sickness Rating (negative), Total Illnesses
(negative), Self Health Evaluation (positiveh

Doctor's Visits (positive).

 

Life Satisfaction will be significantly
related to various measures of Interpersonal

Involvement.

 

Life Satisfaction will be significantly

 

related to various measures of Leisure

Involvement.

 

Hypothesis 5.

26

Consistent with activity theory, there will

 

be a significant negative correlation be-

 

tween life satisfaction and losses, and 81g:

 

nificant positive correlations between life

 

satisfaction and activity, income, and health

 

for respondents under 75. Consistent with

 

disengagement theory, there will be nonsig-

 

nificant correlations between life satis-

 

faction and losses, income, activity, and

 

health for respondents 75 or over.

 

METHODS

Sampling

The respondents used in this study were identified
for the author by various community groups in the Lansing,
Michigan area. In order to get a representative respondent
sample, a diverse group of "referral sources" were used in
identifying potential respondents. The referral sources
for interviewees included: Greater Lansing Area Transpor-
tation Clearing House (G.L.A.T.C.H.) (N = 12), local
churches (N = 28), Tri County Information and Referral
Program (N = 4), Retired Senior Volunteer Program (R.S.V.P.)
(N = 2), Visiting Nurse Association (N = 1), Lansing Senior
Center (N = 6), United Auto WOrkers Union (N = 4), Parks and
Recreation (N = 2), Easter Seal (N = l), and Senior Housing
Offices (N = 10).

Each "referral source" was asked to select and re-
fer a diverse group of elderly individuals over 60 years of
age, who also lived alone. Further, each organization was
asked to select potential respondents who would be repre-
sentative of the entire spectrum of old pe0ple living alone
(i.e., problem stricken individuals to exceptionally

functional individuals). During the course of completing

27

28

this study, 95 respondents were referred to the author for
interviewing. Ultimately, 70 respondents were found who
met the selection criteria of the study and agreed to be
interviewed. Since this number of respondents was seen as
the minimum necessary to justify most of the study's
analyses, no sampling of respondents was done. All 70
potential interviewees were surveyed for the study.
Although the sizable variance present in almost all
the responses obtained in this study would seem to indicate
that efforts at obtaining a diverse sample of old people
living alone were at least partially successful, it should
be pointed out that the generalizability of this study is
limited by its use of a nonprobability selected sample of

respondents (Warheit et al., 1975).

Initial Contact

 

Since each of our referral sources felt they were
sufficiently well known to their referee to simply furnish
names without first contacting individuals, the initial
telephone contact about interviews was made by one of the
studies' two interviewers.

Names and telephone numbers were provided to the
author during the intial phase of the project's gearing up
in June 1975. By July a list of ninety-five (95) potential
interviewees had been put together, which constituted the
source for all the interviews included in this study. Of

the initial list of 95 potential interviewees, nine were

29

unavailable and were never successfully contacted, and 13
refused to participate. The remaining 73 referrals all
agreed to participate in the study and signed informed
consent forms (Appendix A) acknowledging both this per-
mission and their understanding of the purpose of the study.
In the course of the interviewing process, three
(N = 3) respondents were found to be residing with someone
at the time of the attempted interview and thus were ex-
cluded from the interviewing process and this study. At
the completion of the interviewing process, seventy (N = 70)
individuals who were both sixty years of age or older and

lived alone cooperated in this study.

Interviewers

 

Each of the two interviewers for this study sur-
veyed 35 respondents. Interviewer "A," the author, worked
for the Tri County Office on Aging and was also a graduate
student in psychology. Interviewer "B," an undergraduate
student in social work, was hired specifically to assist
in the interviewing. Since interviewer "B" began her
interviewing two weeks after interviewer "A," efforts had
to be taken to insure the equivalence of the content and
. format of their interviewing. This was done by having
interviewer "B" study a tape of one of "A"'s actual inter-
views. Additionally, "B" performed several (5) pilot

interviews with cooperating volunteers (all old people)

30

until these interviews were judged by "A" (the author)

equivalent to his.

The Interview

 

All but three of the scheduled interviews took
place in the home of the respondent. (These three took
place in the Tri County Office on Aging.) Appointments
for the interview were set at a time convenient to the
interviewee, generally in the afternoon, and commenced
after the interviewer showed proper identification and
reacquainted the respondent with the purpose of the study.
The interviewer read each question to the interviewee,
read the apprOpriate response categories and repeated
questions and response categories when requested (the
questionnaire appears as Appendix B). No interpretation
of items on the questionnaire was provided by the inter-
viewers and when so questioned, respondents were instructed
to answer each item as best they could. The length of time

for interviews ranged from 30 minutes to 105 minutes.

Description of Variables

 

This study is based on data collected from respon-
dents with a one hundred and fourteen item questionnaire.
All items were either members of one of several scales used
in the questionnaire or were conceptually members of one of

several specific areas of concern to this study.

31

Table 2 lists the variables and the questionnaire
'location of the individual items used in calculating a

score for that variable.

Demographic Measures.--An assortment of demographic

 

data was collected in the course of administering the
study's questionnaire (items #1-17, Appendix B). The
following demographic data (accompanying numbers correspond
to questionnaire enumeration used in Table 2, and Appendix
B) was collected for all respondents: years residing at
present address (#2), years residing in Lansing area (#3),
type of resident (#4), age (#5), sex (#6), marital status
(#7), years of widowhood (#8), household composition (#9),
education (#10), race (#11), present employment (#12),
occupation (majority of life) (#13), spouse's occupation
(#14), religion (#15), number of living children (#16)

and income (#17). This information was included to assess
how much respondent's individual characteristics determine
behavior or attitudes measured by various questionnaire

scales or items.

Leisure Involvement.--Respondents were questioned

 

about their participation in a number of specific leisure
activities (#18-31) yielding a Leisure Activities Scale.‘
In addition, data was collected on hours of daily leisure
participation (#32) and satisfaction with leisure parti-

cipation (#33).

32

mv .hv .mw .mv
we .Nv .oe .mm
mv .He .am .bm

>uw>wuo¢ pomucoo occammaoe
mufimfl> mo mocmsowum :ufi3 cofluommmwumm
auw>wuo< mcwuwmw>

 

 

 

 

mm .mm .em mode: mm canuw3 mumuuommsm
ucoEm>Ho>sH HmcomummwwucH
mm nuwz :oHuommmwumm
mm muse:
Hmnma oamom >uw>fluo¢ chanson Amo HmnEszv
uucmsm>ao>cH ousmflmq
ha oeoocH
ma couoawno mo umnEsz
ma cowmwamm
vH .MH :owummsooo
NH newshonem
HH comm
ca :oflumosom
m ooon3oow3 mocwm cede
h msumum Hmuwumz
m xmm
m mmfl
v cocooflmmm mo maze
m mammcmq cw oEHB mo numcwq
m mmmuooe ucmmoum um mafia mo sumcmq
"moanmwnmosmo
m xwocommfi so muonssz coflummso moanmaum>

 

muflmccowumoso so ommmmmmfl moanmwum>nl.m canoe

33

boa
woa
moa
woa
moa
NOH
HQH
OOH
mm

mm

hm

mm
em

mm
Hm
om
mm
mm
mm
mm
mm
em
mm
mm

mHmEoumso mafiommamwz mo0fl>uom
coon cam sowuwhusz

oEqu

mofluw>fluo¢ mafia mummm
mowuflcsuuommo ucoESOHmEm
GOHHMGAEfiHomAQ mom
cowumuuommcmue

ocflmsom

sumo spasm:

mEoocH

”mfimanoum

meow oumnm

mama mcfimsom 3oz

mamm mcwmmonm

mnmooama

mama mcwuwmfl>

moansoom Hafioom so woe>ofl
xuoz HomucsHo>
ucmsmoaosm mama
cofiumuuommcmua Honocmw
coed Hmmmq

cH unmsonm mammz

mama mcflcmmau cocoa

meow um anemone Hmowmmnm
omusz oumu meom

mucmEuooue anemooz How cowumuuommcmue

meHnoum new Honmcsoo

 

"mmD OUH>HOW HMHflfimflOm

 

xfiocmmmm so mumnesz coflummoo

moanmwum>

 

.omscwuc00|l.m manna

34

 

amt

maalmoa

om
wk
an
55

molmv

mo
mhlwo
mm .mm
mm

QmD OUH>HOW m50fl>whm

 

”sOAumNHHHuD mow>nmm

mwzlamz

 

"moumum Hmucmz
boo mo mmoq

Honemz mHflEmm mmoHU mo moon
guano: mo mmoq

museum mo moon

”mommoq

«qulamq

 

"cofiuommmfluom omen

muwmfi> m.uouooo mo Honesz

mommmcHHH no qu652
mcwumm mmmchAm

coaumsam>m spasms «How

”nuance

 

xflocmmme so mHoQEdz coflummso

moanmflum>

 

.omscfluc00|l.m canoe

35

Interpersonal Involvement Scales.--As with leisure

 

participation, a scaled multi-item assessment approach was
taken in measuring this aspect of our respondent's life.
Questionnaire items #34-#53 (See Table 2 and Appendix B)
were concerned with various aspects of our respondent's
interpersonal involvement and satisfaction.

Respondents were asked to respond to questions
covering three aspects of their interpersonal involvement:
Supporters within 25 miles (calculated by totaling number
of children (#34), relatives (#35), and friends (#36)
within 25 miles), Visiting Activity (calculated by totaling
visits with children (#37), relatives (#39), friends (#41)
and organizations (#43)) and Telephone Contact Activity
(calculated by totaling telephone contacts with children
(#45), relatives (#46), friends (#47) and organizations
(#48)). Similarly, respondents provided responses consti-
tuting a Satisfaction with Frequency of Visits Scales
(calculated by obtaining their mean satisfaction response
to their visits, with children (#38), relatives (#40),
friends (#42), and organizations (#44)).

Preliminary analysis of the study data indicated
only a small number of respondents (N = 8) had any contact
with organizations so this variable was dropped from the

data analysis.

Health.--Once again, in the interest of comprehen-

siveness a multi-item assessment approach was taken in

36

measuring the health of respondents. This approach seemed
particularly called for here since previous research has
indicated discrepancies between self-evaluation and other
measures of health in predicting life satisfaction scores.
Respondents were questioned about their evaluation of their
present health (#76), number of chronic illnesses (#66-75),
number of doctor's visits (#65) and their sickness rating
(#63, 64). The sickness rating is computed by multiplying
the number of days a respondent reports being sick (#63)

by a weighting factor for the severity of their impairment
(#64). Being "in the hospital" has a weighting of 1.5,
being "in bed at home" has a weighting of 1.0, being "just 43
at home" has a weighting of .5 and "not sick" has a zero
weighting. The month preceding the interview was the

specified time span for computing the sickness rating.

Life Satisfaction.--A modified version of Neugarten,

 

Havinghurse and Tobins (1965) Life Satisfaction Scale was
used and constitutes the principle outcome measure in this
study. The fourteen items belonging to the Life Satis-
faction Scale (#49-62, Appendix B) were statements about
the respondent's life, to which he/she responded either
agree (described my life) or disagree (does not describe
my life). If respondents felt unable to respond either
"agree" or "disagree," they were told to respond with the

answer which best described their situation. This set of

37

items yielded a quite acceptable internal consistency

coefficient of .740 (Alpha).

Losses.--Since the literature suggests that only
recent losses are likely to be of significance to the
affected individual, respondents were questioned about any
recent (previous two years) age related losses they might
have experienced (items #77-80). Specific loss categories
were: death of spouse (#77), death of a close family member
(#78), health loss (#79), and job loss (#80). Job loss
was eventually dropped from the study's analysis since
none of our respondents reported leaving or losing employ-

ment during the specified previous two years.

Mental Status.--The high incidence of mental dys-

 

functioning and impairment in old people discussed in the
previous section is a need related issue itself and can
undermine the validity of responses in a self report
questionnaire like this one, which must assume substantial
verbal and cognitive ability on the part of respondents.
Katz's Mental Status Quotient (1974) (Items #108-113,
Appendix B) was incorporated into this study to measure
this mental health problem area and account for the possi-
bility that elicited responses were artifacts of some
organic or functional brain impairment on the part of
respondents. The six items used to form this scale

required respondents to provide the fOllowing information:

38

location at the time of interview (#108), day of the week
(#109), month of the year (#110), current year (#111),
present president of the United States (#112) and previous
president of the United States (#113). Preliminary tabu-
lations of the study data indicated only two respondents
scored below a perfect 6 on this scale (both of them re-
ceiving a score of 5). Thus this scale was dropped from

subsequent analysis.

Previous Service Use.--The study respondent's present

 

and previous use of services was assessed by asking respon-
dents to name (in Open ended format) (See item #81, Table 2,
and Appendix B) the services they had previously used
(within last year). Regrettably, the open ended elicitation
of the services our respondents used proved confusing and
yielded unreliable and contradictory data. For example,
respondents referred to the study by various service pro-
grams often failed to mention their use of that service.
Therefore, the service use variable was also dropped from

subsequent analysis.

\‘x"

Potential Service Use.--A modified version of

 

Maddox's service needs scale (Katz, 1974) was used in this
study. The response categories were changed from Maddox's
in an effort to make them as behaviorally anchored as
possible. Thus instead of asking respondents if they
thought service "X" would be of help, respondents were

asked to reply to the 15 services (Items #82-97,Appendix B)

39

stating: "definitely would use," "might use," or "probably
would not use." The fifteen services asked about were:
counseling for problems (#82), transportation for medical
treatments (#83), home care nurse (#84), physical therapy

at home (#85), house cleaning help (#86), home meals (#87),
legal aid (#88), general transportation help (#89), employ-
ment finding service (#90), volunteer work (#91), advice on
benefits like Social Security (#92), a home visitor (#93),
home telecare (#94), shOpping assistance (#95), assistance
in finding new housing (#96) and assistance in getting some-
one to share living quarters (#97). (The original purpose in
using this set of items was to simply yield a frequency
tabulation of preferred services, however, the items seemed
conceptually to encompass a single homogeneous factor: 21
gessraLneed-wor-willieasssswts-959. ssrgigeS-> This was sub-
stantiated by its sizable alpha coefficient of .866. Con—

sequently, these individual items were also used to form a

Service Use Scale.

Problems.--Respondent needs were assessed in rela-
tion to a number of Problem Areas. For this 10 item instru-
ment (Items #98-107, Appendix B), respondents replied
either, "No problem for me," "somewhat important problem
for me," or "very important problem for me." The ten
problem areas covered by this scale were: income (#98),
health care (#99), housing (#100), transportation (#101),

age discrimination (#102), employment opportunities (#103),

40

spare time activities (#104), crime (#105), nutrition and
food (#106), and services and businesses misleading their
users (#107). Since the internal consistency of this scale
was also quite respectable (alpha = .688) and seemed to
suggest a Total Problems factor, a scale score for it was

also computed and used in the study's analysis.

Data Analysis

 

Analysis of the data was performed using a varied
assortment of statistical techniques. Simple frequency
counts and descriptive statistics were used with much of the
demographic and direct need assessment questions, yielding
straight forward estimates of central tendency and per-
mitting relative rankings of various services and problems.
In addition, various hypotheses and inferred relationships
among variables were tested by correlative analysis inclu-
ding multiple regression. Finally, the cluster analysis
technique developed by Tryon and Bailey (1970) was used to
examine the underlying structural and associative makeup

of the study variables.

RESULTS

With the possible exceptions that they tend to be
somewhat older, more frequently female and more frequently
widowed than others in their age cohort, old people living
alone when compared to their aged peers are generally unre-
markable in most other characteristics. This was the case
for the group of respondents surveyed for this study.

As expected, the respondents surveyed were con-
siderably older than the minimum criteria age of 60, having
a mean age of 74.67 years (See Appendix C for detailed demo-
graphic breakdowns). In addition, respondents were predo-
minantly female (85.7%) and widows and widowers (78.6%)
although not having recently acquired that role (average
length of widowhood = 10.96 years).

Although detailed breakdowns of the demographics for
senior citizens in the Lansing area were not available,
respondent characteristics seemed almost stereotypically
familiar. Most respondents had gone to but not completed
high school (mean years of schooling = 10.21). Congruent
with local racial breakdowns, respondents were predominantly
caucasion in race (97.1%). Over seventy percent (71.4%)

reported being Protestant, while 24.3% were Catholic. In

41

42

general, most respondents had lived for quite some time at
their present address (mean years = 16.41), had lived in the
Lansing area for the greater part of their life (mean years
= 48.76), and generally owned their own home (55.1%). Of
the 65 respondents who reported ever having been married,
50 reported having living children yielding a very American
average of 2.4 living children per family group (N = 65).

As expected, few respondents reported being employed
(11.4%), with most living on meager gross yearly incomes
(mean income = $3,670.00). Since gross income is a rela-
tively poor indication of a person's socio-economic status
(SES), particularly for an elderly individual past his/her
earning prime, Hollinghead's (1957) two factor Index of
Social Position was computed for all respondents. This
scaling technique produces a single scaled SES value based
on ratings assigned for various occupations and levels of
education. These ratings are then weighted to produce a
value between 11 and 77, 11 being the highest SES score.
On this scale, our respondents showed considerably more
variability in social position than income would suggest.
As indicated in Appendix C, 11.3% of our respondents fell
into Group I (high SES), 27.4% were in Group II, 34.0%
were in Group III, 11.2% were in Group IV and 16.1% were in
Group V (low SES).

Because of the relatively small sample size used in
this study (N = 70), certain variables had extreme

distributions (Race, Sex, and Employment). The resulting

43

small N for blacks, (N = 2), males (N = 10) and employed

(N = 8) thus restricts the generalizability of the results
to these groups. Hence, in order to avoid drawing erroneous
conclusions based on small Ns, analysis of the data using
these variables for cross-tabulations or correlations has
been avoided.

Object 1. To ascertain to what degpee old peOple living
alone in Lansing, Michigan appear to be the victims of

various problems associated with old age.

 

Health.--As was cited in the introduction, health
is consistently reported by old peOple to be an area of
great difficulty. To try to substantiate this, several
aspects of our respondent's health were examined in this
study. Breakdowns of responses to various health related
items are presented in Table 3.

The effects of declining health and the prevalence
of various illnesses among our study sample is quite
striking and points to a real problem and potentially a
real need. An indication of this is that respondents re-
ported an average of 3.04 illnesses. So common were these
various chronic ailments (Diabetes, High Blood Pressure,
Heart Trouble, Stroke, Arthritis, Stomach Ulcer, Emphysema,

Glaucoma, and Cancer) that all 70 of the respondents in the

 

study reported the presence of at least one illness at the

 

time of the interview. Arthritis (67.1%) and High Blood

 

Pressure (47.1%) were the most commonly reported ailments.

44

Table 3.--Breakdown of Health Variables

 

 

Variable Percentage (N) Mean
(Count) Responding (Standard Deviation) Median
Doctor's Visits 12.24 8.20
(per year) (15.35)
(70)
Illness
(70)
-Diabetes 15.7% (11)
-High Blood Pressure 47.1% (33)
-Heart Trouble 41.4% (29)
-Stroke 5.7% (4)
-Arthritis 67.1% (47)
-Stomach Ulcer 11.4% (8)
-Emphysema 8.6% (6)
-G1aucoma 20.0% (14)
-Cancer 8.6% (6)
Total Illnesses ---- 3.04 2.89
(70) (1.37)
1 11.4% (8)
2 25.7% (18)
3 32.9% (23)
4 12.9% (9)
5 12.9% (9)
6 2.9% (2)
7 1.4% (1)
Self Health Evaluation
-Poor 14.3%
—Fair 40.0%
-Good 38.6%
-Excellent 7.1%
Sick Past Month
-Yes (24) 34.3%
-No (46) 65.7%
Sickness Rating:
Past Month
(70)
(number of days 5.06 .261
sick x weighting (9.73)

for degree of
incapacitation

- see items 63-64
Appendix A)

45

Results obtained from the other health measures
seems to substantiate the prevalence of health problems
among our respondents. Fifty-four percent of all respondents
rated their health as either poor or fair on the Self Health
Evaluation item. On the other hand, although health prob-
lems seem wide spread in our sample, so too does frequent
contact with physicians. Respondents reported an average of
12.24 doctors visits per year attesting to the accessibility
of medical care to our population. The frequency of these
visits establishes the physician as the key contact person
with this group of old people, since only one respondent
reported not seeing a physician in the past year.”

Despite the apparent availability and use of
medical care by our respondents, the effects of poor health
seem to persist among our sample, a fact substantiated by
the Sickness Rating scores (based on reports of the number
of days sick during the preceding month multiplied by a
weighting factor determined by whether just sick at home
(.5), sick in bed at home (1.0) or sick in the hospital
(1.5)). Although the mean Sickness Rating was only equal
to 5.06 out of a possible high score of 45, this score
actually indicates that a sizable proportion of our sample
were sick during much of the preceding month. The mean
Sickness Rating of 5.06 was based on over one third of all
respondents (34.3%) reporting they were minimally "sick at

home" some of the previous month. Clearly, despite the high

46

utilization of medical care reported by our respondents
health seems to constitute an important day to day problem

area for our respondents.

Interpersonal Involvement Measures.--Since the mean

 

scale scores for Visiting Activity, Supporters Within 25
Miles and Telephone Contact Activity, were biased upward by
a few socially affluent individuals, median scores are
generally referred to below to give a more accurate estimate
of central tendency.

Breakdowns for all Interpersonal Involvement
measures (scales) are given in Table 4, while a detailed
breakdown of all interpersonal involvement items is given
in Appendix D.

Respondents reported having a median of 15.50
supporters (children, relatives, and friends) available to
them within a 25 mile radius. This statistic, however,
fails to reveal the fact that 12 respondents or 17.1 percent
of our sample had only five or fewer supporters available
to them in the immediate area.

Respondents reported a median of 28.16 visits for
the previous month. Clearly, most respondents seem to fare
quite well on Visits as with our previous measure, Sup-
porters Within 25 Miles. Still in all, there is consider-
able variance on this variable as illustrated by the fact

that nine of our respondents (12.9%) received five or

47

 

 

 

 

Aamv mm.vv +Hm
“may wh.vm ow I am
Amn.mmv ANAL wv.nH on I HH
oo.wv mm.m> Amy wo.ma OH I o
Amaamo Hmuoav
>pa>fluo¢ uomucoo oconmmaoa
Ammo ws.me omflmmwumm sum>
Aamv am.vv oowmmfluom
Amm.v Ase wo.oa owflmmsummmfio
mm.m mm.m on o oowmmAHMmmfla muo>
Abouuomom cofiuommmflumm cmozv
muflmfi> mo
accosomum Qua: cowuommmwumm
Ahmv mm.mm +mm
Away wo.om mm I Ha
Amm.omv Aoav wm.qa OH I w
wH.mm nmm.av Amv wm.NH m I H AOSV
Imusms> sauces
>uw>wuod mcwuwmw>
Ammv $0.0m Hmv I ma
Amv wm.~H ma I HH
Ammav Away wo.om OH I o
om.mH nm.mv Away wH.hH m I H Amov
Ao>fluuommsm Hmuoav usmsw>ao>sH
moaae mm cwnufl3 muouuommsm HmcomuomuoucH
AcoHuMH>oQ choosmumv ocwocommom Aucsoov ousmmoz
cosmos cow: sz mmmucooumm magmanm> oaks

 

mousmmoz meH can ucoEo>Ho>cH ouomwoq .ucoeo>ao>cH accomnomumucH mo :3ooxomumII.v manna

48

 

 

 

 

 

on 0.0 Don M0 mwoa
Amy mm.ma omsomm mo mmoq
Ammo w>.mm m>flumamm mmoao mo mmoq
Ammo am.vm mmoq nuamom mommoq
Ammv w>.mm cmfimmfiumm >uo>
ANMV wh.mv owwmmwumm
“Hm.v Aoav wm.va omwmmfiummmfia
mm.m ma.m Amy wm.v oowmmwummmfio >H¢>
chanson nufl3 sowuommmfluom
mm.VM +0
Amm.Nv av.wv m I m
6.4 mm.m wm.m m I H Isaamav
III chanson mo musom
Amv wH.b va I NH
Ammv mm.mm HA I m
Rov.mv Anny wo.mm m I m
om.h v.5 Amav wv.HN m I m AOBV
Assassz Hmuoev ucofio>ao>cH
mmfiufl>wuo¢ chanson Amquov ousmwoq
AcoHuMA>ma oumocmumv mafiocommom Aucsouv
cosmos cmmz sz ommucoouom canmwum> comm

 

.ooscfiuGOUII.e canoe

49

fewer visits during the previous month and 17 or 27.1 per-
cent of our sample received ten or fewer visits in the
previous month.

The telephone appears to be a commonly used means of
contact with supporters. Respondents reported a median of
48 telephone contacts during the previous month. As with
other interpersonal involvement measures the variance on
this variable was great, with 13.0 percent or 9 individuals
reporting ten or fewer phone contacts during the previous
month.

In general, our respondents seemed quite satisfied
with the frequency of their visiting, reporting a mean 3.35
satisfaction score, placing them somewhere between very
satisfied and somewhat satisfied. Only ten percent or 7
individuals received an overall mean score for this scale
below "somewhat satisfied."

There is no clear indication here, in contrast to
Health reports, of a severe and widespread Interpersonal
Involvement problem. At the same time, there is consid-
erable evidence indicating a certain percentage of our
sample is, relatively speaking, socially isolated.

The degree to which deprivation in one facet of
interpersonal support correlates with deprivation in others

will be examined in the hypothesis testing section.

Leisure Involvement.--Breakdowns of Leisure Activi-

 

ties, Hours, and Satisfaction responses are also presented

50

in Table 4. The majority of respondents had considerable
leisure involvement, spending an average of 5.25 hours per
day, on an average of 7.7 leisure activities. Seemingly
reflecting this preoccupation with leisure, 81.4 percent of
respondents reported being either satisfied or very satis-
fied with their present leisure time participation. Al-
though the vast majority appear well involved and satisfied
with leisure involvement, a sizable minority reported less
than five activities per day (21.4%), two hours or less
participation per day (18.8%) and some dissatisfaction
(18.6%).

Thus as was the case for Interpersonal Involvement,
evidence suggests that there is potentially a problem with
respect to Leisure Time Involvement for some, but by no

means a majority of our study participants.

Losses-—Past Two Years.--This area of inquiry

 

touched upon three Specific age related losses: health,
death of a close relative and spouse's death.

Referring to the specific loss items, thirty-eight
of our respondents (54.3%) reported some kind of recent
health loss, 25 respondents (35.7%) reported losing a close
relative and 9 respondents (12.9%) reported loss of a
spouse during the previous two years (See Table 4).

This data substantiates that the three age related
loss areas inquired about have occurred among a noticeable

portion of our study p0pu1ation. The need implied by these

51

various losses, however, is not readily discernable but
will be further examined in the hypothesis testing.

Objective 2. To ascertain the prioritized importance of

 

various problems to these same respondents.

 

The importance of various problem areas to our
sample was ascertained by asking respondents to categorize
specific problems as being either "no problem for me,"
"a somewhat important problem for me," or "a very important
problem for me." These responses yielded a measure of
"felt" or "unmet need" by the population under study (See
Table 5). The percentages in Table 5 indicate the respon- QQ
dents were very hesitant to classify a problem area a "very K
important problem for me." In the eyes of the study respon-
dents, Transportation (17.1%) and Income (15.7%) are seen
by the most individuals as very important problem areas for
them. Besides these two problem areas, only Crime, Health
and Nutrition are reported as a very important problem by
over 5% (N = 3) of all respondents. These percentages
seem quite low in the context of previously enumerated
signs of problems (i.e., low mean income and poor health).
The small percentage reporting Health as a very important
problem (7.1%) seems the most surprising result, given the
previously cited high prevalence of chronic ailments.

Examining the percentage of respondents calling a \*\
problem area important to them (either "very important" or

"somewhat important") provides an indication of the total

Table 5.--Breakdown of Problem Area Responses

 

Variable
(Count)

Percentage (N)

 

Transportation (70)

-Very important problem
-Somewhat important problem
-No problem

Income (70)

-Very important problem
-Somewhat important problem
-No problem

Crime (70)

-Very important problem
-Somewhat importanthroblem
-No problem

Health (70)

-Very important problem
-Somewhat important problem
-No problem

Nutrition (70)

-Very important problem
-Somewhat important problem
—No problem

Spare Time (69)

-Very important problem
-Somewhat important problem
-No problem

Service Deception

-Very important problem
-Somewhat important problem
-No problem

Housing

-Very important problem
-Somewhat important problem
-No problem

Employment

-Very important problem
-Somewhat important problem
-No problem

Age Discrimination

-Very important problem
-Somewhat important problem
-No problem

17.1
44.3
38.6

15.7
20.0
64.3

\lmob
O O O
l-‘O‘UJ

NNIb
I O 0
wow

(12)
(31)
(27)

(11)
(14)
(45)

(6)
(20)
(44)

(5)
(12)
(53)

(4)
(11)
(55)

(3)
(6)
(60)

(3)
(6)
(61)

(3)
(2)
(65)

(1)
(6)
(63)

(0)
(5)
(65)

 

53

number of respondents in some fashion affected by that
problem-~or the prevalence of the problem. Using this
statistic, transportation (61.4%) was the most prevalent
problem area reported by respondents. Surprisingly, no
other problem was viewed by a majority of respondents as an
important problem for them. In descending order of reported
importance, the other problem areas are: Crime (37.2%),
Income (35.7%), Health (24.2%), Nutrition (21.4%), Spare
Time Activities (13.0%), Service Deception (12.9%),
Employment (10.0%), Housing (7.2%), and Age Discrimination
(7.1%).

There would seem to be some fairly obvious breaks
in the prevalence of reported "important problems." Based
on the responses of our sample, Transportation alone would
warrant being termed an extremely prevalent problem area.
Next, Crime and Income could be grouped as highly prevalent
problem areas, and Health and Nutrition grouped as fairly
prevalent problem areas. Finally, Spare Time Activities,
Service Deception, Employment, Housing and Age Discrimi-
nation could be grouped as problem areas affecting only a
relatively small percentage of the study population. When
one reexamines these percentages in the context of the
previously cited "very important problem" breakdown, it is
noteworthy that although Transportation is clearly the
most prevalent problem for our population under study, it

is only slightly more frequently cited as a "very

54

important problem" when compared with Income (17.1%

versus 15.7%).

Total Problems.—-As the large reliability coeffi-

 

cient for the Total Problems scale would suggest (alpha =
.688), individual problem reports appear to be related to
each other. Respondents reported an average of more than
two problems (2.3) with 38.5% of the respondents reporting
three or more problems (See Table 6). Individuals re-
porting no problems (18.6%) or just one problem (18.6%) are
clearly in the minority. Thus it would seem clear that
respondents who are at risk are confronted with a number of
problem areas not just one. The implications of having a
multiple problem at risk group must be a primary consid-
eration in the formulation of programmatic strategies and
suggests the necessity of a highly coordinated programming
effort.

Objective 3. To identify what services among a large group

 

of "possible" services are viewed as acceptable inter-

 

ventions by these same respondents.

 

Each of the 16 services inquired about and the
percentages of respondents answering in each category are
presented in Table 7. The services are listed from highest
percentage willing to use, to lowest percentage willing to
use (based on "definitely would use 'response'"). When
respondents were asked "Would you use such a service," if

they responded that they would potentially use a service

55

Table 6.--Breakdown of Total Problems Reported

 

 

Number of Problems % (N)
0 18.6 13
1 18.6 13
2 24.3 17
3 15.7 11
4 8.6 6
5 10.0 7
6 1.4 l
7 l 4 1
8 0 0
9 1.4 l

 

but were not presently in need of it, they were asked to
respond to the question, "assuming a need arose?" ‘%
The largest percentage of respondents stated they
"would definitely use" a service which provided Transpor-
tation for Medical Treatments (64.3%). In descending order
of reported (potential) utilization, the other services
were: Home Care Nurse (61.4%), Service Advice Program
(61.4%), Physical Therapy (at home) (58.6%), Legal
Assistance (51.4%), Sh0pping Assistance Service (50.0%),
General Transportation Service (48.6%), Home Cleaning
Service (48.6%), Home Meals (40.0%), Friendly Visitor
Service (34.3%), Telecare (Telephone reassurance) Service

(37.1%), New Housing Assistance (30.0%), House-Mate

56

Table 7.--Breakdown of Responses on Service Utilization.

 

Variable

(Count) Percentage (N)

 

Transportation for Medical Treatment

(70)
-definitely use 64.3% (45)
-might use 17.1% (12)
-probably not use 18.6% (13)
Home Care Nurse

(70)
-definitely use 61.4% (43)
-might use 20.0% (14)
-probably not use 18.6% (13)
Service Advice

(70)
-definitely use 61.4% (43)
-might use 15.7% (11)
-probably not use 22.9% (16)
Physical Therapy (at home)

(70)
-definitely use 58.6% (41)
-might use 21.4% (15)
-probably not use 20.0% (14)
Legal Assistance

(70)
-definitely use 51.4% (36)
-might use 14.3% (10)
-probably not use 34.3% (24)
Shopping Service

(69)
-definitely use 50.7% (35)
-might use 21.8% (15)
-probably not use 27.5% (19)
General Transportation

(70)
-definitely use 48.6% (34)
-might use 20.0% (14)

-probably not use ' 31.4% (22)

57

Table 7.-—Continued

 

Variable
(Count)

Percentage (N)

 

Home Cleaning Service
(70)

-definitely use
—might use
-probably not use

Home Meals
(70)

-definitely use
-might use
-probably not use

Telecare
(70)

-definitely use
-might use
-probably not use

Friendly Visitor
(69)

-definitely use
-might use
-probably not use

New Housing Assistance
(70)

-definitely use
-might use
-probably not use

House-Mate Finding Service
(69)

—definitely use
—might use
-probably not use

Volunteer Work Finding Service
(69)

-definitely use
-might use
-probably not use

48.6%
22.9%
28.5%

40.0%
34.3%
25.7%

37.2%
17.1%
45.7%

34.8%
18.8%
46.4%

30.0%
5.7%
64.3%

21.4%
5.7%
72.9%

18.9%
4.3%
76.8%

(34)
(16)
(20)

(28)
(24)
(18)

(26)
(12)
(32)

(24)
(13)
(32)

(21)
(4)
(45)

(15)
(3)
(51)

(13)
(3)
(53)

58

Table 7.--Continued.

 

Variable

(Count) Percentage (N)

 

Job Finding Service

(70)
-definitely use 15.7% (11)
-might use 4.3% (3)
-probably not use 80.0% (56)
Counselor-Psychologist

(70)
-definitely use 15.7% (11)
-might use 18.6% (13)
-probably not use 65.7% (46)

 

Finding Service (22.8%) , Volunteer Work Service (18.6%) ,
Job Finding Service (15.7%), and Counselor-Psychologist
(15.7%).

These results provide strong evidence of the
acceptability of a wide range of traditionally offered
services. It is interesting to view these results with
attention to the fact that an older person living alone is
considered to be at high risk for institutionalization.
Possibly with this fact in mind, roughly 60 percent of our
respondents stated they would definitely use an assortment
of home health support services (Medical Transportation,
Home Care Nurse, and Physical Therapy at Home). The con-
stituency for these services jumps to the neighborhood of
80 percent of all respondents when one adds in individuals

stating they might use these services. At the same time

 

59

a smaller but still substantial percentage of all respon-
dents, roughly 45-50 percent, stated they would definitely
use a wide assortment of daily living home assistance
services (Shopping Help, General Transportation, Home
Cleaning Service and Meals at Home). The constituency for
these services jumps to the neighborhood of 60 percent of
all respondents when one adds in individuals stating they
"might use" these services. While this study did not deal
directly with the issue of future institutionalization,
the above results suggest that the elderly are willing to
use a variety of home care services that could forestall
or prevent institutionalization.

Although the constituency for informal intepper-

sonal support services (Friendly Visiting and Telecare) is

 

not as large, these services nonetheless are definitely
acceptable to roughly 35 percent of all respondents (and
acceptable to 43-54 percent when "might use" respondents
are included). This contrasts with 15.7 percent who would
definitely use the more traditional helping professional
(i.e., the counselor-psychologist).

Besides the three groups of services just men-
tioned, only Service Advice 61.4 percent (an Information
and Referral Service) and Legal Assistance 51.4 percent,
were reported by over 1/3 of our respondents as a service
they would definitely use. There was only a slight indi—
cation of a willingness to use a House-Mate Finding Ser-

vice, Volunteer Work Service or Job Finding Service.

60

Hypothesis Testing

 

The remainder of the RESULTS will deal with the
relationship of various study variables to Life Satis-
faction, particularly as put forth in the study hypotheses.
The Life Satisfaction Scale used here is a modified version
of Neugarten, Havinghurst and Tobin's (1965) scale.

It is generally good statistical practice to check
correlative relationships to insure observed correlations
are not spuriously related to various demographic variables.
This is all the more important in this study because of the
use of a nonrandomly selected sample. In order to assess
whether any of the correlations about to be reported were
due to spurious relationships, certain demographic vari-
ables were selected and their effects on all hypothesized
associations were partialed out. Table 8 contains the
correlation matrix for Life Satisfaction and the demo-
graphic variables.

The major criteria used for selecting a demographic
variable for the partialing procedure was an observed
significant association with a hypothesis variable. Thus
any demographic variable which correlated highly (p < .05)
with any of the variables in question (Life Satisfaction,
Losses, Health, Interpersonal Involvement and Leisure
Involvement) was chosen for partialing, due to their stati-
stically obvious likelihood for affecting a relationship

between the variables. Based on this criteria, Education

Table 8.—-Correlation of Demographic Variables to Life

Satisfaction.

61

 

 

r 519

Years in Lansing .034 .392
Age .057 .321
Sex .041 .368
Years Widow .151 .133
Education .030 .403
Employed .152 .132
Number of Children .156 .129
Income .218 .046
Residence* —--- ----
Own Home .134 .135
Rent Home -.009 .469
Rent Apt. -.183 .065
Race -.038 .325
SES -.108 .201

 

*Converted into 3 dichotomous variables.

and Income were selected. Although Income's relationship
to Life Satisfaction barely reaches statistical signifi-
cance (p < .05) (Education was significantly related to
Visiting and Satisfaction with Visiting), it's association
is noteworthy in light of the low income of many study
respondents and the previously reported high priority

given Income problems by respondents. Socioeconomic

62

Status (SES), another widely used measure of individual
social status, failed to correlate significantly with Life
Satisfaction (r = -.108) for study respondents. Based on
this difference, one would tend to conclude that for our
sample, many of whom live on marginal incomes, concrete
buying power is a greater determinant of Life Satisfaction
than the SES one associates with education and occupation
levels.

Possibly because of their extreme distributions,
Sex, Race, and Employed did not correlate with any of the
variables of interest above the .05 level of significance
and consequently were not included as partialing variables.
Although age did not correlate significantly with any of the
variables it was included for theoretical reasons, namely
the hypothesis that activity theory would be supported for
persons under age 75 and disengagement theory would be
supported for persons over 75. Further, it should be
pointed out that a common outcome of calculating correla-
tions using a variable which has been restricted in range,
as age was in this study, is to lower the correlation
relative to what one would obtain if the variable was not
restricted. Thus, although age failed to be a significant
correlate of certain variables in this study, does not mean
such would be the case for other samples where age was

free to vary.

63

HYPOTHESIS 1: Life Satisfaction will be significantly

 

related (negatively) to reports of recent (previous two

 

years) age related losses.

 

Health Loss, Spouse Loss and Loss of a Close Rela-
tive were the specific age related losses examined in this
study. As was mentioned earlier, the Job Loss item was
discarded because it was not reported by any respondents.
The zero order and partial correlations of the three loss
variables to Life Satisfaction is presented in Table 9.

None of the expected significant correlations were
found. In fact, only Loss of a Close Relative had a dis-
cernible relationship of any size in the predicted negative
direction (r = -.l49). The association between Loss of
Health (r = -.019) and Loss of Spouse (r = .048) with Life
Satisfaction were only slightly and negligibly greater
than zero.

The prospect that multiple losses might have a
significant effect upon Life Satisfaction was entertained
upon finding these results. However, its nonsignificant
association with Life Satisfaction (r = .054) failed to
support this new post hoc hypothesis (See Table 9).

It seems clear from these results that there is
no support for the hypothesis that recent age related
losses, defined as occurring during the previous two years,
have a significant relationship to Life Satisfaction in the

population under study.

64

 

 

mmm.umflm mvm.umflm mmm.umflm osm.umnm mmm.umsm
vmo. moo. Hmo. mvo. «mo. mommoq Hmuoa
vmo.umflm moo.umflm omo.umam moa.umflm moa.umam
mmH.I mmH.I HoH.I oma.I mvH.I mmoq w>flumamm
mam.umflm mov.umflm mmv.uofim msv.umfim mmv.umflm
vwo. HHo.I oHo.I moo. mHo.I mmoa cuflmmm
mmv.umam mam.umam mom.umflm sHv.umflm sem.nmflm
moo. smo. vvo. mmo. mvo. mmoq mmsomm
"04mm "mflm "@flm "UHM
Ammo w :oflumosoo a mEUUCA Ammo Acowumosom Aoeoocw "mam
mafiaaouucouv mcflaaouucouv ostHouucouv mawHHowucoov Awoouo ouoNV
u H H H H

 

.GOfiHUMmmfime OMflQ OH mmHSmmmz mmog MO

mcowuoaouuoo ooHMAUHmm can Hoowo ouwNII.m magma

65

As Table 9 indicates, there was no meaningful change

in any of the previously reported associations between the
various Losses and Life Satisfaction due to the partialing
procedure. Thus it would seem likely that the nonsignifi-
cant associations just reported are meaningful and not due
to demographics suppressing a larger association.

Even though none of the Loss Measures have demon-
strated any association with Life Satisfaction, procedural
thoroughness warrants examining the internal correlation
matrix of these three variables. Table 10 presents this
matrix. As one might suspect, the matrix does not reveal
any significant inter-Loss correlations although all three
measures are positively (but negligibly) related to each

other.

Table 10.--Correlations Among Loss Measures.

 

 

Spouse Health Relative
Loss Loss Loss
Spouse ---
Loss
Health .101 ---
Loss
Relative .079 .145 --_

Loss

 

66

HYPOTHESIS 2: Life Satisfaction will be significantly

 

related to various measures of Health. (Sickness Rating

 

(negative), Total Illnesses (negative) and Self Health

 

Evaluation (positive), Doctors Visits (positive).

 

Table 11 contains the zero order and partial corre-
lations and significance levels observed between the Health
variables and Life Satisfaction.

Two of the four Health variables correlated signi-
ficantly and in the predicted direction with Life Satis-
faction. Self Health Evaluation (r = .407) and Sickness
Rating (r‘= -.336) both were associated with Life Satis-
faction well above the acceptable significance level
(p < .05). The third Health variable, Total Illnesses,
had a sizable correlation in the predicted negative
direction (r = -.l95), however, it just failed to reach
statistical significance (p = < .053). The final Health
variable, Doctors Visits, correlated negligibly and insig-
nificantly to Life Satisfaction in the predicted positive
direction (r = .016).

Despite the fact that Total Illnesses and Doctors
Visits did not correlate significantly to Life Satis-
faction, it would seem reasonable to accept a modified
version of the hypothesis at hand.

Thus the results substantiate the hypothesis that
Health, as measured by Self Health Evaluation and Sickness
Rating, has a significant relationship with Life

Satisfaction.

67

 

 

omm. u msm ems. I saw «me. I saw mom. I mew mac. n mam
mmo. vHo. mao. hvo. 0H0. mafimw> muouoon
500. u mam moo. I saw moo. I saw poo. I saw moo. I saw
omm. ovm.u smm.u omm.u 6mm.u assumm mmmcxofim
mud. I saw mac. I saw mmo. u mflm mes. a mom mmo. u ohm
nNH.I oom.I an.I oVH.I mmH.I mommosaaH Hmuoe
moo. I saw Hoo. u oflm see. I new moo. u mwm doc. I saw :ofiumsam>m
mom. moo. mav. mom. hov. spammm «How
u mam n mam u mam u mam
Ammo osm coaumosoo Ammo ACOADMUsoo Aosoocfl u mam
oeoosw mafiaaouucoov mafiaaouucoov mafiaaouucoov mcfiHHOHucoov Aumouo OHva
u H H M H

 

cofluomwmwumm swag ou monommoz spammm mo mGOADmHoHHoo oonauHmm cam “mono OHmNII.HH canoe

68

Partialing out the effects of Education, Income,
and Age failed to meaningfully change the significance (or
lack thereof) previously reported between the hypothesized
Health measures and Life Satisfaction. Simultaneously
controlling for the effects of Education, Income, and Age
slightly reduced the correlation reported between Self
Health Evaluation and Life Satisfaction (.407 to .368),
although the resulting correlation would still be highly
significant (p < .002). This reduction appears to be
largely due to partialing out the influence of Income.
Partialing out all three control variables also resulted
in the barely nonsignificant association between Total
Illnesses and Life Satisfaction dropping (-.l95 to —.127)
to a point where it could more meaningfully be labeled
insignificant.

Table 12 presents the observed correlations between
the various Health measures. Examining this matrix should
reveal to what extent these measures constitute one or
several Health factors and help elucidate their varying
associations to Life Satisfaction.

Reviewing Table 12, it is evident that there is a
strong association (-.561) between our two significant
correlates of Life Satisfaction: Self Health Evaluation and
Sickness Rating. Hence, it would appear that Self Health
Evaluation is to a certain extent determined by one's

recent (previous month) health experience (days sick or

69

mo. v m

 

 

o
. m
Ho v n
Hoo. v mm
III mna. oaom. mva.I muwmw> mnouooo
III smo. mamm.u museum mmmaxoflm
mommmcHHH
III nomm.I mo HmnEdz
coaumsam>m
III guano: maom
mufimw> mcwumm mommmcHHH cowumnam>m
muouooo mmwcxowm mo Honasz nuamom «How

 

.mmuzmmmz

spammm moose msOADMHmHHooII.NH canoe

70

well). While the reported correlation between these two
variables is substantial, each variable still only explains
31.4 percent of the variance in the other, leaving a consid-
erable amount of independence between the two measures.

Number of Illnesses, also had a significant but
smaller correlation to Self Health Evaluation (r = .286)
although it had only a negligible association to Sickness
Rating (r = .037). Since Sickness Rating only deals with
the previous month's health, a time during which nonchronic
ailments could be active and chronic ailment in submission,
this low correlation does not seem unusual.

Doctor Visits shows only moderate associations
with all of the other health variables (only reaching sta-
tistical significance with Total Illnesses). This would
seem to indicate that the unhealthy (based on the three
other health measures) only have a slight predisposition
to seeing their physician more frequently than healthier
respondents. Thus for the respondents visiting the doctor
seems to be an accepted regimen of later life with little
relationship to the other parameters of Health or to Life

Satisfactions.

HYPOTHESIS 3. Life Satisfaction will be significantly

 

related (positively) to various measures of Interpersonal

 

Involvement (Total Supporters Within 25 Miles, Total Visits,

 

 

Satisfaction with the Frequency of Visits, and Total

Telephone Calls).

 

71

Table 13 contains the zero order and partial
correlations and significance levels for the Interpersonal
Involvement Variables and Life Satisfaction.

All of the Interpersonal Involvement measures were
found to correlate significantly to Life Satisfaction.

The observed correlations to Life Satisfaction were: Total
Supporters Within 25 Miles, r = .405; Visiting Activity,

r = .371; Satisfaction with Frequency of VIsits, r = .390;
Telephone Contact Activity, r = .309. All observed corre-
lations were highly significant (p < .006)!

The results appear to conclusively support the
hypothesis that Interpersonal Support has a significant
relationship with Life Satisfaction for old people living
alone.

There was no meaningful changes observed in any of
the Interpersonal Involvement measures after the partialing
procedure was performed. All measures remained highly and
significantly correlated to Life Satisfaction even after
Education, Income and Age were controlled for at the same
time.

The matrix of correlations presented in Table 14
indicates that there is considerable overlap among indi-
vidual mechanisms of interpersonal involvement. Supporters,
Visiting Activity, Telephone Contact Activity are all

highly interrelated indicating respondents rating high on

72

 

«Ho. u ofim

moo. u ohm

ooo. u ohm

mac. n mam

ooo. u oam

 

. mHHmU
mom. mam. mom. mmm. mom. ocogmoams Hones
moo. u ofim Hoo. u oflm Hoo. n mam moo. u ohm Hoo. u ohm
mom. omm. mom. mom. Hum. muflmw> Hmuoa
. . I . I . I . I mufimfl> mo
Hoo n oflm Hoo I ohm Hoo I mam Hoo I mom Hoo I saw scamsomum spas
mmm. Hmm. mmm. oov. omm. coauommmfiuom coo:
."mm snow ouww oumflm oumflw
moo A Hoo H Hoo A Hoo . Hoo . moawz mm convex
mmm. mav. vov. omm. mov. muouHOQQSm Hmuoe
"mew "mew "mew "mew
Ammo w :oflumosoo Ammo AGOHumosoo Aosoocw "mom
a oEoocfi msflaaouucouv msflaaouucouv mcflHHouucouv mcwaaouucouv Auoouo ouva
H u u H H

 

consummmflumm moan

on mousmmmz ucmEo>Ho>cH HMGOmMoQHoucH mo mcowumamunou ooaofluumm can ouoNII.MH manna

73

one aspect of interpersonal support generally rated high

on the other two.

Table 14.--Correlations Between Interpersonal Involvement Measures

 

Telephone Satisfaction
Contact Frequency
Activity Visits

Supporters- Visiting
25 Miles Activity

 

Supporters ---
25 Miles

Visiting .524 ---
Activity

Telephone .456a .417 ---
Contact
Activity

Satisfaction .181 .376 .026 ---
With Frequency
of Visits

 

ap < .001

Satisfaction with Frequency of Visits had a signifi-
cant association, as one would expect, only with Visiting
Activity. Although these two variables are significantly
related, the size of the relationship (r = .376) is not as
large as one might expect considering the directly stated
evaluative relationship between the two. The fact that
Visiting Activities only accounts for 14.1 percent of
the variance in Satisfaction with Frequency of Visits
suggests individuals may differ greatly in their desire or

need for visiting.

74

HYPOTHESIS 4. Life Satisfaction will be significantly

 

related to various measures of Leisure Involvement (Number

 

of Leisure Activities, Hours of Leisure and Satisfaction

 

with Leisure).

 

Table 15 contains the zero order and partial corre-
lations and significance levels for the Leisure Involvement
measures and Life Satisfaction.

Both Leisure Activity (r = .318) (number of dif-
ferent activities participated in by respondent during
previous week) and Satisfaction with Leisure (r = .466)
correlated significantly with Life Satisfaction. Surpri-
singly, Hours of Leisure failed to be associated with Life
Satisfaction to any noticeable degree (r = .069). With-
standing the failure of Hours of Leisure to be meaningfully
associated with Life Satisfaction, it would seem reasonable
to accept a modified version of the hypothesis at hand.

Thus, the results substantiate the hypothesis that
Leisure Involvement, as measured by Leisure Activity and
Satisfaction with Leisure, has a significant association
with Life Satisfaction.

Although there were some slight changes in the
size of associations of Leisure Involvement measures with
Life Satisfaction after partialing, none of these changes
altered the significance of any of the previously cited
relationships. Thus the previously reported associations

of leisure Activity, Satisfaction with Leisure (significant)

75

 

 

Hoo. u mew Hoo. I one Hoo. u mew Hoo. u see Hoo. u mam mesmeeq sues
va. moo. mow. mmv. oov. coauommmflumm
emm. I oem Hoo. u oem mom. I ohm mom. I oem mom. I oem see>euo<
Hoo. woo. ono. ooo. moo. mo musoz
oHo. n oem Hoo. u oem eoo. u oem mmo. u oem voo. n mew sue>euoa
vow. mmm. omm. oom. mam. ousmfloq
"mew "mew "mew "mew "mew
oEoocfl mafiaaouucouv mafiaaouusouo msflflaouucoov mafiaaouucoov Anoouo OHoNV
u u u H H

 

mmhawwz pcm8>HO>QH QHDwHO‘H MO

.eoeeummmeemm woes ou

mcofluoaonuoo ooHMAuHom can noose ouoNII.mH magma

76

and Hours of Leisure (nonsignificant) with Life Satis-
faction appear to be meaningful and not spuriously related
to the effects of the examined demographic variables.

Table 16 reveals some apparent inconsistencies in
the relationships observed between various Leisure Involve-
ment measures and their previously cited associations with
Life Satisfaction.

Table l6.--Corre1ations Between Leisure Involvement
Measures.

 

Leisure Activity Hours of Satisfaction

 

 

(Number of) Leisure with Leisure

Leisure ---
Activity
Hours of .301a ——-
Leisure
Satisfaction .185 .305a -_-
with Leisure

ap < .01

The Leisure Involvement measures appear to be
highly correlated with each other with the exception of
Leisure Activity and Satisfaction with Leisure (r = .185).
This is noteworthy when one considers that these two
variables both correlated significantly with Life
Satisfaction.

Thus it becomes quite difficult to infer that a

single Leisure Involvement factor accounts for the

77

significant associations previously previously observed
between both these variables and Life Satisfaction. The
possibility that this association is due to satisfaction
measures naturally correlating highly would seem plausible

but untestable in the context of this study.

HYPOTHESIS 5. Consistent with activity theory, there will
be a significant negative correlation between life satis-
faction and losses, and significant positive correlations
between life satisfaction and activity, income and health
for reepondents under 75. Consistent with disengagement
theory, there will be nonsignificant correlations between
life satisfaction and losses, income, activity, and health
for respondents 75 or over.

The correlations for respondents under 75 years of
age (N = 36) and respondents 75 years or older (N = 34)
used in evaluating this hypothesis are presented in
Table 17.

On twelve occasions, the older group exhibited
lower associations to Life Satisfaction than did the
younger group. Furthermore, while all ten variables
previously shown to correlate significantly to Life
Satisfaction for all respondents remained significantly
correlated for younger respondents, only two significant
correlations to Life Satisfaction were evident for older
respondents. The two variables significantly associated

with Life Satisfaction in the older groups were: Self

78

Table 17.--Comparison of Correlations with Life Satisfaction for
Respondents below 75 Years and Respondents 75 Years

 

 

 

 

or Older.
Significance of
Age < 75 Age > 75 r diff
Variable
r 819(97‘0) r si9(97‘0) sig (p<75 = p>75)
Loss of Spouse -.007 .484 .135 .224 NS
Loss of Health -.158 .179 .185 .148 NS
Loss of Relative -.318 .030 .002 .495 NS
Self Health Evaluation .440 .004 .399 .01 NS
Sickness Rating -.496 .001 -.189 .142 NS
Total Illnesses -.195 .128 -.l96 .133 NS
Doctors Visits -.097 .290 .086 .318 NS
Visiting Activity .447 .003 .266 .064 NS
Supporters - 25 Miles .553 .001 .204 .127 NS
Telephone Activity .419 .007 .135 .223 NS
Satisfaction-Visting .420 .005 .349 .022 NS
Leisure Activity .547 .001 -.007 .484 p < .05
Hours of Leisure .102 .277 —.015 .463 NS
Satisfaction-Leisure .635 .001 .206 .121 NS

Income .358 .019 .125 .271 NS

 

79

Health Evaluation and Satisfaction with Visiting. While one
might entertain the possibility that these two variables
represent real exceptions which invalidate a substantiation
of the hypothesis at hand, another explanation of their
deviance seems plausible. This would be the explanation
described previously, that is, like measures (in this case
subjective measures of a satisfaction-dissatisfaction
trait) will naturally correlate highly. Although this
hypothesis is empirically untestable for the present study
data, it should be considered a viable explanation.
Although the size of the differences between corre-
lations for under and over 75 are, with one exception, not
statistically significant (only the difference between
Leisure Activity's correlations for under and over 75 was
statistically significant), the fact that all the variables
previously found to be significant associates of Life
Satisfaction dropped (when one compares the older cohort
to the younger cohort of elderly) and were nonsignificantly
associated with Life Satisfaction (less the two exceptions
just cited) as the hypothesis predicted, clearly suggests
this older group may have experienced disengagement.
Withstanding the failure of two significant corre-
lates of Life Satisfaction to drop to a statistically non-
significant level for our oldest respondents (they were
reduced however), on balance, the data seems to support the

hypothesis that correlations of Life Satisfaction with

80

various measures of income, activity, and health can better
be explained in the oldest of the elderly by the prediction
of disengagement theory (nonsignificant associations) than

by the predictions based on activity theory (significant

associations).

Overall Prediction of Life Satisfaction

 

Within the context of discussing the major deter-
minants of Life Satisfaction, two further issues would seem
to warrant attention: First, how important (relative to
each other) are the Income, Health, Interpersonal and
Leisure Involvement factors of our respondent's life in
predicting Life Satisfaction; and second, how much total
variance is explained in Life Satisfaction by our assortment
of significant predictors?

This examination should procede with the caution
that terms like "determinants" and "predictors" used here
and in previous references to correlative associations do
not imply any causal direction in the relationships dis-
cussed. With this reservation affirmed, it is possible to
elaborate the relative importance of various variables in
predicting Life Satisfaction and describe the total variance
explained by an assortment of predictors by use of Multiple
Regression equations.

All of the multiple regression tables presented
below identify the individual predictor variables along with

the following information: (1) the order in which they

81

entered the prediction equation, (2) the cumulative multiple
r as each variable entered the equation, (3) R-2 or the
cumulative amount of new variance each variable explained,
(4) R2 change - the change in R2 attributable to each
variable, (5) R - the original Pearson Correlation coeffi-
cient of each variable to the dependent measure, and (6)
standardized Bete--or the standardized regression coeffi-
cient. The Beta weight quantifies how much change will
occur in Life Satisfaction (in standard scores) given a
standard deviation's change in the independent variable.
Beta weights give a measure of the relative important of
each variable in predicting Life Satisfaction since all
other variables in the equation are simultaneously con-
trolled for. The major drawback of using Betas in this
way is that it will only account for unique variance in the
predictor and ignores all variance shared with other
predictors.

The multiple regression equation represented in
Table 18 was performed using a modified stepwise inclusion
approach (best predictor enters first) after Income was
forced to enter the equation first (a means of controlling
for the effects of this demographic variable). The step-
wise inclusion approach used here entailed allowing only
one predictor (the best predictor) from each area or factor
of interest to enter the equation, then eliminating the

other same factor variables from consideration. This was

82

done since having several highly correlated variables from
a single factor in an equation would tend to deflate those
Betas by eliminating more and more shared variance.

The four variables accepted into the Multiple
Regression solution presented in Table 18 (Income, Visiting
Activity, Self Health Evaluation, and Number of Leisure
Activities) explain 27.5 percent (Column 3) of the variance
in Life Satisfaction. It is interesting that because
Income was forced into the equation as the first variable,
Self Health Evaluation, the variable with the largest
Simple R to Life Satisfaction did not enter the equation as
the first variable on the stepwise procedure. It entered
the equation after Visiting Activity. Obviously, Self
Health Evaluation correlates with Income to a large enough
extent that Visiting Activity actually explained the most
new variance and thus became the first variable to enter
the solution (after Income).

When one examines the Betas (Column 6) presented
in Table 18, however, it is obvious that Self Health
Evaluation (Beta = .309), and Visiting Activity (Beta =
.292) are approximately equal in their importance as pre-
dictors. On the other hand, the relatively small Betas
observed for Leisure Activity (Beta = .062) and Income
(Beta = .097) portrays these variables as substantially
less important than the other two as independent deter-

minants of Life Satisfaction. Since Health, Interpersonal

83

 

 

oomoo.o omeam.o oemoo.o emoe~.o msemm.o e aue>euo¢
whomfloq

mooom.o omeoe.o momoo.o eo~e~.o mammm.o m eoeumoam>m
seamen mamm

~oeom.o emaem.o eeema.o mooefl.o Hemme.o m mue>eu04
oeeueme>

mmeoo.o omoH~.o oaoeo.o oaoeo.o omoam.o H msooeH
memo m mqesHm mozmmo omm mmeoom m m memesqoz xz<m mzaz
loo Ame Lev Amo Amo lav mammHm¢>

 

.u:mEm>Ho>cH mucmflwq cam suammm .ucmEm>Ho>cH HocomummnmucH

.oeoocH "coauomwmwumm mmeq mo coflusaom cowmmmumom mamfluHDSII.mH manna

84

Involvement, and Leisure Involvement all had sizable

simple correlations to Life Satisfaction, one must conclude
that Leisure Activity's considerably smaller Beta must be
due to the other three predictors explaining identical
variance. Table 19 presents a correlation matrix of all
four predictors and tends to confirm this fact since Leisure
Activities is highly and consistently correlated to each

of the other predictors.

Table l9.--Correlation Matrix of Self Health Evaluation,
Income, Visiting Activity, Leisure Activity

 

Self Health Income Visiting Leisure
Evaluation Activity Activity

 

Self Health

 

Evaluation ---

Income .281 ---

Visiting

Activity .168 .042 ---

Leisure

Activity .351a .375a .378a
ap < .05

Based on these findings, it would appear that while
Health (represented by Self Health Evaluation) and Inter—
personal Involvement (represented by Visiting Activity)
are equally and highly important in independently pre-
dicting Life Satisfaction, Leisure Involvement (represented
by Leisure Activity) and Income are considerably less

influential.

85

The multiple regression equation represented in
Table 20 was constructed using an orthodox stepwise
inclusion approach (excepting the inclusion of Income)
which examined all significant predictors of Life Satis-
faction as potential "best predictors."

Table 20 reveals that all of the significant asso-
ciates of Life Satisfaction explain 55.3 percent of the
variance in Life Satisfaction. Even when one subtracts
the variance explained by the demographic variable Income
(4.8%) this assortment of Health, Interpersonal Involvement
and Leisure Involvement variables explain about half
(SO-55%) of the variance in Life Satisfaction.

This multiple regression solution also reveals
that a variable not examined in the hypotheses, Total
Problems, entered the equation first on this stepwise
procedure and has a simple r with Life Satisfaction
(Column 5) which is the largest observed for any study
variable (r = .5193). Since this scale is constituted
by items which are similar to the Life Satisfaction
scales, asking the respondents to give specific evaluative
responses (basically on a satisfaction-dissatisfaction

continuum), this finding is not surprising.

Cluster Analysis

 

Although much has been revealed about the asso-
ciation of study variables with Life Satisfaction, and

about some conceptually homogeneous segments of a complete

86

 

aue>euc< oceeeme>

 

vmmdo.o vmabm.o maooo.o ommmm.o mmmqh.o OH
Hmm50.0I Hmmmm.0I vmmoo.o mommm.o ovmvh.o m mcwumm mm0c30fim
hoamo.0I mmhdm.o Nmmoo.o mamvm.o hedvh.o m aufl>flu0¢ musmwma
NmmHH.o Nvomm.o mhmoo.o momflm.o Ohth.o h mcwuflmw> Goauommmflumm
oommH.o Namom.o mHHNo.o omomm.o mmwa.o o hufl>wuo< mconmmama
vamH.o vomov.o NNmmo.o HhmHm.o mamdh.o m mHSmHmA cowuummmwumm
omhmm.o mmhov.o ommmo.o mwom¢.o mammo.o w :Owumnflo>m nudoom waom
omoam.o movov.o voooa.o moomm.o mhmmo.o m mmHHS mm I mumuuommsm
om¢NM.OI ommam.OI vavm.o mmomm.o hommm.o N mEmHnOHm Hmuoa
HvaH.o ommHN.o mquo.o mdmvo.o ommHN.o H mBOocH
damn m mamZHm mozmmo 0mm mm¢30m m m NAmHBADS ¥Z¢m m2€z
Aoo Amy 3; 23 Amy 3. mammal;

 

msoanonm Hmuoa mcwosaosH
wHODOHooMm Hafl "cowuomwmfluom owed mo coausaom scammoumom oamwuas II.om magma

87

correlation matrix, a specialized statistical methodology
is necessary to meaningfully discuss the relationships
present among a large number of variables. In order to
examine all of the variables used in this study, a cluster
analysis (Tryon and Bailey, 1970) was performed on the
study data, producing an empirically based simplification
of the total correlation matrix.

It should be noted here that a cluster analysis can
only produce a solution based on the entered variables.
Since certain areas of this study were examined in greater
detail (i.e., Health and Activity) it is much more likely
that these areas will have independent matrices of vari-
ables associated with them and this must be borne in mind
when explaining the resulting clusters. Cluster analysis
(V-analysis) is accomplished by statistically removing
successive clusters of items with high intercorrelations
from a complete correlation matrix. Although the results
of this technique do not imply causality between various
variables, it reduces the total number of factors which
can reproduce the full array of intercorrelations without
sacrificing the generality of results.

All of the variables used in this study were examined
in the cluster analysis. Since preliminary clusterings
of variables produced self defining clusters which added
little to the understanding of the data, the final analysis
reported below used aggregated or scale variables where

appropriate. Because a subject/variables ratio of less

88

than 2:1 was maintained in the final clustering, caution
should be used in drawing conclusions from the results.

Within each of the clusters the variables are grouped
by conceptual categories for ease of interpretation. The
four clusters, (see Table 21) representing the four most
significant characteristics or properties which underlie
the variables entered, are: (1) Multiple Problems-~Dissatis-
faction, (2) Health, (3) Activity, and (4) Housing (see
Table 21).

Cluster 1, containing the important "outcome measure"
Life Satisfaction, is essentially comprised of three con-
ceptual units. The first--"Mu1tiple Problems"--contains
eight of the ten problem areas queried about in the inter-
view, included were: Employment, Housing, Spare Time,

Age Discrimination, Income, Nutrition, Transportation, and
Service Deception. The loading of the individual problem
areas on this cluster tends to be inversely related to the
prevalence of the various problems, with the more rarely
reported problems generally having higher factor loadings.
The second component of the cluster is a single variable--
low Life Satisfaction--describing someone scoring poorly

on the Life Satisfaction Scale. The final component--"more
use of potential services"--is also a single item indicating
more reported willingness to use services. (This variable

was created by summing across all service item scores.)

89

Table 21.-—V-Analysis of Key Cluster Structure

 

 

Factor
Cluster Loading
Cluster 1 Multiple Problems - Satisfaction
A. Multiple Problems
1. Employment more of a problem (D) .7906
2. Housing more of a problem .5404
3. Spare Time more of a problem (D) .5171
4. Age Discrimination more of a problem .5131
5. Income more of a problem (D) .5061
6. Nutrition more of a problem .4197
7. Transportation more of a problem .4735
8. Service Deception .3857
B. Low Satisfaction
1. Lower Life Satisfaction Score (D) -.7126
C. More Reported Service Use
1. More willing to use services .4252
Cluster 2 Health
1. More sickness past month (D) .7085
2. Health evaluation poorer (D) .7052
3. Health more of a problem .5593
Cluster 3
A. Interpersonal Activity
1. Fewer visits per week (D) .9693
2. Fewer supporters within (D) .5399
25 miles
3. Fewer telephone contacts per week .5053
4. Less Satisfied with frequency of visits .5137
B. Leisure Activity
1. Less satisfied with leisure time .4958
2. Fewer leisure time activities .4784
Cluster 4 Housing
1. Does not rent apartment (D) .8575
2. Owns own home (D) .8279
3. More years at present address .5537

 

(D) - Definer of Cluster

90

All told, the variables within Cluster 1 indicate an
underlying relationship between a large group of problems,
low satisfaction with aspects of the respondents life and
also high reported willingness to use services. The nega-
tive correlations observed between the Life Satisfaction and
problem variables suggests that the reports of these prob-
lems is associated with reduced Life Satisfaction. The
positive correlations observed between the problem variables
with "more willingness to use services" suggests individuals
reporting these problems also report a willingness to make
use of a large number of services. The presence of most of
the least frequently cited problems in this cluster tends
to identify this group of individuals as not only suffering
from common stressors (i.e., Income and Transportation
Problems) but also from uncommon stressors like Spare Time
and Employment Problems (both cluster definers). This fact
might be suggestive of a multiproblem individual who ex-
periences problems rarely reported by the majority of re-
spondents. Such a situation would likely be very exasper-
ating to an already stressed individual and account for
Life Satisfaction loading negatively on this cluster. On
the other hand, one might want to consider to what extent
this group of individuals might represent chronic com-
plainers who find everything a problem and are quite willing
to report this fact.

While both Health and Crime were not included in

this cluster, Health was excluded only because it loaded

91

more strongly elsewhere (Cluster 2 - Poor Health) and not
because it didn't load strongly on this cluster (see Table
22). Therefore, only Crime appears to be reported indepen-
dently of other problems.

Cluster 2, Poor Health, describes an individual who
is less healthy than other respondents; these individuals
tended to have more days and severer sickness, gave them-
selves a lower self health evaluation and reported health as
an important or very important problem.

The clustering of Sickness Rating and Self Health
Evaluation here with Problem: Health affirms that these two
variables are meaningful parameters of a specific need--
health--a fact which would seem consistent with their pre-
viously reported high correlation to the general measure of
need--Life Satisfaction. Similarly the failure of Total
Illnesses and Doctors Visits to load here appears to be
consistent with their failure to correlate meaningfully
with general measure of need--Life Satisfaction.

A review of the problems internal to cluster one
reveals that Nutrition is highly related to this Poor
Health cluster (See Table 22) indicating there is a rela-
tively large association between the occurrence of these
two problems but failing to elucidate the causal direction
between the two problems.

The third cluster, Activity, describes less involve-

ment in both interpersonal and leisure activities; these

92

 

 

 

 

 

HooH.I mooe.I oomH.I mooH. ocHucm mmocxoem
ammo. mmmm. moHH.I omoe.I ousmHoHIGOHuomumHuom
HHem. mmmH. ooHo.I voHH.I whomHoq musom
«who. momm. mmmH.I mmmm.I muH>Hpo< musmHoH
coho. momm. mmvm.I mHMH.I oeoocH
momm. move. mmmH.I mmom.I :oHoHHnO mo Honasz
Hmom. mmmo.I omoo. vao.I GOHmHHom
ommo.I momo. mmHo. mHmm.I oosoon whom»
momo. HemH.I Hooo.I oooH. msuoDm HouHHmz
mHHo.I omoo.I Homo. ommH.I xom
mMHo.I mhmo.I momm. mooH.I 0mm
ommH. ommo. comm. Homo.I mchcoH CH ammo»
Homo.I Homm.I ommm.I oHoH. mmouoo< um muoow
HoumsHO m HoumsHO o HoumsHO H HoumsHU

suHmom >DH>Huo¢ msHmsom COHuoommHuom

msmanee maoeeec:

 

mCHoEoo HoumsHO osoHHno £DH3 moHanum> mo mCOHuoH0HHOOII.mm oHQme

93

 

 

 

 

 

mmmm. vooH. ommm.I omom. moHHz mm I muouuommsm
ommo.I mmmH.I hovo.I mooH. mum
Home. ommv. memo.I omHo. COHuoommHumm oMHH
oome.I meco.I moHH. smmm. ecosuucmc comm
mmvo. mmmo. mono. mHmH. meow ucom
Hmom. mmoo.I momm.I mmoo. mEom s30
mooo.I mmmH.I mmoo.I mmmv. AHmuoav om: 00H>Hom
memH. momH.I mmvo.I vmvo. mmOH nuHmom
oomH.I MHo¢.I vmmo. mmmo. o>HDMHom mo mmoq
mmmH. Hmoo. ooHo. eemo. mmcoom mo moon
mmmH. moon. mmHH. momm. coHumsHo>m nuHmom MHom
mmvH.I momm.I mHmo. mmmH. mommocHHH Houoa
homo. ommH.I moHH.I memo. mUHmH> muouooo
HoDmSHO m HoumsHU v HmumsHO H HmumsHo

suHmmm muH>Huo¢ mchsom COHuommmHumm

msmanoee mHmHuch

 

.omscHucooII.mm mHnma

94

 

 

 

 

 

mEmHnoum mHmHuHsz

MHHH. «mam. HeoH.I emom. coeuoocmo mce>ucm someone
mmo~.I moom.I omeH.I same. comeeeucz someone
mmoo.I memo.I mooo. oaoo. mseeo someone
~oo~.I oomH.I eooH.I HeHm. mafia mucom schoue
HNMH.I been. oomo.I moon. ucmsmoeesm smHaoue
mmoH.I mono. memo. Hmam. comeccesmuomeo woe someone
oHeo. oomm.I moeo. mmse. coeucpuommcmue amHnoum
mmmH.I omeo.I meme. eoem. ocemcom seanoue
memo.I momm.I mHmH.I emoe. spaces smeaoum
--.I ammo.I HoeH.I Hoom. msoccH someone
emHm. oomm. mmmo.I mon. oceumme> comuocmmeucm
mmom. memo. oHH~.I eHem. mee>meoe mcocmmHms
memo. Heme. oeHo. eoem. mue>euo< ocmueme>
HoumsHO m HoumSHU v HoumsHu H HoumsHo

cumcmm mum>muo< ocemoom acmuocmmeucm

 

.omscHucooII.mm oHnma

95

individuals were involved in less visiting, had fewer
supporters available to them, were less satisfied with the
frequency of visiting, had fewer telephone contacts, were
less satisfied with leisure time, and reported fewer
leisure activities.

The fact that the two definers of this cluster were
both descriptive of interpersonal involvement deficits
("Fewer Visits" and "Fewer Supporters Within 25 Miles") and
the leisure measures loaded somewhat lower here, suggests
that the Interpersonal Activity component of the Activity
Cluster is the core of this factor.

Cluster 4, Housing, is actually just a triplet of
self defining residence variables which adds little to our
understanding of the other data. It describes individuals
who didn't rent apartments, owned their own homes, and
lived for longer periods at their present address. It is
noteworthy that none of the other demographic variables
entered into the analysis (including Income) had factor
loadings large enough to include them in a cluster, sug-
gesting this group of variables lacks relevance in deter-
mining our clustered variables. A listing of all variables
excluded from the clustering solution due to communalities
below .20 is included in Appendix E.

An examination of the cluster intercorrelations on
Table 23 reveals that the three deficit or problem oriented

clusters just described all have positive intercorrelations.

96

This would tend to suggest that the relationship of Life
Satisfaction to the problems identified in Cluster 2 (Poor
Health) and Cluster 3 (Low Interpersonal and Leisure Acti-
vity) is not unlike those observed for the problems internal
to Cluster 1. In fact, when one examines the factor

loading of Life Satisfaction with these clusters (Health,

r = -4956; Activity, r = -4751) (ses Table 23) it is
apparent that Life Satisfaction could have satisfied the
lower factor coefficient bounds of both these clusters and
would have been associated with them as expected--nega-

tively.

Table 23.--Inter Cluster Correlations

 

 

Cluster 1
Multiple Problems-
Satisfaction ---

 

Cluster 2
Health .1812 ---

 

Cluster 3
Activity .4428 .2979

 

Cluster 4
Housing .0848 .0610 .1390 ---

 

 

In more closely examining the intercorrelations
among the clusters, it is apparent that there are some
differences in the extent to which the clusters are
positively associated. Multiple Problems and Activity

(Cluster 1 and Cluster 3) have a large association

97

(r = .4428), implying that activity deficits are not sub-
stantially unlike the large group of problems subsumed in
Cluster 1. At the same time, Multiple Problems and Acti-
vity differ in their association to Health (Cluster 2).
The Activity Cluster (#3) correlates moderately to Health
(Cluster 2) while Multiple Problems (Cluster 1) is con-
siderably less associated. This would suggest one thing
which differentiates, activity deficits, from the large
group of problems pulled together in Cluster 1 is the

greater extent to which activity is determined by health.

DISCUSSION

It is important that discussion of the implications
of this study be examined after first touching on some of
its limiting attributes.

Two aspects of the sample used in this study would
seem to limit the generalizability of the findings: the
sample size and the selection process of the sample. A
sample of seventy respondents was used here since cost
considerations precluded drawing a larger number of respon-
dents. Thus it is clearly not possible to use these re-
sults to draw inferences about small subgroups within the
population of old people living alone, such as blacks,
males, and employed elderly.

The use of a nonrandomly selected sample was also
an outgrowth of cost and resource limitations. While a
nonrandomly selected sample also limits the generalizability
of these findings, the method used, informant (agency)
identification of known live alones, would seem likely to
drastically distort only one of the sample's characteri-
stics, degree of isolation. In point of fact, the sample
obtained did appear to approximate expectations, with the

notable exception of Interpersonal Involvement. The fact

98

99

that all respondents surveyed were known to some group or
agency would suggest the present study might underestimate
the extent of problems associated with being isolated. On
the other hand, the presence of a reasonable percentage of
relatively isolated individuals in the study (20-25%)

would likely not misrepresent the reported relationship of

 

Interpersonal Involvement with other study variables. The

 

issue of selectively finding true isolates in a selected
age cohort is a difficult case finding question which could
not be adequately handled within the limited scope of this
study but is definitely worthy of extensive future evalu-
ation. For the purposes at hand, it would seem judicious
to conclude that there is a greater degree of interpersonal
isolation and associated problems extant in the community
than the present study suggests.

Withstanding these avowed qualifications, the study
data provides clarification of the needs of old people
living alone and the factors which contribute to good

aging adjustment.

Multiple Problems

 

The most common problem among old people living
alone is to be stricken with more than one problem. Al-
though this study failed to identify any discrete sub-
groupings of problem areas, it seems quite clear that the
old person living alone with multiple problems is the rule

rather than the exception. Almost two-thirds of all

100

respondents reported more than one problem, slightly more
than reported the most prevalent single problem area, trans-
portation. When considering only those individuals who had
a problem, one finds that these individuals on the average
reported 3 problem areas. In addition, among all study
variables Total Problems exhibited the studies largest
correlation to Life Satisfaction.

The frequency with which problems occurred singu-
larly (17.1%) makes it clear that the neediest in our
target group will have a number of problems afflicting them
simultaneously. While the multiproblem old person is not
made up exclusively of individuals living alone, it would
seem likely to be more common for this group. Substan-
tiating this, a study by the Michigan Office for Services
to the Aging (1975) found problem reports to be more com-
mon for respondents living alone than for any other group.

Although the present data failed to identify any
discrete subgrouping of problems, this finding should not
be construed as meaning that none exists. Multivariate
data analysis requires a large sample to variable ratio in
order to obtain reliable solutions. Obviously, this study
did not provide an opportunity to meaningfully resolve the
issue of distinct problem clusters. However, such an
investigation would seem to be a high priority for future
research. Knowledge of problem clusters and the causal

predominance of one problem with another is the kind of

101

information which would be most useful in allocating limited
resources in a multiproblem situation. Although causal
predominance between various problems was not resolved, (it
is seldom settled by correlative techniques) there is con-
siderable evidence suggesting the primacy of various indi-

vidual needs for the population under study.

Income

While it was not the most frequently reported prob-
lem, Income would appear to be a significant, if not the
most significant, area of need for old people living alone.
It was reported as a very important problem by 15.7 percent
of all respondents, approximately as many as gave Transpor-
tation the same rating.

Providing some validation of the importance of
income as a need is the fact that reported yearly income
was the only demographic variable in the study to be signi-
ficantly associated with Life Satisfaction. Obviously not
every problem area affords the opportunity to examine such
a convergent parameter as reported income. Nonetheless,
this association provides empirical confirmation of the
deleterious effect low income can have on the morale of
an older person living alone.

The importance respondents attach to income prob-
lems, the significant association to Life Satisfaction
exhibited by income and the large number of individuals

who report low incomes (mean income = $3670) serve to

102

identify this aspect of our respondents lives as quite
problematical and probably exacerbating of other problem
areas. Financial security is a key to many consumer ser-
vices in our society such as transportation, food, and
health care and thus it is easy to envision the possibility
for "a spill over effect" providing meaningful improvements
occurred in the target population's income.

While much of the data described here supports the
primacy of income as a need, practical considerations would
seem to preclude effecting much change in this area. Direct
impact on this problem at the local level is probably
beyond the scope and resources of most local agencies and
the obvious mechanism for handling this need for younger
individuals--employment--is neither a high priority prob-
lem area nor a desired service possibility for this group
of individuals. Despite the obvious potential for effec-
ting far reaching change for our target group through
interventions focused on expanded income, change would‘
seem to be more practically effected through mechanisms
which selectively remediated areas which presently drain
our target population's resources.

While direct impact on the income needs of elderly
at a local level seems impractical, the implications of
this need area should not be ignored when considering
interventions for other problems or needs. Regardless of
the targeted need it would seem essential that service

planners and providers insure that no income barriers exist

103

to the utilization of needed services. To institute a
service targeted at the elderly in general or the elderly
living alone in particular which entailed anything but a
minimal cost to the poorest elderly would clearly undermine

the potential impact and effectiveness of the service.

Transportation

 

When considering only the prevalence of reports of
various problem areas, there is little doubt that Transpor-
tation constitutes the greatest area of need. Transpor-
tation was the only problem area viewed as important (re-
sponded either "somewhat important problem" or "very impor-
tant problem") by a majority of respondents (61.4%). In
addition, when one examines only reports of "a very impor-
tant problem" for me, the 17.1 percent citing Transporta-
tion, constituted the largest percentage to so label any of
the problems. The fact that many of those who saw Trans-
portation as a problem balked at calling it "very impor-
tant," would simply imply less seriousness per instance when
compared with income.

A practical consideration would also seem to weigh
toward the importance of transportation. Since case finding
and outreach are so important for the potentially isolated
old person living alone, the large number of individuals
reporting transportation problems and the high percentage
of respondents reporting some willingness to use transpor-

tation services (Medical Related = 81.4%, GENERAL = 68.6%)

104

make transportation related interventions an ideal vehicle

for getting multiproblem elderly into the service network.

Health

The degree to which health is viewed as a problem
by responents would clearly underestimate the amount of
symptomology of impairment reported within our sample of
respondents. One can only conjecture, as was pointed out
in the results section, that poor health is such a common
and expected part of the aging process and the traditional
health provider, the physician, so readily used (on the
average, once a month) that large numbers of respondents
do not feel nor express health as an "unmet need."

Nonetheless, reports of health problems (24.2%)
are not inconsequential by any means, they only seem small
by comparison to the number of individuals reporting chronic
illnesses (mean per respondent = 3.04, 100 percent reported
at least one illness) and those evaluating their health as
fair or poor (54.3%). In relation to other problems,
health rates fourth in total percentage reporting (24.2%),
and fourth in reports of "very important problems" (7.1%).
When one also considers the extent to which Self Health
Evaluation and Sickness Rating correlated with Life Satis-
faction (.407 and -.336 respectively), the importance of
this problem area seems all the more apparent. Although
fewer respondents than expected labeled health a problem,

the degree of apparent disability in the community, the

105

association of health measures with Life Satisfaction, and
the reported high potential utilization of health care
services such as transportation for medical treatment
(81.4%), home care nurse (81.4%), and physical therapy at
home (77.1%) all point to health needs as being considerable
in our study p0pu1ation, although probably more nearly "met"

than either income or transportation needs.

22.122

Reports of Crime as a problem (37.5%) are somewhat
confusing, since to the author's knowledge none of the
respondents who saw it as a problem indicated being crimi-
nally victimized (all respondents were informally asked
this after the interview). This may account for why less
than 25 percent of those calling it a problem (8.6% of all
respondents) actually labeled it as a very important
problem.

A similar finding has been reported in a survey of
needed services recently completed in Tampa, Florida
(Gray, 1977). Within the context of this study, a service
to aid victims of crime was rated as important by more
elderly respondents (93.6%) than any of the other 15 ser-
vices inquired about. In addition, a greater percentage
of old people saw it as important than any of the other
age cohorts. However, when asked which of the same 16
services they or friends and relatives could have used in

the past only 14.9 percent of the elderly named the victim

106

assistance service (making it their 14th most needed ser-
vice). On this parameter, fewer old people reported past
opportunities to use a victim assistance program than other
age cohorts. Clearly, crime's importance as a problem to
the elderly seems to exist independent of past criminal
victimization, suggesting preventive support and counseling
rather than post hoc victim counseling may be the most
needed service strategy.

Old people living alone would seem more likely to
be conscious of this problem and concerned about their
extreme vulnerability to crime victimization than other
elderly. The previously cited Michigan Office on Services
to the Elderly study (1975) tends to substantiate this
hypothesis since it found that 35 percent of the elderly
living alone reported crime as a problem compared to only
29 percent of elderly in other living situations (based on
a sample of 5000 individuals). The fact that Crime is
the only problem area not to load on cluster one (Health
had a loading high enough to load on Cluster 1) in the
cluster analysis would strongly suggest that crime is a
unique area of need in our population as compared to other

problems.

Nutrition

 

Nutrition was identified as a problem by 21.4 per-
cent of the study respondents, making it the fifth most

frequently cited problem. Since it loaded on

107

Cluster l--Multiple Problems with all of the other problem
areas (excepting Crime) it would seem to be associated with
the occurrence of these other problems. It is noteworthy,
however, that Nutrition exhibited the highest loading of any
Cluster 1 problem to the Poor Health cluster.

For an individual experiencing both health and
nutrition problems, the free meal programs currently run
around the Lansing area would be less of an option because
it would necessitate traveling to and from the site of the
meal program. For this less mobile group home meals would
be more convenient. However, since only 34.3 percent of
all respondents expressed a definite willingness to use a
home meal service, this program's present image and oper-
ation may need some revisions before it is acceptable to

large numbers of elderly individuals.

Spare Time

 

The cluster analysis demonstrated that reports of
less social involvement and less leisure involvement are
highly interrelated and are substantially associated with
Spare Time problems. Although only 13 percent of the study
respondents indicated spare time as a problem and only
4.3 percent rated it as a very important problem area, the
truly isolated old person living alone is probably unlikely
to be known to our referral sources (visiting and spare
time problems belonged to the same cluster) therefore

estimates of spare time problems are probably low.

108

The fact that respondents found a friendly visitor
and telephone reassurance programs considerably more
acceptable services than counseling services (54% to 34%)
suggests these options may be the best interventions for
dealing with primarily interpersonal deficits. On the
other hand, none of the services logically related to
leisure, such as volunteer programs or employment help,
demonstrated a wide acceptance (less than 23 percent saw
these as services they would use) among respondents. In
the final analysis, however, strict discriminations be-
tween programming aimed at interpersona1--or-1eisure--spare
time needs would seem counter-productive. It would seem
sufficient to emphasize the importance of interpersonal

opportunities within any number of spare time programs.

Other Problems

 

Service deception, housing, employment, and age
discrimination were reported less frequently than the
other problems just discussed and by less than 13 percent
of all respondents. These problems although obviously of
importance to those affected by them are less prevalent
than the other problem areas. Given the acknowledged
multi-nature of most problem reports, these less fre-
quently occurring need areas could probably be efficiently
handled by integrating appropriate service options (such
as consumer advocacy, house finding, and employment

finding services) into a comprehensive and coordinated

109

service strategy, rather than letting specialized programs

develop independently.

Alternatives to Institutionalization

Reports of willingness to use services indicated
wide acceptance for a number of home support services
although the more survival oriented services (health care,
daily living supports) were acceptable to larger number of
respondents. Although reports of willingness to use a
service are not synonomous with actual utilization, these
reports strongly suggest that the elderly living alone
questioned in this study would accept and use home care
services designed to prevent institutionalization.

This would seem a significant finding. While old
people living with a spouse or a significant other can
generally depend on receiving support during periods of
severe sickness or restricted mobility, such events often
precipitate institutionalization in the elderly living
alone (Rosow, 1967). However, the present evidence suggests
that many of these individuals would be willing to use
at home care and stay in the community and their homes if
adequate levels of home health and daily living support

were available.

 

The acceptability of institutional care to the
study respondents was not examined directly in this study
and thus weakens any conclusion which would imply respon-

dents found institutional care unacceptable. Although this

110

issue should be directly faced by any future assessments

in the Lansing area, the widespread acceptability expressed
by respondents for home care type services leads the author
to conclude this study provides a substantial endorsement
of the need for a well developed and publicized home care
service system.

While continuity of care and coordinated services
are always desirable goals for any human service delivery
system, the prevalence of multiple problems reports among
study respondents must necessarily exaggerate the importance
of achieving these goals. Since an information and referral
service (Service Advice) was the third most acceptable
service in the whole study, with over 75 percent of all
respondents reporting some willingness to use this kind of
service, the author would see a centralized and highly pub-
licized Senior Citizen Information and Referral Center and

telephone listing as a high priority in programming.

Hypothesis Testing

 

Confirmation of four of the five study hypotheses
(three were only partially confirmed) tends to be con-
gruent with the great body of reaearch which has described
the importance of health and activity (interpersonal and

leisure) in determing life satisfaction.

Losses.--Although the various age related losses

inquired about were fairly prevalent among study

111

respondents, the occurrence of these losses (health,
relative or spouse) did not appear to be related to Life
Satisfaction, replicating the findings of Leviton (1976).
While an effort was made to limit the scope of this hypo-
thesis to "recent" age related losses, the author now
believes that the two year period used in this study and
Leviton's study was probably too long and could possibly
have prevented the detection of the effects of very recent
age related losses. The effects on life satisfaction of
age related losses have been described by Kutner (1956) as
temporal. The present results and Leviton's would indicate
if Life Satisfaction is affected by Losses, such changes
are not noticeable after two years' time.

The present lack of support for the effect of age
related losses on Life Satisfaction tempered by the possi-
bility that a time limited change may in fact occur points
up the need for comprehensive longitudinal research on this
theoretical issue. A single point in time study such as
this one cannot hOpe to convincingly prove or disprove the
presence of a short lived alteration in life satisfaction.
While such considerations are theoretically interesting,
practically speaking, the present studies' failure to
reject the null hypothesis regarding age related losses
would suggest there is little need for extensive program-
ming efforts to enhance an elderly population's adaptation

to loss of spouse, relative, or health.

112

Health

Two highly correlated measures of health, Self
Health Evaluation and Sickness Rating, were found to be
significant correlates of Life Satisfaction while Total
Illnesses and Doctors Visits were not found to be meaning-
fully related to the dependent measure.

The high association of Self Health Evaluation
(.561) with (recent) Sickness Rating might suggest that an
individual's recent health-sickness experience is a major
determinant of Self Health Evaluation. While this seems
to be a reasonable theory (and obviously an unreasonable
scientific conclusion since our data is only correlative)
and Sickness Rating at first glance appears to be a rather
objective measure, it is difficult to say for certain
whether the opposite may not be true and one's evaluation
of their health status might not profoundly color a re-
spondent's report of the Sickness Rating response, "Just
Sick at Home." In this case, Sickness Rating would not be
a very objective measure, being determined by the indi-
vidual's evaluation of their health. Thus future research
should include professional appraisals of health as well as
self-report measures (Bultena and Oyler, 1971).

Nonetheless, multiple regression and cluster
analytic examinations of the data in the present study
revealed Self Health Evaluation and Sickness Rating to be

significant and substantially independent predictors

113

(from other variables--not each other) of Life Satisfaction,
replicating previous studies (Bultena & Oyler, 1971;
Lowenthal, 1975; Thompson, 1973; Maddox, 1963) which indi-
cate health, as measured by self health evaluation and
sickness rating, as a major and independent predictor of

Life Satisfaction.

Interpersonal Involvement

 

All of the interpersonal involvement measures
correlated significantly with Life Satisfaction. Based on
the strong associations observed between these variables
and their inclusion in the interpersonally defined cluster
in the cluster analysis it would appear that a single
interpersonal involvement factor accounts for all four of
the observed significant relationships identified in
Hypothesis 3.

Interpersonal and leisure involvement were suffi-
ciently interrelated to belong to the same cluster in the
cluster analysis solution. However, the fact that inter-
personal involvement measures constituted defining vari-
ables in the cluster and the fact that only interpersonal
involvement (Visiting Activity) was demonstrated to inde-
pendently predict Life Satisfaction on a level comparable
to Health (based on Beta weights) in the multiple re-
gression analysis, suggests that the interpersonal facet

of this activity factor is the most important and

114

independent associate of Life Satisfaction for the popu-
lation under study.

Since interpersonal involvement was approximately
equal to health in predicting Life Satisfaction, the
present research suggests both areas are equally important
in promoting life satisfaction for the population under

study.

Leisure Involvement

 

Although the hypothesis that leisure involvement
is a significant predictor of Life Satisfaction was con-
firmed, the present study reveals such involvement is con-
siderably less influential in determining Life Satisfaction
when Health, Income and Interpersonal Involvement effects
are held constant. Thus for individuals of similar
incomes, health and interpersonal involvement, differences
in leisure involvement appears to exert an insignificant
influence on individual Life Satisfaction.

While previous research on Life Satisfaction has
freely and unquestioningly noted the primacy of health in
determining Life Satisfaction, based on the significant
association of Self Health Evaluation measures, the present
study data indicates that other evaluatively oriented
measures related to areas such as interpersonal involvement
and leisure involvement also exhibit similarly large
correlations. This finding would seem to strongly question

the appropriateness of making predictive comparisons across

115

different substantive areas when different kinds of re-
sponses are elicited (i.e., evaluative versus non-evalua-
tive). For instance, to say that health is a better pre—
dictor of Life Satisfaction than leisure involvement
because self health evaluation had a higher correlation to
Life Satisfaction than number of leisure activities (this
was true in this study) would clearly ignore the like
evaluative nature of the Life Satisfaction and self health
evaluation measures. In point of fact, another evaluative
measure used in this study, Satisfaction with Leisure,
exhibited higher correlations to Life Satisfaction than
Self Health Evaluation before and after controlling for
the influence of demographic variables.

While Life Satisfaction appears to be a worthwhile
measure of need, methodological common sense and the idio-
syncratic way in which evaluative responses varied within
this study (i.e., Satisfaction with Leisure showing con-
siderable independence from Leisure Activities while both
significantly predicted Life Satisfaction, and the isolated
significance of Self Health Evaluation and Satisfaction
with Frequency of Visiting with Life Satisfaction for
individuals 75 or over) would seem to preclude seriously
comparing the predictive ability of subjective and non-

subjective measures.

116

Demographic Determinants of
Life Satisfaction

None of the demographic variables examined, except
income, independently exhibited a significant association
with Life Satisfaction. In addition, none of the demo-
graphic variables used in the partialing procedure drama-
tically changed the associations observed between Loss,
Health, Interpersonal Involvement or Leisure Involvement
measures and Life Satisfaction.

For study respondents income or real buying power
and not one's socio-economic status, calculated on the
basis of occupation and education determined Life Satis-
faction. This finding speaks for itself and implies that
the low SES auto trade worker or widow who gets a comfort-
able pension at retirement will probably be better off (all
other factors being equal) with respect to Life Satis-
faction than many of his/her less well endowed but higher
status peers.

As was observed earlier, the economic needs of the
elderly living alone and the income accessibility of any
service strategies which are effected or planned for this
group, must be a crucial consideration in meeting the needs
of old people living alone.

Overall Prediction of
Life Satisfaction

 

The best Health, Interpersonal, and Leisure

Involvement predictors along with Income accounted for

117

27.5 percent of the total variance in Life Satisfaction.
This is slightly more than the 24 percent of variance both
Pallmore and Luikart (1972) and Edwards and Klemmock (1973)
reported explaining in Life Satisfaction with variables
covering similar areas.

Although the results obtained in the first four
hypotheses are predominantly supportive of an activity
theory of aging adjustment, considerable support is pro-
vided through Hypothesis 5 for a disengagement theory of
aging. While the data appears to indicate individuals
with more visits, supporters, leisure activities and better
health reports will have higher life satisfaction scores,
this only seems to hold true for the younger elderly in
this study (under 75). For individuals 75 and over, there
is no evidence that favorable ratings on these same measures
translate into higher Life Satisfaction scores and more
importantly, that low rating on these measures will trans-

late into low life satisfaction scores.

Disengagement and Activity Theory

 

While the findings of the study hypotheses seem to
predominantly support an activity theory perspective of
aging adjustment, testing these identical hypotheses for
only those individuals 75 years or older revealed the
significant association of only two subjectively worded
measures: Self Health Evaluation and Satisfaction with

Frequency of Visits.

118

The findings supporting a disengagement theory
perspective seem convincing: all measures significantly
associated with Life Satisfaction for all respondents
exhibited lower associations among older respondents when
compared to younger respondents. Ten significant asso-
ciations to Life Satisfaction were observed for the younger
cohort, two were observed for the older cohort. While the
difference between scores in older and younger cohorts was
statistically significant for only one measure--Leisure
Activity, the consistently lowered association observed for
older respondents strongly implies some disengagement has
taken place.

Given the homogeneous nature of the population
under study here, old peOple living alone, it seems possible
that this group, possibly due to its limited social network
with its problem solving benefits, finds recourse to dis-
engagement more adaptive or necessary than other segments
of the elderly population who have been previously investi-
gated with respect to this same phenomenon (Tobin and
Neugarten, 1961).

Although these findings are rather persuasive,
they would seem to fall short of a definitive confirmation
of disengagement theory. Such a test could only be
achieved through a well planned longitudinal study which
documented changes in Life Satisfaction for a group of

individuals over an extended period of time. Therefore,

119

the author sees the present research as possibly filling
a small part of the vacuum for those who like Turnball
(1966) have felt that disengagement occurs but not to all
old people.

Despite the support given here to the existence of
the disengagement process, there would seem to be a need
for considerably more work on this area. Clarification of
the ambiguous and illusionary nature of the concept itself,
would seem to be the first order of business for further
studies on the existence of a disengagement process. If
a widely accepted operationalization of the state of dis—
engagement is achieved, researchers could then proceed to
elaborate the applicability of this theoretically attrac-

tive but rather ill defined phenomenon.

CONCLUSIONS

The major findings and conclusions of this study

can be summarized as follows:

1.

While there is some evidence income problems are
seen as very serious by respondents, the impracti-
cality of meaningfully intervening upon this prob-
lem area at the local level argues against tar-
geting this need directly. However, the associa—
tion of income with Life Satisfaction and the low
mean income of respondents dictates that income
barriers to other need related services be examined
and eliminated where possible. The prevalence of
multiple problem reports among respondents provides
evidence of the need for a coordinated and inte-
grated strategy of program deve10pment.
Transportation related problems, the only problem
area named as an important problem by a majority

of respondents, is clearly the most prevalent prob-
lem for the study population and hence must be
considered a high priority need for the population.
The prevalence of transportation problems and the

large number of respondents reporting a willingness

120

121

to use such services would seem to make present

and planned transportation programs an ideal avenue
for casefinding and outreach efforts.

Respondent approval for home health care services,
daily living support services and to a lesser

extent interpersonal support services suggest
elderly live alones would use such services. There-
fore, a comprehensive and integrated home care ser-
vice system should be available to those individuals
who choose community residence over institutional.
Elderly individuals do not appear to exhibit any
meaningful reductions in Life Satisfaction due to
personal or health losses experienced during the
previous two years. On the other hand, an indivi-
dual's current health evaluation, frequency of
involvement with other people, and number of leisure
activities appear to be significant associates of
personal Life Satisfaction.

Amount of social support and leisure activity was
correlated with life satisfaction of respondents
under 75 years of age supporting the activity theory
of aging. However, social support and leisure
activities were not correlated with life satisfac-
tion for respondents over 75 supporting disengage-

ment theory for the eldest of the study sample.

122

Future Assessments of Need

The present study represents a small scale attempt
to assess the needs of a specific at risk subgroup of
elderly individuals in a specific locale, Lansing, Michigan.
While the present study has shed some light on the needs of
the study population, it seems reasonable to state that self
report surveys of limited scope, such as the present study,
are fraught with interpretive difficulties.

The interpretative problems which exist for the
present study can probably be attributed to the incongruity
of trying to generate comprehensive need data within the
confines of a relatively short questionnaire. Since infor-
mation had to be collected about a rather large number of
problem areas and potential services within a short time
frame, questions about these iSSues had to be limited in
number and broad in scope. By limiting the number and
hence the specificity of the individual items used in the
questionnaire, it is difficult to ascertain with certainty
the true validity of the results presented in the fore-
going study.

The validity of the study results could have been
enhanced by doing two things: (1) examining respondent
needs using the multi-trait multi-method approach suggested
by Campbell and Fiske (1959) and (2) questioning respondents
about problem area issues which had more concrete behavioral

specificity than some of the broadly stated categories

123

(like "Leisure Time") which were used here. These
approaches were not pursued extensively in the present
study because they would have geometrically increased the
number of questions asked and the time needed to complete
each survey. In the present instance, logistical consid-
erations of time and money precluded elaborating the
methodological sophistication of the survey instrument to
an extent which matched its breadth of coverage. For
instance, can we say with certainty that the higher impor-
tance given transportation problems over health problems
by study respondents reflects real and not methodological
or interpretative artifacts inherent in the questionnaire
construction or phrasology?

Future assessments of need which are not similarly
constrained by time and resource considerations would
likely reap data of a higher quality than the present
study by examining problem area in depth with respect to
both method of inquiry and specificity. Others, faced with
a similar "trade off" situation as was faced here, may do
well to first consider limiting the breadth of their

study.

APPENDICES

APPENDIX A

SENIOR CITIZEN QUESTIONNAIRE

CONSENT FORM

APPENDIX A

SENIOR CITIZEN QUESTIONNAIRE

CONSENT FORM

1) I have freely consented to be interviewed by Mr.

Denis Gray or Ms. Caroline Clark.

2) This study has been explained to me and I understand
the explanation that has been given and what my
participation will involve. I also understand that
the information given in this interview will be
treated in strict confidence and that I will remain
anonymous. Results of the questionnaire will be

made available to me at my request.

Signed

 

Date

 

(Senior Citizen Program and Michigan State University)

124

APPENDIX B

STUDY QUESTIONNAIRE

APPENDIX B

STUDY QUESTIONNAIRE

The purpose of this questionnaire is to decide on
the need for new programs and services. I will be asking
you questions about different ways you spend your time
and things which have happened to you in order to see
how your needs may be the same or differ from other
senior citizens. Before beginning the actual question-
naire I'll first need some information about you like
your name, age, etc. OK, I'll begin now unless you
have any questions.

Individual or Agency identifying respondent

 

NEED ASSESSMENT FOR ELDERLY LIVING ALONE

 

 

Demographics

#1. Name:

#2. How long at present address ____ yrs.

#3. How long in Lansing area ____ yrs. (code as stated)

#4. Type of residence
Own Apt.
Own Home
Rent Home
Rent Apt.
Rent Room
Other

 

#5. Age:

 

#6. Sex: Male Female

#7. Marital Status: Married Widowed
Divorced Separated
Never Married

 

125

126

#8. If widowed, how long ago

#9. Household Composition: Live Alone Live with
Spouse
Live with other

 

#10. Education (how far did you go in school):
Grade School H.S.
Col. Other

a. How far did Husband go

#11. Race: Caucasian Other
#12. Do you presently have a job? Yes No
a. Job: full time part time

#13. Occupation:

 

#14. Previous occupation:

 

Spouse's occupation:

 

#15. Religion: Protestant Catholic
Jewish Other (spec1fy)

#16. Number of living children:

 

#17. What is your total income from all sources
Social Security
Dept. Soc. Serv.
Employment
Pension
Other (specify)

 

Total

I wish to understand how you spend your time,
therefore I am interested in knowing what you do and
who you see. The first thing I'd like to know is how
you spend your leisure or enjoyment time. I am now
going to read to you some leisure activities and would
like you to tell me if you have participated in each
during the past week.

During the past week have you . . .
(REPEAT before EACH activity) YES NO

#18. Read newspaper

127

#19. Read a book or magazine '____ ____
#20. Watched TV ____ I____
#21. Listened to radio ____. ____
#22. Listened to music ____ ‘____
#23. Worked at hobby (specify) ____ I____
#24. Written letters ____ .____
#25. Participation in church ____ ____
#26. Played cards (other games) ____ ____
#27. Gardened ____ ____
#28. Watched a sports event ____’ ____
#29. Participated in a sport ____ ‘____
#30. Taken rides ____ ____
#31. Gone for walks ____ {____
#32. How many hours would you estimate you spend in an
average day on your leisure activities? hrs.
(Repeat activities performed from above.)_——
#33. How satisfied are you with your present partici-

pation in leisure activities:
very satisfied satisfied
dissatisfied very dissatisfied

The next series of questions relates to various
people you might spend time visiting or might visit you.

#34. How many of your children live within 25 miles
(code as stated)

#35. How many relatives live within 25 miles
(code as stated)

#36. How many friends live within 25 miles
(code as stated)

#37. Just considering this past month, how many personal
visits have you had with your children
(code as stated)

#38.

#39.

#40.

#41.

#42.

#43.

#44.

#45.

#46.

#47.

#48.

128

How satisfied are you with the frequency of the

personal visits you now have with your children

very satisfied ‘ (1) somewhat satisfied (2)

somewhat dissatisfied (3) very dissatisfied
(4)

 

Just considering this past month, hQW'many personal
visits have you had with your relatives (excluding
children (code as stated).

 

How satisfied are you with the frequency of the
personal visits you now have with your relatives
very satisfied somewhat satisfied
somewhat dissatisfied very dissatisfied

 

Just considering this past month, how many visits
have you had with your friends
(code as stated)

How satisfied are you with the frequency of the
personal visits you now have with your friends
very satisfied somewhat satisfied

somewhat dissatisfied very dissatisfied

Just considering this paSt month, how many organi-
zations or club meetings have you attended
(excluding church attendance)

 

How satisfied are you with your participation in
organizations

very satisfied somewhat satisfied
somewhat dissatisfied very dissatisfied

During the past month, how often did you speak to
your children on the telephone

During the past month, how often did you speak to
your relatives (excluding children) on the
telephone

 

During the past month, how often did you speak to
your friends on the telephone

During the past month, how often have you talked
to people on the phone about organizational
activities?

life in general.

129

I am going to read to you some statements about

Applying each statement to yourself,

please tell me whether you agree or disagree with it.

#49.

#50.

#51.

#52.

#53.

#54.

#55.

#56.

#57.

#58.

#59.

#60.

#61.

#62.

Life Satisfaction

 

I have gotten more of the breaks
in life than most of the people
I know.

This is the dreariest time of my
life.

I expect some interesting and
pleasant things to happen to me
in the future.

I often feel lonely.

Getting older doesn't bother me.
Compared to other people my age,
I've made a lot of foolish

decisions in my life.

Compared to other people, I get
down in the dumps too often.

As I look back on my life, I am
well satisfied. ’

Compared to other people my age,
I make a good appearance.

I've gotten pretty much what I
expected out of life.

As I grow older, things seem better
than I thought they would be.

I am just as happy as when I was
younger.

Most of the things I do are boring
and monotonous.

In spite of what people say, the
lot of the average man is getting
worse not better.

Agree

Disagree

130

A person's health is an important part of a
person's life. Therefore, I'd now like to ask you some
questions regarding your health.

#63. How many days have you been sick to the point of
being unable to carry on your regular activities
during the past month .

 

#64. If you were sick during this time, were you mostly:

 

 

a) not sick (0 weight)
b) just at home (.5 weight)
c) in bed at home (1.0 weight)

 

d) in the hospital (1.5 weight)

 

#65. How many times have you seen the doctor during the
past year (code as stated)

Illnesses

 

I am now going to read you a list of illnesses.
Please tell me whether you presently have:

YES __1__ NO __2_

12:2 39
#66. Paralisis (Parkinsons Disease) (1) (0)
#67. Diabetes (1) (0)
#68. High blood pressure (1) (0)
#69. Heart trouble (1) (0)
#70. Weakness from stroke (1) (0)
#71. Arthritis (1) (0)
#72. Stomach ulcer (l) (0)
#73. Emphysema (l) (0)
#74. Glaucoma or (cataracts) (l) (0)
#75. Cancer (1) (0)

#76. How do you rate your health at the present time?
excellent (4) good (3) fair (2)
poor (1)

131

Various events can affect our lives and I am now
going to read a list of different events and would like
you to tell me if each has occurred to you during the
past two years.

During the past two years (repeat for each)

YES 59
#77. Has your spouse passed away? 1 0
#78. Has a close family member passed

away? 1 0
#79. Have you suffered from a serious

injury or illness? 1 0
#80. Have you retired from work? 1 0

#81. Have you, during the past year, received help from
any agencies or social service program in the
Lansing area? Yes No

 

Which ones Type of help

 

 

 

 

 

 

 

 

I am now going to read a list of services, some of
which are available now, some of which are being planned.
Please tell me if you would use each service.

#82. Would you use a service where you talked over
problems with a counselor?
definitely would use , might use ,
probably would not use

 

 

#83. Would you use a service where you were given
transportation for medical treatments?
definitely would use , might use ,
probably would not use

#84.

#85.

#86.

#87.

#88.

#89.

#90.

#91.

#92.

#93.

#94.

132

Wbuld you use a service where you received home
care from a nurse?

definitely would use , might use ,
probably would not use

Would you use a service where you received physical
therapy at home?

definitely would use , might use ,
probably would not use

Would you use a service where you received help
with cleaning and housekeeping?

definitely would use , might use ,
probably would not use

Would you use a service where someone brought meals
to you?

definitely would use , might use ,
probably would not use

Would use a service where you talked to a lawyer
regarding legal problems?

definitely would use , might use ,
probably would not use ,

Would you use a service where you received transpor-
tation for getting around Lansing?

definitely would use , might use ,

probably would not use

 

Would you use a service where you received help
in finding paid employment?

Would you use a service where you received help
in finding volunteer work?

definitely would use , might use ,
probably would not use

WOuld you use a service where you received help and
advice in getting money from Social Security and
other agencies?

definitely would use , might use ,
probably would not use

 

 

WOuld you use a service where you had someone
regularly visit you?

definitely would use , might use ,
probably would not use

 

WOuld you use a service where you had someone call
regularly on the telephone to check on you?
definitely would use , might use ,
probably would not use

#95.

#96.

#97.

133

Would you use a service where you had someone do
shOpping for you?

definitely would use , might use ,
probably would not use

 

Wbuld you use a service where you had assistance
with finding a new place to live?

definitely would use , might use ,
probably would not use

WOuld you use a service which helped match old
people to share their residence with someone
else?

definitely would use , might use ,
probably would not use

 

Now I am going to read you a list of areas which

people say are problems for older Americans. For each
area, please tell me if it is no problem to you, a
somewhat important problem, or a very important problem.
(READ LIST - ROTATE ORDER)

#98.
#99.
#100.
#101.
#102.
#103.

#104.
#105.
#106.

#107.

 

 

Somewhat Very
No Important Important
Problem Problem Problem
Income (money) 1 2 3
Health care 1 2 3
Housing 1 2 3
Transportation 1 2 3
Age discrimination l 2 3
Employment
opportunities 1 2 3
Spare time activities 1 2 3
Crime 1 2 3
Nutrition and food 1 2 3

Services and business
misleading their users 1 2 3

134

This is the last part of the questionnaire. The
answers to the following questions may seem obvious,
however, I'd like you to answer them nonetheless.

#108. Where are we now?

 

#109. What day of the week is it?

 

#110. What month is it?

 

#111. What year?

 

#112. Who is president of the U.S.?

 

#113. Who was president before him?

 

APPENDIX C

BREAKDOWN OF DEMOGRAPHIC

CHARACTERISTICS OF

RESPONDENTS

135

 

 

 

Amm.mv Amm u zv
oo.m mm.oa III zoowz .mumow
now u zv
Aomvw>.mm mHmEmm
Aoavwm.qa mam:
xmm
Amn.nv Ao> u 2V
om~.v> hm.v> III mm<
Ame u 21
Aaaewm.ma ucmsunmmm uamm
Aomvwo.mm 050m ucmm
Ammvwa.mm mEOm G30
mocmoflmmm
Amh.wav th u zv
oo.oa av.ma In: mmmuood .mumm»
Avv.mav Amo u zv
moCm who Q? .lll mCngQH smHmmN
Acofiumw>mo osmocmumv mcwocommmm Aucooov
cmflomz com: sz mmmucwoumm manmwum>

 

mBZszommmm m0 mUHBmHmmaodmde UHmmdmwozma m0 zzoamdmmm

U xHQmem4

136

“on u zv

 

 

 

Amavwv.am Hmcuo
Ammvwm.mn Soowz
msumum Hmuflumz
Amucmocommmu
omfiuums Ho>mv
Amo.mv Amm u zv
mm.a ov.m In: cmuoawso mo Hmnaoz
th u zv
Amvwm.v Hmsuo
Awavwm.em oaaonumu
Aomvwv.ah ucmummuoum
coamwamm
Ammv
Amvwm.v made Haom
Amvwa.n mafia uumm
Aamvwm.mm wcoz
pcwEMOHmEm
A05 n zv
Amvwm.m xomam
Ammvwa.hm muagg
momm
Amm.Nv Ron u Zv
oo.oa HN.OH III cowumooom
ACOHDMH>OQ oumocmumv mcflocommwm Ausooov
GMflomz com: sz monocoOHmm manmwum>

 

137

coaufimom Hmfloom swan u H4

 

 

 

 

1mm u zv
loavwa.ma m
lhvwm.aa v
Aamvwo.vm m
Inacwv.u~ N
Anvmm.aa H
mmm«
1mm u zc
Aavwm.ma mu
Aamcwm.vm mu
“GOAHOSO
mflumum Hmpfimz
Am.s-.~v loo u ze
o.v~o.m oo.onm.mw In- meoocH
ACOHflMHNrOO OHMUGMHmV mfiflflcommwm ADGDOUV
cmaomz

ommz

sz ommuomoumm mannaum>

 

APPENDIX D

BREAKDOWN OF INTERPERSONAL INVOLVEMENT

ITEMS BY SOURCE OF

INVOLVEMENT

 

 

III mam.h mm. ma.v rosy Ignace “may muflmfi> I
III om.m om. mH.m Aonv mmawe mm canvas I
mm>wumamm
Anuooa Hmmv
III mn.ma om.om om.mm Aomv maamo mcocmmamu I
mmm oofimmHDMm huo>
wmm omwmmADMm
wm omflmmwummmwo
wm omfimmfiummmwo huo>
Aomv cowuommmwumm I
II- Hmm.vH omn.e meH.mH Ammo lapses ammo muflmn> I
III mv.a mm.a mm.H Ammv mmHHE mm cflnufi3 I
g
omnpcmoumm Guano: com: 2 manmanm>

 

BZMZH>AO>ZH ho momDOm Mm mszH

BZQSW>QO>ZH AdzommmmmMBZH m0 Z3OQM¢mmm

XHQmem<

138

139

Anacoe Hmmv

III Hm.am om.o~ sq.mm lone maamo mconmmamu I

wm.mm omwmmwumm hum>

wm.Hm owammflumm

wo.ma _ omwmmwummmfio

wv.a omwmmfiummmfio aum>
cofluomwmwumm I

III mv.am om.oa s~.o~ lone lapses ammo muflmfl> I

III mm.ma m~.oa oa.ma Ammo mmaae mm cwcufl3 I

mmmmmmm

Anacoe Homo

III vo.~a oo.~ mm.m 4 Ammo mason mcogmmamu I
w~.mv omwmmflumm aum>
ma.mm omflmmflumm
wm.m oowmmflummmflo
w~.m omfimmwummmwo >Hm>

Ammo coauommmflumm I

 

mmmucmoumm cowomz com: 2 manmwum>

 

140

 

Anacoa Homo

 

 

III no.m NH. mh.m Ammo mHHmo muonmoamu I
ma.ha 30c muoe
wm.vm 30: mean
ww.m¢ 3oz mmma

Amv mom mocwmv

mufimfi> ow mmcmso I
wo.a> omwmmflumm hum>
mb.ha omflmmfiumm
mh.m omflmmwummmwo
wm.a omwmmHDMmmHo >Hm>

Ammv cofluomMmHDMm I

III wm.m RH. mH.H lone Amuomucooc muflmfi> I

mcowumeqmmHO
mmmucmoumm cowomz com: 2 manoeum>

 

APPENDIX E

VARIABLES EXCLUDED FROM CLUSTER SOLUTION
BECAUSE OF COMMUNALITIES

BELOW .200

APPENDIX E

VARIABLES EXCLUDED FROM CLUSTER SOLUTION
BECAUSE OF COMMUNALITIES

BELOW .200

Years Residing in Lansing
Age

Sex

Marital Status
Years Widowed
Religion

Number of Children
Income

Hours of Leisure
Doctors Visits
Total Illnesses
Loss of Spouse
Loss of Relative
Health Loss

Socio Economic Status.

Problem Crime

141

BIBLIOGRAPHY

BIBLIOGRAPHY

Anderson, Barbra, G. The adaptive task of aging: a psy-
chosocial analysis Fund Report, NIMH Grant, MA
12492, 1968.

 

Bennett, R. Social Isolation and Isolation Reducing Pro-
grams. Bulletin of the New York Academy of
Medicine, 49, No. 12, 1973, 1143-1163.

Bultena, Gordon, and Oyler, Robert. Effects of Health and
Morale on Disengagement and Morale. Aging and
Human Development: 2, 1971, 142-148.

 

 

Butler, Robert N. Why Survive Being Old in America.
Harper & Row Publishers, N.W. 1975.

Campbell, D.T., and Fiske, D. W. Convergent and discrimi-
nant validation by the multi-trait--multimethod
matrix. Psychological Bulletin, 56, 1959, 81-105.

 

Cumming, Elaine and William, Henry. Growing Old. New
York: Basic Books, 1961.

 

Dohrenwend, B. P. and Dohrenwend, B.S. Social Status and
Psychological Disorder: A Causal Inquiry. New
York: John Wiley & Sons, 1969.

 

Edwards, J. and R. Klemmock. Correlates of Life Satis-
faction: A re-examination. Journal of Gerontology.
1973, 28, 497-502.

 

Fairweather, G. W. Methods for Experimental Social Inno-
vation. New York: Wiley, 1968.

Fowler, F. J., McCalla, M.E. Correlates of Morale among
aged in greater Boston. Proceedings of the 77th
Annual Convention of the APA, 1969.

 

 

Gray, D. 0. Community Service Preference in North Tampa,
unpublished manuscript, 1977.

Hollingshead, A. B. Two Factor Index of Social Position.
New Haven, Yale Press, 1957.

142

143

Holmes, T. H., and Rake, R. H. Social Readjustment Rating
Scale. Journal of Psychosomatic Research. 1967, 5,
115-122. "'

 

Institute for Interdisciplinary Studies (IIS). Indicators
of the Status of the Elderly in the United States.
1972.

 

 

Katz, Sidney. Patient Status Instrument Project for
Functional Outcome Measures, Part II of Final
Report. 1974. College of Human Medicine, Michigan
State University.

 

 

Koos, E. L. Families in Trouble. New York: Kings Crown
Press, 1946.

 

Kutner, B. et a1. Five Hundred Over Sixty: A Community
Survey on Agin . New York, Russell Sage
Foundation, 1 56.

 

 

Locke, B., Kramer, M., Passamanick. "Mental Diseases of the
Semi Mid-Century." First Admissions to Ohio State
Public Mental Hospitals. American Journal of
Public Health, 1960, 50, 998-1012.

 

 

Leviton, B. Social Support Systems and the Aged Widow
and Widower. MA Thesis, MiChigan State University,
1976.

 

 

Lowenthal, M. F. Antecedents of isolation and mental ill-
ness in old age. Archives of General Psychiatry,

 

Lowenthal, M. F. and D. Boler. Voluntary and Involuntary
Social Withdrawal. Journal of Gerontology, 20,
1965b, 363-371.

 

Maddox, G., and C. Eisdorfer. Some correlates of activity
and morale among the elderly. Social Forces 1962,
40 (March): 254-260.

 

Maddox, G. R. Activity and Morale: A Longitudinal Study
of Selected Elderly Subjects. Social Forces,
1963, 42, 195-204.

 

Maddox, G. L. Disengagement Theory: A critical evaluation.
Gerontologist, 4: 80-82, 1964.

 

Maddox, G. L. Persistence of Life Satisfaction Styles Among
the Elderly: A Longitudinal Study of Patterns of
Social Activity in Relation to Life Satisfaction.
Proceedings, 7th International Congress of Geron-
tology, Vienna, 6, 1966, 309-311.

 

144

Maddox, George. The Future of Aging and the Aged. SNPA
Foundation Seminar Books. Atlanta, 1971.

Michigan Office of Services to the Aging. Michigan Aging
Citizens: Characteristics, Opinions, and Service
Utilization Patterns. Ann Arbor: Institute of
Gerontology, 1975.

 

Morrison, D. E. and Kristjanson, H. Personal Adjustment
Among Older Persons, Technical Bulletin No. 21,
Agriculture Experiment Station, South Dakota State
College of Agriculture and Mechanic Arts, 1958.

 

Neugarten, B. Personality in Middle and Late Life. N.Y.
Atherton, 1964.

Neugarten, B., Havinghurst, R., and Tobin, S. The Measure-
ment of Life Satisfaction. Journal of Gerontology,
20, 1965.

 

Pallmore, E. and Leukart, C. Health and Social Factors
related to Life Satisfaction. Journal of Health
’and Social Behavior. 1972, 13, 68-80.

 

 

Passamanick, Benjamin. "A Survey of Mental Disease in an
Urban Population VI: The Approach to Total Pre-
valence by Age," Mental Hygiene 1962, 46, 567-572.

Riley, M., Foner, A. (eds). Aging and Society. Volume
One: An Inventory of Research Findings. New York,
Russell Sage Foundation, 1968.

 

Rosow, I. Social Context of the Aging Self. Gerontologist,
3(1), 1973.

 

Rosow, Irving. Social Integration of the Aged. New York:
Free Press, 1967.

St. Lawrence Hospital Community Mental Health Center, Out-
patient Utilization Statistics, 1974.

St. Lawrence Community Mental Health Center. Outpatient
Utilization Statistics, 1975.

Scott, Frances and Ruth Brewer (eds). Perspectives in
Aging II. Operational Focus. Continuing Education
Publications, 1971.

 

Siegel, L. M., Attkinson, E. C., Cohn, A. H. Mental Health
Need Assessment: Strategies and Techniques.
National Institute of Mental Health Reports. San
Francisco: Langley Porter NeurOpsychiatric Insti-
tute, 1974.

 

145

Sjogien, Torsten, and Tage Laisson. "The Changing Age Struc-
ture in Sweden and its Impact on Mental Illness."
Bulletin of World Health Organization, 1959,

21, 569-582.

Srole, Leo et a1. Mental Health in the Metropolis: The
Midtown Manhattan Study, 1, 1962. New York:
McGraw-Hill Book Company, Inc.

 

 

Strieb, Gordon. "Morale of the Retired." Social Prob-
lems, 1956, 3, 270-276.

 

Thompson, G. Adjustment in Retirement: A Causal Inter-
pretation of Factors Influencing Morale in Retired
Men. Dissertation Abstracts, 1973.

 

Tobin, S. and Neugarten, B. Life Satisfaction and Social
Interaction in the Aging. Journal of Gerontology,
1961, 16, 243-245.

 

Tryon, R. C. and Bailey, D. E. Cluster Analysis. New
York: McGraw-Hill, 1970.

 

Turnstall, Jeremy. Old and Alone. Routledge and Kegan
Paul: London, 1966.

 

U. S. Bureau of Statistics, Census Data, 1970.

 

United States Department of Health, Education, and Welfare
(USDHEW), Office of Human Development, Administra-
tion on Aging. Assessing the Status and Needs of
Older Americans. Utilization Manual. Washington,
D.C.: Government Printing Office, 1975.

 

 

United States Department of Health, Education, and Welfare
(USDHEW), Public Health Service. Mental Disorder
of the Aging. Washington, D.C.: Government
Printing Office, 1968.

 

 

Waddell, F. (ed.) The Elderly Consumer. Baltimore:
Human Ecology Center, Antioch College, 1976.

 

Waheit, G. J., Bell, R. A., Schwab, J. J. Planning for
Change: Need Assessment Approaches. National
Institute of Mental Health Report. Tallahassee:
University of Florida, 1976.

 

 

 

Wohlhill, J. Methodology and Research Strategy in the
Study of Developmental Change. In Life-Span
Developmental Psychology, Goulet, L., Boltes, P.
(eds), New York: Academic Press, 1970.

 

 

"‘IIIIIIIIIIIIIII