4‘ 1° ' ‘e' H. . I .\ .
a. 1 . 1.1, I .32“. ‘I A \N:
...?3f13'3,$g:.. g I‘A‘,,:¢:, I" I?

ALTERNATIVE MODES 0F CONDUCTING OUTREACH
TO LOW INCOME ELDERLY:
AN EXPERIMENTAL EXAMINATION

Thesis for the Degree of M. A.

MICHIGAN STATE UNIVERSITY

MARTIN GREGORY KUSHLER
1977

-- 7.4—

3 1293

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII ~_ ~ ~ [’ {g {-5 '2?" .

’ALILA

 

./

 

 

17" A153

 

ABSTRACT

ALTERNATIVE MODES OF CONDUCTING OUTREACH
TO LOW INCOME ELDERLY:
AN EXPERIMENTAL EXAMINATION

By

Martin Gregory Kushler

The purpose of this study was to experimentally evaluate different
methods of conducting outreach to low income elderly, in an attempt to
inform them of services available to meet their needs. Four modes of
contact (informational mail, personal mail, telephone and in—person)
were analyzed for their effectiveness in reaching and influencing a
target p0pulation of lower income elderly in a three county area in
southwestern Michigan.

A sample of 325 low income elderly (age 65 and over), having no
prior contact with the regional information and referral network, were
selected as a target sample and randomly assigned to groups receiving
one of the four abovementioned modes of contact or to a no-treatment
control group. The goals of the outreach contact were designed to
correspond to three of the outcome measures commonly used by that
existing regional information and referral network. Specifically, the
outcome criteria included: whether or not the person registered with
the local center, whether or not the person requested to sign up to

receive the center's "newsletter,"

and whether or not the person actually
received a service from the center.

The results revealed that the in—person mode of contact was the most

effective in terms of all three outcome criteria used in this study and

Martin Gregory Kushler

was clearly the preferable mode to utilize if a maximum impact was desired.
In addition, however, the results indicated that the lower cost modes of
mail and telephone contact would achieve some degree of success with this
target population, particularly for non-complex or non-threatening out-
reach tasks such as providing basic information, creating name awareness
and fostering a positive attitude toward an agency. Hence, it appears
that the choice of a particular mode of outreach should depend on such
situational variables as the purpose for which the contact was intended
and the amount and nature of resources available to the organization. It
is felt that the results of this study and the discussion of their
implications for policy would prove interesting to anyone involved with

the provision of services to the elderly.

ALTERNATIVE MODES OF CONDUCTING OUTREACH
TO LOW INCOME ELDERLY:

AN EXPERIMENTAL EXAMINATION

By

Martin Gregory Kushler

A THESIS

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

MASTER OF ARTS

Department of Psychology

ACKNOWLEDGMENT

I wish to thank the members of my committee, Dr. William S.
Davidson II, Dr. Esther Fergus and Dr. Cyril Akpom, for their guidance
and support in this research. Their advice and their patience over the
extended length of this field project are much appreciated. Also, I
would like to thank Dr. Robert Calsyn, who was very instrumental in
helping me plan and initiate this project.

Thanks must also be expressed to Jake Miklojcik and the Michigan
Office of Services to the Aging for their sponsorship of this project,
and to Bob Dolsen and the Region IV Area Agency on Aging for their tre-
mendous cooperation, assistance and appreciation of the need for true
experimental field research. Special thanks are included for all the
people who make up the information and referral network in Berrien, Cass
and Van Buren Counties. Clearly, this research would not have been
possible without the support and assistance of all of these people.

Thanks are extended to John Jeppesen, who was also pursuing research
in the Region IV area, for providing good company and being a good co-
driver through literally thousands of miles of snow, wind, rain and sun.
Happily, it all seems worth it now.

Thanks also go to the other faculty and students in the ecological
psychology program for their support and advice and for being a part of
the total educational experience allowing me to pursue such a project.

Finally, and most importantly, thanks to Charlene for her support,
encouragement, assistance, patience and love. Without generous amounts

of each of those, this project could never have been completed.

ii

TABLE OF CONTENTS

Page

LIST OF TABLES .................................................. vi

LIST OF APPENDICES .............................................. viii
Chapter

I. INTRODUCTION ............................................ 1

Background: The Elderly as a Problem Population .... 1

Special Characteristics of the Elderly ......... ..... 2

Biological Aspects of Aging ................ ....... 3

Psychological Aspects of Aging............ ......... 5

Situational Aspects of Aging ...................... 8

Rationale for Social Services ....................... lO

Assumptions Underlying Government's Role .......... l4

Legislation, Structures and Agencies .............. l7

Demographics and Needs of the Elderly .. ............. 24

The Problem of Linkage .. ..... .......... ........ ..... 28

Information and Referral ........................... 30

Outreach Strategies ............................... 31

Modes of Outreach Contact ......................... 32

Mail ............................................. 33

Telephone... ..................... . ............... 35

In-Person ......... ' ................. . ............. 35

The Project ........................................ 36

Rationale .......................................... 37

II. METHOD .... ............................................. 39

Subjects.. ............. . ............................. 39

Design ............................................... 40

Procedure ............................................ 41

Experimental and Control Conditions ................ 41

Control Group............. ....................... 41

Informational Mail Contact Group...... ........... 42

Personal Mail Contact Group..... ................. 42

Telephone Contact Group.......................... 43

In-Person Contact Group....... ................... 43

Training and Supervision of the I & R Personnel.... 44

Measurement of the Dependent Variables ............ 44

The "Client Card" ............. ..... .............. 45

"Newsletter Sign-Up"..................... ........ 46

"Number of Pe0ple Receiving Services" ..... . ...... 47

The FolloweUp Survey. .......... . ................ ..... 48

Construction of Scales ............................. 49

Family Contact Scale ............................. 50

Other Social Contact Scale ....................... 51

iii

Chapter Page

Knowledge of Social Services Scale............... 51

Life Satisfaction Scale.......... ............... . 51
Further Analysis of the Survey Items. .............. 51
The "Problems" Scale ..... ..... ...... . ....... ..... 52
Other Areas of Interest .......... . ..... .... ........ 53

Concluding Operations................................ 54

III. RESULTS.............. ............. ....... ..... . ......... 56
The Sample ........ ... ..................... . .......... 56
Subject Mortality.......... ........................ 56
Effectiveness of Randomization ..................... 57
Treatment Effects.................................... 58
Primary Outcome Measures................... ........ 58
ANOVA to Test for Population Differences and
Interaction Effects....... ..... ........... ....... 60
Secondary Outcome Measures......................... 61
The Effects of Receiving a Service................. 62
Correlational Analyses .......... ........ ..... . ....... 64
Relationship of the Survey Variables to the
Three Dependent Variables..... ........ ..... ...... 64

Summary of the Predictability of the Three
Dependent Variables Using the Survey Variables
as Predictors.................................... 68

Predictability Including Type of Contact ........... 7O
Predictability Within Type of Contact ...... ... ..... 72
In-Person Contact Recipients..................... 73
Telephone Contact Recipients....... .............. 74
Mail Recipients (Informational and Personal
Letter Groups Combined).................. ...... 76

Reasons Given by Interviewees for Response or
Non-Response to the Outreach Contact............... 78

IV. DISCUSSION.. ............................................ 80
Experimental Outcomes...... ..... ............ ......... 80
The Client Card Registration Variable........ ..... . 80

The Variable of Receiving a Service................ 81

The Newsletter Dependent Variable.................. 84

The Informational Mail vs. Personal Mail Contacts.. 86
Results of Attempts at Identifying Predictors

of Positive Response............................... 87

The Client Card Registration Variable: Who
Responds?........................................ 87

The Variable of Receiving a Service or Not: Who
Receives?........................ ..... 90

The Newsletter Dependent Variable: Who Signs Up?.. 92
Implications for POI—icy. O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 95
Relative Costs ...... . ..... .... ...... . ............ .. 96

iv

Chapter Page

For the Client Card Registration Variable ........ 96
For the Variable of Receiving a Service .......... 96
For the Variable of Signing Up for the
Newsletter ....................... . ............. 96
Use of the Lower Cost Modes of Contact............. 97
The No-Contact Control Group .......... . ............ 99
The Mass Media ............ ......... .............. 100
Other Miscellaneous Sources of Information....... 101
The Use of the In—Person (Home Visit) Mode ......... 101
The "Passive" Recipient .......................... 101
Conclusion .......................... . ................ 104
APPENDICES ....................................................... 107
LIST OF REFERENCES ............................................... 142

TABLES

10.

ll.

12.

13.

14.

15.

l6.

17.

LIST OF TABLES

Experimental Design ....................................
Correlations Among the Scales ............ . .............
Project Outline ........................................
Final Sample Composition: Subject Mortality by

Contact Mode..... ..... . .................. ... ...........
Client Card Registrations by Type of Contact ...........
Receiving a Service by Type of Contact .................
Newsletter Sign-Ups by Type of Contact .................

Analysis of Variance for Client Card Registrations

by Type of Contact and Assistance Group ..... .... .......

Analysis of Variance for Receipt of Service by

Type of Contact and Assistance Group.......... .........

Analysis of Variance for Newsletter Sign-Ups by

Type of Contact and Assistance Group ..... . .............

Pearson Correlations of Demographic Variables

with the Three Dependent Variables .....................

Pearson Correlations of Survey Variables with

the Three Dependent Variables............. .............

Pearson Correlations of Client Ratings of

Outreach Contact with the Three Dependent Variables...

Multiple Regressions: The Three Dependent Variables

with the Significant (p< .10) Survey Variables .........

Multiple Regressions: The Three Dependent Variables
with Type of Contact and the Significant (p< .10)

Survey Variables ........................... . ...........

Multiple Regressions: The Three Dependent Variables
with the Significant (p<;.lO) Survey Variables (For

Those Receiving an In-Person Contact)... ...............

Multiple Regressions: The Three Dependent Variables
with the Significant (p<L.lO) Survey Variables (For

Those Receiving a Telephone Contact) ...................

vi

PAGE

41

52

55

56

58

59

59

60

6O

61

65

67

68

7O

72

74

75

TABLES PAGE

18. Multiple Regressions: The Three Dependent Variables
with the Significant (p< .10) Survey Variables (For

Those Receiving a Letter Contact) ...................... 77
Ael Demographic Characteristics of the Project Sample ...... 107
E-l Number of Referrals by Type of Contact ................. 116
G—l Internal Consistency of the Five Rational Scales ....... 127

H-l Results of Analyses on 21 Variables by Type of
Contact. ...... ...... ..... . ............. . ........ . ...... 130

I-l Program Eligibility Requirements ....................... 131

I-2 Results of Analyses on 19 Variables by "Assistance
Group" (SSI or State Medical)..... ..... . ............... 134

J—l Analysis of Variance: Knowledge of Social Services
Scale by Type of Contact... ............. ...... ......... 135

J-2 Analysis of Variance: Problems Scale by Type of
Contact ............................ .......... .......... 135

J—3 Analysis of Variance: Life Satisfaction Scale by
Type Of contaCtO.......OOOOOOOOOOO.......OOOOO... ...... 135

K—l Results of Analyses for the Client Card Registration
Variable (Responders vs. Non-Responders) ............... 136

K—2 Results of Analyses for the Variable of Receiving
a Service (Service Recipients vs. Non-Recipients)...... 137

K—3 Results of Analyses for the Newsletter Variable

(Responders vs. Non—Responders) ........................ 138
L-l Estimated Costs of Completing One Outreach Contact ..... 139
L-2 Estimated Costs Based on Effectiveness Data ............ 140
M—l Client Ratings of Outreach Contact.. ................... 141

vii

LIST OF APPENDICES

APPENDIX
A. Demographic Characteristics of the Project Sample ........
B. Example Contact Formats..... ....... ......................
C. Response Recording Devices ...... ..... .......... . .........
D. Memorandum of Agreement .................. . ...............
E. Number of Referrals by Type of Contact...................
F. The Follow-Up Survey Questionnaire .................. .....
G. Internal Consistency of the Five Rational Scales .........
H. Effectiveness of the Randomization Procedure.............
I. Comparison of the Two Assistance Groups.. .............. ..
J. Three Rational Scales as Secondary Outcome Measures ..... .
K. Relationship of the Demographic and Social Service
Related Variables to the Three Dependent Variables.......
L. Outreach Cost Estimates..................................
M. Client Ratings of Outreach Contacts ........... ...... .....

viii

PAGE

107

108

115

116

117

127

128

131

135

136

139

CHAPTER I

INTRODUCTION

The topic area of social services and the elderly is indeed an ex-
tensive one. Recognizing this, and in an effort to provide some broad
conceptual background, this paper will begin with a fairly lengthy
literature review. For those not desiring such information, the latter
part of this chapter provides the specific background for the actual
experimental project conducted. For those who would prefer some general
background, this review will include a discussion of the problems and
needs of the elderly, the services available to meet those needs and,
most particularly, the role that information and referral/outreach can

play in linking needy elderly with services available.

Background: The Elderlygas a Problem Population

 

There can be little doubt that the elderly, as a subpopulation of
American citizenry, face a long list of critical problems. The problems
many of the elderly face are not new. Income, housing, crime, health
care and transportation, to name a few, have long been concerns espec-
ially crucial to the elderly. However, a series of social, economic
and demographic developments have served to heighten the impact of
these problems. First, such social developments as the decline of the
extended family and the relegation of the rural society to a backstage
role have left many of today's elderly lacking in the traditional means
of support and frequently isolated from the mainstream of modern life.

In addition, they have the added handicap of being old in a society that

2
places a premium on youth, attractiveness and vitality. Hence, there
frequently results a very real physical and social isolation. Second,
the economic developments of the past few years have severely compounded
the already difficult problems of income and expenditures. Inflation
has hit the elderly, most of whom are on fixed or relatively fixed in-
comes such as pensions, social security, etc., particularly hard. They
find themselves with little or no increase in income, facing raging
inflation, expecially in such areas of crucial importance to the elderly
as food and medical care. Clearly, many are caught in a desperate
squeeze. Finally, the increase in the number of elderly, both in
absolute and in percentage terms, has led to a proportional increase
in the sheer volume of such problems in America. The 1970 census
showed slightly over 20 million "aged" (i.e. over age 65) Americans,
accounting for almost 10% of the U.S. population. This is in comparison
to only 4% of the population in 1900. Projections are that the percen-
tage of elderly will continue to climb as birth rates decline or remain
level and as increased medical effectiveness, nutrition, etc., prolong
life (Council of State Governments, 1973; Bild and Havighurst, 1976).
As more and more Americans have the opportunity for an extended life
span, the question then becomes what is to be the quality of that life?
The manner in which social services for the elderly are designed and

delivered will have a substantial impact on the answer to that question.

Special Characteristics of the Elderly

 

Any approach to the problem area of social services for the elderly
must necessarily take into account the special biological, psychological
and situational characteristics which frequent that population (Tibbetts,

1960; E. Palmore, et al., 1975; etc.). This is particularly important

both in identifying needs of the elderly and in designing programs to
meet those needs. The following summaries, though far from comprehen—
sive, help to illustrate some of the problems the elderly face in these

areas and some of the strategies that can help to ameliorate them.

Biological Aspects of Aging

 

It is well established that many biological changes occur as one
ages (Shock, 1962; Blumenthal, 1962; etc.). There are numerous theories
attempting to explain the aging process, such as "Exhaustion Theory,"
"theory of accumulation of deleterious materials," etc. (Busse &
Pfeiffer, 1969). However, from the standpoint of social services, it
seems most appropriate to be concerned with the practical limitations
that biological aging puts on the elderly person's functioning and with
the practial needs that such aging creates.

Adriaan Verwoerdt (in Boyd and Cakes, 1973) does an excellent job
of detailing in layman's terms some of the biological changes that occur
in the elderly and some of the practical implications of those changes.
For example, age-related changes in the skin and connective tissues,
such as loss of elasticity and altered pigmentation, result in changes
in external appearance of the elderly. This, of course, can have im-
plications for the self image of the elderly person, resulting in ten-
dencies to avoid social contact, etc. In addition, the elderly person
is particularly prone to chronic skin sores from falling or other
accidents or from prolonged immobility. Accident prevention and main—
tenance of the ability to walk are thus important. Changes in the mus—
cular and skeletal system, such as stiffened joints, weakened or altered
bone structure and loss of muscle strength may result in changes in pos—

ture and reduction of mobility. These developments can be painful as

well as impair the ability to perform many vital functions. Many of
these undesirable changes can be eliminated or minimized through
appropriate exercise (especially walking). Various health and recrea-
tion programs could be of value in these areas, and where mobility is
too severely restricted, such programs as friendly visitors and home
services could assist in various household chores. Changes in the car—
diovascular system include degeneration of tissues, simple "wear and

' and narrowing of the blood vessels (arteriosclerosis) resulting

tear,’
in increased blood pressure. It is important to remember that the aging
heart is less capable of responding to the demands involved with the
stress of heavy work. Thus, home services should again be available.
Also, prOper diet, exercise, and adequate medical supervision are
necessary for optimal preventive practices. If blood pressure problems
exist, drug therapy should be considered. Numerous age-related changes
in the gastrointestinal system also occur. They range from a decrease
in sense of smell and taste and problems with loss of teeth, to reduced
motility of stomach and intestines and a decrease in production of
digestive juices. (It should be kept in mind that many social-
psychological factors, such as loneliness, also interrelate with these
problems and influence diet.) Practical problems which manifest them-
selves in the gastrointestinal system include under or overeating,
various nutritional deficiencies, increased intake of sweets, decreased
fluid intake, constipation and/or bowel incontinence and hemorrhoids.
The major objective in this area would be to assure a proper diet. This
could be accomplished through a variety of means, including providing

appropriate nutritional information, assuring adequate resources for

food (i.e. food stamps, etc.), or through more direct means, such as

"meals on wheels" or "friendly visitor" programs. Where appropriate,
of course, medical care must also be available.

The above material is by no means an exhaustive list of the bio-
logical changes which affect the elderly person. It does, however,
present some indication of the problems to be faced. The most appro-
priate role for social services to take would seem to be to provide to
those who require them, some means to prevent or minimize the aversive
impact of age-related biological changes.

Psycholggical Aspects of Aging

 

In a general sense, the findings in the area of the psychology of
aging appear to be important to those inolved with social services to
the elderly in two major respects. First, it is important to be aware
of what psychological characteristics frequent the elderly population
and what practical needs thus exist; and second, it is crucial to
realize that many of the dysfunctional psychological patterns observed
in the elderly are not, as is often popularly assumed, a necessary char-
acteristic of old age and can, in fact, be modified or eliminated with
appropriate situational intervention.

Some of the more obvious psychological changes occur in the area
of sensation and perception. Vision suffers a decline with age in
several respects, including general visual acuity, adaptation to dark-
ness, and accommodation to different distances (Riley and Foner, 1968).
Age also brings about a loss in hearing, both in auditory threshold
(Corso, 1959) and in ability to discriminate speech (Melrose, et al.,
1963). Kinesthetic perception declines with age (Carp, 1973, in Boyd &
Oakes) as does reaction time (Welford, 1959). Such changes can have

a marked impact on the well-being and social functioning of elderly

individuals. The relationship between sensory losses and resultant
behavior is not direct. Many people devise ways of compensating. How-
ever, such losses often tend to make the individual hesitant and unsure
of him/herself, particularly among other people and may result in
varying degrees of social withdrawal (Carp, 1973) or other dysfunctional
symptoms such as personality changes or deteriorating interpersonal
relationships (Harris, 1975). These sensory changes are examples of
psychological changes which cannot, of course, be eliminated, but proper
diagnosis and intervention can help to minimize their effect on the
lives of the elderly.

The more serious psychiatric disorders that affect the elderly fall
into two major groups: those appearing in the absence of brain damage
(psychogenic) and those appearing in the presence of brain damage
(organic). Although clearly disabling mental disorders appear to .
afflict slightly less than 10% of the 65 and over p0pulation, approxi-
mately one-third of which are under institutional care (Riley and Foner,
1968), these disorders are important to those in social services
because they are the most severe that must be dealt with. The role of
social services in these situations is primarily one of ensuring detec-
tion and appropriate treatment. However, with the increasing evidence
of social and environmental factors playing a role in all mental dis-
orders of the elderly (particularly with respect to social deprivation),
social services may often perform a preventive function as well. This,
in fact, is one rationale behind such services as the "friendly visitor"
programs. In addition, some feel the actual treatment of age—related
brain disease can be facilitated by the treatment of various exogenous

factors (Kobrynski, 1973).

The above developments suggest an expanded role for multifaceted
social services in both the prevention and treatment of mental dis-
orders in the elderly. This seems particularly true for the less
severe and more common psychological ailments afflicting the aged.

There is mounting evidence that depression in the elderly should be
generally understood as a reaction to real losses and stresses in the
biological, psychological, and social aspects of life (Gordon, 1973;
Levin, 1963; Amster and Krauss, 1974). Hence, services that minimize
these stresses can be seen as minimizing the likelihood of depression.
Other age-related psychological factors that are closely tied to the
individual's past and present educational and social experiences in-
clude learning ability, motivation, morale, and anxiety (Carp, 1973).
These areas, too, may be open for positive intervention by those in
social services.

It is important to grasp the implications of this situational
impact on the psychological functioning of the elderly. Many behaviors
and mental states observed in older persons are simply assumed to occur
because of age when, in fact, they may occur for many other reasons,
including social-environmental changes that may be associated with
aging (Amster and Krauss, 1974). Misconceptions about aging, even among
mental health professionals, often lead to a fatalistic attitude toward
the elderly (Kobrynski, 1973), which must be overcome if positive inter—
vention is to occur. Busse and Pfeiffer discuss the question of situa—
tional impact in terms of adaptation, concluding that successful psycho—
logical adaptation by elderly persons does indeed depend to a great ex-
tent on functional or dysfunctional patterns of adaptation deve10ped
earlier in life and on various biological, social and economic factors

in the individual's present environment (Busse & Pfeiffer, 1969).

Situational Aspects of Aging

 

As alluded to in the previous section, there are several areas in
which the circumstances of the aging individual changes rather drama-
tically as s/he enters his/her elder years. A few of these situations
will be referred to here.

The elderly typically experience a substantial drop in income over
their younger years (Butler, 1975). Inadequate income is considered by
many to be the greatest problem facing the elderly (U.S. Office of
Economic Opportunity, 1972). The problem is particularly severe for
those elderly that comprise 20% of the nation's poor (Butler, 1975)
but, in fact, exists for far greater numbers of elderly than those who
are below the strict poverty level. For the nation as a whole, the
median income of the retired elderly couple is roughly one-half that
of those still in the labor force. In more revealing terms, in early
1975, the median income of an American elderly individual was only $52
per week (Butler, 1975). Also, a shortage of income has direct rami-
fications on other situational problems of particular concern, such as
lack of nutrition, lack of health care, inadequate housing, etc., and
should be considered in these terms as well. A number of combinations
of income and "in kind" assistance could conceivably provide for these
needs (for example, the presently existing income programs of social
security and S.S.I., and the in kind programs of food stamps, Medicare
and Medicaid), if appropriately funded and administered.

Retirement is a situational transition faced by almost all elderly
and involves several interrelated problems in addition to the above—
mentioned drop in income. Included are such components as changes in

social role and status, which are closely related to one's occupation

in this country. Tangential to this shift, there often occurs de-
moralization, a decline in social participation and a loss of feeling
of productivity and self-worth (Blau, 1973). Finally, there is the
problem of excess free time that must be put to some constructive use
(Buhler & Kleemeier, 1961). The cumulative impact of these factors
may have an adverse effect on the mental health of the elderly indi-
vidual (Lowenthal, 1968). (Although there is a lack of literature on
the subject, one might assume that "retirement" from the occupation of
"housewife" might include all of the above—mentioned problems.)

Other significant situational changes for many elderly take place
in the related areas of family structure and living arrangements. The
process for many can be briefly described as separation and isolation.
More than 25% of the nation's elderly live alone or with nonrelatives
(Council of State Governments, 1973). Reflected in that figure is the
additional problem of death of many loved ones, including frequently the
spouse. Nationwide, almost 55% of aged women are widows, and approxi-
mately 40% of aged men are widowers (Council of State Governments, 1973).
Separation and isolation can have many negative consequences for the
elderly, including role loss (Blau, 1973) and poor social adjustment
and cognitive functioning (Bennett, 1973). However, the effects of
isolation may be reversible through various resocialization programs,
such as friendly visitors (Bennett, 1973).

Finally, and not to be overlooked, is simply the fact of finding
oneself to be old in a society which quite clearly is designed for and
emphasizes youth. This realization requires a redefinition of one's
identity and social role and obviously provides a situation where

problems of adjustment can arise.

10

In summary, one can see that the typical elderly individual ex-
periences a series of fairly dramatic situational changes, each of which
can have significant implications for the behavior and well-being of the
individual. The need for and demands of adjustment are great. To those
working with the elderly, these situational changes are some of the

problems toward which social service programs must address themselves.

Rationale for Social Services

 

After surveying some of the biological, psychological, and situa-
tional aspects of aging and the resultant problems of adjustment, it
becomes apparent that the elderly as a group face numerous challenges
that, without access to apprOpriate resources, can mean substantial
hardship. Indeed, this has been the findings of numerous investigations
into the subject ("Project FIND," U.S. Office of Economic Opportunity,
1972). Specifically, in 1969 it was estimated that approximately half
of the nation's elderly, or about 10 million individuals, were living
below the poverty level (O.E.0., 1972). By 1972, it was still found
that 25% of the households below the poverty line were headed by persons
65 and over, and another 11 million persons were being kept just over
the "extremely strict low standards of the official measure of poverty"
by their social security benefits (Robert M. Ball, Commissioner of
Social Security, 1972). In terms of health, it has been estimated
that 81% of the over-65 group suffer some chronic illness, and slightly
under half of the elderly are at least somewhat disabled due to a
chronic illness (S. Brody, 1973, from the National Center for Health
Statistics, 1969). Similarly, studies have shown that the elderly for
a variety of reasons often do not receive adequate nutrition (Jordon,

1954; LeBovit, 1965; Swanson, 1964; O.E.O., 1972).

11

Clearly, in terms of both the eXpected processes of aging and the
observed conditions of the aged, there exists substantial need within
the elderly population. The question thus becomes, who is going to
meet those needs?

Most people would agree that the preferred situation would be for
the individual him/herself, or the individual's family, to be able to
care for the elderly person. Some peOple would insist on this situation.
However, it should be obvious from the preceding material that for both
physical and financial reasons the elderly individual is often unable
to completely care for him/herself. Even the best laid plans may fall
prey to physical infirmity, health problems, the effect of inflation on
fixed incomes, or even the startling fact that approximately 40% of
private retirement programs fold and never meet their obligations
(Butler, 1973). As for family care, the most convincing argument
against this source being solely relied upon is that for many elderly
it is simply unavailable. Less than 25% of elderly live with their
children (Shanas, et al., 1968). Even more revealingly, less than 3%
of elderly live with children and grandchildren (U.S. Bureau of Census,
1970). Hence the extended family appears quite inadequate as a wide-
spread source of care. Finally, there is no evidence that having the
elderly live with one's children is the preferred pattern for either
the elderly or their children (Kaplan, 1975).

Although it has been fairly widely publicized that 75% of elderly
live within a 30 minute travel distance of a child (Shanas, et al.,
1968), this should not be interpreted as implying that an equal number

of elderly should be able to rely on their offspring to meet their needs.

12

It is safe to assume that of that 75%, many of the children lack the
financial resources to meet all the economic needs of the elderly per-
son, particularly if one accepts the logical correlation between
mobility and increasing resources in this country (i.e. intergenera-
tional mobility being positively correlated with income, those off—
spring with less financial resources would tend to live in closer
proximity to their parents). Further, even among those who are able
to meet the economic needs of their parents, many may not be able to
fill other important needs. For a variety of reasons, such as lack of
time, lack of interest, or personal incongruence, the children may very
likely be unable to provide all the social and supporting services
required. Thus, it seems unrealisitc to assume (or require, as some
have proposed) that the individual and his/her family should be the sole
source of support for all or even most of the elderly in this country.
Hence, there appears to be a legitimate role for public social services.
At this point it seems to be apprOpriate to respond to two often
raised criticisms of social services for the elderly. First, and per-
haps with the most merit, is the objection to the ”sick" stereotype of
the elderly that is often fostered. Hopefully without stereotyping, it
is important to realize that the elderly, as a group, do indeed have
substantial needs. In the words of Elaine and Stanley Brody (1974):
"The widespread view of old pe0ple as sick, isolated, feeble and senile
is inaccurate. Most persons age successfully and do not fit such stereo-
types. Nevertheless...the elderly easily qualify as a vulnerable group...
(they are) poorer economically, have more health problems both physically
and mentally, experience more interpersonal losses, and lose status and

social roles more than any other population group."

13

One way to conceptualize this situation is that numerous age—
related environmental demands, magnified even further by declining
physical and psychological resources, can place a tremendous burden of
stresses on the elderly individual. However, various mediating factors
can enter the relationship to modify the effect of these environmental
stresses on the actual well-being of the aged person (Doherwend &
Doherwend, 1974). The resources possessed by the individual and his/her
family, as referred to previously, can function as mediators in this
relationship and are preferred if adequate. When needed, however, social
services are another set of mediating factors that can reduce the aver-
sive impact of situational stresses, particularly when the elderly
individual is lacking in other appropriate resources.

A second objection to the use of social services, is the hypoth—
esized negative impact on the independence of the recipient; perhaps
this is a remnant of the philosophy of rugged individualism. In order
to answer this criticism, it is necessary to consider the capacity for
independence in the elderly person. Aging, almost by definition, is a
process where adaptive resources (physical, mental, perceptual, etc.)
are being lost (Goldfarb, 1959). Hence aging typically becomes, to one
degree or another, a time of increasing dependence on others (Turner,
1965). This is not to imply that all elderly are unable to meet their
own needs. Certainly some fortunate elderly can maintain true "indepen—
dence." However, most elderly fall somewhere on the continuum of need
between the absolutes of no need and total dependence (Berg, Atlas &
Zeiger, 1974). Social services must recognize this variability and
provide flexible services that can be tailored to the true needs of the

individual.

14

Rather than view this situation as a chronically passive dependence,
however, it may be much more useful to consider a "search for aid" as an
appropriate adaptive response on the part of an elderly individual whose
own capacities are declining (Goldfarb, 1964). If care is taken to avoid
the fostering of passive dependence, a variety of social services can
enter this framework to increase the repertoire of resources to which
the elderly individual can turn when needs arise. Services of income
maintenance and health care are probably the most commonly accepted for
this purpose. However, even such potentially independence-threatening
services as homemaking programs (Berg, Atlas & Zeiger, 1974); mental
health services (Burr, 1971) can have positive effects on the overall
self—sufficiency of the elderly.

Assumptions Underlying Government's Role in Social Services

 

Before detailing the major legislation. and structures serving the
elderly, it seems appropriate to briefly mention some of the assumptions
underlying the involvement of the U.S. government in the provision of
services for the elderly.

Consistent with the concerns discussed in the previous section,
official government policy traditionally reflected the expectation that
the individual and family were responsible for support, with private
charities to be relied upon as a last resort. With the occurrence of
the New Deal era and the passage of the Social Security Act of 1935,
the federal government at last acknowledged that the traditional means
of support were ofttimes inadequate. Although this legislation placed
a heavy emphasis on equity (i.e. one pays in to an insurance type program

so one can draw out later), it also allowed for the concept of social

15

adequacy (whereby payments can be made on the basis of need). In the
late 19603, medical care was added to the role of government, once
again in the dual form of insurance (Medicare) and assistance (Medi—
caid). This division of philosophies often creates problems for
recipients both in confusion about services available and in meeting
conflicting eligibility guidelines (Ozawa, 1973). The main function
such a division seems to serve is to at least partially preserve,
through its equity emphasis, the concept of individual responsibility.
A second discernable policy orientation is in the choice of pro—
gram content. Income and health programs have been the dominant
concern, often to the exclusion or neglect of related social services
(S. Brody, 1973). Further, even the health care programs are in
reality income rather than service programs, where the government
provides for payment of certain costs incurred rather than actually
ensuring the provision of any services. This approach has hindered
the access to health care of many individuals, particularly in rural
or other areas where transportation or health services themselves may
be lacking (Harris, 1975). This avoidance of direct service provision

"proper" role of government has been

gives some indication of what the
conceived to be. It should be noted, however, that some movements
toward the provision of a wider variety of services (such as trans-
portation, home care, etc.) have begun with the passage of the Older
Americans Act of 1965 and its amendments (although funded to a much
lesser degree than the income and health payment programs, and still
provided and administered at the local level).

A final policy orientation deserving of note is the national

emphasis on alternatives to institutionalization, particularly visible

16

in the last decade or so (Kaplan, 1973). As stated in the 1973 amend-
ments to the Older Americans Act, one of the goals of the legislation
is to provide for "maximum independence and dignity in home environment
for older persons capable of self—care with appropriate supportive
services" (Section 301). Hence, it appears that the valued concept of
"independence" has been redefined somewhat, as Goldfarb and others might
applaud, to include the appropriate use of some social services by the
elderly person. Such a slight philosophical shift might help explain
how government policymakers could in good conscience expand the role of
the federal government (at least as a funding source) to include the
provision of a much wider variety of services than income and health
care payments (as the Older Americans Act and amendments have done).
The question of motivation for these policies, particularly the
increasing involvement of government over time, is open to varying
interpretation. The most favorable interpretation would be simply that
an increasing elderly population made the problems of the elderly more
visible and thus brought about government intervention in response to
need. Many other more Machiavellian interpretations are available,
including that government increases social services to quiet unrest
(Piven & Cloward, 1971) and, more specifically, that many policies
reflect attempts to woo the currently estimated 20 million elderly
voters (Butler, 1974). Louis Lowy (1974) was probably closest to the
truth when he observed that basic expediency, rather than any real
desire to integrate the aged into the mainstream of American life, has
characterized the responses of government to the problems of aging.

Having considered some of the assumptions underlying the government's

l7

role in social services, one can now take a fairly descriptive look at
legislation and agencies serving the elderly.

Legislation,gStructures and Agencies

 

The two primary pieces of federal legislation providing for the
elderly are the Social Security Act of 1935 and the Older Americans Act
of 1965. The Social Security Act of 1935 brought the federal govern-
ment into the position of providing for the elderly for the first time
on a large scale by creating the programs of "Old Age Assistance" or
direct financial assistance to needy elderly, and "Old Age Survivors
Insurance" or the program of payroll and paycheck tax financed social
security payments to eligible elderly or survivors (OASDI). These pro-
grams were intended to provide some base minimum income for elderly
persons but were never intended to provide the sole supporting income.
In fact, however, for many needy elderly, social security is the sole
source of income. For that purpose, however, it is marginal at best and
usually inadequate.

The Social Security Administration has recognized the fact that
social security benefits are often "just too low to meet needs" (Ball,
1972) and has attempted to remedy that situation somewhat through the
formation of the Supplemental Security Income program (SSI) in 1973
(see below). Meanwhile, the social security program has also come
under fire from the political right for spending too much money, being
a non-voluntary program, etc. Despite criticisms and funding problems,
however, social security is a widely used program and indispensable to
millions of Americans. For example, a recent survey conducted in the
state of Michigan revealed that 84% of the elderly contacted were

receiving some kind of social security payments (Office of Services to

18

the Aging, 1975). Further, the Social Security Administration estimated
that by 1973, 93% of all people becoming 65 in America would be eligible
for a social security benefit (Ball, 1972). In terms of sheer numbers,
the Social Security Act is the price of federal legislation which has
had the most profound impact on elderly Americans.

In addition to the income programs of OASDI and SSI, two amendments
to the Social Security Act have produced medical programs affecting the
elderly. In 1965, Title 18 created the Medicare program which provides
for payment for medical services in the categories of outpatient and
diagnostic tests, hospitalization, and post-hospital care. Unfortunately,
the list of things Medicare does not provide for is almost as impressive,
including drugs, dentistry, podiatry, medical appliances, mental health
care, and home health services (Harris, 1975). Title 19 created the
Medicaid program, which, in contrast to Medicare, is an "assistance"
program to the needy. Medicaid can at times fill some of the gaps left
by Medicare coverage; however, it is a state administered program and
coverage varies widely from state to state (HEW, 1973). Finally, it
should be reiterated that neither of these programs actually provides
medical services, but rather provides a means for payment for certain
medical services.

In 1973, amendments combined the programs of Old Age Assistance,

Aid to the Blind, and Aid to the Disabled into one program entitled
Supplemental Security Income (SSI). The major reason for this shift,

as referred to above, was to "take a giant step toward adequacy" in the
provision of income for the elderly (Ball, 1972). This was to be
accomplished through such means as slightly loosened eligibility standards

and slightly higher payments. It was also intended, through procedural

19

changes (such as moving the whole program from the Departments of Social
Services in the states to the Social Security Administration (SSA) it-
self), to increase convenience and decrease the "welfare" stigma (Ball,
1972). However, many now contend that this procedural change has also
had some adverse impact on receipt of services by the elderly, including
maintaining the conflicting ideology and eligibility procedures involved
with the simultaneous administration of an insurance type program (social
security) and an assistance type program (SSI) (Ozawa, 1974). There is
some evidence that the new SSI policies have lead to increased hardships
and frustrations for the elderly client (Moses & Zander, 1975). Also,
moving Old Age Assistance from the Department of Social Services (DSS)

to the income oriented SSA has often served to isolate many needy elderly
from contact with the more service oriented DSS (Region IV Area Agency
on Aging, 1975). Hence, it remains for further evaluation to determine
whether the SSI legislation, with its commendable attempt to move toward
social adequacy, is in fact proving to be the best alternative for

income maintenance for the elderly.

The "Older Americans Act" of 1965, in response to the increasingly
visible problems of the elderly, created at the federal level the
Administration on Aging (AOA) within the U.S. Department of Health,
Education and Welfare (HEW). The AOA was given the primary role for
administering the Older Americans Act and, thus, for planning, funding,
and advocacy for the nation's elderly. Reflecting the emerging policy
of promoting independence by avoiding institutionalization, the legis—
lation provided means for funding a much wider variety of services than
the basic income and health programs provided by SSA. The Act itself

listed under "Title I" ten broad "national objectives" for older

20

Americans (including the areas of income, health, housing, etc.) and
emphasized their adoption by all levels of government, thus giving impetus
for organization at the state level as well. The five other Titles under
this Act, briefly described, provided for the following: Title II
created the Federal Administration on Aging (AOA); Title III provided for
grants for community planning, services, and training (and has been, in
practice, one of the most consistent sources of funding from programs for
the elderly), and also furthered the intent discussed in Title I by
requiring states to designate a single state agency to administer the
Older Americans Act programs and assume responsibility for state planning;
Title IV authorized research and deveIOpment grants; Title V authorized
training grants; and Title VI established an advisory committee to the
HEW, consisting of a chairman and 15 members appointed by the secretary
of HEW.

In 1973, amendments added increased strength to the Older Americans
Act. First, the AOA was elevated in status with HEW by transferring it
out from under the Social and Rehabilitation Service directly to the
office of the secretary. Second, the "Federal Council on Aging" was
created, consisting of 15 members appointed by the President and confirmed
by the Senate. This body was given the responsibility to: assist and
advise the President on the needs of elderly Americans, make recommen—
dations to federal agencies and officials, and increase public awareness
of problems of the aged (presumably through publications, etc.). Third,
a new Title was added to the Act. This was "Title VII," which provided
funding for a federal nutrition program for the elderly (and has resulted
in such programs as "meals on wheels" as well as centrally provided

nutrition programs). Fourth, the amendments created the concept of

21

"Area Agencies on Aging," by requiring the states to develop substate
planning units, thus emphasizing the need for local planning and coor—
dination for services to the elderly.

The Older Americans Act and amendments have been fairly successful
in fostering a wide variety of service programs for the elderly. A

publication by the AOA entitled Let's End Isolation (1971) describes

 

numerous programs, both federal and non-federal, existing nationwide
(such as nutrition programs, friendly visiting, transportation programs,
etc.). In addition, the legislation has succeeded in causing states to
create state and local agencies concerned with aging, which have been
particularly useful in coordinating services, providing information and
referral, and gathering data about the elderly population. (Information
concerning federal legislation and agencies was obtained primarily from
a publication of the "Council of State Governments" entitled 93 Growing
'Qld (1973) and from copies of the legislation itself obtained from

United States Statutes_a£ Large (1939 and 1965).

 

At the state level, the Michigan Comprehensive Plan on Aging (1975)
gives a history of structures and agencies serving the aged in Michigan
and details the present structure of the state Office of Services to
the Aging and the 13 associated Area Agencies on Aging. Michigan has
had some form of agency involved with the aged since 1955. From 1955
to 1960, it was called the "Legislative Advisory Council on Problems on
Aging." From 1960 to 1973, it was called the "Commission on Aging" and
was organized under the governor's office. Thus, when the Older
Americans Act of 1965 was passed emphasizing the creation of state agen-
cies on aging under each state's executive branch, Michigan was already

in substantial compliance. Then, when the 1973 amendments to the Older

22

Americans Act broadened state functions by requiring the development of
substate planning agencies, Michigan passed Public Act 106, creating a
state "Office of Services to the Aging" (OSA). The OSA then designated
13 area regions as distinct planning and service areas for services to
the elderly. In 1974, the governor further strengthened the OSA in two
significant ways. First, he transferred responsibility for adminis-
tering the Older Americans Act from the DSS to the OSA. Second, he

' consisting of repre—

created a state "Interagency Council on Aging,’
sentatives of all principle state agencies serving the elderly. The
membership of this group provides a quick rundown of most of the other
state agencies involved in any significant respect in services to the
aging and includes the Departments of Social Services, Public Health,
Mental Health, Education, etc. As a final note, the OSA was created as
a "terminocracy" with an expected life span of two years. However,
Public Act 146, passed in July 1975, has extended the life of the OSA
for an additional two years; presumably the OSA will one day be given
permanent status.

The functions of the OSA are essentially the statewide planning,

coordination and evaluation of services to older citizens. As speci—

fically stated in the State Plan, they include:

 

--to prepare, in cooperation with other state agencies, a com-
prehensive plan for services to older citizens;

——to approve services for older citizens funded by state funds
or state administered funds;

--to coordinate education and public information programs con-
cerning the elderly;

—-to evaluate the effectiveness and efficiency of current state
statutes on the life-styles of the elderly;

--to coordinate the development of performance standards for the
licensing of residential and medical facilities for the elderly;

——to supervise the establishment of demonstration programs;

—-to make recommendations to the governor and the legislature on
budget and grant requests for services to older citizens (p. 4).

23

At the local level, there are 13 distinct Area Agencies on Aging
(AAA), with the expressed mandate of providing for the planning and coor-
dinating of services and the provision of information and referral ser—
vices at the local (area) level. Aside from information and referral,
these Area Agencies are not direct service providers. (They can, however,
under certain circumstances, purchase direct services by contract or

grant.) The AAA duties, as outlined in the State Plan, are:

 

1. Develop an annual plan detailing priorities and objectives
for services within the region;
2. Coordinate existing services to ensure their accessibility
by older citizens;
3. Identify and pool untapped resources in the region;
4. Provide access to information and referral to assist elderly
in obtaining knowledge of and services from service providers
in the area;
5. Establish linkages with other service providers in the region.
Structurally, the Region IV Area Agency on Aging is centered around
the Executive Director, who supervises the general network of the Agency.
He receives technical assistance from the state OSA and is advised by an
Advisory Council, composed of local residents (83% of whom currently are
over age 60, thus providing a source of input for local elderly). Under
the director, the Area Agency's major components are the Information and
Referral, Services and Housing, and Senior Aides Programs, with a
separate coordinator heading each of those areas. The Information and
Referral (I & R) component is of major importance to this report and
includes a Senior Information and Referral System (SIRS), consisting of
eight local I and R centers distributed throughout the three county area.
These local centers perform outreach, information and referral, and data
gathering functions.

Other service providers operating at the local level include the

county DSS offices, the local SSA office, any other local government

24

agencies that may exist, volunteer groups, church groups, and various
non-profit organizations that may be involved in providing services.

In summary, the approach that seems to be emerging emphasizes the
assessing of needs, planning, coordinating, and setting of objectives at
the local level, all within a guiding framework of federal legislation.
This would seem to allow for appropriate flexibility within a national

policy of concern for the problems of the elderly.

Demographics and Needs of the Elderly

 

At the federal level, information about the demographics of the
elderly is primarily available through various summaries of U.S. census
material. This can be obtained through U.S. government publications or
in most general texts on the elderly (e.g. Tibbitts, 1960; Riley & Foner,
1968; Woodruff & Birren, 1975; etc.). In terms of needs of the elderly,

The Golden Years: .A Tarnished Myth (U.S. Office of Economic Opportunity,

 

 

1972) is an excellent source. This publication represents the findings
of "Project FIND," a study of 50,000 elderly individuals in a total of
twelve communities nationwide. It includes excellent demographic and
needs data, particularly in terms of poverty and the elderly. (Specific
data at the national level will not be included in this paper for reasons
of brevity. However, the situation at the national level can be inferred
fairly accurately by considering the state and local situation.)

At the state level, two excellent sources of information were ob-
tained to provide data on Michigan's elderly. These sources were Th3

Michigan Comprehensive Plan 92 Aging, developed by the Offices of Ser-

 

vices to the Aging (OSA), and Michigan Aging Citizens, a comprehensive
survey of Michigan elderly, deve10ped by that same office. The data ob—

tained from these sources was in most cases accurate through January 1975.

25

Paralleling the national situation, the elderly population is the
fastest growing segment of the p0pulation within Michigan. By 1970, there
were 1.1 million citizens over the age of 60, accounting for 12% of the
state's total population. The OSA estimates that the elderly segment
will continue to grow, accounting for at least 15% of the total popula-
tion by 1990. Hence, problems connected with the elderly will likely
increase, at least in proportion to the population increase.

Of this elderly population, data on several demographic variables
are available. For example, the proportion of elderly who are women
increases dramatically with age. Of the under 65 population, 51% are
male and 49% are female; for the 65-74 age bracket, 55% are female; for
75—84, 59% are female; and for 85 and over, 63% are female. In terms
of geographic location, 61% of the "over-60" population live in urban
areas of 50,000 or more, as compared to 64% of the total population.
However, within those cities, the central city vs. suburbs distribution
for the elderly is 56% vs. 44%, respectively, while for the total popu-
lation the percentages are exactly reversed. Thus, one arrives at a
situation where a disproportionate number of elderly today live in
either a poor urban or poor rural setting. Racial data are available
in only two categories, with the over-60 population being 92% white and
8% black as compared to 88% and 12% in the total population.

More specifically addressing the concept of "needs" was a survey
conducted by the OSA in the spring of 1974 of 3000 non-institutionalized
aged. This survey examined the major categories of: neighborhood,
housing, transportation, nutrition, health, social support, employment/
retirement, earnings and expenditures, government services, consumer

protection, general problems and demographic characteristics. For the

26

entire sample, the following major problem areas were identified. They
are presented in rank order, with the percentages representing percen—
tage of those surveyed who considered that problem to be a "problem for
older Americans." Thus, the question was presented in "surrogate" form.
(This surrogate technique is intended to be less threatening than a
direct question and is intended to help avoid the problems of denial or
concealment of need as well as intentional overstatement of need. How—
ever, it also leaves one without any direct indication of personal need
by the individual. Although the use of this technique may have distorted
the relative percentages somewhat, the major problem listed in the high—
est five ranks parallel closely those identified in other studies such
as in Illinois (Shore & Schilsky, 1975) and in twelve major communities

nationwide (National Council on Aging, 1972).

Income 38%
Crime 31%
Health Care 24%
Transportation 21%
Nutrition and Food 15%
Consumer Protection 12%
Spare Time Activities 11-12%
Housing 11%
Employment Opportunities 8%
Age Discrimination 8%
Getting More Education 5%

The top five spots contained the same five problem areas with minor
variations in rank, across all demographic subgroups. Income was con-
sistently ranked as the major concern with the exception of two subgroups:
those living in Detroit, and those with an income of over $6,000, each of
whom ranked crime first and income second.

Two other topics merit mention at this time. First, the "living
arrangement" of the respondent had a definite interaction with problems

perceived. As might be expected, those living alone reported the highest

27

frequency of problems facing older Americans, while those living with
spouses reported the lowest. Second, as age increased, apparent dis-
content tended to decrease with one major exception, that being trans-
portation. Transportation problems increased with age, becoming
particularly apparent after age 79.
This paper has outlined some of the demographic characteristics

and perceived needs of the Michigan elderly population. For more
detailed information, particularly in regard to the "SON" survey,

please refer to the Michigan Comprehensive Plan 93_Aging (December,

 

1974).

At the local level, the Region IV Area Agency on Aging éEEEHElflfl
1975-76 was utilized. (Region IV encompasses the tri—county area of
Berrien, Van Buren, and Cass counties in Michigan.) Here the elderly
population and problems of the elderly generally reflect the data
presented in the Michigan Comprehensive Plan. Two of the counties
(Cass and Van Buren) are predominantly rural, while Berrien County
includes the urban centers of Benton Harbor and St. Joseph. Within
these areas, the elderly pOpulation includes a wide spectrum of demo-
graphic variables including income, age, living arrangement, race,
community size, etc.

The total population of this three county area, as of the 1970
census, was approximately 263,000. Of these, approximately 36,000 or
14% were over the age of 60. To further break down those figures: of
the elderly, approximately 29% were aged 60 to 64; 42%, aged 65 to 74;
and 29%, aged 75 and over. In racial composition, just under 10% of
the elderly were black and 89% white, with the remainder composed of

American Indian, Spanish American and Oriental, with not more than 1% of

28

each. Perhaps the most revealing statistic of all, however, is that, of
the 36,000 plus elderly in this region, over 10,000 or approximately 28%
had incomes classified as below the poverty line. Further, as can
logically be expected, problems of poverty often signal problems in
many related areas. The data compiled in Region IV seems to bear that
out, as a similar percentage to those found wanting in income also had
"identified need" for health, housing, transportation, and "social and
reassurance" services and about half that number had "identified need"
for nutrition, employment and legal services. (For actual percentage

breakdowns, refer to the Rggion Four Area Plan, 1975-76.)

 

The Problem of Linkage

 

Clearly, there exists a substantial population of identified needy
elderly in this three county region. In addition, other studies suggest
that it can safely be assumed that an equal or perhaps greater number of
elderly live in a near-poverty or a low-income situation (National
Council on Aging, 1975). Unfortunately, it is also true that for one
reason or another, many of the most needy elderly are simply not getting
the services they require, even when those services are currently
available.

This problem is most strikingly evident from the results of the
nationwide "Project FIND" study. Of the 50,000 elderly individuals sur-
veyed, 28,079 cases of need were discovered where referrals could be
made to existing services (as well as 24,124 cases of need for unavail-
able services). Thus, one can observe that there is a problem not only
in providing services for the elderly but also in linking needy elderly

with services provided.

29

A second linkage problem has occurred as a result of the shift of
Old Age Assistance from the Department of Social Services to the Social
Security Administration (Region IV Area Agency on Aging, 1975). As
referred to previously, this move from the more service-oriented DSS to
the income-oriented SSA (which is basically not involved in compre-
hensive service provision--Ball, 1972) has tended to isolate many
elderly from other supporting services.

A third linkage problem, also involving SSI, is the apparent dif-
ficulty in linking needy elderly with the SSI program itself. At the
inception of the SSI program in 1973, federal officials of the Social
Security Administration insisted that over 200,000 elderly would be
eligible for SSI in Michigan. Yet three years later, less than
120,000 elderly are enrolled in the program (OSA, 1975).

At the local level, each of the above-mentioned problems are con-
cerns of the Area Agency on Aging (Region IV Area Agency on Aging,
1975). For example, most of the 10,000 plus elderly in that region
with incomes below the poverty level are likely eligible for Old Age
Assistance (under the new SSI payments). Yet a recently compiled list
of SSI recipients for that region shows only slightly more than 2300
people are actually receiving SSI payments. Part of this discrepancy
may be due to difficulties in implementation of the relatively new SSI
system, but most of the lack of participation is presently unaccounted
for (OSA, 1975).

In addition, as discussed previously, there exists substantial need
for services among elderly of the tri-county region. As the Region IV
Area Plan (1975) indicates, the location and linkage of those needy

elderly with available services is one of the two essential functions of

30

the Area Agency on Aging (the other being the gathering of information
about the needs of elderly in the area, which this study will also
attempt to perform).

The previous examples illustrate some of the problems requiring the
linkage of needy elderly with available services. In so doing, they help
emphasize the need for some means to facilitate that linkage.

Information and Referral

 

To help solve the problem of linkage, the concept of an "information
and referral" service has been utilized. However, as Brumfield, Fox and
Goldman (1968) point out, the problem with information and referral is
that the client must initiate the request for services. Hence, most of
the factors that prevent the client from contacting the service providers
directly also prevent him/her from contacting the information and
referral (I & R) service. These factors are hypothesized to include:
client unaware of service; general fear of or avoidance of unnecessary
outside contact; denial of any problem (Gaitz, 1974); client apathetic
toward service; general misinformation of many kinds (such as "I'm not

' etc.); and the desire to avoid "welfare" (Alexander & Podair,

eligible,’
1969).

To help overcome these obstacles, the concept of "outreach" has been
added to information and referral. When the information service reaches
out and contacts the client directly, it is hoped that most or all of the
above-mentioned obstacles can be remedied. This is the approach that the
Area Agencies on Aging have developed in their "Information and Referral/
Outreach" Service. However, they still face the problem of a limited

amount of resources and a large eligible client population. The challenge

thus becomes one of getting the maximum amount and quality of information

31

and referral from the existing I & R/Outreach and related structures.

Outreach Strategies

 

Very little literature exists specifically addressing the problem
of how best to provide information and referral/outreach services to the
elderly. This may reflect the fact that area agencies on aging are a
very recent phenomenon, appearing in force only after the 1973 amendments
to the Older Americans Act. However, the studies that do exist, combined
with findings from related subject areas, can at least help delineate
some important variables to consider. For example, from the area of
business and marketing, an article by Klippel and Sweeny (1974) dis—
cusses their findings concerning the use of information sources by the
elderly. Two conclusions appear particularly relevant. First, it
appears that elderly consumers rely heavily on informal information
sources in decision-making. This would seem to indicate the usefulness
of utilizing peers or other forms of information contact to facilitate
outreach. Second, their findings showed that the technique of product
sampling was an important one to use in influencing the elderly consumer.
If the analogy can be drawn to the "consumption" of I & R/Outreach ser-
vices, it might be wise, where feasible, to ensure that needy elderly
experience some favorable contact with the I & R center and/or personnel.

In terms of the design of any persuasive messages used, two findings
from the area of social psychology seem particularly applicable. Infor-
mation provided in a persuasive communication should be very specific
(Katz & Lazarsfeld, 1955), and the act of comparing oneself to others can
often be an impetus toward change (Lippitt, 1958). Each of these tech-

niques might well be considered in designing a persuasive outreach contact.

32

Modes of Outreach Contact

 

One mode of communication that naturally comes to mind is that of the
mass media. However, for a variety of reasons the use of the mass media
for outreach may not be particularly desirable in this situation.

Unlike mail, telephone, and in-person contacts, the mass media are
not personal media. They do not allow the selection of a specialized
audience nor do they allow the delivery of a personalized message
(Havelock, 1971). This would seem to hamper the effectiveness of the
mass media in presenting meaningful persuasive messages to a target group
of needy elderly.

The question of mass media use by the elderly is examined in a paper
by Rush and Kent (1974). Their findings indicate that, although mass
media usage by the elderly as a whole resembles the population at large,
a certain subgroup of the elderly have more highly developed communica-
tion use skills. As might be expected, those most able to utilize mass
media communications are typically not the ones I & R/Outreach services
would be directed toward. Their findings reveal they are "more highly
educated, have higher income levels, are frequent meeting attenders, get
out frequently to talk to friends, and in this study are more often
white than black."

Bergner and Yerby (1968) arrived at similar conclusions with respect
to low income groups, indicating that such groups are often unable to
utilize the mass media effectively to obtain health service information.

Hence, it would appear that mass media use may not be a particularly
effective strategy for reaching and influencing the target group of needy
elderly. (In addition, in this particular region, mass media campaigns

concerning I & R services have already been conducted during the past

33

few years. There is little reason to believe that those elderly who have
not so far responded to these campaigns would be any more likely to do so
in a subsequent effort. For these reasons, as well as a lack of proper
controls, no mass media component was directly manipulated in this experi-
ment. However, by inference, the results of the "control" condition may
reveal something about the relative efficacy of the mass media in this
situation.)

In more practical terms, three of the most basic modes of outreach
available to the typical I & R center are: direct mail, telephone and
in—person contact.

.Mgil. Direct mail can be an effective, persuasive medium. In ad—
dition to its low cost, it has a fairly high contact rate and can allow
personalized content (Barton, 1964). Also, in contrast to the mass
media, one can be highly selective about who receives one's message.
Thus, it is not surprising that results have shown that a direct mail
campaign is most efficient when it contains a specific message and is
aimed at a specialized target population (Barton, 1964; Havelock, 1971).

For example, in the political arena Eldersveld and Dodge (1954)
used direct mail contact as one persuasive technique and found it to be
more effective than a "mass media only" technique both in terms of in-
fluencing people to vote and in influencing "undecided" Voters to vote
for a particular proposal.

There is some reason to believe that a direct mail campaign may
increase its effectiveness with the elderly if a more "personal" approach
is taken. The Klippel and Sweeny article (previously cited) seems to,
support this contention through its recommendation of an informal com-

munication strategy. Somewhat similar to Klippel and Sweeny's findings

34

are the results of several studies in the area of agricultural innova-
tion acceptance, which indicate that the individual frequently uses
other people similar to him/herself as referents when considering the
adoption of innovations (Lionberger, 1960). Thus, there may be some
usefulness in providing (in case such are not already available) in-
dividual referents or a referent group with positive attitudes toward
the use of I & R services. This could be accomplished in a letter
situation by including a personalized message from several senior
citizens who do use the Center and its services.

The use of such a technique is also supported by Bergner and Yerby
(1968) who suggested involving those who are already using a service in
persuasive attempts aimed at non-users.

One problem that must be taken into consideration in using a mail
contact with an elderly population is the effect that poor vision may
have on the reading of mail. Goodrow (1975) cited poor vision as one
of the three most prevalent factors (along with home responsibilities
and lack of transportation) limiting participation of the elderly in
educational programs. He recommended such policy decisions as minimal
dependence on printed material, flexible scheduling of contacts, and
some kind of transportation arrangement.

However, in spite of possible visual decline, numerous studies have
shown that the elderly population as a whole does maintain an active
interest in reading, particularly for local news and serious, as opposed
to entertaining, content (Riley & Foner, 1968). Hence, it appears that
a written persuasive message cannot be ruled out as an effective com-
municator for reaching the elderly, particularly if such conventions as

large print and simple wording are followed.

35

Telephone. The telephone is a major vehicle of personal communica-
tion for all age groups, including the elderly and should be utilized to
help facilitate the acquisition and retention of information (Rue, 1973).
Similar to direct mail, it has a low cost and can allow personalized con—
tent. In addition, due to the "two—way" nature of the medium, one can
be certain that contact has been established and can receive immediate
feedback (Havelock, 1971).

Although very little has been written about the use of the telephone
in outreach work, there is some indication from the field of advertising
that it can be useful as a persuasive communication medium, particularly
in a supporting role to other media efforts (Roens, 1961; Barton, 1964).
Devoe (1954) devotes an entire book to the use of the telephone in sales
and promotion and states that effective telephone use can cut time and
costs up to 25—30% over face-to-face contacts.

In an area important to the elderly, that of public health,
Brumfield, Fox and Goldman (1968) recommend the use of the telephone in
outreach efforts, especially when a shortage of resources makes home
visits difficult. Despite a lack of evaluation of the effectiveness of
this medium, it seems that it also cannot be ruled out as a potentially
effective means of reaching the elderly population.

In—Person Contact. Like mail and telephone contacts, an in-person

 

contact is a personal medium and can allow specific selection of clients
and a personalized message. In addition, this mode of outreach has the
advantage of offering actual face-to—face contact, thus enabling a
maximum amount of feedback and interaction.

Eldersveld and Dodge (1954), in their study of political persuasion,

found that a personal persuasive visit was more effective than sending

36

literature through the mail or than a mass media only strategy.

In the area of health services, Bergner and Yerby (1968) advocate
outreach in the form of personal visits for maximum effectiveness in
reaching low income groups.

Goodrow (1975), in his study of factors limiting participation of
elderly in educational programs, recommended interpersonal contact as
the best means of communicating with the elderly.

The implications of the above comments with respect to in—person
communication are strongly supported by the results of "Project FIND."
Those involved in that project concluded that "a vigorous outreach pro-

" was the most effective in reaching the

gram involving home visiting...
isolated, needy elderly (National Council on Aging, 1972).

The Project

 

The project described in this paper was designed to address the
problem of linkage by actually experimentally evaluating different
methods of conducting outreach operations. Alternative modes of mail,
telephone and in-person contact were conducted throughout a three county
area to examine them for effectiveness in reaching potential clients.

The above studies would seem to suggest that an in-person mode of
contact would be the most effective in reaching and producing a positive
response among the elderly. The present study examined that premise
through the major manipulation of the three communication modes.

The readings would also seem to suggest that the content of a per-
suasive communication should be specific and serious (as opposed to
entertaining) in nature, and that the format or delivery should be per-
sonal and informal, and perhaps include a chance for the person to com-

pare him/herself with others. These premises of specific content were

37

followed with respect to the informational content of the outreach con-
tacts made, and the use of dual letter conditions allowed the testing of
a "personal" vs. an "impersonal" communication in the same medium. Also,
the chance for "comparison with others" (belonging to a service-using
"referent group" created for the individual) was provided to some extent
by the use of the "names" strategy in the personal letter condition.
(See Chapter II.)

Finally, the mentioned concern of transportation appears to be an
important one. To avoid non—response due to this inhibiting factor,
each communication mode emphasized that transportation was available for
meeting with I & R center personnel and also that suitable alternatives
could be arranged (such as an outreach worker visiting the home).
Rationale

Although the above studies suggest the superior effectiveness of
the in—person outreach mode, the importance of an actual experimental
evaluation should not be underestimated. Despite the apparent drawbacks
of the other two modes of contact, they each would in most cases be
substantially less expensive to conduct than personal contacts. Hence,
an experimental evaluation of the relative effectiveness of these tech—
niques could be combined with appropriate cost data to reveal the most
efficient mode of outreach given various budgetary Constraints. Further,
by considering specific client data, alternative modes of outreach for
different client types might be discovered.

Finally, this study also attempted to gather information about the
central question of why many needy elderly apparently don't request, and
subsequently aren't receiving, available supporting services. This problem

will be approached by examining the data from a follow-up survey

38

questionnaire soliciting opinions, attitudes toward social services,

comments about peer networks, etc.

CHAPTER II

METHOD
Subjects

A population of approximately 2000 needy elderly (aged 65 and over)
in the Berrien, Cass, and Van Buren County area was identified by means
of two lists provided by the Michigan Department of Social Services.

(This number represents 8% of the total elderly population and approximate—
ly 20% of the estimated 10,000 elderly living at or below the poverty
level in this region--Area Agency on Aging, 1975.) The first list con-
tained the names of all those elderly receiving income under the "aged"
category of the Federal Supplementary Security Income (SSI) program. The
second list, mutually exclusive, identified those elderly receiving assis-
tance under the state medical program called the "Old Age Related Medical"
(a state assistance program). With the exception of income (low income

is an eligibility requirement of these programs--see Appendix I), these
lists provided a wide demographic variety of elderly, representative of
the diverse pOpulation of this three county area. The only additional
restriction that was placed on the selection of subjects for this study
was that they be non-institutionalized at the time of contact.

The lists obtained for the three county area were broken down into
regions roughly conforming to the seven service areas covered by the seven
Senior Information and Referral Centers of the Region IV Area Agency on
Aging. These resulting sublists were then given to the center directors
for them to remove from the lists the names of all those elderly with
whom their centers had had previous contact. (This was made possible

by the comprehensive record-keeping system that was maintained by all of

39

40

the centers in the Region IV area.) The remaining "revised lists" of
previously uncontacted elderly thus became the final target population
(N=approximately 1000).

After receiving the revised lists of needy elderly who had had no
recorded contact with the Senior Information and Referral System, the
experimenter then randomly selected 25 names from each list (SSI and
state medical) for each corresponding I & R Center. From this pool of
50 names for each center, 10 names (5 SSI and 5 state medical) were ran-
domly assigned to each of the four experimental groups and to the con—
trol group. Hence, this study incorporated random selection from the
target pOpulation and random assignment to treatment conditions.

Finally, the results of a follow—up survey revealed that the final
sample selected had a mean age of 75 years, was 75% female, 78% white
and 22% black, and had an average of an eighth grade education. (See

Appendix A for more complete demographic information.)

Design

The experimental design was essentially a 2x5 analysis of variance
design. (See Table 1.) There were two levels of population (SSI and
state medical) and 5 levels of treatment condition (control, informa-
tional mail, "personal" mail, telephone, and in-person contact). Equal
Es were randomly assigned to each of the 5 treatment conditions. The
two subject categories had almost identical_gs, with slight differences
as described below.

To use the terminology of Campbell (1957), the experiment was a

post-test only design. There were three dependent variables.

41

Table 1

Experimental Design

Mode of Outreach

 

Control Informational Personal Telephone In—Person Total

 

 

 

Mail Mail
Population
SSI n=35 n=35 n=35 n=35 n=35 175
State
Medical n=30 n=30 n=30 n=30 n=30 150

 

Total N = 325*

*Note. The original plan called for 50 subjects (25 SSI and 25 state
medical) to be selected from each of the 7 I & R centers, thus giving

a total N of 350 persons. However, due to a limited population of
eligible subjects in two areas, one center received only 10 subjects
assigned to the state medical category, and another center received only

15 in that category. Thus, the n per cell was adjusted as shown above,
and the total N became 325.

Procedure

Experimental and Control Conditions

 

The experimental and control conditions in this study consisted of
the following:

Control Group. The individuals in this group served as a control by

 

which to measure the relative effectiveness of each of the treatment cate-
gories. Hence, the individuals in this category received no contact
directly resulting from I & R/Outreach services. They, of course, were
not prevented from exposure to any incidental information concerning
available services (i.e. by word of mouth, from other groups, by self-
inquiry, from the media, etc.) that they would otherwise have been able

to receive. In this manner, the only difference between those in the con-

trol category and those in the other experimental conditions, in terms of

42

information about I & R services, should be the outreach contact
specified in each of the experimental treatment categories.

Informational Mail Contact Group. Individuals in this group

 

received a direct mail letter from I & R/Outreach, identifying some basic
services available through the I & R center and inviting the individual
to contact the center either by phone, letter or in person if s/he would
like any further information or would like to request services. It also
noted that an outreach worker could be sent to the individual's home if
desired. The overall tone of the letter was courteous, but primarily

"services to meet needs"

informational in nature, and emphasized a
approach. The final statement of the letter was a request for the in-
dividual to call or write the I & R center if s/he would like to sign up
to receive the I & R center's "newsletter." The letter was standardized
for all centers with the exception of information pertinent only to a
particular center such as names, phone numbers, etc. (See Appendix B for
a sample letter.) The mailing of the letters was performed by Area
Agency personnel through simultaneous batch mailings for all centers,
according to a pre—determined schedule established by the experimenter

(see p. 44 ).

"Personal" Mail Contact Grogp. Individuals in this group received

 

a direct mail contact with the same informational content as the letter
described for the informational mail group. However, the style of the

letter was made more personalized through the consistent use of "you,"

"your neighbors," "we," etc. when presenting the material and particu-

larly by the listing of four or five names of elderly persons who do

use the I & R center. The names were listed in a personalized request

to contact the helpful people at the I & R center (who were referred to

43

by their first and last names, i.e. Barb Jones, etc.). In contrast to
the needs/services tones of the "informational" letter, this letter
presented a friendly, informal approach. Once again, the letter followed
a standard format for each center. (See Appendix B for sample "personal"
letter.) The mailing procedure was identical to that described for the
"informational" letters.

Telephone Contact Group. The individuals in this group received a

 

telephone call from an I & R/Outreach worker, actually establishing ver—
bal contact and presenting essentially the same information as presented
in the letter conditions. To as much an extent as possible, the phone
contacts followed a standardized format, beginning with an introduction
and explanation of the I & R center services and winding up, as in the
letter conditions, with a separate request to call or write the I & R
center if they would like to sign up for the newsletter. (See Appendix
B for sample phone format.) In addition, this mode of contact allowed
for immediate feedback to any client-initiated questions, requests, etc.
In the instances where this occurred, outreach workers then handled the
individual as they would during any "normal" outreach contact.

In—Person Contact Group. The individuals in this group received a

 

personal visit by an I & R/Outreach worker. The worker established
actual face-to-face contact with the individual and provided essentially
the same information as provided in the mail and phone contact. The
format was similar to the phone format and was standardized to as much
an extent as possible (recognizing that such an interpersonal situation
requires a degree of flexibility). Once again, this contact allowed for
immediate feedback to any client-initiated questions, requests, etc.

When such occurred, the outreach worker handled them as in any "normal"

44

outreach contact. As in all other conditions, the contact included at
the end a request that the individual call or write the I & R center to
sign up for the newsletter.

Training and Supervision of the I & R Personnel

 

As soon as the final target subjects were identified and random
assignment to conditions had taken place, a training session was held to
explain to outreach personnel the protocols to be followed in conducting
the telephone contacts and personal visits. Also, the procedures for
data collection were reviewed and apprOpriate checklists were distributed.
(See Appendix C for example of checklist.)

Upon completion of this training, the assigned client lists were
distributed and center personnel were instructed to begin outreach
Operations. To avoid possible biasing effects of time, weather, etc.,
contacts were proportionately staggered (for example: 5 of each type of
letter contact, 5 phone calls and 5 personal visits per center, per week)
throughout the outreach phase. Administrative personnel from the Region IV
Area Agency on Aging supervised daily project Operations. (See Appendix D
for memorandum of agreement.) In addition, the project researcher was
available for telephone consultation whenever needed, as well as per—
sonally monitoring operations on a weekly basis.

Measurement of the Dependent Variables

 

The experimental manipulation used in this study was intended to
test various modes of conducting outreach. Unfortunately, the dependent
variable of "response to outreach" is something that might be defined in
a variety of ways. Recognizing this fact, this study utilized three
dependent measures, each of which is presumed to reflect one operational

definition of "response to outreach."

45

The "Client Card." Whenever I & R/Outreach contacts someone for the

 

first time and a "successful" outreach contact is made, appropriate demo-
graphic and needs information is recorded on a standardized form called

a "client card." (See Appendix C for example card.) The rationale behind
using this measure as an indicator of a successful outreach contact is
that, if the outreach worker has established enough rapport and inter-
action with the elderly person to get him/her to provide the information
required on that form, then the worker has presumably been able to observe
and assess the major needs of that person. Following such an assessment,
any necessary referrals or service provision can be arranged. This
locating and serving of needy elderly encompasses the major function of

I & R outreach. In fact, the number of client cards completed is one of
the indices by which outreach operations are evaluated by the Area Agency
on Aging. Thus, in addition to making use of the existing record—keeping
system (all seven I & R centers use the standardized client card system);
this particular definition of the dependent variable corresponds to the
definition of the "success" of the I & R centers currently used by the
Area Agency on Aging.

For this dependent variable then, a "positive" response to contact
was operationally defined as any outreach contact (mail, phone, or in
person) that resulted in a client showing a willingness to participate
with the I & R center by providing the information necessary for a client
card to be filed in his/her name (i.e. if there was a client card recorded
for that person by the end of the experimental phase, it was considered
a "positive" response).

The other category of this dependent variable was, of course, "non—

positive" response. This included any response short of a client card

46

being filed in the client's name.

These two categories of response, as defined above, were used in the
experimental analysis.

It should be pointed out that the personal contact mode, and to a
lesser extent the telephone mode, had the advantage of allowing a
"positive" response and recording to take place during the initial con-
tact. This is a legitimate advantage of these modes and, in fact, occurs
in ”normal" outreach operations. Since the purpose of this experiment
was to test the effectiveness of actual methods of outreach in reaching
potential clients, such immediate responses, when they occurred, were
fully allowable as "positive" responses in terms of the dependent variable.

The above operational definition of "response to outreach" is concep-
tualized in terms of the perceived function of I & R/Outreach. However,
as mentioned previously, other possible indicators of response could also
be obtained. For this purpose, a second dependent variable was utilized,
conceptualizing "response to outreach" more in terms of response to
various modes of persuasive communication.

"Newsletter Sign—Up." Each mode of outreach concluded the initial

 

contact with a request for the elderly person to call a certain phone
number or write the I & R center to be placed on the mailing list for the
I & R "newsletter." (Each center publishes a monthly newsletter with bits
of information and local news of interest to the elderly.) In contrast

to the client card dependent variable, this measure removed the "immediate
response" possibility from the phone and personal contact modes. This
measure required the same specific, time-delayed response in each out-

reach category. Although this type of response is less reflective of the

actual purpose of an outreach contact, it is nonetheless a relevant

47

indicator of effective persuasive communication with the elderly. In
particular, signing up for a newsletter may represent a less "threaten-
ing" intermediate response for those elderly who may not be ready to
acknowledge a need or may not have a need at the present time, but who
might still wish to establish some connection with the I & R center. As
such, it provides a chance to measure another dimension of the elderly
person's response to outreach contact.

Although no specific form, such as a client card, exists to measure
this dependent variable, recording positive responses was a fairly sim—
ple procedure. Each center was provided with a checklist containing the
names of all those elderly to be contacted in the study. Then, as calls
or letters came in requesting the newsletter, the appropriate date, mode
of contact, etc. were indicated on the checklist. For this dependent
variable, a positive response was operationally defined as any outreach
contact that resulted in a client calling or writing to sign up to
receive the I & R newsletter. The other category, i.e. non-positive
response, was, of course, defined to be when the client did not call or
write to sign up.

"Number of People Receiving Services." Since the linkage of needy

 

elderly with available services is one of the main functions of I & R/
Outreach, a third dependent variable ("number of peOple who received
services as a result of I & R contacts") was also utilized. This
provided a measure of whether type of initial contact had any effect on
the number of people who would receive services out of a given sample of
elderly. The number of new services received through I & R center
efforts was easily obtainable due to the standard practice of the I & R

centers to record services provided directly and to "follow—up" all

48

referrals made to other agencies. Examples of the services received
include the following: food and nutrition (e.g. "meals on wheels"),
transportation, blood pressure checks, legal aid, social and companion-
ship services, home repair and various miscellaneous services (i.e.
getting a pair of glasses repaired, etc.).

Originally, it was planned to utilize the variable of "number of
referrals” as a fourth dependent variable and to examine, first, the
differences between those groups contacted by each method in terms of
number of referrals received and, second, the differences between groups
in terms of those not showing up for referred appointments. This plan
was abandoned when the results revealed that a very small total number of
referrals were made to outside agencies and, further, that all persons
referred did show up for their appointments. (See Appendix E for data
on number of referrals.) Hence, it was decided to focus on the number of
people receiving services as a result of I & R contact, directly and by

referral, as the third dependent variable.

The Follow-Up Survey

 

Approximately one month after initial contact, all subjects in each
of the four experimental conditions, as well as the control condition,
were visited by an outreach worker from the local I & R center for a
follow-up interview. The standard procedure followed was fairly simple.
The outreach workers identified themselves as being from the I & R center
and, when applicable, referred to the earlier contact the person should
have received from the center. They also identified themselves as being
authorized by the Michigan Office'of Services to the Aging to survey

Opinions Of Michigan senior citizens (in order to gather information to

49

be used in state planning). They, thus, presented themselves as perfor-
ming a dual role of (l) seeking to assist the elderly person in any area
where s/he might require service and (2) seeking to gather some informa-
tion about the person and his/her attitudes, etc. If the person agreed,
a survey questionnaire was administered to the respondent. (The survey
questionnaire included a wide variety of demographic, attitudinal and
self-report items. See Appendix F for a copy of the instrument.) Over-
all, the approach used proved quite successful, as only 18 of the 210
people contacted (approximately 9%) refused to complete the interview.1
The major function of the follow-up phase was to provide necessary
background data on the elderly persons involved in the project. The
data gathered in the follow-up survey were used essentially in post—
experimental interpretation of the "response to outreach" findings and
in the examination of any relationships between demographic, attitudinal,
self—report and outcome variables. The data were analyzed primarily by
correlational techniques, particularly for any relationship between the
survey-generated variables and the dichotomous dependent variables.

Construction of Scales

 

In an effort to simplify the interpretation of the follow-up data
and to increase the reliability of the variables identified from that
data, several scales were created from the items in the questionnaire.
These scales were formed by a combination of logical and empirical

processes.

 

1It might also be interesting to note that chi square analyses
revealed that there were no significant differences between the five
types of initial contact (p=.40) or between the seven I & R centers
(p=.34) in terms of percentage of persons who refused to complete the
follow-up interview.

50

When the questionnaire was originally constructed, numerous items
were included to reflect some basic content areas of interest in this
study (social contact, knowledge of social services, "life satisfaction,"
etc.). After the data were collected, the item intercorrelations were
examined to check on the feasibility of constructing scales. Whenever
it was found that a set of items seemed to group together both logically
and empirically, a tentative scale was formed and subjected to further
testing. The following criteria were used: (1) each item must logically
fit with the content of the scale; (2) each item must correlate with its
scale total significantly (p<(.001); (3) the correlation of each item
with the total score of its scale must be higher than with the other
scales; (4) the scale reliability (in terms of Cronbach's Alpha) must be
sufficiently high (.50 was selected as the minimum acceptable level);
and (5) the scale reliability must not be increased or decreased drama-
tically by the removal of an item.

After several repetitions of this selection process, four fairly
distinct scales were created. These four scales are briefly outlined
below. (The obtained scale reliability, in terms of Cronbach's Alpha,
is shown in parentheses for each scale. For more complete information
about the internal consistency of the scales, see Appendix G.)

"Family Contact" Scale (alpha=.74). This scale is composed of

 

items #18, 20, 30, 31 and 32 from the follow-up questionnaire. (See
Appendix F.) Items 18 and 20 are self-report items about frequency and
amount of contact with the person's family and relatives, while items
30, 31 and 32 involve specifically the number of and amount of contact
with the person's children. This scale is intended as an indicator of

the individual's contact with family and relatives outside of the

51

immediate household. (Item 17, it should be noted, is used to gauge
social contact within the household.)

"Other Social Contact" Scale (alpha=.56). This scale is composed
of items #19, 21, 22, 23 and 23a from the questionnaire. It is inten—
ded to measure non-family social contact outside of the immediate house-
old. (The questions include self—report items about frequency of social
contact with friends, neighbors, clubs and through church attendance.

"Knowledge of Social Services" Scale (alpha=.68). This scale is

 

composed of items #38, 39 (i.e. the number of social services the person
knows about and the number the person has actually tried) and the four
parts of item 41 (i.e. whether or not a person can identify a social ser—
vice for each of the four problem areas mentioned). This scale is inten-
ded to measure the person's level of awareness of social services avail—
able.

"Life Satisfaction" Scale (alpha=.57). This scale is composed of

 

items #28, 36, 52, 53, and 56 from the questionnaire. It is intended to
measure an overall concept of current satisfaction with one's life by
measuring satisfaction in each of three presumably key areas affecting
an elderly person's life (loneliness, income adequacy, and health) along
with two more generally worded items about satisfaction with life.

(See items 36 and 56.)

Further Analysis of the Survey Items

 

To determine whether or not further data reduction might be possible,
a factor analysis was performed on the survey items. In addition to the
four previously discussed scales, one other empirically related cluster
of items appeared which seemed to represent a logical content area. To

determine whether or not it formed a usable scale, the same tests were

52

performed as were outlined for the prior scales. In this manner, the
following rational scale thus emerged.

"Problems" Scale (alpha=.66). This scale is composed of six items

 

which assess the client's self—rating of level of need in a variety of
areas. The items are taken from question #51 of the follow—up survey

and cover the following problem areas: housing, health care, income,

crime, nutrition and food, and transportation.

In all, five logical scales were created from the follow-up survey
data. As expected, there was some degree of relationship between the
scales. (Nevertheless, further attempts at data reduction, including
factor analysis, provided no rational or empirically sound combinations
of these variables.) The following Pearson correlation matrix illus-
trates the relationships among the five scales. (Note: for the purposes
of this matrix, all the scales are coded such that a "low" score is a
"desirable" score, i.e. high family contact; high social contact; fewer
problems; higher knowledge of social services; and greater life satis-

faction.)

Table 2

Correlations Among the Scales

 

 

l 2 3 4 5
1. Family Contact
2. Other Social Contact .12
3. Problems .13 .02
4. Social Service 4
Knowledge .04 .13 -.21

5. Life Satisfaction .10 .25 .31 .04

 

53

As one can see, several of the scales are empirically related.
(Correlations of approximately .25 or greater are "significant" at
p‘(.01 due to the large sample size.) However, it should be noted that
the correlations are between rational scales which would logically be
expected to be related. For example, one would expect that both social
contact and lack of problems would be positively related to life satis-
faction. Similarly, it is not surprising that those with fewer problems
tend to have less knowledge of social services.

In view of these rational relationships in content, the obtained
empirical correlations were not judged to be prohibitively high. Further,
it was felt that the usefulness of these scales in post-experimental data
analysis was not diminished (particularly since these five scales were
all found to be internally consistent). In fact, somewhat intercor-
related rational scales are here regarded as much more useful than or—
thogonal, but uninterpretable, empirical "factors." Hence, for the pur-
pose of subsequent data analysis, it was decided that these five scales
would be used as summary variables representing key content areas from
the follow-up survey.

Other Areas of Interest

 

Finally, in addition to the five logical scales outlined above and
the basic demographic information (i.e. age, sex, race, marital status,
number of persons in household, education, etc.), the follow-up survey
also sought to shed some light on the central question of why some needy
elderly respond to services available and seek assistance while other
needy elderly do not. Toward this end, certain questions on the follow-
up survey were intended to solicit client opinions about social services,

client reactions to being contacted by I & R personnel, client responses

54

about his/her peer network and his/her peers' Opinions about welfare,
etc. These items were also analyzed primarily for their relationship to

the three dependent variables.

ConcludingfiOperations

 

Following the experimental and follow-up phases of the project, the
entire original lists of elderly were provided to the centers for their
use in normal outreach operations. In this way, none of the elderly were
denied the Opportunity to learn of services available through the I & R
centers. In addition, a copy of a report of findings was provided to
the Area Agency on Aging for distribution to all centers that partici-
pated in the study. This was followed by personal consultation to insure

complete explanation of the findings.

55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N Havana: on I
o Hmm mm z
Houuaoo
Hwowvoz on I
No Ho Hmm mm 2
uwma> Hmaomuom
HOOHOOz on
No #0 Hmm mm 2
ozonmoaoe
N H HOOHvOz on n
o o Hmm mm 2
Ham: Hwaomumm
HOOHOOZ om uz
Hmm mm
«0 Ho
Ham:
HOOOfiumauomaH
.OOO«>OHQ
somnvoom muwv ho>u3m .m .vo>oaou
paw muasmou moanawuwoaon .q A.uomuaou mead: vouomu
Hmucmafiummxm mooa>umm .m Hmcawwuo A.uomuaoo mo . . Iaoo hamsow>
.muouaoo unu nouuoamama .N Hound mmmv Ouwv scum mmmv A.mxooa N maouma loam .wouw
ou voauau sumo“ omV .aawwn om wcaaoad<v Iaxouaaw uwqu .ucmaummuu an OBOO noxoun
Ion nomad luso cu uncommon .H muumuaoo .momaommou Ham .aawon mcowu Ou poems mo uaoa mauovao mo
oumaaaou.mwmzawam dump aawmm QOISOHHom vuoumu mumucmo Imuomo numwuuso Iawwmmm aovcmm mumwfl HmauHaH
mcoauwuono .
uvsaoaoo mwmhadaa mama mansoaaom momaonmou Oxoa nuomuuso Gamma vanaaovaum .Oumaa HmuuHaH

onwfluao qunoum

m manna

CHAPTER III

RESULTS

The Sample

 

Subject Mortality

 

Due to the inevitable procedural problems associated with such a
field experiment, the final sample size deviated somewhat from that des—
cribed in the projected plan. (See Table 3.) All randomization pro-
cedures for the selection and assignment of names were kept intact,
however, so that the integrity of the experimental design was not
violated. The final number of subjects assigned to each mode of contact,
as well as the subject mortality within each mode, are presented in the

table below.
Table 4

Final Sample Composition: Subject Mortality by Contact Modea
Type of Contact

 

 

 

 

Informational Personal In—
Control Mail Mail Telgphone Person

Final Con-

firmed Sample 41 45 38 43 43 210
Removeg from

Sample 19 10 l9 17 17 82
Total Initial

Assignment 60 55 57 6O 6O N=292

 

ax2=4. o7 df=4 (p < .30)

bReasons for removal included: person could not be located, person was
deceased, person was in nursing home, etc. This "subject mortality"
occurred prior to treatment. The "final confirmed sample" (N=210) rep—
resents the actual "treated" sample.

cThe discrepancy between these values and those projected in the original
plan is due primarily to the fact that one center lost approximately
half of its assignment lists. This table, then, is based on the correc-
ted total of N=292 persons initially assigned.

56

57

The top row of the table represents the confirmed number of subjects
in the final sample (i.e. confirmed to be living at the registered loca-
tion and physically capable of making the responses measured as the de-
pendent variables). As one can see, these totals represent a fairly
sizeable reduction from the initially assigned sample. This was quite
as expected, however, and the important finding is that there was no
relationship between "subject mortality" and experimental condition
(p<:.30). Further, the sample size was still quite large and well dis-
tributed across conditions. Hence, it appears that in the above ways,
the final sample composition was quite satisfactory.2 Although these
findings were very encouraging, one more characteristic of the sample
was still considered prior to examining the experimental results.

Effectiveness of Randomization

 

In order to examine for effects of treatment in this experiment, it
was important to try to determine whether or not the five contact groups
were indeed "equivalent" on all salient variables other than type of con—
tact. One would suspect that this would be the case with this large of
a sample, due to the laws of probability when a randomization procedure
is utilized. Nonetheless, it is preferable to check for equivalency
whenever possible, particularly in this case since, as expected, a
sizeable portion of the initial list of targets had to be removed from
the sample.

In order to check for equivalency, a total of 21 demographic and

descriptive variables were examined for differences between the five

 

2In addition, it should be noted that chi square analyses on the final
sample of 210 subjects showed no significant difference in distribution of
types of contact by I & R center (p=.9l). Also, they revealed no signifi—
cant difference in distribution of SSI vs. state medical recipients by
type of contact (p=.65), nor by I & R center (p=.28).

58

groups. The results reveal that the randomization procedure apparently
was quite effective, as only one of the 21 variables was found to vary
between the groups at less than the .05 level of significance. In
addition, further analysis revealed that the one discrepant variable had
no strong relationship to any of the dependent variables. (See Appendix H
for a list of the variables tested and a discussion of the procedures
used.) Hence, it appears that the randomization procedure did indeed
produce functionally equivalent groups for the purposes of this experiment.

With this in mind, the experimental results can now be examined.

Treatment Effects

 

Primary Outcome Measures

 

A chi square analysis was performed to test for significant effects
between the independent variable, type of contact, and each of the three
dependent variables: (1) client card registrations; (2) receiving a ser—
vice; and (3) newsletter sign—ups. The results are shown below in
Tables 5, 6, and 7.

Table 5

Client Card Registrations by Type of Contact8

 

 

 

Informational Personal In-
Control Mail Mail Telephone Person
Client Card 0 1 4 12 34 51
No Client 41 44 34 31 9 159
Card
Total 41 45 38 43 43 N=210

ax2=99.47 df=4 (p<.001)

59

Table 6

Receiving a Service by Type of Contact3

 

 

 

Informational Personal In-
Control Mail Mail Telephone Person

Received

Service 0 2 4 8 10 24
Did Not

Receive

Service 41 43 34 35 33 186
Total 41 45 38 43 43 N=210

ax2=15.94 df=4 (p=.003)

Table 7

Newsletter Sign-Ups by Type of Contact8

 

 

 

Informational Personal In-
Control Mail Mail Telephone Person
Signed Up 0 0 3 l 5 9
Did Not
Sign Up 41 45 35 42 38 201
Total 41 45 38 43 43 N=210

ax2=11.19 df=4 (p=.024)

As one can see, all three dependent variables show a significant
relationship to type of contact, with the client card dependent variable
showing the most pronounced effects. In all three cases, the invperson
mode of contact achieved the highest rate of effectiveness. A more
detailed interpretation of these results will be presented in the

"Discussion" section.

60

Analysis of Variance to Test for Population Differences and Interaction
Effects

 

To examine the difference between the two client "populations" tested
(SSI recipients and state medical recipients) in terms of the dependent
variables and also to test for any interaction between population and type
of contact, a two—way analysis of variance was performed for each of the
dependent variables. The results reveal that the main effects of "type
of contact" were again significant for each dependent variable (client
card registration, p=.001; receipt of service, p=.006; newsletter,
p=.003). As for pOpulation differences, however, none of the main effects
were significant and none of the interaction effects of population with

type of contact were significant. (See Tables 8, 9 and 10.)

Table 8

Analysis of Variance for Client Card Registrations
by Type of Contact and Assistance Group

 

 

Source DF Mean Square F Significance of F
Type of Contact 4 4.93 45.9 .001
Assistance Group 1 .01 .06 .807
2—Way Interaction 4 .01 .09 .967
Residual .10
aSSI or state medical

Table 9

Analysis of Variance for Receipt of Service
by Type of Contact and Assistance Group

 

Source DF Mean Square F Significance of F
Type of Contact 4 .390 3.72 .006
Assistance Group 1 .203 1.93 .166
2-Way Interaction 4 .035 .33 .855

Residual .105

 

61

Table 10

Analysis of Variance for Newsletter Sign-Ups
by Type of Contact and Assistance Group

 

Source DF Mean Square F Significance of F
Type of Contact 4 3.37 4.08 .003
Assistance Group 1 .04 .05 .833

2—Way Interaction 4 .59 .71 .587

Residual .83

 

The absence of any main effects or interaction effects due to "popu—
lation" (SSI or state medical) raised the question of whether or not these
two sample groups of low income elderly really represent distinct popu-
lations. To be on the safe side, they were originally classified as such
because they originated from two distinct lists of aid recipients. How-
ever, after examining (1) the lack of difference between these two groups
in terms of outcome measures, (2) the convincing lack of difference
between the two groups in terms of the follow-up survey variables, and
(3) the similarity of eligibility requirements for the two programs as
outlined by Michigan Department of Social Services manuals, it was
decided that, for the purposes of this study, the two groups in fact
represent a single population. (See Appendix I for a rationale for this
decision and for a description of the data examined.) Hence, in all per—
tinent data analyses, the data from the groups were treated as that from
a single sample, randomly selected from a target population of low-
income elderly with no prior contact with the I & R network.

Secondary Outcome Measures

 

It was felt that the responses to certain of the follow-up survey
variables could conceptually have been influenced by the experimental

treatment (i.e. by having received an outreach contact of one type or

62

another). In particular, the scales, measuring knowledge of social
services, problems the person had and life satisfaction, were thought
of as potentially related to type of contact. To examine this pos—
sibility, these scales were analyzed as secondary outcome measures. A
one-way analysis of variance, using type of contact as the independent
variable, was performed on each of the scales. (Scale scores were cal-
culated by summing the scores of the scale items.) The results of these
analyses revealed that there were no significant differences on any of
these three variables between the five experimental groups. (See
Appendix J.) Thus, it appears that the I & R/Outreach contact, regard-
less of mode, was not a substantial enough event in the lives of the
elderly persons to have produced a significant change in these more
global life dimensions. (The scales of "family contact" and "other
social contact" were also examined for any differences between the five
groups. Again, no significant differences were found.)

The Effects of Receiving a Service

 

One additional area of interest that might be regarded as a pos-
sible outcome of this study was the question of what effect, if any,
receiving a service might have had on an elderly individual. Although
this question was not directly addressed in this study, several
variables were examined to see if any possible effects could be noted.
To accomplish this, T—tests were conducted between service recipients
and non-service recipients on a number of variables.

Once again it appears that, for the most part, the intervention did
not produce a significant effect in terms of the more global life dimen-
sions measured by the rational scales. For example, service recipients

were still the highest scorers on the "problems" scale (p=.015). This

63

is as would be expected in that their high level of need is one of the
factors that determined they receive a service. (Presumably receiving
a service did not increase their problems in the areas measured by that
scale.) Whether or not receiving a service may have improved the
"problems" situation from a previously even greater discrepancy is un—
known. Unfortunately, due to the lack of a pre-test, the question of
change brought about by receiving a service is unanswerable.

As for the scales measuring family contact and "other social con-
tact," no significant differences were observed (p=.464 and p=.708,
respectively). For the scale of life satisfaction, the service
recipients did score higher, although not significantly (p=.284). The
fact that the service recipients scored higher is somewhat surprising
given the problems situation discussed previously and may relate in
some way to the recipients' involvement with the I & R center. Again,
however, such a statement can only be speculation in this case.

One scale that did show a significant difference favoring the ser-
vice recipients was the social service knowledge scale (p=.017). Here
the logical connection between receiving a service and the score on
this scale is somewhat stronger. This could plausibly be an indicator
of outcome to some extent. Finally, as might be expected, those who
received a service more Often indiCated that they would call the I & R
center in the future if a problem occurred (p=.001). (In fact, every
person who received a service indicated that they would contact the
center in the future if a need arose.) This may indicate that recip—

ients were satisfied with the service received from the I & R center.

64

Correlational Analyses

 

A substantial portion of the correlational analysis of the survey
data, that of the empirical analysis of items and the creation of
logical scales, was described previously in Chapter II. What remains,
primarily, is to see how the major survey variables relate to the
dependent variables used in this experiment.

Relationships of the Survey Variables to the Three Dependent Variables

 

Of principal importance in this study is the attempt to identify
those factors that may help explain why certain needy elderly respond to
a notification of available services and other needy elderly do not. To
pursue this question, a wide variety of demographic, attitudinal, and
need-related variables were examined for their relationship to the
three dependent variables.

To begin, eleven demographic variables (age, sex, race, education,
marital status, number of persons in the household, with whom the person
resides, health as rated by the outreach worker, memory/ability to pay
attention as rated by the outreach worker, number of living children,
and number of children living within a one hour drive) were examined
for their relationship with the three dependent variables. Simply
stated, chi square analyses and T—tests revealed that only two of the
demographic variables significantly (p‘(.05) differentiated positive
and non-positive responders on any of the dependent variables. (See
Appendix K for a table of the variables and their respective signifi-
cance levels.) The only significant findings out of 33 relationships
tested were between positive and non-positive responders on the news-
letter dependent variables, as differentiated by the demographic variables

Of the client's memory/attention and health ratings. (Positive

65

responders were ashown to have both higher memory and higher health
ratings, p ( . 01.)

To further examine for any relationships, simple Pearson correlation
coefficients were calculated for those variables allowing such a level
of analysis. (See Table 11.) Here, only the variables of age, educa—
tion, and memory achieved any significant correlations. (For age:
r=.13, p=.042 with the client card dependent variables; for education:
r=.15, p=.021 with the newsletter variable; for memory, r=.l4, p=.03l
with the newsletter variable.)

In summary, the results show that very few demographic variables
were significantly correlated with response or non-response on the
dependent measures. Further, the actual magnitude of even the signifi-
cant relationships was quite small. Hence, it appears as though the
demographic variables reveal very little in terms of answering the ques-
tion of why some needy elderly respond positively to outreach contact

while others do not.

Table 11

Pearson Correlations of Demographic Variables
With the Three Dependent Variables

Dependent Variables

 

Demoggaphic Variables Client Card Service Newsletter
Age -.13"‘1 -.O8 -.06
Sex —.11 -.05 .01
Race .06 -.07 .05
Education .01 .00 .153
Number of persons in household -.11 -.06 .00
Health .02 .08 .11
Memory —. 09 .04 .14a
Number of living children —.10 -.12 .07
Number of living children within

1 hour drive —.09 —.Ol .03

a . . . . .
Correlation 13 Significant at p( .05.

66

The next step in the process was to examine five categorical variables
measuring various facets of the person's interactions with, and attitudes
toward, social service. (See Questions #25, 37, 48, 50, and 54.) Once
again, chi square analyses revealed that none of these items were signi-
ficantly (p4(.05) related to any of the dependent variables. (See Appen—
dix K for a list of variables and their significance levels.) Pearson
correlation coefficients were then calculated between the five variables
and the three dependent variables to further examine for any relationships.
Again, none were found to be significant. (See Table 12.)

The third step in the process was to examine another major category
of variables from the follow-up survey, that of the five rational scales
(family contact, other social contact, knowledge of social services,
problems perceived, and life satisfaction). Pearson correlation co—
efficients were calculated between the five scales and each of the
dependent variables. For the dependent variable of client card registra-
tions, only the "problems" scale was significantly correlated (r=.l7,
p=.015). For the dependent variables of receiving a service, the prob-
lems scale was again the only one significantly related, with a slightly
larger correlation (r=.18, p=.01). For the newsletter dependent variable,
none of the scales showed a significant correlation. (See Table 12.)

The fourth step in the process was to examine the other major
category of items in the follow-up survey, which was a set of questions
asked specifically about how the person perceived the outreach contact
s/he received. Questions #5 through 8 from the survey (see Appendix F)
were examined by Pearson correlation for any relationship to the depen-
dent variables. For the dependent variable of client card registration,

none of the four items were significantly related. Two of the items came

67

close: question #5 (i.e. how clear was the meaning and content of the
contact) and question #8 (i.e. how much need the person had for their
services), with correlations of .16 (p=.05) and .11 (p=.l3), respec—
tively.

For the dependent variables of signing up for the newsletter, the
situation was similar. Question #5 again was slightly related (r=.15,
p=.08) and so was question #7 (extent to which the center seemed like it
would be a friendly place) with a correlation of .13 (p=.10).

For the dependent variables of whether or not the person received
a service, however, the relationships were much stronger. Question #7
was almost significantly related (r=.15, p=.08); question #6 (i.e. did
the person think that the I & R center would be a helpful place) was
significantly related (r=.20, p=.02); and question #8 (i.e. need for
their services) was quite strongly related (r=.36, p=.001) to this

dependent variable. (See Table 13.)

Table 12

Pearson Correlations of Survey Variables
With the Three Dependent Variables
Dependent Variables
Social Service Related Variables Client Card Service Newsletter

 

Question #25 (would friends accept

foodstamps) .14 .08 .04
Question #37 (would person accept
foodstamps) —.01 .03 -.09

Question #48 (positive, neutral or

negative comment about social

services in general) —.01 .06 .05
Question #50 (how long client has

been associated with Department

of Social Services) .05 .08 .12
Question #54 (does client perceive

any welfare stigma) —.05 .04 -.14

68

Table 12 (cont'd.)

 

The Rational Scales Client Card Service Newsletter
Family Contact .04 .06 .08
Other Social Contact .05 .03 -.05
"Problems" .163 .18a —.02
Social Service Knowledge .10 .13 .08
Life Satisfaction .02 .08 .08

aCorrelation is significant at p'(.05.

Table 13

Pearson Correlations of Client Ratings of the
Outreach Contact with the Three Dependent Variables
Dependent Variables

 

Client Rating Items Client Card Service Newsletter
Question #5 (how clear) .16 .04a .15
Question #6 (center helpful) .06 .20 ,09
Question #7 (center friendly) .06 .15b .13
Question #8 (need for services) .11 .36 .04

aCorrelation significant at p< .05.
bCorrelation significant at p< .01.

Summary of the Predictability of the Three Dependent Variables Using_the
Survey Variables as Predictors

 

The four major categories of variables from the follow-up survey (the
demographics, the social services related items, the five rational scales,
and the four items measuring reactions to the outreach contact) were
examined for their relationship to the three dependent variables. Those
that were found to be somewhat significantly related to any of the
dependent variables (p< .10) were selected for use in the final overall
regression analyses. This was done in an effort to examine the usefulness
of these variables as predictors of response or non-response in a group
of elderly receiving an outreach contact.

For the client card dependent variable, the following variables were

69

entered into a "stepwise" regression analysis: age, sex, question #5
(i.e. how clear the meaning of the outreach contact was to the person),
number of persons in the household, number of living children the person
has, and the person's score on the problems scale. Together, these six
variables accounted for only 9% of the variance in the dependent vari-
able (p=.390). The single most important variable, the problems scale,
accounted for only just less than 3% of the variance (p=.160). As one
can see, the survey variables do little to explain differences between
positive and non-positive responders in terms of this dependent variable.

For the dependent variable of signing up for the newsletter, the
results are quite similar. For this dependent variable, the following
survey variables were used in the regression analysis: health, memory,
question #5, question #7 (i.e. to what extent the I & R center seemed
like it would be a friendly place), education, question #50 (i.e. when
the person first had contact with the Department of Social Services),
and question #54 (i.e. whether or not the person feels people think
poorly of those who receive government assistance). Here the results
reveal that these seven variables account for only 6-1/2% of the vari-
ance in the dependent variable (p=.670) and that the most important
variable, education, accounts for only 2% of the variance (p=.l90).

For the dependent variable of receiving a service, the results are
somewhat more encouraging. For this dependent measure the following
survey variables were used: the problems scale, question #6 (i.e. to
what extent did the I & R center seem like it would be a helpful place),
question #7, and question #8 (i.e. to what extent did the person feel
s/he had a need for the I & R center services). The results show that

these variables account for nearly 14% of the variance in this dependent

70

measure (p=.016). However, most of the variance explained (13%) is

accounted for by question #8 alone (p=.002). (See Table 14.)

Table 14

Multiple Regressions: the Three Dependent Variables
nflth the Significant Q)(.10) Survey Variables

 

 

 

 

Mult. R Simple Overall
R Square R F Signif.

For the Client Card Variable:

Variables in the Equation

"Problems" scale .165 .027 —.l65 2.02 .160
Question #5 (how clear) .218 .048 —.l60 1.77 .178
Sex .245 .060 -.107 1.49 .224
Number of living children .272 .074 -.101 1.35 .251
Age .293 .086 .131 1.27 .283
Number of persons in household .295 .087 .106 1.07 .390
For the Newsletter Variable:

Variables in the Equation

Education .149 .022 .149 1.63 .190
Memory .190 .036 .137 1.33 .270
Question #50 .220 .049 —.128 1.19 .320
Question #7 (how friendly) .230 .053 .131 .96 .433
Health .247 .061 .105 .89 .493
Question #54 .252 .064 .142 .76 .604
Question #5 (how clear) .254 .065 .146 .65 .670
For the Service Variable:a

Variables in the Equation

Question #8 (how much need) .360 .130 .361 10.77 .002
Question #7 (how friendly) .368 .135 .145 5.56 .006
Question #6 (center helpful) .368 .136 .204 3.67 .016

aVariables mentioned in the text which are absent from the multiple regres-
sion tables were excluded because their additional impact on the R—square
value was negligible.

Predictability Including Type of Contact

 

Since the major independent variable of type of contact demonstrated

obvious significant treatment effects, it was decided to examine what the

7l

inclusion of this variable does to the overall level of predictability.
To accomplish this task, the categorical variable of type of contact
was transformed into a dummy variable and forced into the regression
equations prior to the respective survey variables just discussed.

For the client card registration variable, it was discovered that
the type of contact accounted for an extremely high 49% of the variance
Q3<.CKHJ. Once this portion had been explained, the remaining survey
variables accounted for an additional 5%, giving an overall level of
54% Of variance explained (p( .001).

The results were quite different for the newsletter dependent
variable. Here the type of contact accounted for only 3% Of the variance
(p=.715). Once type of contact had been entered, the remaining survey
variables still accounted for approximately 7% of the variance, with
education still the most important variable of the group. Together,
type of contact and the survey variables accounted for just over 10% of
the variance (p=.699).

For the dependent variable of receiving a service, the type of con-
tact demonstrated a somewhat stronger relationship, though still not
quite significant. Type of contact alone accounted for slightly over
8% of the variance (p=.l96). Once type of contact had been entered,
the remaining variables picked up an additional 14% of variance, giving
a total of 22-1/2% explained (p=.008). Once again, question #8 accounted

for most of that additional 14%. (See Table 15.)

72

Table 15

Multiple Regressions: the Three Dependent Variables with
Type of Contact and the Significant (p(_.10) Survey Variables

 

 

 

 

 

Mult. R Simple Overall

R Square R F Signif.
For the Client Card Variagle:
Variables in the Equation
Type of contact .688 .480 ---- 15.49 .000
"Problems" scale .718 .516 .165 14.48 .000
Number of living children .727 .529 -.10 12.55 .000
Sex .734 .539 .160 11.03 .000
Question #5 .737 .543 —.107 9.67 .000
For the Newsletter Variabée:
Variables in the Equation
Type of contact .167 .028 —--- .493 .715
Education .237 .056 .149 .810 .547
Memory .273 .074 .137 .896 .503
Question #50 .299 .089 .127 .922 .495
Health .307 .094 .105 .844 .568
Question #7 .314 .098 .131 .779 .636
Question #54 .321 .103 .142 .724 .699
For the Service Variable:a
Variables in the Equation
Type of contact .288 .083 ---— 1.55 .196
Question #8 (how much need) .470 .221 .361 3.86 .004
Question #7 (center friendly) .474 .224 .145 3.23 .008

aVariables mentioned in the text which are absent from the mutliple
regression tables were excluded because their additional impact on the
R-square value was negligible.

Predictability Within Type of Contact

 

As a final means of examining the usefulness of the survey variables
in predicting response or non—response, it was decided to examine the
appropriate relationships within each type of contact. In other words,
for example, it might be that age has a differential effect on response

depending on the type of contact used. To check this possibility, the same

73

four categories of survey variables (demographics, social service related
variables, the five scales, and the four items measuring reactions to the
outreach contact) were examined for their relationship to the dependent
variables within the groups receiving each type of contact.3 Once again,
those variables found to be somewhat significantly related (p<;.10) to
any of the three outcome measures were selected for use in regression
analyses.

In-Person Contact Recipients. For the client card registration

 

dependent variable, the survey variables of number of persons in the
household and education were entered into the regression equation. To-
gether they accounted for 13% of the variance in the dependent variable
(p=.210), with education accounting for 9% alone (p=.l35).

For the newsletter dependent variable, the survey variables of age,
education and question #37 (i.e. whether or not the person would accept
help from the government, such as foodstamps) were entered in the regres-
sion equation. Together these three variables accounted for 19% of the
variance (p=.213), with question #37 alone accounting for 14% (p=.067).

For the dependent variable of whether or not the client received a
service, the survey variables of memory, question #6, question #7, ques-
tion #8, and education were entered into the equation. Together these
five variables accounted for 21-1/2% of the variance (p=.421), with

question #7 accounting for lO-l/2% alone (p=.ll6). (See Table 16.)

 

3Because of the small number of positive responders in the informa—
tional mail and personal mail categories, these two groups were combined
into an overall "mail contact" category for the purpose of the "within
contact mode" regression analyses. Hence, the three types of contact dis-
cussed here will be simply: mail, telephone, and personal visit. In
some cases, as noted, the number of positive responses may still be so
low as to make the regression outcomes somewhat tentative in nature.

74

Table 16

Multiple Regressions: the Three Dependent Variables
with the Significant (p( .10) Survey Variables
(For Those Receiving an In—Person Contact)

 

 

 

 

 

Mult. R Simple Overall

R Square R F Signif.
For the Client Card Variable:
Variables in the Equation
Education .307 .094 .307 2.39 .135
Number of persons in household .364 .132 —.247 1.68 .210
For the Newsletter Variable:
Variables in the Equation
Question #37 (accept foodstamps) .372 .138 -.372 3.68 .067
Education .409 .167 .248 2.21 .133
Age .435 .189 —.l48 1.63 .213
For the Service Variable:
Variables in the Equation
Question #7 (how friendly) .323 .104 .323 2.67 .116
Education .376 .141 .284 1.81 .188
Question #6 (center helpful) .443 .196 .233 1.71 .196
Question #8 (how much need) .456 .207 .248 1.31 .300
Memory .464 .216 .200 1.04 .421

Telephone Contact Recipients. For the client card dependent variable,

 

the survey variables of question #5, question #8, number of persons in

the household, education, question #48 (i.e. whether the person had any
complaints or compliments toward social service agencies in general, rated
on a three point scale), the social service knowledge scale, and the prob-
lems scale, were entered into the regression equation. Together these
seven variables accounted for 67% of the variance in the dependent
variable (p=.020). The problems scale alone accounted for 24% of the
variance (p=.023). The first four variables (problems, social service

knowledge, education, and number of persons in the household) accounted

75

for 61% of the variance (p=.003).

For the dependent variable of receiving a service, the survey vari-
ables of question #6, question #8, education, question #37, question #54
(i.e. whether or not the person feels that people tend to think poorly
about those who accept help from the government), and the problems scale
were read into the regression equation. Together these six variables
accounted for 52% of the variance in the dependent variable (p=.074).
Question #8 alone accounted for 37% of the variance (p=.003). The first
three variables (question #8, education, and question #37) accounted for
49% of the variance (p=.008).

Table 17
Multiple Regressions: the Three Dependent Variables

with the Significant (p( .10) Survey Variables
(For Those Receiving a Telephone Contact)

 

 

 

Mult. R Simple Overall
R Square R F Signif.
For the Client Card Variable:
Variables in the Equation
"Problems" scale .494 .244 .494 6.14 .023
Number of persons in household .589 .347 -.231 4.79 .022
Social service knowledge scale .699 .489 .371 5.43 .008
Education .781 .611 -.308 6.27 .003
Question #8 (how much need) .803 .644 .257 5.43 .005
Question #48 (attitude toward
social services) .815 .664 .220 4.61 .009
Question #5 (how clear) .816 .666 .305 3.71 .020

For the Newsletter Variable:
Variables in the Equation

 

(No regression was performed due to an insufficient number
of positive responders.)

76

Table 17 (cont'd.)

 

 

 

Mult. R Simple Overall
R Square R F Signif.

For the Service Variable:
Variables in the Equation
Question #8 (how much need) .611 .373 .611 11.29 .003
Education .679 .461 —.389 7.70 .004
Question #37 (accept food—

stamps) .701 .491 .284 5.46 .008
Question #6 (center helpful) .708 .501 .492 4.01 .019
"Problems" scale .714 .510 .349 3.12 .040
Question #54 (welfare stigma) .719 .517 .277 2.49 .074

As can be noted in the above table, there were not enough positive
responders (n=l) to permit any analyses for the dependent variable of
signing up for the newsletter.

Mail Reeipients (Informational and Personal Letter Groups Combined).

 

For the client card dependent variable, the survey variables of question
#8 and the problems scale were read into the regression equation. Ques—
tion #8 accounted for 12% of the variance in the dependent variable
(p=.069). Since the problems scale is highly correlated with question #8
(r=.47), the additional variance picked up by the problems scale in this
equation was negligible.

For the newsletter dependent variable, the survey variables of age,
health, memory, and question #48 were read into the equation. Together
these four variables accounted for only 6% of the variance (p=.821), with
the variable of health alone accounting for 3% of the variance (p=.345).

For the dependent variable of receiving a service, the survey vari-
ables of race, question #8, question #50 (i.e. when the person first had
contact with the Department of Social Services), and the problems scale

were read into the regression equation. Together these four variables

77

accounted for 20% of the variance (p=.251), with question #8 accounting
for 12% of the variance (p=.069). (See Table 18.)
Table 18
Multiple Regressions: the Three Dependent Variables

with the Significant (p< .10) Survey Variables
(For Those Receiving a Letter Contact)

 

 

 

 

Mult. R Simple Overall
R Square R F Signif.

For the Client Card Variable:
Variables in the Equation
Question #8 (how much need) .348 .121 .348 3.58 .069
"Problems" scale .349 .122 .215 1.73 .198
For the Newsletter Variable:
Variables in the Equation
Health .185 .034 .185 .92 .345
Age .225 .051 .143 .67 .522
Question #38 (attitude toward

social services) .244 .059 .273 .51 .682
Memory .249 .062 .086 .39 .821
For the Service Variable:
Variables in the Equation
Question #8 .348 .121 .348 3.58 .069
Question #50 (how long a D88

service recipient) .433 .187 .215 2.88 .075
Race .447 .200 -.219 1.99 .142
"Problems" scale .448 .201 .192 1.46 .251

In summary, performing the regression analyses within each group of
contact recipients had some effect on which variables were related to the
outcome measures and, as expected, resulted in accounting for a higher
percentage of variance. However, with the exception of the telephone con—
tact recipients, the results of the regression analyses still failed to
achieve statistical significance. More about the interpretation of these

findings will be presented in the "Discussion" section.

78

Reasons Given by Interviewees for Response or
Non-Response to the Outreach Contact

 

 

A final area of the questionnaire, intended to provide information
about why some needy elderly respond to an outreach contact while others
do not, was a set of questions asking positive responders why they did
respond and non-responders why they did not. (See items #9-12 of the
questionnaire.) For those categorical items, the following frequencies
of response were obtained.

Of those who had a positive response in terms of the client card
registration variable, 66% said they did so because of the services avail-
able, 3% because of the social activities, 6% out of curiosity about the
center (i.e. wanted to become involved and find out more about the center),

22% because the center "seemed friendly,"

and 3% for other, nonspecified
reasons.

Of those who did not respond positively in terms Of the client card
registration variable, 41% did not because they had no need for the I & R
center or its services, 16% because they weren't sure what the contact
was talking about, 4% felt the center was too far away, and 39% just
never got around to it.

For those who signed up for the newsletter, 56% did so because it
sounded interesting, 22% because they wanted to know more about the I & R
center, and 22% because they wanted to know more about the services avail—
able.

For those who failed to sign up, 16% flatly said it was because they
didn't want it, 18% said they didn't know they could sign up, 2% lost the

phone number to call, and 64% again simply never got around to it.

To see if these reasons for response or non—response varied depending

79

on the type of contact, a chi square analysis was done on each of the four
items. The results of those analyses revealed that there were no signifi-
cant differences between types of contact in terms of reasons for response
or non—response (p=.927 for question #9; p=.360 for question #10; p=.370
for question #11; p=.365 for question #12). Additional comments con-
cerning these items, as well as other survey variables, will be presented

in the "Discussion" section.

CHAPTER IV

DISCUSSION

The primary purpose of this study was to experimentally examine the
relative effectiveness of four alternative modes of conducting outreach
contacts (i.e. informational letter, personal letter, telephone call, and
home visit) with low income elderly persons. "Effectiveness" of the
contacts was operationally defined in terms of three fairly distinct
criteria: whether or not the person registered with the I & R center
(by having a client card filled out); whether or not the person received
a service (directly from the center or by referral); and whether or not
the person called, wrote, or visited the center to sign up for the news—
letter. The relative effectiveness of the four modes of contact was
tested by means of a longitudinal field experiment conducted in a three

county area in southwestern Michigan (as described in Chapter II).

Experimental Outcomes

 

As can be seen from the data presented in the preceding chapter,
these four modes of contact do demonstrate significant differences in
effectiveness in terms of each of the three dependent variables selected.

The Client Card Registration Variable

 

As the data indicate, there was an extremely strong relationship
between type of contact and response or non-response in terms of this
variable. The home visit mode of contact demonstrated a very high (79%)
rate of positive response, followed by telephone contacts (28% positive);

personal letter contacts (11% Positive); informational letter contacts

80

81

(2% positive); and control group (0% positive). Since the similarity of
these five groups on all available variables other than type of contact
has been previously established, one can have a fairly high degree of
confidence in these findings. Hence, efforts at interpretation of these
results should focus on the nature of the contact modes themselves.

One plausible interpretation that draws support from the literature
(Havelock, 1971; Klippel & Sweeny, 1974; etc.) is that the level of in-
teraction between source and target allowed by the contact mode is
related to the type of response obtained; that is, that those modes
allowing direct contact, immediate feedback, 2-way communication, etc.
are the ones that demonstrated the greatest success (i.e. the in—person
and telephone modes). This interpretation draws additional support from
the fact that another key variable identified in the literature, that Of
informational content (Katz & Lazarsfeld, 1955; Barton, 1964; Havelock,
1971), was standardized to as much an extent as possible across the
modes of contact and, thus, should not have greatly influenced the out-
comes observed. Finally, the in—person mode of contact (presumably the
mode allowing the highest level of the above-mentioned attributes of
immediate feedback, direct contact, etc.) was clearly the most effec—
tive. In this respect, the findings for this dependent variable appear
to be quite as would be expected from the literature reviewed in
Chapter I (Eldersweld & Dodge, 1954; Bergner & Yerby, 1968; Havelock,
1971; National Council on Aging, 1972; etc.). The only surprise, per-
haps, might be at the magnitude of the relationship observed.

The Variable of Receiving a Service

 

Once again, the data demonstrate a strong relationship between type

of contact and response or non—response in terms of this variable.

82

Further, the order of effectiveness of the modes of contact is identical
to that for the client card variable. However, the magnitude of dif-
ference between the types of contact is greatly reduced (i.e. home visit
resulted in 23% positive response; telephone in 19% positive; personal
mail in 11%; informational mail in 4%; and control group in 0%). The
interpretatiOn that seems most plausible in this case is that, although
the "high interaction level” communication modes are still more effective,
the relative advantage attributable to those modes has diminished.

One possible explanation for the change in relative level of success
may involve the nature of the responses required by each of the dependent
variables. For example, the client card registration variable involves
a response that is not particularly threatening but, on the other hand,
is not of high saliency in terms of benefits (i.e. it involves a general
assessment of one's situation, including one's needs, and functions as
sort of an initial step toward "membership" with the I & R center). In
these respects, it is a decision that requires a relatively low amount of
personal investment. Hence, one might expect that the nature of the mode
of contact itself might have a proportionately greater degree of impact
on the type of response Obtained (perhaps more of an impact than does
the particular situation of the individual). In fact, this appears to
be what has occurred. Those modes of contact allowing the greatest level
of persuasive interaction, and also being most facilitative to a "passive”
positive response (i.e. in-person and telephone modes), demonstrated the
highest level of positive response. (As will be discussed shortly, the
particular characteristics of the individual contacted did not appear to
have much impact in terms of response or non-response to this variable.)

In contrast, a positive response in terms of requesting and receiving

83

a service tends to be both a threatening situation (as discussed in
Chapter 1; Alexander & Podair, 1969; Gatz, 1974; etc.) and of more direct
benefit in terms of client needs, than is a positive response to the
client card registration variable. For these reasons, a decision to re-
quest and/or accept a service very likely requires a higher degree of
personal investment than does the client card response. Thus, one would
expect that the particular mode of contact might have a relatively
smaller impact, and the person's particular situation (needs, problems,
etc.), a relatively larger impact than was true for the client card
dependent variable.

Once again, the results reveal a pattern that makes this explanation
plausible. The margin of superiority between contact modes is indeed
much smaller than for the client card variable. (As will also be dis-
cussed shortly, the particular characteristics of the individual seem to
play a much larger role in terms of response and non-response for the
service dependent variable.)

Finally, in contrast to the client card variable (for which, by
definition, no one had previously registered so everyone was "eligible"),
one cannot assume that everyone in the sample had an existing need/
eligibility for a service that the I & R center could provide. Thus,
there may have been a ”ceiling effect" which could have held down the
absolute number of positive responses available to each mode of contact.
Hence, this factor may also have contributed to the smaller margin of
superiority demonstrated by the in-person and telephone modes for this
variable. Nevertheless, assuming equally distributed need (and the
findings of the follow-up survey give no reason to doubt that assumption),

one can see that the results show an obvious and statistically significant

84

order of effectiveness among the modes of contact in terms of this
dependent variable.

The Newsletter Dependent Variable

 

For this dependent variable, the effects of type of contact were
even less pronounced, although still statistically significant. The home
visit mode of contact resulted in a 12% positive response, followed by
personal mail contact (8% positive), telephone contact (2% positive),
informational mail contact (0% positive), and the control group (0%
positive). Perhaps the most striking feature of these results, as con-
trasted with the other two dependent variables, is the overall low level
of positive response. In addition, it is worthy Of note that the personal
letter mode surpassed the telephone mode in effectiveness.

Each of these findings may be explained to some extent by the nature
of this dependent variable and the response it required. The newsletter
dependent variable was included in this experiment primarily for two
reasons. First, by requiring all responses to this variable to be time-
delayed (i.e. all persons were required to recontact the center on their
own in order to sign up for the newsletter), it was intended that the
"immediacy" advantage of the in-person and telephone contact modes would
be removed. Second, it was felt that this variable might represent a
less-threatening alternative to asking for a service or registering with
the I & R center.

The results seem to indicate that this strategy succeeded in one
instance but may not have in the other. The newsletter variable does
appear to have removed some of the advantage of the in—person mode of con—
tact and much of that of the telephone mode. Although the in—person mode

was still the most successful, its margin of superiority was greatly

85

reduced. The telephone contact effectiveness was so low it was actually
surpassed by the personal mail mode. Hence, in this respect the news-
letter variable performed as expected and resulted in a lower level of
response for those modes previously enjoying an immediate response advan-
tage. However, in terms of the intention of producing a less threatening
response alternative and thus encouraging more responses from all groups,
particularly the two letter groups, the results seem to be negative.

Indeed, the overall response level for all groups was extremely low.
To what extent this was due to a failure to reduce the level of perceived
threat is not known (although the reasons given for non-response seem to
make this unlikely, i.e. 64% of non-responders indicated that they just
"never got around to it"). What does seem to be a plausible explanation
is that, in general, the perceived level of benefit was not enough to
provide the motivation necessary to produce the active response required
for this dependent variable. (More about the question of "active"
response will be presented later.) At any rate, the in-person mode of
contact remained superior in terms of this dependent variable. This
would seem to demonstrate that some advantages of the face-to-face con—
tact persisted in spite of the delayed response requirement.

As for the improved relative effectiveness of the personal letter
contact, it should be kept in mind that this particular mode of contact
emphasized a friendly, informal style and content. In this sense, the
outreach contact itself was quite similar to the "product" it was
attempting to promote (i.e. the newsletter). Therefore, it is not sur—
prising that those to whom the outreach letter itself appealed might tend

to be attracted to the possibility of receiving the newsletter.

86

The Informational Mail vs. Personal Mail Contacts

 

For most of the remainder of this discussion, the primary distinc-
tions referred to in terms of types of contact will be between the major
conceptual divisions of in—person, telephone, and mail contacts. However,
it should be emphasized that in spite of the overall low level of response
to mail contacts in general, the results reveal that the "personal" mail
contact format consistently produced a higher level of positive response.
Although the numbers of positive responders were too small to allow
significance tests, the personal mail mode clearly had a higher level of
response than the informational level mode on all three dependent
variables.

In addition, in terms of the client ratings of the outreach modes
they received, personal letter recipients rated the center as likely to
be more friendly (though not quite significantly, p=.l90) and rated the
outreach contact as being much clearer (p=.004) than did the informa-
tional mail recipients. Further, these differences in performance and
in client ratings occurred in spite of the fact that there was virtually
no difference between the groups in terms of their perception of the
extent to which the center would be a helpful place or their rating of the
extent of their own need for I & R services. (In fact, the informational
letter group rated the contact slightly, though non-significantly, higher
on each of those two variables.)

It is uncertain whether the superior performance of the personal
letter contact is due to the more friendly, informal tone of the letter
(as Klippel & Sweeny, 1974, might suggest); the presence of the person-
alized invitation and the names of others provided as a "referent group"

(as Lionberger, 1960, or Bergner & Yerby, 1968, might suggest); or to some

87

other variable. What is clear, however, is that the use of the combination
of strategies composing the "personal" letter contact mode did produce ob-
servable differences on several important criteria.

Results of Attempts at Identifying
Predictors of Positive Response

 

 

The experimental results in terms of the relative effectiveness of
the various modes of contact are clearly demonstrated by the data. Having
observed these results, one can now turn to the central question of what
led certain needy elderly to respond to an outreach contact while others
did not. To examine that question, possible predictors that have been
identified will be discussed for each of the three dependent variables used
in this experiment.

The Client Card Registration Variable: Who Responds?

 

The results outlined in Chapter III reveal that from the total pool
of persons contacted, by far the most important variable in determining
who will positively respond is the type of contact the person received.
This may be due in part to the more "passive" nature of the response to
this variable, as discussed previously, or perhaps to other qualities of
the contact such as its immediacy of feedback. At any rate, in-person
contact recipients are much more likely to be positive responders,
followed by telephone contact recipients, personal mail recipients, and
informational mail recipients, in an almost geometrically descending rate.
(Those in the control group had no positive responses at all.) Other
than type of contact, several variables demonstrated slight, but non-
significant relationships as possible predictors. To summarize these:
positive responders tended to have more problems as measured by the prob—

lems scale; tended to more clearly understand the outreach contact

88

(regardless of mode); tended to be female and slightly younger than the
overall average; tended to have fewer living children; and, finally,
tended to have fewer persons living in the immediate household. Once
again, however, these variables had nowhere near the impact that type of
contact did in terms of their relationship to response or non—response.

Within each type of contact, the relationships tend to become some—
what more selective. For those receiving a home visit, only two variables
appear to have any relationship to response or non-response. Furthermore,
regression analysis reveals they are not significant. This lack of iden-
tifiable predictors may likely be due to the fact that the home visit
mode of contact is dominant enough to overcome most individual variables
that might otherwise determine response or non-response. At any rate, as
for tendencies, positive client card responders in this contact category
tend to be somewhat better educated and tend to have fewer persons
residing in the immediate household. (One could speculate that this
might indicate such features as companionship, recreation and/or cultural
activities, etc. might be acting as motivators for positive response for
this group.)

For those receiving a mail contact (informational and personal letter
groups were combined to provide a larger number of positive responders),
there were again only two related variables that emerged. However, these
variables indicate that a much different process was occurring within
this contact group. The two variables were question #8 (i.e. how much
need they had for the I & R center services) and the score on the problems
scale, each of which was positively related to the dependent variable
(i.e. higher need and more problems result in greater likelihood of

positive response). In contrast to the home visit group, one might

89

speculate in this case that the salient motivator for positive response
for mail recipients might have been the services available through the
I & R center.

For those receiving a telephone contact, several variables appeared
which were significantly related to type of response. Positive respon—
ders scored higher on the problems scale and also on the social services
knowledge scale. Positive responders were also lesser educated and had
fewer persons residing in the household. Finally, positive responders
tended to perceive more need for the I & R center services, tended not to
feel that persons thought poorly of those that receive help from the
government, and tended to more clearly understand what the outreach con-
tact was about (i.e. question #5). Here the interpretation is somewhat
more complicated than for the other contact types. It seems clear that
the positive responders perceive themselves as having problems/needs, and
this may be acting as a motivator. In this respect, the positive respon—
ders to the telephone contact resemble their counterparts in the mail
contact group. Also, there appears to be a familiarity with and/or
knowledge of social services and lack of concern over any stigma
resulting from service use.

On the other hand, the variable of fewer persons in the household
once again appears in a relationship similar to that in the home visit
group. This Opens the possibility of companionship related variables as
motivators for positive response. In view of these multiple relationships,
it seems as though the telephone contact mode may lie somewhere on a con-
tinuum between home visits and mail contacts and that variables active
predominantly in one or the other of those two modes are combined to some

extent in helping to determine the response to telephone contact (although

90

such an interpretation is necessarily speculative at this point). Finally,
the variable of education, for some unexplained reason, demonstrates a
negative relationship to positive response (almost exactly the reverse of
its relationship in the home visit group).

In summary, it appears evident that positive response, in terms of
the client card registration variable, is a phenomenon that is dependent
in large part upon which type of contact is being employed. In addition,
the particular individual variables that relate to positive response vary
depending on the type of contact.

The Variable of Receiving a Service or Not: Who Receives?

 

In contrast to the client card registration variable, "positive
response" in terms of this outcome measure is much less dependent on the
type of contact the person received (although a fairly strong significant
relationship between those two variables still exists). Instead, features
of the individual's situation and the individual's perception of the out-
reach contact seem to play a slightly larger role in predicting response
or non-response for this variable. Not surprisingly, the person's
”problems score" and perceived need for the I & R center services are the
dominant variables that emerge as significantly related to positive
response. Following these and of lesser importance than the type of con-
tact itself are the variables of: extent to which the I & R center
seemed like it would be helpful; and, extent to which the center seemed
like it would be a friendly place (i.e. questions #6 and #7 rating the
original outreach contact). To summarize, from the total sample, the
major identifying characteristics of positive responders were: the
presence of a high level of problems as rated by the problems scale; a

high level of self-rated need for I & R services; a perception of the

91

I & R center as a place that would be helpful; and a perception of the
I & R center as a place that would be friendly.

Within each type of contact, the same core of significant predictors
remained, but in each case a couple other variables also joined in the
relationship. For the in—person contact recipients, positive responders
once again perceived the center as a friendly place and as someplace that
would be helpful. Also, they had a higher self-rating of need for I & R
center services. (In these respects, positive responders to this
dependent variable present a notably different picture than in-person
positive responders to the client card variable.) Finally, positive
responders tended to be better educated and have a better memory rating.
Overall, however, the regression analysis for this dependent variable
within the in—person contact group revealed that these relationships did
not achieve statistical significance. (This seems to suggest that, even
for this dependent variable, the in—person mode of contact itself may
have had more to do with determining response or non-response than the
situational and individual characteristics of the persons contacted.)

For telephone contact recipients, in contrast, the regression
analysis did indeed produce significant results. This was due primarily
to the relationship between question #8 (i.e. self-rating of the extent
of need for I & R services) and the dependent variable. In addition to
having a high level of need for I & R services, positive responders
tended to perceive that the center would be a helpful place, tended to
be lesser educated and to have a high amount of problems and, finally,
tended to have no objection to accepting help from the government,
feeling that people did not think poorly of those who did. It seems

clear that, for the telephone contact, personal characteristics

92

(particularly level of perceived need) did play an important role in
distinguishing positive responders from non—responders in terms of this
dependent variable.

For mail contact recipients, the client's self—rating of need for
the I & R center services was once again by far the most dominant
variable (though not quite significant as a predictor). In addition to
a high level of perceived need, the positive responders also tended to
be black, to be more recently acquainted with the Department of Social
Services, and to have a higher score on the problems scale. Overall,
however, the results of the regression analyses were not statistically
significant.

In summary, it appears as though positive response in terms of the
"receipt of service" variable is still somewhat dependent on the type of
contact received. However, to a much greater extent than with the client
card variable, response or non-response does seem to relate to the situa—
tional and individual characteristics of the persons contacted. (This
is particularly true in terms of the level of need for the I & R center
services that the person perceives.)

The Newsletter Dependent Variable: Who Signs Up?

 

Of the three dependent variables utilized in this experiment, the
newsletter variable has the least amount of relationship between type of
contact and type of response. As discussed previously, this was somewhat
as expected due to the nature of the response required. (However, it
should be noted that a statistically significant relationship between
type of contact and response or non-response still exists for this depen-
dent variable.) In addition, the newsletter variable also demonstrates

the least amount of relationship between individual and situational

93

variables and response or non-response. Hence, in terms of percent of
variance accounted for, the results appear to reveal little about who
responds to this dependent variable and why. (This may in part be due
to the difficulty of examining such relationships when there were so few
positive responders.) Nevertheless, those non-significant relationships
that are revealed are quite interesting. Positive responders tended to
be better educated, to have better memory ratings, and to have better
understood the outreach contact (i.e. question #5). They also tended to
have perceived the I & R center as a friendly place, to have better
health ratings, and to be more recently acquainted with the Department of
Social Services. Finally, in contrast to positive responders to the
service variable, they tended to feel that people do think poorly of
those who accept help from the government. In terms of the total sample,
these results seem to suggest that there may be some fundamental dif-
ferences between newsletter variable responders and service variable
responders, particularly in regard to their need situation and their
perception Of social services.

Within type of contact, some further interesting differences appear.
For those who received an in-person contact, positive responders tended
to be better educated (similar to in-person responders in terms of the
client card variable) and slightly younger than non-responders. But some—
what surprisingly, the dominant relationship (almost significant in the
regression) was that positive responders to the newsletter stated that
they would not accept help from the government, e.g. foodstamps (i.e.
questions #37). Again, in this respect the newsletter respondents seem
to demonstrate some fundamental differences from the service variable

responders.

94

For the mail contact recipients,4 the relationships are much dif-
ferent. Here positive responders tended to have good things to say about
social services (i.e. question #48), tended to be Older and in better
health, and tended to have a somewhat better memory rating than non—
responders. Although they appear to be somewhat different than service
variable responders (particularly in that needs and problems are not
emphasized), there does not seem to be as dramatic a difference as was
evident for the in—person contact recipients.

In summary, although the results are not conclusive enough to make
any definitive statements, it seems as though some curious interactions
occurred between type of contact, purpose of contact (i.e. dependent
variable examined) and type of respondent.

In general, the in-person mode of contact demonstrated the highest
level of effectiveness for each purpose of contact (i.e. for each of the
three dependent variables). However, although type of contact had a very
large impact on response in terms of the client card variable, it had a
smaller impact on the response in terms of the service variable, and even
less for the newsletter variable. Further, the results reveal that, for
each of the dependent variables, the characteristics of those who respond
positively tend to differ. In general, those who respond positively to
the service variable are distinguished by having a higher "problems"
level and by having more self-expressed need for the I & R center ser—
vices. In contrast, those who respond positively to the newsletter, and
to a lesser extent the client card variable, tend to be distinguished by

being younger, in better health, better educated, etc. Finally, however,

 

4The telephone mode of contact could not be examined in terms of this
dependent variable due to an insufficient number of positive responders.

95

even those generalizations of respondent characteristics within purpose
of contact tend to vary according to the type of contact utilized.

In conclusion, it appears as though some interesting relationships
have been identified which may be of considerable use in understanding
the larger question of why some needy elderly respond to an outreach
contact while others do not (as well as in understanding the more complex
question of who responds to what type of contact for what purpose).
However, it must be kept in mind that many of the results of the predic-
tion attempts did not achieve statistical significance, and hence, should
be interpreted with care. Similarly, the direct experimental results,
although of solid statistical significance, should properly be regarded
as tentative until further replications increase confidence in those

findings.

Implications for Policy

 

In spite of the necessarily tentative nature of the inferences drawn
from this study, it is felt that the results obtained from such a large-
scale, fairly well-controlled experiment can indeed be of use to policy-
makers, administrators and others involved in fields related to service
provision for the elderly. In particular, the most obvious usefulness of
the results of the study would be in deciding what mode(s) of contact to
employ in planning an outreach strategy. Toward this end, the actual
results in terms of absolute and relative effectiveness of the various
modes of contact have been presented and discussed previously. An addi-
tional input of information that one might want to consider, however, is

an estimate of relative costs incurred through the use of each method.

96

Relative Costs

 

In addition to effectiveness information, it may also be useful to
know the relative costs of the three alternative modes of contact. The
costs per unit of contact Of each type incurred during this project, as
well as the relative costs per unit of successful response, are esti-
mated in Appendix L.

As one might reasonably expect, a letter is shown to be the least
expensive method per unit, followed by a telephone call and then by a
home visit. However, to achieve an estimate of relative costs if each
method were to be actually used, the hypothetical costs of completing one
contact must be multiplied by the effectiveness rate (i.e. 79% effective-
ness means one successful registration every 1.3 contacts, etc.). This
means of estimating relative costs was used for each of the three out-
come measures, with some rather interesting results.

The Client Card Registration Variable. The estimated cost per suc-

 

cessful registration was found to be almost identical for all three
modes. (The informational letter mode was not included due to its very
low success rate.) The lower cost of the telephone and personal letter
modes was just enough to make up for their lesser effectiveness.

The Variable of Receiving a Service. The personal letter mode

 

actually demonstrated the lowest cost per person serviced. The telephone
mode was close behind, costing approximately 30% more per person served.
Finally, the mode of home visiting was the most expensive, costing almost
three times as much as the personal letter mode per person served and
more than twice as much as the telephone mode.

The Variable of Signing Up for the Newsletter. The personal letter

 

mode was again the least expensive, followed by the home visit mode, which

97

was five times as expensive. Finally, the telephone mode was nine times
as expensive as the letter mode per positive response.

In using such cost information to make policy decisions, however,
one fundamental question must be considered, i.e. which is more important,
the absolute number of positive responses or the unit cost per positive
response? The variable of whether or not the client receives a service
provides probably the best example of this dilemma. Although the letter
and telephone modes are less expensive per unit of service, they also
result in fewer services provided for a given sample of elderly. Depen—
ding On the particular situation involved (i.e. how crucial is the service,
what are the budgetary constraints of the I & R center, etc.), each admin-
istrator would presumably have to make this decision for him/herself.

Use of the Lower Cost Modes of Contact

 

After examining the cost data, one conclusion does appear to be
warranted, i.e. that in certain circumstances (as Eldersveld & Dodge,
1954; Barton, 1964; Havelock, 1971; etc. seem to suggest) a personal
letter mode of contact can indeed be competitive with other more expensive
modes of contact (particularly if one's outreach budget is limited). The
areas where this would seem to be most true would be in non—crucial ser-
vice areas, such as in promoting newsletter sign-ups or in providing
simple information (such as the name, location, phone number and brief

description of the local senior citizen center). Indeed, the results show

 

5Looking at the cost results for this dependent variable, one realizes
they are unimportant in a literal sense since it would have been much less
costly to simply mail out the newsletter every month to everyone on the
list. However, this dependent variable is intended to be regarded as il-
lustrative of the relative costs of seeking to persuade elderly persons
to make some active response similar to calling and signing up for the
newsletter. In that sense, these results are informative.

98

that, in spite of the low rate of actual response, 57% of those sent a
personal letter remembered the letter and associated it with the senior
center at the time of the follow-up survey (at least one month subsequent
to the mailing).

Similarly, the telephone mode of contact may also be competitive
with the in-person mode for some purposes. The survey also revealed that
78% of telephone contact recipients remembered the phone call and
associated it with the center at the time of follow-up. Hence, it does
appear that the more limited contact, low-cost methods of outreach are
successful at least in terms of the transmission of fairly simple infor-
mation.

In addition, the results obtained in terms of client rating of the
outreach contacts reveal that personal mail and telephone contact recipi—
ents were not significantly lower than home visit recipients in their
rating of the anticipated usefulness or friendliness of the I & R center
or in the clarity of the contact itself. (In fact, personal letter
recipients rated these variables somewhat higher than home visit recipi—
ents. See Appendix M.) Thus, in addition to successfully conveying
information, it appears that these lower cost modes of contact can also
instill in their recipients an impression of the I & R center that is as
positive as that fostered by the in-person contact.

Finally, however, certain limitations of these lower cost modes of
contact must be pointed out. For example, in order for the methods of
telephone contact or personal letter to be used, addresses and/or phone
numbers must be available. In many circumstances this limitation might

severely restrict the applicability of these modes of contact. Other

99

important factors to consider include the ability of the potential client
to read a letter or hear a telephone call. Although no data were
specifically gathered on the ability to hear, only one incident (out of
approximately 40 contacts) was reported to the experimenter where an
inability to hear on the part of the recipient resulted in the contact
attempt failing. Thus, it would appear that hearing problems would not
present a major obstacle to the use of phone contacts. As for the ability
to read, the follow-up survey did specifically gather this information
from the entire sample. The results showed that, according to verbal
reports, approximately 20% of the sample could not read a simple letter
due to lack of education or poor vision. However, this figure is
moderated by the fact that, of that 20%, two-thirds did have someone to
regularly read their mail to them. Hence, in terms of the inability to
read, a letter contact would be definitely inappropriate for only about
7% of the total sample.

The No-Contact Control Group

 

In discussing the policy implications of this study, some mention
also needs to be made of the results for the no-contact control group.
The lack of any positive responses on any of the three dependent variables
for this group provides a stark contrast to the four groups receiving the
various types of contact. Indeed, this lack of "success" for the control
group should be underscored by stating that none of the centers received
any recorded contact whatsoever from any of the control group members
during the monitoring phase of the project. Yet, as the follow—up survey
revealed, this group of persons did not differ from the other groups in

terms of any of the observed personal or situational characteristics.

100

Hence, one is left with the inescapable conclusion that it was the out-
reach contacts that were responsible for the differences in performance
in terms of the dependent variables.

The Mass Media. Implicit in the results portrayed by the control

 

group is the failure of the mass media to produce positive responses in
these domains. Although it must be emphasized that the media were not
experimentally analyzed in this study, there were periodic media cam-
paigns (newspaper, radio and television), occurring throughout this
three county area (and had been for months, and in some cases years,
previously). Yet none of the control group responded to these campaigns
and contacted the center. Further, the overall impact of the media in
terms of awareness of the centers was small. Only about 20% of the con-
trol group said they had heard of the I & R center through the media.
(Incidentally, this was the same percentage cited by the total sample,
thus indicating, as one would expect from having used a randomization
procedure, that there was no differential impact in terms of this
variable.)

In defense of the mass media, one should recall (as discussed in
Chapter II) that this particular target population was a very resistent
one. Indeed, various center personnel stated that media campaigns have
at times resulted in many new persons contacting the center. Neverthe—
less, the results with this target population seem to demonstrate, as
Rush and Kent (1974) suggested, that the mass media may not be a
particularly effective strategy for reaching and influencing into

action a target group of low-income, relatively needy elderly.6

 

6The possibility of there being an interaction effect brought about
by having heard of the center through the media and then receiving an

101

Other Miscellaneous Sources of Information. Finally, it should also

 

be noted that various other miscellaneous sources of information were
active within this three county area (i.e. church announcements, bake
sales, raffles, informal word-Of—mouth contacts, etc.). By inference,
these various activities also produced no positive responses among the
control group members (even though one can assume they must have produced
positive responses from the population at large in the past). Once again,
a plausible explanation for the lack of response seems to be the resis—
tent nature of the target population of this experiment.

The Use of the In—Person (Home Visit) Mode

 

Having discussed the use of the lower cost modes of contact (i.e.
telephone and letter) as well as the option of no direct outreach con—
tact at all, this topic will conclude with a discussion of and, in a
sense a recommendation for, the in-person mode of contact.

It is true that the lower cost modes of contact seem well-suited
for many purposes, as discussed previously. However, for the more
crucial functions of an I & R type agency (such as the assessment of
needs, the advocacy of services, etc.), it seems that the higher ab-
solute number of persons reached and served would be more important than
the per unit cost of those persons served. Whenever this is the case,
the in-person mode of contact would be recommended.

The "Passive" Recipient. The distinction between types of contact

 

seems to be an important one for those in this target population, given

 

outreach contact was also examined. This was done by comparing the
responses of those who had heard of the I & R center through the media
with those who had not. The results showed no significant differences
between the two groups (p=.520 for the client card variable; p=.992 for
,the receipt of service variable; and p=.245 for the newsletter variable).

102

their apparent tendency to avoid making positive, active requests for
service, information, assistance, etc. A couple of examples should
serve to illustrate this apparent reluctance. First, an interesting
situation emerged with the newsletter dependent variable. As originally
structured (i.e. with the delayed response requirement), only 6% of the
total sample made the required effort and signed up for the newsletter.
Yet in the follow-up survey, only 16% of the non—responders indicated
that they did not sign up because they did not want the newsletter.
Non-responders were given the chance to sign up at that point, and a
full 76% of those not previously signed up did so right then. Of course,
the newsletter could be regarded as a low priority item and, as men-
tioned previously, the perceived rewards may not have been strong

enough to encourage the active response required during the project.
Still, the discrepancy is rather startling between the number of per-
sons who positively responded On their own initiative vs. the number who
positively responded in a situation allowing immediate response with a
minimum of effort and threat.

A second example involves a more serious service situation, as
opposed to the possibly entertaining or less important newsletter. The
survey revealed that 87% of those in the sample were Medicaid recipients.
When asked how they came to be signed up for Medicaid, only 12% of the
respondents indicated that they had signed up through their own
initiative. The remainder were signed up at the initiative of rela-
tives, friends, a doctor, or some social service agency (42%). Once
again, the image that emerges is one of a fairly passive recipient
rather than someone actively seeking the services/benefits to which s/he

is eligible and entitled.

103

The reasons for this lack of self—advocacy may be numerous, possibly
including some of those discussed in Chapter I, i.e. general fear of or
avoidance of unnecessary outside contact, or the denial of any problem
(Gaitz, 1974); general apathy toward the service; continued misinforma-

' etc.); and the desire to

tion of many kinds (such as "I'm not eligible,‘
avoid "welfare" or the stigma of welfare use (Alexander & Podair, 1969).
Not much information was gathered in this study relating to these points,
with the exception of the latter concern about "welfare." In response

to the survey, only 14% of the sample flatly stated they would not

accept help from the government (such as foodstamps). On the other hand,
only one-third of the total sample could agree with the statement that
people in general do not think poorly of those who accept help from the
government. Hence, although the adamant refusal to accept help is fairly
uncommon, a more widespread underlying stigma of welfare use does seem

to exist and may be acting to minimize self—advocacy in terms of the use
of services or assistance. At any rate, whatever the underlying reasons,
it does seem fairly clear that this target population as a whole demon-
strates a passive/receptive tendency rather than an active/self-advocate
approach to the use of social services. If this is indeed the case, it
is not surprising that the method of home visiting produced the highest
levels of response.

In conclusion, the in—person mode of contact has been demonstrated
to be superior in terms of all three of the dependent variables used in
this experiment. Although the lower cost modes of contact appear to be
quite useful in some situations, whenever somewhat threatening or complex
tasks (such as registering with the senior center, signing up for a ser-

vice, etc.) are required, then a home visit seems to be the recommended

104

mode of contact for the reasons discussed above. This may be due to a
number of factors, including possibly the ability of the outreach worker
to allay fears and provide immediate reassurance, the increased per-
suasiveness of a face-to—face contact, or the ability of the outreach
worker to immediately assist in the completion Of necessary tasks. What-
ever the causal factors, the results of this study seem to support the
conclusions of the National Council on Aging (1972) in that the mode of
home visiting was indeed found to be the most effective method of con—
ducting outreach to this target group of low income relatively needy
elderly.

Conclusion

 

After examining and discussing the results of this experiment, it
appears as though some general statements can be made regarding several
aspects of the study. To begin, in terms of the primary focus of this
experiment (i.e. the relative effectiveness of the five alternative out—
reach strategies), several tentative conclusions seem apprOpriate. First,
and simply stated, the in-person mode of contact appears to be the most
effective means of conducting outreach. Although, as previously dis—
cussed, the relative advantage of the in-person mode seems to vary some-
what depending on the purpose of the contact, it has demonstrated a
superiority across all three dependent variables used in this experiment.

Second, although their overall effectiveness was lower than the in-
person mode, the lower cost modes of telephone and personal letter appear
to have potential usefulness in certain circumstances (i.e. where out-
reach budgets are limited and/or where the goals of outreach are non-
threatening and non-complex, e.g. the transmission of fairly simple

information, the generation of name awareness, etc.).

105

Third, when a letter strategy is selected as a mode to be utilized,
it appears that a more "personal,” ”friendly" format, including possibly
the names of other service users in a personalized "invitation," will be
more effective than a more straightforward, information—oriented format
(at least when the actual informational content is the same).

Finally, it is interesting to note that the expressed opinions of the
elderly persons in the control group (as to what type of contact they
would have preferred to receive had they received one) paralleled the
effectiveness outcome results. The majority (56%) of the persons pre-
ferred the in—person mode of contact, followed by 28% who chose a tele—
phone contact, and 16% who picked a letter contact.

In terms of the other major area of interest in this experiment (i.e.
the question of why certain low—income elderly respond positively to an
outreach contact while others do not), it appears that once again some
general statements can be put forward. First, it appears that the type
Of contact a person receives may often have an impact over and above the
personal and situational characteristics of the person receiving a con-
tact. Indeed, the effects of type of contact were strongly demonstrated
and remained fairly constant in terms of order, across all three depen-
dent variables used in this study. (However, it must be acknowledged
that the effect of type of contact in determining response or non-
response does seem to vary a great deal depending on the purpose of con-
tact one considers as well as the personal and situational characteris-
tics of the individual.)

Second, the purpose for which the outreach contact is intended also
appears to have an impact on response or non-response. For example, the

results suggest that a purpose requiring an active time-delayed,

106

self~initiated response (i.e. the newsletter variable in this case) will
produce a low level of positive response among a given target population

of low—income elderly. In contrast, a similar purpose requiring only a
somewhat passive and immediate response will generate a much higher level
of positive response (as was demonstrated by the greatly increased response
to the newsletter in the follow-up survey).

Third, the particular personal and situational characteristics of
the contact recipient also seem to be related to response or non-
response. Without going into full detail, it appears that the extent of
self—perceived need and self—perceived problems in various areas of life
are generally the most important determinants of response or non-response.
Also, the person's reactions to the outreach contact (in terms of under-
standing it, perceiving the center as helpful, and perceiving the center
as friendly) seem to be important factors.

Finally, within certain types of contact, and for certain purposes,
such variables as age, memory rating, education, and attitudes toward
social services also appear to relate to client response.

In summary, the results of this experiment suggest that the type of
contact, the purpose of the contact and the personal/situational charac—
teristics of the target, all play a role in determining response or non-
response to outreach. Although these variables interact to some extent,
the most successful mode of outreach appears to be the home visit. The
"typical" positive responder (to the extent that such a generalization
is possible) tends to be someone who has self—perceived need, who under-
stands the meaning of the outreach contact, and who perceives the I & R

center as potentially friendly and helpful.

APPENDICES

APPENDIX A

DEMOGRAPHIC CHARACTERISTICS OF

THE PROJECT SAMPLE

Demographic Characteristics Of the Project Sample

Table A—1
Mean age 75.4 years (range: 65 to 91 years)
Sexual distribution 75% female; 25% male
Racial distribution 78% white; 22% black

Average number of living children 2.9

Marital status Presently married 27%
Single (never married) 5%
Widowed 64%
Divorced/separated 4%
Residential status Lives with spouse 25%
Lives with child 19%
Lives alone 42%
Other 14%

Average number of persons

living in household 1.9
Education Less than 8th grade 34%
Completed 8th grade 26%
Some high school 22%
Completed high school 14%
Some college 4%

107

APPENDIX B

EXAMPLE CONTACT FORMATS

Example Contact Formats
(Informational Letter)

Dear

The Central County Center for Senior Citizens is a multirpurpose center
serving the needs of individuals aged 60+ (and their spouse) that live
in Royalton, Sodus, Pipestone, Baroda, Oronoko and Berrien Townships.

We began with 23 people attending a community event sponsored by the
Berrien Springs Jaycees in October of 1973. In July of 1974 we opened
our doors as a federally funded drop-in center for senior citizens
sponsored by the Berrien County Council on Aging. In July of 1975 we
decided to go on our own and became a private, non—profit incorporation
with a board of directors comprised of senior citizens and their spouses.
We now provide services to over 1000 people, with about 400 being active
members in our organization.

Services we provide include:

SIRS — (Senior Information and Referral Service) a county-wide program
offering information to senior citizens about services available,
employment, job training, volunteer positions, health care, etc. When
information alone is not enough, we can arrange a direct contact with an
agency through a referral.

Nutrition - we serve 40 meals each weekday, providing good nutrition to
senior citizens who may not be eating properly because of loneliness,
or physical or financial inabilities.

Transportation - we provide transportation to senior citizens from 9:30
to 4:30 each weekday. Please call at least 24 hours in advance.

 

Social/Educational - programs provided throughout the year, including
BINGO, cards, crafts, trips, movies, plays and various informational
programs. All activities are planned in advance and notification is
given through our newsletter.

 

If you have questions about the services provided by our center or our
activities, please call the center at 471-2017 from 9:00 to 5:00
weekdays. We can also send our outreach worker out to visit you if you
would like. We will do our best to serve your needs.

Sincerely,

Barbara J. Williams
Director

P.S. Also, whether or not you have any needs at this time, please call
our center at 471-2017 or write us at 608 South Mechanic St., Berrien
Springs if you would like to receive the monthly newsletter, which
describes upcoming activities and includes bits of news for senior
citizens.

108

Example Contact Formats
(Personal Letter)

Dear
Hello!

All of us at the Central County Center for Senior Citizens would like
to invite you to join us...

The Central County Center is a multi-purpose center used by your friends
who are 60 or older and live in the townships of Royalton, Sodus,
Pipestone, Baroda, Oronoko and Berrien.

What is a multi—purpose center? It's a place where older people can go
to either give or receive services such as:

SIRS - (Senior Information and Referral Service) we will try to help you
with problems such as taxes, part—time employment, health care, etc.,
and when we can't, we will put you in touch with someone who can.

Nutrition — we serve 40 great meals every weekday. These meals provide
good nutrition to peOple who may not be eating properly because of
physical or financial inability--or just because it's no fun to eat alone.

Transportation - is provided between the hours of 9:30 and 4:30 weekdays
by a group of volunteer drivers using the center vehicle. We take
doctor appointments as a first priority and require at least 24 hours
notice to plan rides. We will consider any transportation need you have,
however.

 

Social/Education Programs - are the fun part of the center. There are
cards, crafts, trips, movies, plays, etc. which are planned in advance
and listed in our newsletter each month. Cost is kept small, and many
of our programs are free.

 

If you have any questions about our activities or our services, please
feel free to call us at 471-2017. We can also send our outreach worker
out to talk to you about our center if you would like.

We began with 23 people attending an ice-cream social sponsored by the
Berrien Springs Jaycees in October of 1973. In July of 1974 we opened
our doors as a federally funded drop—in center for senior citizens
sponsored by the Berrien County Council on Aging. In July of 1975 we
decided to go on our own and became a private, non-profit incorporation
with a board of directors comprised of senior citizens and their spouses.
We now provide services to over 1000 peOple, with about 400 being active
members in our organization.

Here are some of your neighbors in the Central County Senior Center area
who would like to invite you to come join us. We believe "there are

109

no strangers here, only friends we haven't met."

Grace Robinson Ivan French

 

 

Art Robinson Pearl Fryman

 

 

We hope to hear from you soon.

Sincerely,

Barb Williams
Director

P.S. Also, whether or not you wish to join in on center activities at
this time, please call 471-2017 or write us at 608 South Mechanic St.,
Berrien Springs, if you would like to receive our monthly newsletter
which describes the activities planned for the coming month and includes
bits of news for senior citizens.

110

Example Contact Formats
(Phone Call Format)

Check as step is completed: Client name

 

--after identifying party—-
"Hello, M . My name is and I am from the

Senior Citizens Center."

 

"I'm calling some of our neighbors in the area, and I would
like to tell you about our center."

"We at the Center provide information and services to senior
citizens. We also provide social activities and recreation. Would
you like to hear about some of our services and activities, M ?"

Step 1: a) If yes, briefly describe the same services as in
the letters. (SIRS, Nutrition, Transportation,
Social) Follow-up with: "Would you be interested
in any of these?" Then go to Step 2.

b) If no, go to Step 2.

Step 2: "Would you like to have one of our outreach workers
come talk to you about our center?"

a) If yes, confirm address and set up time. Then go
to Step 3.

b) If no, encourage person to contact the center if
s/he ever has need for a service. Then go to
Step 3.

Step 3: "I'd like to give you our phone number and address if
you would like to write them down." Provide phone
number and address to person, having him/her write
them down and encourage him/her to call or write to
sign up for the newsletter.

If s/he asks to sign up now say: "I'm sorry but I don't have our record
book mailing list right now. If you will call the main desk at the num—
ber I gave you after we hang up or any time it's convenient for you, the
center will sign you up, or you could write the center at the address I
gave you, whichever is most convenient for you.”

Step 4: Remind the person to contact the center if s/he ever
has a need. Thank him/her for talking to you.

—-or, if a home visit was arranged--
Step 5: Confirm home visit and thank him/her for talking to you.

Note: If at any point the person wants to discontinue the conversation,
try to get him/her to write down the center phone number and address in
case the person should ever want to talk in the future, and ask him/her
to call or write to sign up for the newsletter.

lll

Example Contact Formats
(Format for Outreach Visits)

Client name

 

1. Introduce self and explain purpose of contact.

a) If possible, proceed as in any successful outreach contact (i.e.
fill out client card with person, assess needs, discuss services,
etc.). At conclusion of interview, go to Step 2.

b) If client does not wish to talk to you, go immediately to Step 2.

2. Ask the person to write down the name, phone number and address of
the center on a slip of paper. Thensz the person to call or write
the center if s/he would like to receive the monthly newsletter.

*Note* Do not sign the person up for the newsletter at that time!
If they ask to do so, tell them you don't have the necessary
list with you and ask them to call or write the center at
their convenience.

3. Thank the person for his/her time and remind him/her to call the
center if s/he ever has a need.

We know that each one of you has your own special ways of talking to the
elderly people you contact on outreach, so we are not asking that any
special approach be used. However, we would like the above steps to be
followed at a minimum so that we know that all of the elderly contacted
got at least that same information. In addition, we would like for you
to go through the following list of items immediately after an outreach
contact (for example, as soon as you return to your car) and check off
all those things that did occur in that particular interview. If the
interview was cut short for any reason, indicate at what point the
contact stopped.

Check if occurred:

Spoke to the correct person. (If this is "no," explain briefly.)
. Introduced self.

. Explained purpose of contact. (Mentioned the center.)

Described some services and/or activities.

Gave person address and phone number of center.

. Filled out client card with the person.

\IONU'IJ-‘LONH

. Reminded person to call or write to sign up for the newsletter.

112

APPENDIX C

RESPONSE RECORDING DEVICES

113

 

 

 

 

 

 

 

 

 

 

 

 

 

.c .¢ .e .c .c .< .c .e
on on on on on In on on
.N .N .N .N .N .N .N .N
.H .H .H .H .H .H .H .H
.q .e .¢ .¢ .G .Q .G .q
.m .m .m .m .m .m .m .m
.N .N .N .N .N .N .N .N
.H .H .H .H .H .H .H .H
cc 0“ 0* 0*. Ce 0* cc 0“
.m .m .m .m .m .m .m .m
.N .N .N .N .N .N .N .N
.H .H .H .H .H .H .H .H
.e .c .e .q .q .q .q .q
.m .m .m .m .m .m .m .m
.N .N .N .N .N .N .N .N
.H .H .H .H .H .H .H .H
.c .q .c .G .q .q .q .q
.m .m .m .m .m .m .m .m
.N .N .N .N .N .N .N .N
.H .H .H .H .H .H .H .H

AuOLOOV guano: coaumosvm coaumuuoqmaaus Hmuuoumm .wsu>«o uOquHmsoz smucwu

\Hmfioom IOOaOOEMONGH usumoavom wouowuaoo
uaowao unmade
omcomwom umusou m a H Omaoamwm udOHHu
.uao OOHHHM ma puma
ucoaao m cog: can: any so
Oumv Onu can :mr wmm.m yam .N
.on OOOO
onxm mafia some Ada aouv
A.quH .anna .uOquHv uwuauo
Onu ou OOOO«H mconuuwo Onu mo sumo you menu onv onu muumucoo uaoaao Ono
.ao>«w as ooa>umm O£u mafia name must onu nouam www.mnu was Ouav uuuau .H
omaomwom pounce m a H you maoauosuuuau monommum ucouao wow OdouuosuuuaH

nouw>un wuwvuouom omaoauom

mamz
m.cmuauwu
poacow

Response Recording Devices

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[ YEAR OF BIRTH 1 CLIENT IDENTIFICATION NUMBER

CLIENT CARD ......n ,,

Client 0 Date

0 N“ I -—
LAOV nest IIIYIAL
Address Phone
Soc. Sec. U — —

I. VITAL O‘I'ATISTICS
Birth: Dete__ / /_ Piece U.S. Citizen? 1
Rece: I )Caucasien I )Negro I )Spen.Leng. I )Am.Indien I )Orientei I )Other:

ngueoe restriction?
lndicete one: Mela/Female Ferm/Non-ienn Rural/Urban accident/Niger“
unitelStetue: I )Merried ( )Singie I )Widowed ( )Divoroed/Sepereted

c. MOVIE!"

Employed? I )No I )Yee Phone:
Retired? ( )No ( )Yee Veteran? I )No I )Yes

D. FMIAL REMCES

Source oi Income: I )Pension I )lnsurence I )Seilfy I )Sociei Security I )Priveteincome ( )Supportirom reietivee
I )Other:

Supplementery: ( )SSI ( Ioss I )Food Stamps I )Medicere I )Mediceid I )Other:

Approx. income: 8 _____I ) Below Poverty Levei I ) Ne. Poverty Level

I. we

independent I )Owne Home I )Rente ( )Apertment I )Room I )Room/Boerd I )Retirement Home ( )Pubiic Low-Rent
( )Mobiie Home
nt: I iHospitel I INuningi-iome ( )Foster Cere Home I )HomeoiReietiveorFriend ( )Other:
Livee with: ( )Spouee ( )Alone I )Friend I )Reletive

F. ”MTV
Mobile: ( )Drivee I )Vlelke I )Pubiic Trenep. I )Privete Trenep.
impaired: I )Welker ( )Crutchee ( )Wheeicheir ( )Bedridden I )Houeebound I )Other:

e. mm [Indicete (Gigood. (FHeir. (P) poor] I Ismm I )Speech I )Heering I walking I teeter-I
Note eny chronic heeith probleme:

H. MIKEY mm '00 m I. LWINO 6“.)le ADEQUATE WHAT!
Treneportetion I ) I ) Lodging I I I I
Moots I i I I Heeling ( i I i
Lewdry I I I I Ventiietion ( ) I I
i-ioueekeepim I i I ) Lighting I I I I
Home meintenence I ) ( ) Accessibilitytoout-oi-doore ( ) I )
Shopping I I I I Privacy I I I I
Personal Hygiene I I I I Plumbing ( I I I
MomoIOuIIook I i I ) Hot Water I ) I )
Personal Seiety I I I I

J.

SERVICES NEEDED:

 

 

 

 

 

 

 

114

APPENDIX D

MEMORANDUM OF AGREEMENT

Memorandum of Agreement

On the part of Robert L. Dolsen, Executive Director, and Patricia J.
Hohnstein, Projects Manager, of the Region Four Area Agency on Aging:

Agree to allow access to data necessary for evaluation Of the
project.

Agree to supervise and require cooperation of I & R personnel in
outreach operations and data collection required by the project.

Agree to insure the provision of postage, telephone, and travel
costs as required by the project.

Agree to allow I & R personnel sufficient time to conduct outreach
operations and data collection as described by the researchers.

With the assurance of confidentiality,

Agree to permit use of project data for educational requirements
and publication by the researchers.

 

On the part of Martin Kushler and John Jeppesen, Project Researchers:
Agree to protect confidentiality of all data obtained in the project.

Agree to design and supervise the implementation and operation of
the project.

Agree to insure that the project promotes the goals of the I & R
network and the best interests of the client population.

Agree to provide, for the sole use of the respective I & R centers,
all lists of potential clients as soon as project operations are
concluded.

Agree to take responsibility for analysis and interpretation of the
data, after insuring appropriate consultation with Area Agency
personnel.

Agree to act as liaison between the Office of Services to the Aging
and the Region Four Area Agency on Aging for purposes of this
project.

Agree to provide a written report summarizing the results and
findings of the study to the Region Four Area Agency on Aging, for
their use in planning, funding requests, etc.

 

 

Robert L. Dolsen, Exec. Dir. Martin Kushler

 

 

Patricia J. Hohnstein, Proj. Mgr. John Jeppesen

115

APPENDIX E

NUMBER OF REFERRALS BY TYPE OF CONTACT

Received
Referral

Did Not
Receive
Referral

Total

ax2=6.85

Number of Referrals by Type of Contact3

Table E—l

Type of Contact

 

 

 

 

Informational Personal Telephone Personal
Control Mail Mail Call Visit
0 l 2 3 5 11
41 44 36 40 38 199
41 45 38 43 43 N=210
df=4 (p=.l44)

116

APPENDIX F

THE FOLLOW—UP SURVEY QUESTIONNAIRE

The Follow-Up Survey Questionnaire

1. Have you ever heard of the Senior Center?

 

1. yes 2. no 8. other

If yes:

2. How did you hear about the Center?

 

1. television 6. I & R/Outreach contact
2. radio (letter, phone or visit)
3. newspaper 7. don't remember
4. friends 8. other
5. relatives

3. Were you contacted by a (letter, phone call, outreach worker) from

the Center?

1. yes 2. no 3. can't remember 8. other

If yes:

4. What did you think about the (insert whichever way person was con—
tacted, i.e. "the letter from the Center," "the phone call from the
Center," etc.)? Open-ended--record all responses.

(Read to respondent)
I would like to ask you some questions about the (letter, phone call,
visit):

 

not at a some- quite very
all little what a lot much
5. Was it clear what the
was talking about? 1 2 3 4 5
6. Did you think the Center
could be helpful to you? 1 2 3 4 5
7. Did the Center seem like it
would be a friendly place
(from the contact)? 1 2 3 4 5
8. How much need did you have
for their services? 1 2 3 4 5
If positive responder:
9. What influenced you to cooperate with the Center?

 

117

(or ”What interested ou about the Center?")
y

l. the services available 4. they seemed friendly
2. the social activities 8. other
3. curiosity, I wanted to know

more about the Center

 

 

If not a responder:

 

10. What influenced you to not contact the Center?
1. had no need 4. not sure what they were
2. don't believe in help talking about
like that 5 too far away
3. seemed unfriendly 6. never got around to it
7. other

 

What influenced you to (or not to) call or write to sign up to receive
the newsletter?

11. Did sign up, 12. Did not sign up

 

 

 

 

 

1. sounded interesting I. didn't want it
2 wanted to know about the 2. didn't know I could
Center 3. I lost the number
3. wanted to know about 4. never got around to it
services 8. other
4. wanted to know about
activities
8. other
If hasn't signed up for newsletter:
13. Would you like me to sign you up now?
1. yes 2. no
14. What are the most important things that the Center

 

might be able to help you with? (list below)

15. In the future, would you consider calling the Center
if problems arise?

 

1. yes 2. no 8. other

118

(Form B-—Controls Only)

 

 

 

 

 

 

 

1. Have you ever heard of the Senior Center?
1. yes 2. no 8. other
If yes:
2. How did you hear about the Center?
1. television 5. relative
2. radio 6. I & R/Outreach worker contact
3. newspaper 7. don't remember
4. friend 8. other
4a. If you were to be contacted by someone to tell you about our
Center and you had never heard about it, which type of
contact (letter, phone call, or visit by someone in person) would
you prefer?
1. letter 2. phone call 3. personal visit 8. other
13. Would you like me to sign you up for our newsletter?
1. yes 2. no 3. already receives 8. other
14. What are the most important things that the Center
might be able to help you with? (list below)
15. In the future, would you consider calling the Center

 

if problems arise?

1. yes 2. no 8. other

 

119

l6.

l7.

18.

20.

22.

23.

23a.

Are you:
1. married 3. widowed
2. single (never married) 4. divorced

8. other

 

How many people live in your immediate household? (enter exact
number)

 

How often did you visit in—person with a member of your family,
friends or neighbors last week? (either at your house or theirs)

Family 19. Friends or Neighbors

 

1. every day 1. every day
2. a few times 2. a few times
3. once 3. once

4. not at all 4. not at all

How many close friends or relatives do you have in this neighborhood?

Relatives 21. Friends
1. 4 or more 1. 4 or more
2. three 2. three
3. two 3. two
4. one 4. one
5. none 5. none

About how often last week did you talk to friends, relatives or
others on the telephone?

every day

. several times last week
once last week

. not at all

J—‘UONH

About how often do you go to meetings or activities of clubs, other
organizations, or information meetings of groups?

a few times a week

once a week

2 or 3 times a month

once a month

less than once a month or never

U1-I-\UJNI-'

How often do you attend a church?

. every week

almost every week

about once a month

a few times a year - holidays, etc.

. never (haven't gone in a year or more)

LJ‘I-I-‘LAJNH

120

24.

25.

26.

27.

28.

29.

30.

31.

32.

When you need help around the house, such as moving a heavy object,
who do you contact for help? (check only one)

 

l. someone in household takes 5. hire someone
care of it 6. do it myself
2. neighbor 7 don't know of anyone
3. relative 8 other
4. friend

Do you think most of your friends, if they came on hard times, would
accept help from the government, such as foodstamps?

1. yes 2. no 3. don't know 8. other
Do you feel the same way as your friends about things like that?

1. yes 2. no 3. don't know 8. other

 

Do you think you see your friends, relatives and neighbors as much
as you would like?

1. see them too much (would like more privacy, etc.)
2. see them enough (satisfied)
3. would like to see them more

How often do you find yourself feeling lonely? (Read items)

1. never 4. fairly often
2. hardly ever 5. very often or always
3. sometimes but not too often

How do you see yourself now as compared to before you were 65 in
terms of involvement with friends, neighbors, relatives, in groups,
etc.? (Read items)

1. much more active now 4. somewhat less active now
2. somewhat more active now 5. much less active now
3. about the same

How many living children do you have? (including adopted or step-
children) (record exact number)

 

Do you have any children living in this part of the state? (within
a one hour drive) (record exact number)

 

How Often do any of your children call or visit you or do you call
or visit them?
Call (phone) Visit

 

1. once a week or more 1. once a week or more

2. 2 or 3 times a month 2. 2 or 3 times a month

3. once a month 3. once a month

4. less 4. less (but at least once per year)
5. never 5. less than once a year or never

121

33.

34.

35.

36.

37.

Where do you get most Of your information about what goes on in the
community? (check only one)

1. television 5. relatives
2. radio 6. neighbors
3. newspapers 7. don't know
4. friends 8. other

 

What was the last grade of schooling you completed?

no schooling at all
some elementary (1—8)
completed 8th grade
some high school

completed high school
some college
completed college
other

buNI—J
mNC‘sLfl

 

What was your occupation (or the occupation of your spouse) before
retirement? (If not retired, enter present job and note "not
retired.")

Compared to other elderly people, do you think your life today is:
(read items)

1. much better than others 4. somewhat worse
2. somewhat better 5. much worse

3. about the same

If you needed it, would you accept help from the government, such as
foodstamps?

1. yes 2. no 8. other

 

Now let's talk about social services.

38.

39.

Please tell me all Of the community services or social services for
older people that you have heard about. (have person list as many
as can and write them below.)

Now, please tell me all the community or social services you have
tried at one time or another. (list below)

(If the above responses indicate that the person knows of services s/he
has not tried, ask #40. Repeat for each service heard of but not tried.)

122

40.

41.

42.

42a.

43.

44.

What are the reasons you have not tried those services of
? (first service heard of but not tried)

1. don't need them 4. don't know how to apply
2. not eligible 5. don't believe in welfare
3. haven't gotten around to it 6 other
For 2nd service not tried? (enter one of above answers)
For 3rd service not tried? (enter one of above answers)

If you found that you had a need for each of the following, who or
where would you call for help?

(First time through) (Re—check)*
Where Would Call Social Service

 

 

Food

Income (money)
Medical help
Transportation

 

 

 

 

 

 

-‘----“

 

 

(If client doesn't know where to call, write "don't know.")

*(Interviewer: for each "need" for which the person did not identify
a social service the first time through, go back and ask them if they
know a service or agency for that purpose and record that in the
right hand column.)

 

Do you have a Medicare card?
1. yes 2. no 8. other U.
DO you have a Medicaid card?

1. yes 2. no 8. other

 

Where did you first hear about Medicaid?

l. newspaper 5. relative
2. radio 6. social worker
3. television 7. doctor
4. friend 8. other
9. don't remember

How did you come to be signed up for Medicaid?

1. applied for it myself 5. Department of Social Services
2. friend/relative helped me signed me up
3. social worker contacted me 6. Social Security signed me up
4. doctor helped me 7. don't know

8 other

 

123

45.

46.

47.

48.

49.

50.

51.

If something went wrong with receiving your check, for example, is
there someone who you could call at the Social Security Department?
(a person s/he usually talks to, for example)

1. yes 2. no 8. other

 

Was (or is) there someone who you can call at the Department of
Social Services? (including specific "operator number")

1. yes 2. no 8. other

 

Which agency, the Department of Social Services or Social Security,
gives you the best personal attention?

1. Department of Social Services 4. don't know (but uses both)
2. Social Security 8. other
3. no difference between the two

 

(Interviewer: If the person doesn't know the difference between the
D88 and the Social Security Agency, please write that down here.)

Do you have any complaints or suggestions in general about the way

senior citizens are treated by social service agencies? (note all
responses)

Do you think the people at the Department of Social Services have
changed much since the first time you ever contacted them?

1. yes 2. no 3. don't know 8. other

 

As near as you can remember, when was the first time the Department
of Social Services ever provided you with a service?

1. 1950-1960 4. since 1970 exact ear if
2. 1960-1965 5. never has known y
3. 1965-1970 6. other

 

Now, I am going to read you a list of areas which people feel are
problems for older Americans. For each area, please tell me if it
is no problem for you, somewhat of an important problem, or a very
important problem. (Read list; rotate order.)

 

 

 

 

No Problem Somewhat Very Important
Housing 1 2 3
Employment 1 2 3
Health Care 1 2 3
Income 1 2 3
Crime 1 2 3
Getting More Education 1 2 3
Nutrition and Food 1 2 3
Transportation 1 2 3
Loneliness l 2 3

 

124

52.

53.

54.

55.

56.

57.

How well do you think your income and assets (including those of
spouse and dependents which you can use) satisfy your needs? (Read)
1. very well 4. barely
2. well 5. poorly
3. adequately

Compared to other people your age, would you say your health is:
(Read items)

1. much better than others 4. somewhat worse
2. somewhat better 5. much worse

3. about the same

Do you think a lot of people tend to think poorly about people who
receive help from the government, such as foodstamps or Medicaid?

1. yes 2. no 3. don't know 8. other

 

How do you usually do your shOpping?

1. walk 5. public transportation (bus,
2. drive myself dial-a-ride, etc.)
3. ride with friend/neighbor 6. don't go shopping (explain)
4. ride with relative 7. don't know

8. other

 

In general, how do you feel about your life now?

feel very satisfied about my life now
feel fairly satisfied

neither satisfied nor dissatisfied
feel fairly dissatisfied

feel very dissatisfied

m-I-‘UJNH
on...

Do you know anyone in the area who might be interested in our Center
or some of the services we can provide?

125

Interviewer's Page

 

 

 

 

 

Date of Interview Interviewer

Approximate Length of Interview (minutes)
Age Sex______ Race______ Name

Client Card (yes or no and number, if applicable)

 

Was anyone else present at the interview?

 

Did they assist in answering questions?

 

The client's health appeared to be:

. very good
fairly good
satisfactory
somewhat poor
. very poor

LII-book)?“

The client's memory and ability to pay attention appeared to be:

. very good
fairly good
satisfactory
somewhat poor
very poor

U14>LANH

Could the client read? (his/her mail, for example)

. yes, well

yes, somewhat

no, poor vision

no, lack of education
. don't know

UDWNH

(If "no," does someone read his/her mail to him/her? 1. yes 2. no)
Did the client have a phone? 1. yes 2. no

Did the client have a television and radio?

 

Television Radio
1. yes 1. yes
2. no 2. no

Please write any additional notes of importance on the back of this page.

126

APPENDIX C

INTERNAL CONSISTENCY OF THE FIVE RATIONAL SCALES

Internal Consistency of the Five Rational Scales

Scale

Family contact

Other social
contact

"Problems"

Knowledge of
social services

Life
Satisfaction

 

127

Table G-l
Item-Total Alpha if Overall
Items Correlation Item Deleted Alpha
.74
Question #18 :48 .70
Question #20 .55 .69
Question #30 .58 .69
Question #31 .59 .66
Question #32 .42 .73
.56
Question #19 .51 .40
Question #21 .26 .55
Question #22 .36 .49
Question #23 .23 .55
Church attendance .31 .52
.66
Housing .21 .67
Health care .53 .57
Income .39 .62
Crime .28 .67
Nutrition/food .50 .59
Transportation .47 .59
.68
Question #38 .63 .56
Question #39 .55 .60
Question #41 (food) .41 .66
Question #41 (income) .43 .65
Question #41 (medical) .42 .65
Question #41 (trans-
portation) .21 .69
.57
Question #28 .18 .63
Question #36 .55 .39
Income 0K .28 .55
Self-rated health .29 .54
Question #56 .48 .46

APPENDIX H

EFFECTIVENESS OF THE RANDOMIZATION PROCEDURE

Effectiveness of the Randomization Procedure

In order to check for the effectiveness of the randomization pro-
cedure, a total of 21 demographic and descriptive variables were examined
for differences between the 5 experimental groups. Only one of the 21
variables was found to vary between the groups with a significance level
less than(p=.05). This was the variable of age (p=.049).

Although finding only one barely significant difference out of such
a wide range of variables lends quite strong support to the notion that
the five groups were functionally equivalent, the variable of age was
further examined to see where the differences occurred and how this
variable related to the outcome measures. In this manner, it could be
determined if the differences observed might have added bias to the
analysis of the effects of type of contact.

A Duncan's multiple range test was administered as a post hoc
analysis of the difference between groups. It was found that the sig—
nificant difference was attributable to the fact that those in the tele-
phone and in-person groups (mean age 73.3 and 74.0, respectively) were
younger (p<:.05) than those in the personal mail group (mean age 78.3).
Those in the control group and informational mail group (mean age 74.9
and 76.2, respectively) form a homogeneous subset with either the tele-
phone and in—person group or with the personal mail group.

The available evidence indicates that the impact of these differences
upon the outcome measures was probably small. A quick glance at the
ordering of the groups in terms of "success" in each of the outcome
measures (see "Treatment Effects") reveals that they are not at all the

same as the ordering of the groups in terms of age. Indeed a

128

correlational analysis shows that age has a low correlation with each of
the three dependent variables. Younger age is slightly positively
correlated with client card registration (r=.13, p=.04) and virtually
non-correlated with signing up for the newsletter (r=.06, p=.20) and with
receiving a service (r=.08, p=.l3). Although the first correlation
achieves significance at below the .05 level, this is primarily due to
the large sample size. As can be seen, the actual magnitude of each
relationship is small. Indeed, the r2 value reveals that age accounts
for only 1.6%, .4% and .7% of the variance of the three respective
dependent variables. Finally, T-tests show that positive responders
taken as a group did not differ from non—positive responders in terms
of age. In summary, it appears that the observed differences in mean
age between the groups should not add any substantial bias to the inter—
pretation of the outcome results.

In conclusion, having examined a wide range of available data, the
weight of the evidence seems clearly to demonstrate that the randomiza—
tion procedure did indeed produce functionally equivalent groups for the

purposes of this experiment.

129

Effectiveness of the Randomization Procedure

Table H-l

Results of Analyses on 21 Variables

by Type of Contact

Analyses of Variance

 

Variable

Age

Number of persons in household
Number of living children

Number of living children within a 1 hour drive
Health as rated by outreach worker
Memory as rated by outreach worker
Family contact scale

"Other social contact" scale
"Problems" scale

Social service knowledge scale
Life satisfaction scale

Chi—Squares Analyses

 

Variable

Sex

Race

Marital status

With whom person resides

Education

Response to follow-up survey

Source of community information (Question #33)

Does person know someone who might be interested
in I & R services

SSI or State Medical distribution

Which I & R center contacted person

130

 

 

F—Value Signif.
2.88 .025
1.45 .221

.56 .690
.70 .594
.37 .776
2.27 .066
.74 .570
1.50 .223
.10 .960
.99 .416
1.73 .145

2 .

x Value Signif.
4.92 .295
3.57 .467
7.81 .090

21.38 .164
9.36 .313
4.06 .398
8.69 .369
1.97 .741
1.29 .862

15.34 .910

APPENDIX I

COMPARISON OF THE TWO ASSISTANCE GROUPS

Comparison of the Two Assistance Groups

When it became evident that the three dependent variables demon—
strated no main effects or interaction effects attributable to the SSI
vs. State Medical group distinction, the question arose as to whether or
not these two sample groups did, in fact, represent two distinct popula-
tions. In order to pursue this question, the Michigan Department of

Social Services' Medical Assistance Eligibility Manual (1976) and SSI

 

Handbook (1976) were obtained. These sources reveal that the eligibility
requirements for the two programs are quite similar. The major eligi-

bility guidelines for the two programs are compared below.

Table I-l

Program Eligibility Requirements

 

 

SSI ("aged" category) State Medical ("aged" category)
Age Over 65 Over 65
Income $2004/yr. For 1 person $2688/yr. For 1 person
$3012/yr. For 2 persons $3540/yr. For 2 persons
Non—exempt* $1500 For 1 person $1500 For 1 person
Property $2250 For 2 persons $2250 For 2 persons

*The two programs each consider one house, one automobile, and most house-
hold goods and personal effects as exempt from property value computations.

As one can see, in terms of the criteria of income and available
assets, these two groups are, by definition, quite similar (and also
quite poor).

As the third, and most important, criterion, the two groups were
compared in terms of the actual data distributions gathered in the
follow-up survey. The two groups were examined for any differences on a

total of 19 variables, including the five logical scales discussed in the

131

"Methods" section. A complete listing of the variables and their sig-
nificance levels are presented in Table I-2.

As one can see, the results revealed that there were only two sig-
nificant (p<:.05) differences between the SSI recipients and the state
medical recipients among these 19 variables. These were with respect to
the variables of age and race. The presence of only two significant
differences between these population sub-groups on such a wide range of
variables would seem to indicate that, for most practical purposes, the
two groups were, in fact, equivalent. In order to strengthen this con—
clusion, however, the two variables which distinguished the two groups
were examined for their relationships to the dependent variables.

Regarding the variable of race, it appeared that a similar conclusion
was merited. The variable of race was not significantly related to any
of the three dependent variables (p').50 for each of the three variables).
Hence, it appeared that this difference also should not have biased the
relationship between these two groups and the dependent variable.

Finally, as a further test of the similarity of the two population
sub-groups, the variances of response were compared on numerous variables
from the follow-up survey, including the five rational scales. None of
the variables examined showed a significant difference in variance
(P<.05) between the two groups.

In summary, it appeared as though there were only two minor dif-
ferences between the two assistance recipient groups selected for use in
this study. Those differences consisted of the fact that the state
medical group had a higher percentage of black persons (32% as compared
to 14%) and was slightly younger (Ee74.1 years as compared to 76.6 years)

than the SSI group. Other than that, however, comparisons on a wide range

132

of descriptive, demographic and attitudinal variables demonstrated that
the two groups of low—income elderly were quite similar. This was also
true in terms of the three dependent variables used as outcome measures
in this experiment. With the weight of this evidence in mind, the two

groups were collapsed into one large sample for the purpose of all per—

tinent data analyses.

133

Comparison of the Two Assistance Groups

Table I—2

Results of Analyses on 19 Variables
by "Assistance Group" (SSI or State Medical)

Analyses of Variance

 

 

Variable F—Value Signif.
Age 2.23 .026
Number of persons in household 1.956 .165
Number of living children .060 .953
Number of living children within 1 hour drive .530 .600
Health as rated by outreach worker .776 .311
Memory as rated by outreach worker .697 .405
Family contact scale 1.58 .117
"Other social contact" scale .150 .882
"Problems" scale .520 .606
Social service knowledge scale 1.09 .277
Life satisfaction scale 1.84 .068

Chi-Squares Analyses

 

 

Variable x2 Value Signif.
Sex 2.81 .094
Race 9.13 .003
Marital status 3.21 .360
With whom person resides 3.71 .446
Education 1.21 .546
Response to follow—up survey 1.76 .184
Source of community information (Question #33) .71 .698
Does person know someone who might be interested

in I & R services 1.22 .270

134

APPENDIX J

THREE RATIONAL SCALES AS SECONDARY OUTCOME MEASURES

Three Rational Scales as Se

Table

Analysis of Variance:

condary Outcome Measures

J—l

Knowledge of Social

 

Services Scale by

Type of Contact

 

 

 

 

 

 

 

 

 

 

 

Source .22 Mean Square F_ Significance of F
Main Effects 4 4.35 .536 .710
Type of Contact 4 4.35 .536 .710
Residual 133 8.11
Table J-2
Analysis of Variance: Problems Scale
by Type of Contact
Source _DF Mean Square _F Significance of F
Main Effects 4 .206 .996 .411
Type of Contact 4 .206 .996 .411
Residual 170 .207
Table J-3
Analysis of Variance: Life Satisfaction Scale
by Type of Contact
Source D§_ Mean Square F_ Significance of F
Main Effects 4 .654 1.59 .178
Type of Contact 4 .654 1.59 .178
Residual 186 .411

135

APPENDIX K

RELATIONSHIP OF THE DEMOGRAPHIC AND SOCIAL SERVICE
RELATED VARIABLES TO THE THREE DEPENDENT VARIABLES

Relationship of the Demographic and Social Service
Related Variables to the Three Dependent Variables

Table K—l

Results of Analyses for the Client Card Registration Variable
(Responders vs. Non—Responders)

 

 

 

Variable F—Value Signif.
Age 1.91 .058
Number of persons in household 1.21 .229
Health Rating .18 .859
Memory rating 1.85 .068
Number of living children 1.40 .164
Number of children within 1 hour drive 1.53 .128

Chi—Squares Analyses

 

 

Variable x2 Value Signif.
Sex .82 .364
Race .49 .486
Marital status 3.52 .318
With whom person resides .86 .931
Education .09 .954
Question #25 (friends accept foodstamps) 4.80 .091
Question #37 (self accept foodstamps) .04 .846
Question #48 (comments about social services) 1.83 .608
Question #50 (when first used DSS) 2.05 .563
Question #54 (perceives welfare stigma) .11 .741

136

APPENDIX L

OUTREACH COST ESTIMATES

Outreach Cost Estimates

Table L—l

Estimated Costs of Completing One Outreach Contacta

Mode Outreach Worker + Related Costs = Cost Per Contact
Labor
Home Visit $2.55 .45 (mileagec) $3.00
Telephone Call $.85 .17 (telephone costs) $1.02
Personal Letter $.22 .22 (postage and $.44
materials)

8Note: the cost estimates provided in this table are taken from the
records of the Region IV Area Agency on Aging in southwestern Michigan,
from the period of October 1, 1975, through June 30, 1976. Estimates
based on other regions and for other time periods may require adjust-
ments according to local cost considerations.

bOutreach worker labor costs are estimated at an average of $2.55 per
hour and are based on a single outreach worker making the contact. If
outreach workers travel in pairs, labor costs must be increased propor-
tionately.

CA rate of 15¢ per mile was used to estimate mileage costs. Actual
mileage costs vary from region to region and can be expected to be
higher in rural areas. To the extent that mileage required increases,
the relative desirability of home visits may decrease.

139

Outreach Cost Estimates

Table L—2

Estimated Costs Based on Effectiveness Data

For the Client Card Dependent Variable

 

 

Cost of 1 Effectiveness = Estimated Cost Per
x .
Contact Factor Successful Registration
Home Visit $3.00 1.3 $3.90 + overhead
Telephone $1.02 3.6 $3.67 + overhead
Personal Letter $.44 9.0 $3.96 + overhead

For the Dependent Variable of Receiving a Service

 

 

Cost of 1 Effectiveness = Estimated Cost Per

Contact Factor Successful Registration
Home Visit $3.00 4.0 $12.00 + overhead
Telephone $1.02 5.0 $5.10 + overhead
Personal Letter $.44 9.0 $3.96 + overhead

For the Newsletter Dependent Variable

 

 

Cost of l x Effectiveness = Estimated Cost Per

Contact Factor Successful Registration
Home Visit $3.00 8.3 $24.90 + overhead
Telephone $1.02 50.0 $51.00 + overhead
Personal Letter $.44 12.5 $5.50 + overhead

140

APPENDIX M

CLIENT RATINGS OF OUTREACH CONTACTS

Client Ratings of Outreach Contacts

 

 

 

 

 

Table M—la
Mean _£;_ Significance
Question #5 (how clear) 2.96 .036
informational mail 2.95
personal mail 4.11
telephone call 3.89
home visit 3.78
Question #6 (center helpful) .46 .711
informational mail 2.90
personal mail 2.84
telephone call 2.48
home visit 2.70
Question #7 (center friendly) .68 .566
informational mail 3.79
personal mail 4.26
telephone call 4.00
home visit 4.09
Question #8 (need for services) .34 .798
informational mail 2.26
personal mail 2.21
telephone call 2.35
home visit 2.03

 

8These questions were rated on a five point scale, from lowest to highest.
(see Questions #5 through #8 in Appendix F.)

141

LIST OF REFERENCES

LIST OF REFERENCES

Administration on Aging. Let's End Isolation. U.S. Government Publica-
tion, Department of Health, Education and Welfare, 1971.

 

Administration on Aging. Social Indicators for the Aged. Minneapolis,
Minnesota: American Rehabilitation Foundation, 1970.

 

Alexander, Raymond S. and Podair, Simon. "Educating New York City
Residents to the Benefits of Medicaid," Public Health Reports, 84,
September 1969, p. 767-772.

 

American Public Welfare Association. Program Planning Through Public
Welfare Agencies. Chicago, Illinois, 1967.

 

 

Amster, Leslie E. and Krauss, Herbert H. "The Relationship Between Life
Crises and Mental Deterioration in Old Age," International Journal
9£_Aging and Human Development, Vol. 5, No. 1, Winter 1974, p. 51—
55.

 

 

Ball, Robert M. "Income Maintenance for the Aged and Handicapped,"
Public Welfare, Vol. 30, No. l, 1972, p. 33-38.

 

Barton, Roger. Media in_Advertising. New York: McGraw—Hill Book Com-
pany, 1964.

 

Bennett, Ruth. "Living Conditions and Everyday Needs of the Elderly with
Particular Reference to Social Isolation," International Journal gf
Aging and Human Develgpment, Vol. 4, No. 3, 1973, p. 179-184.

 

 

Berg, William E., Atlas, Lucille, and Zeiger, Joan. ”Integrated Home-
making Services for the Aged in Urban Neighborhoods," Gerontologist,
Vol. 14, No. 10, 1974, p. 388-392.

 

Berguer, Lawrence and Yerby, Alonzo S. "Low Income and Barriers to Use
of Health Services," New England Journal g£_Medicine, Vol. 278,
part 1, March 1968, p. 541-546.

 

Bild, Bernice R. and Havighurst, Robert J. "Senior Citizens in Great
Cities: The Case of Chicago," Gerontologist, February 1976, part 2.

 

Blau, Zona Smith. Old Age lama Changing Society. New York: New View—
points, 1973.

 

Blumenthal, Hermant (ed). Medical and Clinical Aspects_gf Aging. New
York: Columbia University Press, 1962.

 

Bowers, John W. Designing the Communication Experiment. New York:
Random House, 1970.

 

142

Boyd, Rosamonde Ramsay and Oakes, Charles G. (eds). Foundations gf_
Practical Gerontology, 2nd ed. revised. Columbia, South Carolina:
University of South Carolina Press, 1973.

 

 

Brody, Elaine and Brody, Stanley J. "Decade of Decision for the Elderly,"
Social Work, Vol. 6, No. 2, 1974, p. 544-554.

 

Brody, Stanley J. "Comprehensive Health Care for the Elderly: An
Analysis," Gerontologist, Vol. 13, No. 4, 1973, p. 412-417.

 

Brumfield, William A., Jr., Fox, Ruth 1., and Goldman, Jack J. "Reaching
the Target Population," Public Health Reports, July, 1968, p. 597-
602.

 

Buhler, Charlotte and Kleemier, Robert W. (eds). Aging and Leisure. New
York: Oxford University Press, 1961.

 

Burr, James J. "Protective Services for Older Adults: A Demonstration
Project," Welfare ig_Review, Vol. 9, No. 6, 1971, p. 1-6.

Busse, Ewald W. and Pfeiffer, Eric (eds). Behavior and Adaptation ig_Late
Life. Boston: Little, Brown and Company, 1969.

 

Butler, Robert N. "How to Grow Old and Poor in an Affluent Society,"
International Journal gf_Aging and Human Development, Vol. 4, No.
3, 1973, p. 277-279.

 

 

Butler, Robert N. "Pacification and the Politics of Aging," International
Journal gf_Aging and Human Development, Vol. 5, No. 4, 1974, p. 393-
394.

 

 

Butler, Robert N. "Taking Stock of the Status of the Elderly," Inter—
national Journal gf_Aging and Human Development, Vol. 6, No. 2,
1975, p. 179-182.

 

 

Carp, Frances M. "The Psychology of Aging," Foundations of Practical
Gerontology, 1973.

 

 

Conner, K.A. and Powers, E.A. "Structural Effects and Life Satisfaction
Among the Elderly," unpublished, 1973.

Corso, John F. "Age and Sex Differences in Pure-Tone Thresholds," Journal
g£_the Acoustical Society gf_America, Vol. 31, 1959, p. 498—507.

 

Council of State Governments. QE_Growing Old. Lexington, Kentucky, 1973.

 

Cumming, Elaine and Henry, William E. Growing Old: The Process gf_Dis-
engagement. New York: Basic Books, Inc., 1961.

 

 

Davis, Donald L. and Taylor, Walter. "The Senior Community Service
Project: A Manpower Model for the Older Disadvantaged," Industrial
Gerontology, Vol. 2, No. 2, 1975, p. 122-134.

 

 

143

Dervin, Brenda and Greenberg, Bradley S. "The Communication Environment
of the Urban Poor," Current PerSpectives in_Mass Communication
Research. London: Sage Publications, 1972.

 

 

Devoe, Merrill. Successful Telephone Selling. Eaglewood Cliffs, New
Jersey, 1956.

 

Dohrenwend, Barbara Snell and Dohrenwend, Bruce Phillip. Stressful Life
Events: Their Nature and Effects. New York: Wiley-Interscience,
1974.

 

 

Economic Associates, Inc., "Services for the Elderly," Federal Aid
Reporter, November-—December, 1967, p. 1-16.

 

Eldersveld, Samuel J. and Dodge, Richard W. "Personal Contact or Mail
Propaganda: An Experiment in Voting Turnout and Attitude Change,"
Public Opinion and Propaganda. New York: Holt, Rinehart and Win-
ston, 1965.

 

Erikson, Erik. Identity and the Life Cycle. New York: International
Universities Press, 1959.

 

Erp, Sue H. and Freeman, Jean D. Aging: Psychological Changes. Corvallis,
Oregon: Continuing Education Publications, 1971.

 

Gaitz, Charles M., "Barriers to the Delivery of Psychiatric Services to
the Elderly," Gerontologist, Vol. 14, No. 4, June 1974, p. 210-214.

 

Goldfarb, A.I. "Depression, Brain Damage, and Chronic Illness of the
Aged: Psychiatric Diagnosis and Treatment," Journal gf Chronic
Diseases, Vol. 9, No. 3, 1959, p. 220-233.

Goldfarb, A.I. "Patient-Doctor Relationship in Treatment of Aged Persons,"
Geriatrics, Vol. 19, No. 1, 1964, p. 18-23.

 

Goodrow, Bruce A. "Limiting Factors in Reducing Participation in Older
Adult Learning Opportunities," Gerontologist, Vol. 15, No. 5, 1975,
p. 418-422.

 

Gordon, Susan Krauss. "The Phenomenon of Depression in Old Age,"
Gerontologist, Vol. 13, No. 1, 1973, p. 100-105.

 

Graney, Marshall J. "Happiness and Social Participation in Old Age,"
unpublished, 1973.

Harris, Raymond. "Breaking the Barriers to Better Health Care Delivery
for the Aged," Gerontologist, Vol. 15, No. 2, 1975, p. 52-56.

 

Havelock, Ronald G. Planning for Innovation: Through Dissemination and
Utilization_9f Knowledge. Ann Arbor, Michigan: Institute for
Social Research, 1971.

 

 

 

144

Jette, Donna W., Lebawitz, Barry D., and Pendleton, Pete. "Social Par-
ticipation and Policy Alternatives: Studies in Quality of Life,"
unpublished, 1975.

Jordan, Maryrose, et a1. "Dietary Habits of Persons Living Alone,"
Geriatrics, Vol. 9, 1954, p. 230.

 

Kaplan, Jerome. "Services Without the Facts," Gerontologist, Vol. 13,
No. 2, 1973, p. 132.

 

Kaplan, Jerome. "The Family in Aging," Gerontologist, Vol. 15, No. 5,
1975, p. 385.

 

Katz, Elihu and Lazarsfeld, Paul F. Personal Influence. New York: The
Free Press, 1955.

 

Katz, Elihu, et a1. Studies gf_1nnovation and Communication £9_the Public.
Stanford, California: Institute for Communications Research, 1962.

 

 

Kent, James A. and Smith, Harvey C. "Involving the Urban Poor in Health
Services: The Employment of Neighborhood Representatives," American
Journal 2f_Public Health, Vol. 57, No. 6, 1967.

 

Kline, F. Gerald and Tichenor, Phillip J. Current Perspectives in Mass
Communication Research. London: Sage Publications, 1972.

 

 

Klippel, Eugene R. and Sweeney, Timothy W. "The Use of Information
Sources by the Aged Consumer," Gerontologist, Vol. 14, No. 2, 1974,
p. 163-166.

 

Kobrynski, Borys. "Innovations in Programs of Care for the Elderly,"
Gerontologist, Vol. 13, No. l, p. 50-53.

 

LeBovit, Corrine. "The Food of Older Persons Living at Home," Journal 2:
the American Dietetic Association, Vol. 16, April 1965, p. 285—289.

 

Leonard, Lois E. and Kelly, Ann M. "The Development of a Community Based
Program for Evaluating the Impaired Older Adult," Gerontologist,
Vol. 15, No. 4, 1975, p. 114—118.

 

Levin, S. "Depression in the Aged: A Study of Salient External Factors,"
Geriatrics, Vol. 18, 1963, p. 302—307.

 

Lionberger, Herbert F. Adoption of New Ideas and Practices. Ames, Iowa:
Iowa State University Press, 1960.

 

Lippitt, Ronald, Watson, Jeanne, and Westley, Bruce. The Dynamics 2:
Planned Change. New York: Harcourt, Brace and Company, 1958.

 

 

Lowenthal, M.F. "The Relationship Between Social Facts and Mental Health
in the Aged," Aging in_Modern Psychiatry. New York: American
Psychiatric Association, 1968.

 

145

Lowy, Louis. "The White House Conference on Aging: Two Years Later,"
International Journal 9f_Aging and Human Development, Vol. 5, No.
2, 1974, p. 205-211.

 

 

Maynard, Marianne. "Three Theoretical Aging Frames of Reference: Implica-
tions for a Rehabilitation Counselor's Model," Journal 2£_Applied
Rehabilitation Counseling, Vol. 5, No. 4, 1974, p. 207-214.

 

McKinney, John C. and deVyver, Frank R. Aging and Social Policy. New
York: Appleton-Century—Crofts, 1966.

 

Melrose, Jay, Welsh, Oliver L., and Luterman, David. "Auditory Responses
in Selected Elderly Men," Journal gf_Gerontology, Vol. 18, 1963, p.
267-270.

 

Michigan Department of Social Services. "Medical Assistance Program
Bulletin," No. 11, December 22, 1975.

Michigan Office of Services to the Aging. Michigan Comprehensive Plan_gn
Aging, 1975.

 

Moses, Beth and Zander, Mary. "Impact of SSI on Public and Private
Agencies Serving the Aged," unpublished, 1975.

National Council on Aging. The Myth and Reality gf_Aging in_America.
Washington, D.C.: National Council on Aging, 1975.

 

National Council on Aging and U.S. Office of Economic Opportunity. The
Golden Years: A Tarnished Myth ("Project FIND"), 1972.

 

 

Nuckols, Robert C. "Personal Interview Versus Mail Panel Survey," Jour-
nal gf_Marketing Research, Vol. 1, February 1964, p. 11-16.

Ozawa, Martha N. "SSI: Progress or Retreat," Public Welfare, Spring
1974, p. 33-40.

 

Palmore, Erdman (ed). Normal Aging_ll: Reports from the Duke Longi—
tudinal Studies, 1970-1973. Durham: Duke University Press, 1974.

 

 

Pivan, Frances Fox and Cloward, Richard A. Regulating the Poor. New
York: Random House, Inc., 1971.

 

Postema, Leonard J. Reminiscing, Time Orientation and Self-Concept in
Aged Men. Ph.D. Thesis, Michigan State University, 1970.

 

Region IV Area Agency on Aging. Area Plan 1975-1976, 1975.

 

Reichard, Suzanne, Livson, Florine and Petersen, Paul G. Aging and
Personality: ‘A_Study 9f Eighty-Seven Older Men. New York: John
Wiley and Sons, Inc., 1962.

 

 

Riley, Matilda White and Foner, Ann. Aging and Society, Vol. 1, New York:
Russell Sage Foundation, 1968.

 

146

Roens, Burt. "Using Research in Advertising Decisions," Proceedings, 7th
Annual Conference, Advertising Research Foundation, 1961, p. 65-70.

 

Rosenzweig, Norman. "Some Differences Between Elderly People Who Use
Community Resources and Those Who Do Not," Journal 2£_the American
Geriatrics Society, Vol. 23, No. 5, p. 224—233.

 

 

 

 

Rue, Vincent M. "Retooling Information Systems for Aging," International
Journal 9f Aging and Human Development, Vol. 4, No. 3, 1973, p. 361-
374.

Rush, Ramons R. and Kent, K.E.M. "Communication Channel Selection Con-

sideration for Reaching Older Persons," unpublished, 1975.

Scammells, Brian. The Administration.gf_Hea1th and Welfare Services.
Manchester, England: The University Press, 1971.

 

 

Schaie, K. Warner and Bribbin, Kathy. "Adult Development and Aging,"
Annual Review_g£ Psychology. Palo Alto, California: Annual
Reviews, Inc., 1975.

 

 

Shanas, Ethel, et a1. 01d Pegple in Three Industrial Societies. New
York: Atherton Press, 1968.

 

 

Shock, Nathan W. (ed). Biological Aspects gf_Aging. New York: Columbia
University Press, 1962.

 

Shore, A.L. and Schilsky, S.E. "The Senior Action Center of Illinois:
An Evaluation," unpublished, 1975.

Stewart, James C. and Hood, William R. "Using Workers from 'Hard-Core'
Areas to Increase Immunization Levels," Public Health Reports,
Vol. 85, p. 177-185.

 

Swanson, Pearl. "Adequacy in Old Age, Part 11: Nutrition Education
Programs for the Aging," Journal gf_Home Economics, Vol. 56,
December 1964, p. 732.

 

Tibbets, Clark L. (ed). Handbook 2£_Socia1 Gerontology. Chicago,
Illinois: University of Chicago Press, 1960.

 

Turner, Helen. "Personality Functioning in Later Life: Implications for
Practice," Paper presented at the Training Institute for Public
Welfare Specialists on Aging, Cleveland, Ohio, June, 1965. Pub-
lished by the U.S. Department of Health, Education and Welfare.

United Community Services of Detroit. "Well-Being Project for the Aging:
Final Report on a Three Year Demonstration Project," December, 1967.

Verwoerdt, Adriaan. "Biological Characteristics of the Elderly,”
Foundations 9f_Practical Gerontology, 1973.

 

 

Welford, A.T. "Psychomotor Performance," Handbook g£_Aging and the
Individual. Chicago: University of Chicago Press, 1959.

 

147

Westley, Bruce H. and MacLean, Malcolm 8., Jr. "A Conceptual Model for
Communications Research," Communication and Culture. New York:
Holt, Rinehart and Winston, 1966.

 

Woodruff, Diana S. and Birren, James E. Aging: Scientific Perspectives

 

.and Social Issues. New York: D. Van Nostrand Company, 1975.

 

148

ICHIGQN STQTE UNIV. LIBRQRIES

31293100658750