.,. .. .un-

 

 

 

 

 

 
 

 

 

 

 

nu. "1.:

“nun.-

 

$273050?!

 

 

”3:223... ‘nn‘ua‘mmgmm‘mm/mm
University 76 2242

EXPLORING ASPECTS OF FAMILY HEALTH:

PARENTS AND ADOLESCENTS

presented by

Joan E. Wood

has been accepted towards fulfillment
of the requirements for

PhD Family & Child

de in
gree Ecology

 

 

W «A m

Major professor

 

Date 1-30-90

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

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before duo due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

air—7
|___J-L_J

EDP

usu Is An Atttnn-ttvo ActioNEqual Opportunity Institution
chna-nt

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EXPLORING ASPECTS OF FAMILY HEALTH:

PARENTS AND ADOLESCENTS

By

Joan E. Wood

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Family and Child Ecology

1990

C7 A a

C‘w
we’d

be

ABSTRACT

EXPLORING ASPECTS OF FAMILY HEALTH:

PARENTS AND ADOLESCENTS

By

Joan E. Wood

The influence of the family on the health behavior of its
members was investigated in this cross-sectional exploratory study. A
conceptual model which included components of current health behavior,
family ecology, and child development frameworks and models provided
study guidance. A total of 206 respondents, i.e., 60 dual parent
families composed of father, mother and adolescent, and 13

female-headed single parent families composed of mother and

adolescent, completed pen and pencil questionnaires. The presence of

a family characteristic, a previously unknown or minimally
investigated family phenomena, was determined by the congruency of
family members’ reports on health status, intrinsic motivation, health
behavior (psychosocial and behavioral), and family functioning. The
Discrepancy Score and ConjunCLtve Models (Klein, 1984) and their
respective statistical approaches, i.e., derived discrepancy score and
t-test and analysis of variance with repeated measures and F-test,

determined congruency. Pearson correlations accompanied by

frequencies of the coeficients were used to examine within—family
relationships between the supported characteristics for dual parent
families. Analysis of variance was used to examine differences
between family types on the supported family characteristics.
Descriptive Pearson correlations examined supported characteristics by
family type and income.

Both approaches supported health status as a family
characteristic, neither supported intrinsic motivation, and there was
conflicting support of health behavior and family functioning. Other
findings were: a positive relationship between health status and
family functioning from the within-family perspective, no differences
between dual and single parent characteristics, and non-significant

correlations between the characteristics and income by family type.

ACKNOWLEDGEMENTS

This study was completed with the assistance and support of many
individuals including the participating families, the religious
leaders of numerous congregations in the Greater Lansing Community,
professional colleagues, representatives of local and state
organizations, my doctoral committee and my family.

I wish to take this opportunity to express my appreciation and
thanks -

To the families for their genuine interest in this study and
taking precious time from their individual and family activities
to participate.

To the religious leaders for their interest in this study and
their willingness to identify and contact families in their
congregations who met the study criteria and share this essential
information with me.

To the many representatives of local and state organizations who
guided the development of the Health Practices Survey used as
part of this study. Special thanks are extended to: Don Diskey,
Director of Health Education, Ingham County Health Department;
Dee Fitzgerald, Educator, East Lansing Public Schools; Mary Gray,
Program Director, Health, Safety and Youth Services, American Red
Cross, Lansing; Gabrille Kende, Nutritionist, Michigan State
University, College of Nursing; William Mehrens, Professor,
Michigan State University, College of Education; Pamela Minkley,
Supervisor, Sleep Disorder Program, Ingham Medical Hospital,
Lansing; Harry Perlstadt, President, Michigan Public Health
Association; Jim Perkins, Past President, Lansing Area
Congregations Together in Service; Don Sweeney, Project Office
for the Michigan Model of Comprehensive Health Education,
Michigan Department of Public Health; Jack Thrush, Chief, Health
Surveilance Section, Michigan Department of Public Health; and
Paul Vogel, Project Director for Measurement and Evaluation,
Michigan Model for Comprehensive Health Education.

To Diane Harte, Program Manager, Michigan Capitol Girl Scout
Council, Lansing, who facilitated the identification of leaders
and troops in the Greater Lansing Community who participated in
the development of the study questionnaire.

To the girl scout leaders and the scouts in troops #170, #215,

iv

#553, #555, #570, #600, who critiqued the Health Practices Survey
and the Self Care Inventory and assisted in pilot testing the
Health Practices Survey.

To Lee Anne and Jim Roman for their support, encouragement and
assistance with the development of the enclosed figures.

To Judy Pfaff for her valuable assistance with data entry, data
retrieval, and statistical consultation.

To the members of my doctoral committee: Dr. Margaret Bubolz,
Dr. Robert Griffore, and Dr. Thomas Luster, Department of Family
and Child Ecology faculty, and Dr. Cheryl Cox, College of
Nursing, University of Illinois at Chicago for their interest,
availability, guidance, and expertise.

To the chairperson of my doctoral committee and director of the
dissertation, Dr. Linda Nelson, for the interest and enthusiasm
she displayed for this study, for her encouragement and
commitment that I keep my proposed program schedule, for her
willingness to guide my progress, and for her many valuable
suggestions regarding my program plan, the organization and
content of the dissertation, and meetings with my committee. Her
support and expertise are greatly appreciated.

To my the members of my family, Gordy, Dana, and Irma and William
H. Sitz, my mother and father, for their love and support during
the past five years as I progressed toward completion of my
doctoral work. This goal is shared with them; their
contributions were many and very highly valued.

TABLE OF CONTENTS

LIST OF FIGURES .

LIST OF TABLES

CHAPTER I - INTRODUCTION.
Statement of the Problem .
Purpose of the Study .
Specific Research Objectives .
Hypotheses . . . .
Conceptual Framework .

The Ecological Family Framework .

The Interaction Model of Client Health Behavior .

The Development of Health Behavior in
Children

The Proposed Conceptual Framework .
Theoretical Assumptions.
Methodological Assumptions .
Theoretical Definitions.
Operational Definitions.

Significance of the Study.
CHAPTER II - LITERATURE REVIEW.
Conceptual Issues.

Theoretical Frameworks and Models .
The Epidemiological Perspective.
The Sociocultural Perspective.

vi

.xi

. xii

P0

.11

.14

.14

.16

.18

.19

The Psychosocial Perspective .
The Inter-relational Perspective .
The Community Perspective.
Dimensions and Variables.
Demographic or Personal Characteristics.
Social Influences.
Parental teaching of health behavior
through modeling, demonstration and/or
verbal behavior . . . . .
Parental influence on the development
of motivational attributes of their
offspring .
Family variables.
Environmental Influences .
Health Behavior.
Health Outcome .
Unit of Analysis.
Family Characteristics.
Methodological Issues.
Health Behavior Measurement .
Health Outcome Measurement.
Family Characteristic Measurement .
CHAPTER THREE - METHODOLOGY .
Research Design.
Sampling Procedure.
Data Collection .
Human Subjects Protection Procedures .
The Measurement Instruments.
The Health Self-Determinism Index (HSDI).

The Self-Care Inventory (SCI)

vii

.27
.29
.30
.31
.32

.33

.34

.37
.38
.40
.40
.42
.42
.43
.44
.44
.45
.46
.52
.52
.53
.57

.61

.63

.65

The Health Protective Behaviors Questionnaire
(HPBQ~Part 2)

The Health Practices Survey .

The Michigan Model for Comprehensive Health
Education. . . . . . . . . . .

Consultation from Health Professionals,
Educators, and Evaluators.

Pilot Testing.

The General Functioning Subscale of the McMaster
Family Device . . . . . . . . . . . . .

Demographic Information .
Scoring Procedures .
The Statistical Methods.
Data Analysis Procedures .
CHAPTER IV - RESULTS OF DATA ANALYSIS.
Description of the Study Sample .
Summary of Major Characteristics of the Sample.
Testing of the Hypotheses .

Hypotheses 1 through 4: Congruence of Family
Members’ Measures. . . . . . . . . . . .

Ho 1.

Ho 2.

Ho 3.

Ho 4.
Hypotheses 5 through 13: Relationships Between
Variables which are Congruent Within the Context
of the Family System .

Ho 5.

Ho 6.

Ho 11 .

Ho 12 .

viii

.67

.70

.72

.72

.74

.77

.79

.79

.80

.83

. 91
. 91
.105

.106

.107
.108
.108
.111

.113

.113
.114
.115
.115

.115

Ho 13 .
Hypotheses 14 through 18: Relationships Between
Selected Variables and the Family Realm Variables
of Family Type and Family Income .

Ho 14 .

Ho 16 .

Ho 17 .

Ho 18 .

Summary of the Hypotheses Testing .

CHAPTER V - OVERVIEW, DISCUSSION, CONCLUSIONS, AND
IMPLICATIONS.

Overview of the Study .
Discussion of the Findings.

Objective 1.

Objective 2.

Objective 3.
Conclusions .

Objective 1.

Objective 2.

Objective 3.
Implications of the Study .

The Continued Investigation of this Data Set .

Future Studies .

APPENDIX A - Introductory Letter to Religious Leaders.
APPENDIX B - Family Confirmatory Letter.

APPENDIX C - Consent Form.

APPENDIX D - List of Dates and Locations .

APPENDIX E - The Health Self-Determinism Index .

APPENDIX F

The Self-Care Inventory .

ix

.116
.116
.116
.118
.119

.120

.125
.125
.127
.127
.134
'.136
.138
.138
.139
.139
.140
.140
.143
.146
.148
.149
.150
.152

.154

APPENDIX G - The Health Protective Behaviors
Questionnaire-Part 2 .

APPENDIX H - The Health Practices Survey .

APPENDIX I - The General Functioning Subscale of the
McMaster Family Assessment Device

APPENDIX J - Demographic Information .

APPENDIX K - Preliminary Data.

LIST OF REFERENCES .

.159

.161

.172

.174

.177

.179

LIST OF FIGURES

Figure

1. The Interaction Medel of Client Health
Behavior . . . . . . . . . . . .

2. A Conceptual Framework for the Development of
Positive Health Behavior in Young Children .

3. A Conceptual Framework for Family and Individual
Characteristics Related to Health Behavior .

4. Diagrammatic Representation of the Study Design .

xi

Page

.10

.12

.15

.53

Tables

1.

10.

ll.

12.

13.

LIST OF TABLES

Summary of Church Participation by Low,
Medium, and High Categories of Percent of
High School Graduates. . . . . . . .

Summary of Statistical Procedures

Reliability Analysis of the HSDI, the FAD-7,
the SCI, the HPBQ-Part 2, and the HPS and
its Subscales by Subsamples. . . .

Marital Status of Parents in Dual and Single
Parent Families. . . .

Number of Family Members as Reported by
Mothers in Dual and Single Parent Families .

Family Income by Dual and Single Parent
Families .

Age in Years as Reported By Respondents in
Dual and Single Parent Families.

Ethnicity of Respondents in Dual and Single
Parent Families. . . . . . . . . . . .

Years of Education as Reported by Parents in
Dual and Single Parent Families.

Current Occupation of Respondents in Dual
and Single Parent Families .

Religious Preference of Respondents in Dual
and Single Parent Families .

Frequency of Attendance at Religious Services
or Functions During Past Five Weeks by

Respondents in Dual and Single Parent Families .

Perceived Overall Health Status of Respondents
in Dual and Single Parent Families .

xii

Page

.57

.81

.87

.92

.92

.94

.95

.96

.97

.98

.99

.99

100

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Long Term Impairment or Health Problem as
Reported by 41 Respondents in Dual and Single
Parent Families. . . . . . . . . . .

Comparison of Suggested Heights and Weights to
Reported Heights and Weights of Parents in Dual
and Single Parent Families . . . . . . . .

Comparison of Suggested Heights and

Weights by Age to the Heights, Weights, and
Ages Reported by Female Adolescents in

Dual and Single Parent Families.

Comparison of Suggested Body Heights and
Weights by Age to the Heights, Weights, and
Ages Reported by Male Adolescents in Dual and
Single Parent Families . . . . . . . . .

The t-test for Determining Congruence (Aggregate

Descriptive Statistics) on the Major Variables
by Dual and Single Parent Families .

Analysis of Variance with Repeated Measures for
Determining Congruence (Within-Family Factor)
by Dual and Single Parent Families .

Frequency of Pearson Correlations for Dual
Parent Families on Combinations of Major
Variables with Support for Congruence.

Family Mean Scores on Family Functioning,
Health Behavior, and Health Status by Family
Type . . . . . . . . . . . . . . . . .
Pearson Product Moment Correlation Coefficients
Between the Major Variables with Support for
Congruency and Income by Family Type.

Result of Hypotheses Testing.

xiii

101

102

. 104

105

109

. 110

. 117

118

.120

.121

CHAPTER I

INTRODUCTION

The adoption of a healthy lifestyle has been emphasized by the
media, educational programs, and health professionals during the
1980s. This is based upon growing empirical evidence that the
regular use of positive health practices, such as adequate nutrition,
sleep and exercise, sound dental practices, reduction of stress, and
the avoidance of smoking and alcohol will enhance personal health and
well-being and increase longevity (Belloc & Breslow, 1972; Palmore,
1970; and Pratt, 1971). The health behavior research conducted
during this period utilized divergent frameworks and models and
examined an array of potential influencing factors. It was assumed
that if the significant factors which promote the adoption of a
healthy lifestyle and the regular use of positive health practices
were identified, more effective and perhaps efficient intervention
strategies could be developed and implemented. These strategies
could then enhance the acquisition of more positive health practices,
increase the frequency of those already practiced, and facilitate a
higher level of health.

One influential factor frequently reflected in the health
behavior frameworks and models was the social unit of the family.

1

This conceptualization is consistent with family and child
development theorists and researchers (Andrews, Bubolz & Paolucci,
1980; Bruhn & Cordova, 1977; Bruhn & Parcel, 1982a; Crawford, 1971;
Doherty & McCubbin, 1985a; Kandzari, Howard & Rock, 1981; Upheld &
Harper, 1986) who view the family as the primary environment
responsible for the psychosocial, physical and spiritual development
of its individual members. Although the family has been identified
as an important factor, past investigations have generally fecused on
the role of the family when a specific health related problem exists,
e.g., the family and the diabetic child or the family and the
alcoholic adult. The influence of the family on a broad range of
psychosocial and behavioral health behaviors, e.g., practices related
to stress reduction, nutrition, sleep, recreation, substance use, the
home, and the use of the automobile, has been minimally investigated.
Investigative efforts have been hampered by inadequately developed
conceptual frameworks and models, by the lack of agreement about the
variables of interest, by a lack of clarity regarding the unit of
analysis, by measurement techniques of questionable reliability and
validity, and by a lack of analytical approaches appropriate when
data from two or more family members are obtained (Larzelere & Klein,
1987; Upheld 8: Harper, 1986).
Statement of the Problem

This research examines selected family and individual factors
which influence the adoption of health behavior by family members by
investigating the following questions:

1. Do the family characteristics of family functioning, family

motivation, family health behavior, and family health status

exist?

2. If they do exist, what are the relationships between these

characteristics within the context of the family?

3. If they do exist, what are the relationships between these

characteristics for different family types?

4. If they do exist, what are the relationships between these

characteristics and family income?

Purpose of the Study

The purposes of this study are: (a) to utilize a theoretical
framework reflecting family ecological, health behavior development,
and client-interaction perspectives, (b) to assess the presence of
family characteristics, based upon individual member characteristics,
by utilizing analytical approaches appropriate for relational data,
and (c) to strengthen the current data base regarding the influence
of selected aspects of the family realm (Beutler, Burr, Bahr &
Herrin, 1989) on the health behavior of its members. The following
research objectives and accompanying hypotheses resulted from the
investigator’s review of the health behavior and family research
literature, the need for exploratory studies as suggested by
theorists and researchers, and the investigator’s observations of
families as a community health nurse since the early 1960s.

Specific Research Objectives

1. Determine if there is congruence of family members’ measures

of health status, intrinsic motivation, health behavior

(psychosocial and behavioral), and family functioning.

2. Investigate the relationships between the family mean scores

of the supported family characteristics, e.g., health status,

intrinsic motivation, health behavior (psychosocial and

behavioral), and family functioning, from the perspective of the

individual family.

3. Describe the relationships between the family mean scores of

the supported family characteristics, e.g., health status,

intrinsic motivation, health behavior (psychosocial and

behavioral), and family functioning, and the family realm

(Beutler et al., 1989) variables of family type and family

income.

Hypotheses

Hypotheses 1-4 are directed at establishing the presence or
absence of the proposed family characteristics; the presence of a
family characteristic is supported when congruence of perceptions or
behavior is established. The investigation of hypotheses 5-18
depends upon the support or lack of support for these proposed family
characteristics.

Hypotheses for Research Objective 1 related to family members’
measures are:

1. Health status measures of the parent(s) and young adolescent

within the same family will be congruent.

2. Intrinsic motivation measures of the parent(s) and young

adolescent within the same family will be congruent.

3. Health behavior measures (psychosocial and behavioral) of

the parent(s) and young adolescent within the same family will

be congruent.

4. Family functioning measures of the parent(s) and young

adolescent within the same family will be congruent.

Hypotheses for Research Objective 2 related to each family are:
5. The family health status mean score of each family will
correlate positively with the family’s psychosocial health
behavior mean score.

6. The family health status mean score of each family will
correlate positively with each of the family’s two behavioral
health behavior mean scores.

7. The intrinsic motivation mean score of each family will
correlate positively with the family’s psychosocial health
behavior mean score.

8. The intrinsic motivation mean score of each family will
correlate positively with each of the family’s two behavioral
health behavior mean scores.

9. The intrinsic motivation mean score of each family will
correlate positively with the family’s health status mean score.
10. The intrinsic motivation mean score of each family will
correlate positively with the family’s family functioning mean
score.

11. The family functioning mean score of each family will
correlate positively with the family’s psychosocial health
behavior mean score.

12. The family functioning mean score of each family will
correlate positively with each of the family’s two behavioral
health behavior mean scores.

13. The family functioning mean score of each family will
correlate positively with the family’s health status mean score.

Hypotheses for Research Objective 3 related to all families are:

14. Dual parent families will have higher family mean scores on

health status than female-headed single parent families.

15. Dual parent families will have higher family mean scores on

intrinsic motivation than female-headed single parent families.

16. Dual parent families will have higher family mean scores on

family functioning than female-headed single parent families.

17. Dual parent families will have higher family mean scores on

health behavior (psychosocial and behavioral) than female-headed

single parent families.

18. The family mean scores of the family characteristics, i.e.,

family health status, family intrinsic motivation, family health

behavior (psychosocial and behavioral), and family functioning ,

will correlate positively with family income.

Conceptual Framework

Due to the purposes and objectives of this study and inadequate
conceptual frameworks specifically focusing on the influence of the
family on the health behavior of its members, components of three
conceptual models: (a) the ecological family framework (Andrews et
al., 1980), (b) the Interaction Model of Client Health Behavior (Cox,
1982), and (c) the development of health behavior in children (Bruhn
& Parcel, 1982a), were selected and combined to guide this study.
Each of these conceptual models is described and the selected
components identified. The resulting conceptual framework is
described in detail. The description includes a diagrammatic
representation, the associated theoretical assumptions and
definitions, and the selected methodological assumptions and

definitions.

The Ecological_§gmily,Framework

The ecological perspective refers to looking at the relationship
between organisms and their environment. The adaptation of this
perspective to the analysis of the family is referred to as family
ecology. Although various family ecological frameworks have been
proposed, the framework developed by Andrews et a1. (1980) was
identified as a significant contribution to the conceptualization of
the family and its members, and their relationship to the larger
community (Thompson & Chin, 1985).

Andrews et a1. (1980) acknowledged that a family ecosystem
framework incorporates general systems theory, the ecological
perspective, and the research findings from family scholars.
Therefore, their framework views the family as an ecosystem which
incorporates the interdependence of family members or subsystems, and
the interdependence of the family with the environment. Their
framework is not a singular general conceptual approach but a series
of conceptual segments for studying the family. This suggests that
the segments need further delineation so that their conceptualization
can be more fully developed.‘ While visualizations of some of these
segments were provided, this investigator believes the stated
assumptions rather than the visualizations are the strength of this
framework. Therefore, the visualizations are not included.

In general, the relationships between the major concepts are
logical, consistent and intertwined; the definitions are conceptual
and not operational in nature. Examples of assumptions from this
framework which are appropriate to the present study are: (a) the

family is an environment for its members and influences their

development, (b) the family exhibits family characteristics or
attributes in addition to being composed of members with individual
characteristics or attributes, and (c) the family and its members
each contribute to the external environment which in turn contributes
to the family’s and members’ level of well-being, e.g., income and
health.

Andrews et a1. (1980) suggest their framework has potential for
operationalization and for explaining and predicting family
relationships at the individual, family, and societal levels. They
indicate this outcome can be achieved by identifying the presence of
additional family characteristics or attributes, by exploring the
significant factors and processes involved, and by developing related
intervention strategies.

Thg_lgter§gtiog Model of Client Health Behavior

The Interaction Model of Client Health Behavior (Cox, 1982)
focuses on three areas: (a) the client as an individual, (b) the
client-professional interaction process, and (c) the health care
outcomes resulting from the interaction. While applicable to a broad
spectrum of health professionals, it was specifically developed to
provide the nursing profession with a predictive theoretical
framework which would facilitate the delivery of client care and
advance nursing knowledge through nursing research.

The Interaction Model of Client Health Behavior incorporates a
set of psychological, environmental, and sociological variables which
have been empirically supported by nonnursing perspectives on client

health behavior. However, its emphasis is on process, i.e., impact

of the client and professional encounter on the client’s health care
behavior.

Basic assumptions incorporated into the model are: (a) the
client is capable of making informed, independent, and competent
choices about personal health care behavior, (b) the choices are
affected by various aspects of the ”client’s singularity", and (c)
the client should have access to maximum control within the context
of the environment in determining personal health behavior and health
status (Cox, 1982, p. 3-4). Client singularity represents an
amalgamation of variables, i.e., demographic (personal)
characteristics, social influences, previous health care experiences,
environmental resources and the relationships between and among them.
It also incorporates intrinsic motivation, cognitive appraisal, and
affective response. The model suggests the influence of the major
elements and variables by incorporating feedback mechanisms utilized
in general systems theory. Figure 1 represents a visualization of
the model.

Two elements of the model have been selected and incorporated
into the proposed framework, i.e., the background variables
demographic characteristics, social influence, and environmental
resources) and intrinsic motivation. Cox (1982) states "...the
background variables operate over time within each client to produce
a specific health behavior" (p.6); they are considered to be
antecedents to subsequent variables. Motivation is viewed as an
important aspect of the model. Its incorporated variables of
competency and selfedeterminism are viewed as causal factors in

health behavior.

10

r

. use...
._ 2... we: .30. 5.3.13 :55: ezmso mo moo: maeoémhz. . m

 

 

 

 

.
'1
1‘ x83 $83252 “
- .
u " «3:63: A rV
" “ Eco—ececfcm ‘1
a " All].
. . .
. 8351950 3:88: .
28 5.? . m. " 35p A - 3.622 A; n 0
88523 . 35.32.. .
. .
. ooceceaxm eao‘n .
n m 5.8: Sosa... J m.
fits... I ..§_ "
03 D0 . . .
owsosscsg " .2"; o_ I t' 3992 A m
H 3.280 n
. l . n 0803:. r
c n 3.8” ‘1 N
.32.. " c825... " .
E
030 5.001 . m. H 53.01 H
.0 533 n u
u I c8930: "
. SS .
. «33.5 0.9... All.
E236:- oBSw . N . 0380.? " 33.28820 ‘1 . N
5.3: son... .6 u . " 028.858 AL
" " m03¢tfl> gsosmfiflm
- .
82)..” g .
03° ””58““ o N A u
.o co: ..
c3082:— accessed
2:830 .

5.8: .0 asses

_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 82.0 .o 8:956

_

 

 

 

 

 

28.395 .88 a season

 

 

 

11

The Development of Health Behgyior in Children

In 1981, 25 researchers involved in the investigation of health
behavior of children attended an invitational conference to identify
areas requiring future research, and to achieve consensus on
definitions, variables, and methodology to guide the research (Bruhn
& Parcel, 1982a). While these goals were not achieved,
interdisciplinary perspectives were shared and the need for more
rigorous research methods and innovative designs was recognized.
During the conference, four task forces were formed; each examined
one aspect of children’s health behavior: (a) family influences, (b)
developmental and psychosocial characteristics, (c) health behavior,
and (d) health status indicators.

The task group on family influences identified several
assumptions about the family to guide their discussion, developed a
conceptual framework on the development of positive health behavior
in children (Figure 2), identified demographic background variables
to assess or control in studies, and selected family variables which
could influence health behavior in children. While the framework is
incomplete, it represents the only available framework which '
specifically focuses on the development of health behavior in
children. Demographic background variables identified by the task
force which have been incorporated into the current study as sample
descriptive characteristics are: educational level of parents, gender
of child, ethnicity, birth order, employment status of parents,
religion, and age. Although income was identified as a demographic
background variable by the task force, it has been identified as a

family realm variable for this study. Family variables identified by

12

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Family '
Influences
Health Health Status
—'>I Behavior —" Indicators for
of Child the Child
Child —
Developmental
and Psychological
Characteristics
<1
Predisposino. Enabling. and
Reinforcim Factors Behavioral Factors Oumome Factors

 

 

 

FIGURE 2. A CONCEPTUAL FRAMEWORK FOR THE DEVELOPMENT OF
POSITIVE HEALTH BEHAVIOR IN YOUNG CHILDREN

[Bruhn & Parcel, 1982, p.143 - 241 ].

13

the task force which have been incorporated into the current study
are modeling and interaction patterns as evidenced by the frequency
of parental and child health behavior practices, family type, and
family functioning.

The task force on developmental and psychological
characteristics identified several assumptions to guide their
discussion about the cognitive and affective dimensions which could
influence health behavior in children. Intrinsic motivation was
identified as a specific important personal attribute. This position
further supported its inclusion as a major variable in the current
study.

The health behavior task force also identified several working
assumptions to guide its discussion. The group listed nine areas of
health behavior with each area containing a list of specific
practices that could be learned by children under 14 years of age.
”These behavioral areas were selected because there is
epidemiological data to indicate that young children are currently at
risk or the behaviors are likely to relate to future life-style and
risk factors" (Bruhn & Parcel, 1982a, p. 250). This list of nine
areas of health behavior and the corresponding list of specific
practices guide the definition of health behavior and thus data
collection.

The health status task force attempted to identify variables or
measures that would indicate the desired outcome had been achieved.
The group assumed that health behavior results in social, emotional,
cognitive, and physical outcomes. While several variables were

identified in each category, the variables of perceived health

l4

status, height and weight were selected for the current study; each
can be measured with accuracy and has been used in previous health
behavior research. Height and weight have been incorporated as part
of the sample’s descriptive characteristics; perceived health status
is investigated from the perspective of a potential family
characteristic.
The Proposed Conceptual Frgpgwork

The proposed model suggests that the family is a powerful
determinant of the health behavior of its members, that multiple
mechanisms of influence exist, that specific variables about the
individual are related to health behavior and in turn to personal
well-being, and that family characteristics or patterns develop and
therefore have the potential for repetition in future generations.
A diagramatic representation of the proposed framework which
incorporates assumptions, elements, and variables from the three
previously presented frameworks and models is presented in Figure 3.
The two sets of assumptions, theoretical and methodological, which
form the basis for this study follow.
Theoretical assppptions

l. The family ecosystem is composed of the environed unit,

i.e., the family system or the family, the family realm (Beutler

et al., 1989), the external environment, and the pattern of

interactions between them.

2. The family system is composed of interacting and

interdependent persons, each with their own distinctive

characteristics, who share some common goals and resources, and

share living space for a part of the life cycle.

15

THE FAMILY ECOSYSTEM

 

 

 

EXTERNAL ENVIRONMENT

 

FIGURE 3. A CONCEPTUAL FRAMEWORK FOR FAMILY AND INDIVIDUAL
CHARACTERISTICS RELATED TO HEALTH BEHAVIOR.

 

l6

3. The family system is a whole with identity, actions, and
characteristics of its own.
4. The maintenance of the family system and the development of
skills and shared or common components are facilitated through
the interaction of the family sytem and its subsystems with the
external environment.
5. The family system includes structure, process, and outcome.
6. The family realm (Beutler et al., 1989) influences the
development of its individual members or subsystems.
7. The family realm (Beutler et al., 1989) shapes attitudes,
values, expectations, decision-making, and behavior.
8. The family realm (Beutler et al., 1989) is the context for
the development of health behavior and influences health.
9. Family type and the family income are factors which
influence the family realm (Beutler et al., 1989).
10. Family health status, family intrinsic motivation, family
health behavior (psychosocial and behavioral), and family
functioning are unique characteristics of the family which can
be constructed and assessed.

Methodological assppptions
1. Respondents can accurately evaluate and report their health
status, the extent of their sense of competency and
selfedeterminism in health matters, and the psychosocial and
behavioral practices they implement on a daily or regular basis.
2. Respondents can accurately evaluate and report their
perception of the family’s ability to work together, i.e.,

family functioning.

17

3. Respondents can accurately evaluate and report

their personal characteristics of age, gender, religiosity,
ethnicity, marital status, educational level, occupation, and
employment status.

4. Family characteristics can be determined from the self
reports of the family’members.

5. By the age of 13, individual family members can learn a wide
variety of health behavior practices and perform them on a daily
or regular basis (Bruhn & Parcel, 1982a; Vogel, 1987).

6. The statified randomsampling of religious congregations of
different denominations in a specified geographic area will
yield a population of families with different religious
affiliations.

7. Individual family member paper and pencil questionnaires
provide privacy of response, delete the possibility of
interviewer biases, and facilitate data quantification.

8. The study questionnaire is developmentally appropriate for
the selected family members.

9. The Discrepancy Score and Conjunctive Models (Klein, 1984)
are conceptually appropriate for the investigation of the
presence of family characteristics related to health behavior.
10. Congruence of the family members’ measures on health status,
intrinsic motivation, health behavior (psychosocial and
behavioral), and/or family functioning indicates a family

characteristic is present.

18

11. The specific analytic approaches selected are guided by the

intent of this study, the data, and statistical and financial

resources.

12. The creation of a summary score, i.e., individual member

mean score or family mean score, reflects the variable of

interest and is a reasonably good measure of the variable.
Theoretical definitions

1. Family. A system composed of two or more interrelated

interdependent individuals (i.e., members) who share common

goals and resources including living space for some part of the

life cycle (Andrews et al., 1980).

2. Health behavior. These personal psychosocial and behavioral

practices implemented by a family member on a daily or regular

basis which influence well being (Beutler et al., 1989; Harris &

Guten, 1979; Parcel, Bruhn, & Murray, 1984).

3. Intrinsic motivation. The extent of the family member’s

sense of competency and self-determinism in health matters (Cox,

1985).

4. Health status. The family member’s current assessment of

personal overall well-being (Ware, 1976).

5. Family characteristic. A unique property of the family as a

unit resulting from the relationships between the members and

reflected in the shared or common (congruent) components of the

members (Andrews et al., 1980).

6. Contextual variables. Descriptors of the family and its

members.

19

Operational definitions

la. Dual parent family. The biological mother and father and
their offspring, the eldest being between 13 and 16 years of age
or the adoptive mother and father and their adopted children,
the eldest being between 13 and 16 years of age and adopted
during the child’s first year of life.

lb. Single parent family. The biological mother and her
offspring with the eldest being between 13 and 16 years of age
or the adoptive mother and her adopted children, the eldest
being between 13 and 16 years of age and adopted during the
child’s first year of life.

2a. Psychosocial health behavior. The evaluation of the family
member’s psychosocial practices as determined by the score
attained on the SelfeCare Inventory (Hagar, 1984).

2b. Behavioral health behavior. The evaluation of the family
member’s behavioral practices as determined by the score
attained on each of two measures: the Health Protective
Behaviors Questionnaire - Part 2 (Harris & Guten, 1979; Salovey,
Rudy & Turk, 1987) and the Health Practices Survey (Wood, 1989).
3. Intrinsic motivation. The evaluation of the family member’s
sense of competency in health related matters as determined by
the score attained on The Health SelfaDeterminism Index (Cox,
1985).

4. Health status. The evaluation of the family member’s
perception of personal well being as measured by excellent = 4,
good = 3, fair = 2 or poor = 1 (Were, 1976; Ware, Brook, Davies

& Lohr, 1981).

20

5a. Family intrinsic motivation. The summary mean score
resulting from the combination of family members’ mean scores
within the same family (Discrepancy Score Model) or the
within-family factor (Conjunctive Model) as measured by The
Health SelfeDeterminism Index (Cox, 1985).

5b. Family psychosocial health behavior. The summary mean score
resulting from the combination of the family members’ mean
scores within the same family (Discrepancy Score Model) or the
within-family factor (Conjunctive Model) as measured by the
SelfeCare Inventory (Hagar, 1984).

5c. Family behavioral health behavior. The two summary mean
scores resulting from the combination of the family members’
scores within the same family (Discrepancy Score Model) or the
two within-family factors (Conjunctive Model) as measured first
by the Health Protective Behaviors Questionnaire-Part 2 (Harris
& Guten, 1979; Salovey, et al., 1987), and second by the Health
Practices Survey (Hood, 1989).

5d. Family functioning. The summary mean score resulting from
the combination of the family members’ scores within the same
family (Discrepancy Score Model) or the within-family factor
(Conjunctive Model) as measured by the General Functioning
Subscale of the McMaster Family Assessment Device (Epstein,
Baldwin & Bishop, 1983).

6a. Family income. The highest response given by an adult,
mother or father, in the dual parent family or the mother in the

single parent family, as determined by intervals of $5,000

21

beginning at "Less than $10,000", and ending with "More than
$100,000".

6b. Family type. The classification of the family as determined
by the presence of either: (a) the mother and father referred to
_as the dual parent family, or (b) the mother referred to as the
single parent family.

6c. Age. The chronological age as determined by the birthdate
provided by the family member.

6d. Gender. The sex of the family member as measured by their
written response of female or male.

6e. Religiosity. The family member’s selection of religious
affiliation from the following categories: Catholic, Protestant,
Jewish, and Other (Specify), and the specified frequency of
attendance at religious services or functions at intervals of
one from ”0" to ”more than 5” during the month preceding the
study.

6f. Ethnicity. The family member’s selection from the following
categories: White not Hispanic, Black not Hispanic, Hispanic,
Asian, Pacific Islander, American Indian, or Other (Specify).
6g. Marital status. The family member’s selection from the
following categories: Married, Divorced, Widowed, Separated,
Never married, or Other (Specify).

6h. Educational level. The family member’s selection of the
last year of school completed as measured at yearly intervals
from kindergarten through post doctoral study.

6i. Occupation. The category, e.g., professional, management,

proprietor (small business), clerical worker, sales worker,

22

skilled craftsman or foreman, operative unskilled laborer,

student, farmer, homemaker and other, as determined by the Job

title and description given by the family member.

65. Employment status. The family member’s selection from the

following categories: self-employed full-time, selfeemployed

part-time, employed by other full-time, employed by other

part-time, unemployed, student part-time, student full-time, and

other (Specify).

6k. Height. The family member’s self report of how tall he or

she is without shoes as measured in feet and inches.

61. Weight. The family member’s self report of how much he or

she weighs without shoes as measured in number of pounds.

Significance of the Study

Past investigations regarding health behavior of the individual
and the family suggest that further exploration of selected potential
influencing factors be pursued. Specifically, the investigation of:
(a) health status (Hunt, McEwen & McKenna, 1986; Ware et al., 1981),
(b) intrinsic motivation (Cox, 1985; Cox, Miller & Mull, 1987), (c)
health behavior (Bruhn & Parcel, 1982a; Doherty & MCCubbin, 1985b;
Pratt, 1976), (d) family functioning (Pratt, 1973; 1976), (e) family
type (Duffy, 1988; LovelandrCherry, 1986), and (f) family income
(Bruhn & Parcel, 1982a), could provide important additional
information. One approach proposed by family researchers (Fisher,
Kokes, Hanson, Phillips & Rudd, 1985; Klein, 1984; Schumm, Barnes,
Bollman, Jurich & Milliken, 1985) to investigate the influence of the
family as a unit is the statistical combining of data from multiple

members in the same family. Minimal evidence of the application of

23

this approach to data describing family functioning and health
behavior exists; there was no evidence of its application to data
describing intrinsic motivation in health related matters or personal
health status. Utilization of this exploratory approach could reveal
the presence of previously unknown unique family characteristics or
family attributes. Confirmatory results could clarify present
conceptualizations of the influence of the family on the health
behavior of its members and provide direction for future family

focused health behavior research.

CHAPTER II

REVIEW OF LITERATURE

Systematic investigations conducted since the 1960s have
revealed that the incorporation of positive health behavior into an
individual’s lifestyle will be reflected in improved personal health
and increased longevity (Belloc & Breslow, 1972; Palmore, 1970; and
Pratt, 1971). Although this relationship has been established, the
critical factors which promote the adoption of positive health
behavior are still being clarified and identified by researchers from
a broad spectrum of disciplines, e.g., medicine, nursing, sociology,
health education, and psychology. The breadth of these
investigations will be reviewed from the perspective of the
associated conceptual and methodological issues.

Conceptual Issues

Conceptual issues identified and discussed include: the
variation of current health behavior theoretical frameworks and
models, the dimensions and variables, health behavior, health
outcome, unit of analysis, and the potential presence of family

characteristics related to health behavior.

24

25

Theoretical Frameworks and Models

Since the early 1960s various conceptual models and frameworks
have been utilized in investigations to account for individual health
behavior. They ”differ considerably in their theoretical
perspective, the types of health behavior they wish to explain and
the labels employed to describe their respective dimensions and
variables" (Becker & Maiman, 1983, p.558). Initially many of the
models reflected a health service perspective and were directed
toward increasing understanding of illness behavior of the
individual, fer example, the responses of the newly diagnosed
diabetic patient to suggested dietary changes. However, many of the
current models incorporate a health promotive perspective, i.e., the
factors which influence positive health behavior and personal health.
In general, the frameworks and models employed to predict and explain
the presence or absence of health behavior reflect one or a
combination of the following perspectives: epidemiological,
sociocultural, psychosocial, inter-relational, and community.
Selected frameworks and models utilizing these perspectives are

reviewed to demonstrate their diversity.

The Epidemiological nggpectivg

 

The standard factors of the epidemiological model, i.e., host,
agent and environment, were modified by Suchman (1975) in his
investigation of the acceptance or rejection of an accident
preventive measure among Puerto Rican sugar cane cutters. The host
factor represented personal characteristics which influenced
readiness. The negative and positive attributes of the preventive

measure represented the agent, and the social factors, e.g., mass

26

media and social participation, represented the environment.
Such-an (1975) stated that the model modification enabled him to
investigate the probl- of public acceptance of health innovations.
The Sociocultural Perspggtive

Fabrega (1976) developed a behavioral framework for studying
human disease. His model demonstrates how an individual ”processes
information about illness and makes decisions designed to alleviate
his condition" (Fabrega, 1976, p.200). Assumptions of the model are:
(a) there is a relationship between the behavior of the person and
the cause of the disease, (b) the behavior resulting from the disease
can be evaluated systematically, (c) persons with the same disease
may exhibit different behaviors, (d) personal orientation directs the
individual’s action regarding treatment, and (e) the disease can be
viewed within the context of personal adaptation which occurred
through an evolutionary process. Culture, the mechanism for the
transmission of rules, beliefs and strategies to the individual or
group, influences the individual’s definition of disease, treatment,
and health. Although culture is acknowledged as a significant
influence, Fabrega (1976) also recognized that patterns change and
thus produce stress to the individual or group. Therefore, he
proposed the use of a system perspective to demonstrate the
interrelatedness of the individual and the disease, and the
individual and other social systems. Three paradigms for approaching
disease were proposed: (a) task action-the daily routine activities
of the individual, (b) role enactment-the impact of social position
and role on the individual’s obligations and duties, e.g.,

occupational, familial, recreational and religious, and (c)

27

psychological-—the individual’s values and cultural assumptions.

Nine stages of decision-making that an individual goes through during
an illness were also described. Becker and Maiman (1983) indicated
that although the variables may be difficult to operationalize, that
use of this model in future studies could assist in the

identification of the key variables.

The Psychosocial Perspggtive

0

Theorists and researchers have been most supportive of the
psychosocial perspective, therefore, three different examples of are
provided.

The Health Belief Model (HEM) has been a.maJor organizing
framework for explaining and predicting acceptance of health care
recommendations (Hosenstock, 1974). The model was developed in the
early 1950s by a group of social psychologists at the 0.8. Public
Health Service to understand why people did not utilize preventive
and screening tests available for the early detection of disease
(Hosenstock, 1974). The maJor areas of emphasis of the model are: (a)
personal desire to avoid illness or to get well, (b) the individual’s
perceived threat of illness, and (c) the impact of personal action to
reduce the threat. The dimensions of the model are perceived
susceptibility, perceived severity, perceived benefits and perceived
barriers. According to Hosenstock (1974, p. 332) "The combined
levels of susceptibility and severity provided the energy or force to
act and the perception of benefits (less barriers) provided a
preferred path of action." However, a cue to action, internal to the
person (symptoms) or external to the person (e.g., interpersonal

interactions or mass media communication) was believed to be

28

‘ necessary befbre the decision-making process was initiated. Although
this model focused primarily on personal attitudes and beliefs, it
also recognized that demographic, sociopsychological and structural
variables influence perception (Becker &,Maiman, 1983; Janz & Becker,
1984).

Janz and Becker (1984) reviewed 46 investigations conducted
between 1974 and 1984 which used the HPM. They concluded that there
was substantial evidence which supported the HBM’dimensions as
important contributors to the explanation and prediction of
individual health behavior. They also noted that these studies
focused on the major model dimensions and therefore provided very
limited information about the influence of other variables.

Another example of a model in this category was proposed by
Kulbok (1985). This model, the Resource Model of Preventive Health
Behavior, recognizes that the investment of personal social and
health resources into an individual’s lifestyle facilitates good
health habits. Two major assumptions guide this model: (a)
preventive health behavior is influenced by personal health and
perceptions of health, and (b) health is a positive multidimensional
process. Social resources are identified as the achieved education
while income and health resources are defined as the multidimensions
of health which include physical, mental and social well-being.
Preventive health behavior is defined as the voluntary actions an
individual undertakes regardless of the perceived or actual health
status.

A third model reflecting this perspective was proposed by

Doherty & McCubbin (1985a). The Family Health and Illness Cycle

29

Model represents a biopsychosocial portrayal of the family’s efforts
to reduce illness risks and to manage and adapt to illness. Phase 1,
family health promotion and risk reduction, emphasizes the influence
of the fellowing factors: environmental, social, psychological and
interpersonal. These factors within and surrounding the family
promote health and reduce associated risks of family members. Phase
2, family vulnerability and illness onset, refers to life events and
experiences which could initiate an episode of illness, e.g., divorce
or death. Phase 3, family illness appraisal, denotes the efforts of
the family to understand the member’s symptoms, i.e., to determine
the seriousness of the complaint. Phase 4, family acute response,
addresses the ”immediate emotional and interactional aftermath for
the family of the illness experience and the family’s appraisal of
it” (Doherty & McCubbin, 1985a, p.8). Phase 5, family and the health
care system, focuses on the decision of the family to handle the
illness or seek external assistance. Phase 6, family adaptation to
illness, addresses the illness’s long-term impact upon the family.
According to Doherty & McCubbin (1985a), the terminology of this
symptom-oriented approach can be altered from the primary focus of
illness to the area of health promotion. They speculate that the
phases of the model represent the areas of future research on the
family and health care.
Thevlnter-relgtiongl Pergpective

A good example of this perspective is The Interaction Model of
Client Health Behavior (Cox, 1982, 1985). Although described
previously in relation to the conceptual framework for this study, a

brief review follows. This model was developed to explain the

30

relationships between the individuality of the client, the
client-provider relationship, and the client outcomes and behaviors.
The model assumes that clients are capable of making informed,
competent and independent decisions about their health behavior.
Client characteristics, the background variables (demographics,
social group influence, previous health care experiences and
environmental influences), intrinsic motivation, cognitive appraisal,
and affective response, are viewed as antecedents to the interaction
and outcome. The client-provider relationship includes the variables
of affective support, health information, decisional control and
professional/technical support; variables specified in health outcome
are utilization of health care services, clinical healthestatus
indicators, severity of health care problem, adherence to the
recommended-care regimen, and satisfaction with care (Cox, 1982,
1985) .
The Community Perspggtive

Rodgers’ (1984) multivariate model proposed that a systematic
process be utilized in the assessment, prediction, and intervention
in communities with respect to aggregate health risks. Health was
viewed as a dynamic concept reflective of an ecological relationship;
the community was viewed as the frame within which health and risk
occur. Although the model was intended as a tool to assist nurses to
improve health care, it could be applied by or in conjunction with
other individuals and groups interested in health and the community.
The process outlined follows: (a) obtain information about the
community of interest, (b) analyze the data, assets and liabilities,

(c) classify problems identified by formulating community or

31

aggregate diagnoses, (d) develop alternative strategies for resolving
diagnoses, (a) select and implement the best strategy, and (f)
evaluate the impact of the intervention. Specific information to
acquire for the assessment includes: (a) population characteristics,
e.g., age, sex, marital status, education, income, causes of death,
and (b) community characteristics, e.g., government, public services,
sources of energy, economic status, communication vehicles,
utilities, recreation, religion, transportation, health/medical
services, social services, educational services, and power and
influence.
Dimensions and Variables

Current frameworks frequently identify different dimensions and
variables. Dimensions refer to the large fecal areas of the
frameworks. In general, in health behavior research, these
dimensions are reflected as: (a) demographic or personal
characteristics, (b) social influences, (c) environmental influences,
(d) health behavior, and (e) health outcome. Within each of these
dimensions, the variables of interest are specified. Based upon a
thorough examination of current health behavior frameworks, Becker
and Maiman (1983) concluded that the variables of interest within
each of these dimensions tend to differ with each framework. They
further noted that some of the variables included had empirical
support while others had no empirical support. In addition, they
found that some of the dimensions omitted variables which had
empirical support. Therefore, the comparison of study findings is

complicated.

32

Although many different variables have been investigated, this
study focuses on selected variables appropriate to its purpose and
research questions. They have been organized under the general
dimension categories identified.

Demogppphic or Personal Characteristics

The demographic or personal characteristics of the individual,
e.g., level of education, gender, marital status, occupation, age,
and religiosity, have been recognized as important in the
investigation of health behavior (Berk-an & Breslow, 1983; Campbell,
1987; Roth-an, 1978). Socioeconomic differences are associated with
differences in health behavior (Berk-an & Breslow, 1983; Coburn &
Pope, 1974; Fuchs, 1974; Kirscht, 1983). The influence of income
will be reviewed in greater detail under "family variables". A near
linear relationship between years of formal education, occupational
"level" and dental check-ups was reported for male adults (Kasl &
Cobb, 1966). According to Coburn and Pope (1974), education and age
were noted as two of the best predictors of frequency of physical
examinations, dental checkups, and obtaining polio vaccinations by
adult males. Kirscht (1983) reported that smoking has a negative
association with educational achievement while good nutritional
practices are positively associated with socioeconomic status.
Studies of gender differences reveal that women, rather than men,
typically obtain dental care, immunizations, and asymptomatic
check-ups (Lairson & Swint, 1978; Nathanson, 1977). Married adults,
age 55 and older, who attended activities at a senior citizens’
center scored significantly higher on health practices than their

nonmarried counterparts (Hubbard, Muhlenkamp & Brown, 1984). Black

33

men and women were found more likely to engage in highrrisk practices
than other ethnic groups (Berk-an and Breslow, 1983).
Social Influences

The dimension of social influences recognizes that the health
behavior of the individual is influenced by social relationships
(Bruhn, 1988; Mullen, 1983). Although this dimension is routinely
incorporated into the frameworks, past investigations have focused
primarily on personal characteristic variables rather than social
relationship variables. Investigations of social relationship
variables, such as parents, siblings, the fuily as a unit, other
relatives, individuals in the broader community, the media,
organizations, and institutions have been a recent interest of
researchers. While each social influence variable is recognized as
important, the influence of the family has been selected as the key
social influence variable in this investigation.

The family and specifically the parents have been recognized as
playing a significant role in the development of health behavior in
children (Bruhn, 1988; Bruhn & Cordova, 1977; Bruhn & Parcel, 1982a;
Crawford, 1971; Doherty & McCubbin, 1985b; Haydon, 1987; Holman,
1983; Handzari et al., 1981; Mullen, 1983; Sallis & Nader, 1988).
While family and parental influence are believed important, the
conceptualization of the factors to be investigated has been unclear.
As a result, the investigation of their impact has been limited and
the findings inconsistent (Duffy, 1988). However, specific
mechanisms of social influence by the family or parent(s) which have
been identified for further investigation include: (a) parental

teaching of health behavior through modeling, demonstration, and/or

34

verbal instruction, (b) parental influence on the development of
motivational attributes of their offspring, and (c) family variables

(Bruhn & Parcel, 1982a; Mullen, 1983).

Parental teaching of health behavior thropgh modeling,
demonstration, anglor verbal instruction. Study results about
parental influence are mixed. Mechanic (1964) assessed the degree of
influence mothers had in shaping their children’s health attitudes
and behaviors. His data on 350 mother-child pairs suggested that
maternal influence on both was considerably less than hypothesized.

A lack of support for the assumption that parental modeling behavior
affects children’s health behavior was also noted by Bruhn and Parcel
(1982b). They interviewed low income, predominantly ethnic minority
single mothers of 202 children ranging in age from two to four years.
Their analysis showed no relationship between the mothers’ health
behavior, mothers’ locus of control, or the mothers’ value of health
and the children’s practice of health and safety behavior.

However, 59 female single parents who identified behaviors which
promoted health or prevented disease indicated that the positive
health behaviors of a balanced diet, rest, and exercise originated in
their childhoods (Duffy, 1986). Also, a follow-up study by Mechanic
(1979) of childhood symptomatology revealed that the young adults
reported fewer symptoms and that they remembered their parents
emphasized selfecare and health promotive practices. Pratt (1973 and
1976) used self reports of two parents and one offspring between the
age of 9 and 13 to investigate selected personal health practices,
i.e., sleep, exercise, elimination, dental hygiene, smoking, alcohol

consumption, and nutrition. She reported that one of the factors

35

influencing children’s positive health is the positive health
behavior of the parents. She also documented that a large majority
of parents purposefully explain personal health procedures to their
children, i.e., proper foods to eat, how to keep clean to maintain
body health, and the effect of adequate sleep.

Further support for the position that the development of a
child’s resources and capabilities early in life are significant
factors for later health practices was provided by Dielman, Leech,
Becker, Rosenstock, Horvath & Radius (1982). Their examination of
the relationship between health beliefs and health behaviors of
parents and their children, age 6 to 17 years, suggested that at
least two areas of children’s health behavior - cigarette mmeking
and snacking between meals - are influenced by parental behavior and
to a lesser extent by parental beliefs.

Another example of the influence of parental smoking behavior on
children follows. The social factors related to cigarette smoking
among 2,156 school aged children between the ages of 12 and 18
residing in Muscatine, Iowa, were surveyed via a confidential
self-report questionnaire by Lauer, Akers, Massey & Clarke (1982).
Abstinence and smoking by school aged children was closely related to
the smoking behavior of the child’s parents and friends. They
further stated: "The group of respondents with parents who smoke had
approximately twice the proportion of regular smokers, compared with
those respondents whose parents never smoked” (Lauer et al., 1982, p.
423).

There is also growing documentation that parents exert influence

over dietary and physical activity patterns (Haydon, 1987). Mothers

36

have been found to play a significant role in matters of nutritional
habits and attitudes (Litman, 1974). Hertzler (1988) suggested that
family dynamics can enhance or deter the use of food and nutrition
information by children. She stated that the presence of aggression,
marital dishanmony or contempt between parents is the basis for the
inappropriate or inconsistent use of food and thus influences the
child’s food behavior and coping skills. Specific examples of
parental influence on dietary patterns follow.

Kintner, Boss & Johnson (1981) conducted an exploratory study
utilizing mailed questionnaires and follow-up telephone interviews of
42 young families to determine the strength and direction of
relationships between Moos Family Environment Variables and family
food intake. While they concluded that their findings were tentative
and that further investigation was needed, they indicated that the
eating behaviors of family members revealed information about the
nature of the family environment. Children’s preferences were
related to fathers’ preferences (Bryan & Lowenberg, 1958) and
familiarity with foods, as controlled by the family, was identified
as a determinant of a child’s eating patterns (Philips & Kolasa,
1980).

Gottlieb & Chen (1985) and Butcher (1983) noted in their
multivariate studies that parental factors were the strongest
correlates of a child’s physical activity. Ross, Pate, Caspersen,
Damberg & Svilar (1987) also reported that there is a strong
correlation between the exercise habits of parents and those of their
children. A survey of physical activities was administered to the

parents of children in the first through the fourth grades who

37

participated in the National Children and Youth Fitness Study II.
Parents rated their children’s activity levels, hours of television
watched, and involvement in community sponsored physical activity
programs. Parents also provided personal information regarding a
rating of their own activity level, number of days per week they
engaged in vigorous activity for 30 or more continuous minutes, and
the number of days each week they exercised at least 20 minutes with
their child. The self-rated activity level of the parent(s)
resembled the teachers’ ratings of the activity levels of the
children, their students. This study also revealed that fewer than
30 percent of the mothers and fathers participated in appropriate
physical activity and that the frequency with which parents exercise
with their children is directly related to the frequency of parental
exercise.

Parental influence on the developpgnt of pptivational attributes
of their offspring. The influence of the family and parent(s) on the
development of health beliefs, locus of control, and behavioral
attribution has also been acknowledged as important (Bruhn & Parcel,
1982a). However, when beliefs, locus of control and attribution are
used as motivational constructs in studies on the individual, they
have not been found to predict consistently who does or does not
practice positive health behavior (Cox et al., 1987). During the
1980s there has been growing support for the position that intrinsic
motivation may be a significant contributor to an individual’s health
behavior (Cox, 1985; Cox et al., 1987; Deci, 1980). In particular,
health self-determinism as described by Cox (1985) has been

identified as a correlate of positive health behavior, health status,

38

and general well-being in adults. There is also evidence that health
self-determinism reflects socioeconomic patterns. For example, older
adults, less-educated individuals and males demonstrated less
intrinsic motivation and were responsive to external forces which
promote positive health behaviors (Cox et al., 1987; Cox & Wachs,
1985). When measured with children, intrinsic motivation varies
systematically with children’s learning styles and achievement (Cox,
1988).

Family variables. Relationships, those between and among family
members, those between the members and the environment external to
the family, and those between the family as a unit and the external
environment, have also been acknowledged as important factors in
facilitating health behavior (Mullen, 1983). Pratt (1976) suggested
that personal health practices depend upon the structure of the
family and the relationships among all the family members. She
further concluded that the pattern of family relationships is
significantly related to the composite health practices of the family
group. She reported that positive health practices were present when
regular and varied interaction occurred among family members, when
parents encouraged personal autonomy, and when family members
participated in activities outside the home. A study by Laskey &
Eichelberger (1985) revealed that self-care practices were displayed
by children whose parents provided a progressive, developmentally
appropriate, transference of health self-care decision making. .
Although the influence of family relationships, also referred to as
family functioning, has been identified as an important contributing

factor to health behavior, the inclusion of a family functioning

39

measure has been virtually absent in studies examining health
behavior (Campbell, 1987).

The variables of family type, birth order, and number of
children have also been identified as contributing influences of
health behavior (Campbell, 1987; Mullen, 1983). However, only one
study comparing positive health behavior by family type, i.e.,dual
parent and single parent, (LovelandrCherry, 1986) and only one study
examining positive health behavior of the female-headed single parent
family (Duffy, 1986) could be located. LovelandrCherry (1986) used a
convenience sample of 41 white middle-class families, 21
female-headed single parent families and 20 dual parent families,
with at least one child between the ages of 8 and 14. She found that
the children’s scores, the mother’s scores and the total family unit
scores on personal health practices did not differ by family type.
However, a greater degree of score variance for children from single
parent families existed. Although Duffy’s (1986) investigation of
the health behavior of the single parent family focused on the
mother’s report of specific practices rather than on the comparison
of practices between mother and child, she noted that the mothers
identified social support as a major determinant of the their
practice of health behavior. A recent publication by Duffy (1988)
urges researchers to further investigate health behavior from the
perspective of family type. The literature suggesting that the
single parent family experiences different stressors supports this
position (Herman & Turk, 1981; Coletta, 1983; Herman, 1977; Horowitz

& Purdue, 1977; McLanahan, 1983; McLanahan & Booth, 1989; Smith,

40

1980). A comparison of health behavior by family type could suggest
needed resources specific to family type.
Environmental Influences

A range of environmental influences (social, political, legal
and economic factors) external to the individual and family have been
identified in relation to health behavior. Social institutions,
e.g., schools, churches, political and social organizations, the
health care industry, and the work setting, influence the individual
and the family (Mullen, 1983). Since the focus of this study is the
family’s influence on health behavior, many of these environmental
influences are not appropriate. However, one economic variable,
family income, the total income available to the family, has been
selected because it exerts significant effects on the health and
health behavior of family members, e.g., the type and quantity of
contact with health care providers, the use of preventive health
services, control of environmental hazards, and children’s
selfeesteem (Child Trends, Inc., 1980; Select Panel for the Promotion
of Child Health, 1981; U.S. Department of Health and Human Services,
1980). Income relates positively with the adoption of positive
health behavior (Berk-an & Breslow, 1983; Coburn & Pope, 1974; Kasl &
Cobb, 1966; Kirscht, 1983) and the lack of financial resources has
been cited as an important deterent of positive health behavior by
female headed single parent families (Duffy, 1986).
.Health Behavior

Surveys undertaken on various populations provide much of the
information known about health behavior. Findings can be related to

health-care—based preventive practices or to those practices which

41

are selfeinitiated. Health-care-based preventive practices include
asymptomatic routine physical and dental examinations, pap tests,
immunizations, and a variety of health screenings, e. g., breast
examination by a health professional. Practices outside the health
care system which are selfeinitiated and suggest personal control
include safety practices, e.g., household, automotive and pedestrian
practices and the reduction of environmental hazards, nutritional
habits, personal care and hygiene, sleep and relaxation, physical
exercise, and the avoidance of substance use, e. g., drugs, tobacco
and alcohol (Bausell, 1985; Berkman a Breslow, 1983; Harris & Guten,
1979; Kasl & Cobb, 1966; Hirscht, 1983; Salovey et al., 1987). As a
result, past investigations have focused on one or multiple, similar
or dissimilar, practices. The studies previously described support
this conclusion.

Discussion among and between health behavior researchers has
focused on the dimensionality of health behavior. The evidence is
mounting in support of multidimensionality (Green, 1970; Harris &
Guten, 1979; Kulbok, 1985; Langlie, 1977). For example, Green
(1970), and Langlie (1977, 1979) found that medically based
preventive behaviors are modestly related (correlation from .2 to .4
range). However, Kulbok (1983) stated that a review of 30 relevant
studies revealed that the concept of health behavior was often
ambiguous and that little attention was given to its dimensionality.
Health behavior is conceptualized from both perspectives in this
study. The nine dimensions or factors of health behavior and their
associated practices identified through concensus by a group of

researchers interested in the development of health behavior in

42

children (Bruhn & Parcel, 1982a) and one additional dimension,
personal care, identified by the investigator as a result of the
literature review reflect the multidimensional perspective. The
unidimensional perspective is reflected in the second approach to
assess health behavior, a 30 item scale (Harris & Guten, 1979).
Health Outcome

The dimensions and specific variables previously discussed
contribute to the health outcome, the end result. Since there is no
concensus regarding the definition of health outcome, wide variation
of the health outcome variables, e.g., mortality, morbidity (severity
of illness), risk status, and personal health status, exists across
health behavior studies. This lack of conformity in health outcome
variables has also contributed to the use of different measurement
approaches. Both subjective and objective assessment of these
outcomes has occurred: (a) selfereport of symptoms, perceived
well-being and health practices, and (b) measures of physical,
psychological and social functioning, e.g., personal height, weight,
blood pressure, coping mechanisms, and relationships. Approaches
used to measure health outcome are discussed under Methodological
Issues.

Unit of Analygis

The individual adult has been the primary unit of analysis in
past investigations of health behavior. Although researchers have
attempted to focus on the influence of the family and specify that
the family is the unit of analysis, data have generally been sought
from one individual, usually the wife or mother. In those instances

when comparisons between parent-child behaviors have been made, the

43

information on both mother and child has been obtained from the
mother and the comparison restricted to the mother-child dyad.
Rarely has infonmation about the health behavior of three or more
family members been sought.

Family researchers question the accuracy of viewing individual
and dyad data as representative of the family. They propose that if
the family is to be the unit of analysis than information from and
about multiple members needs to be secured (Schumm et al., 1985).
Only two studies, LovelandrCherry (1986) and Pratt (1976), could be
located in which three family members, both parents and their child,
each provided information about a broad array of personal health
practices. One additional study (Blaxter & Paterson, 1982) was
located in which three family members representing three different
generations provided information about their respective health
practices.

Family Characteristics

In the 1980s researchers interested in the family suggested that
the family may reflect unique characteristics. They proposed that
these characteristics, also referred to as family properties, family
attributes, family scores, aggregate family components, or family
interactive measures (Brown & Kidwell, 1982; Ezell, 1982; Gillis,
1983; Klein, 1984), reflect the existence of a pattern between family
members. However, the identification of family characteristics has
been limited. The characteristics of family functioning and marital
quality/satisfaction have been the primary focal points of past
investigations. Ezell (1982) extended the focus of the measure of

social well-being by investigating ”quality of life” as a family

44

characteristic. The investigation of family functioning, health
behavior, health status, and intrinsic motivation as family
characteristics could provide additional insight into the influence
of the family. Although previous studies (LovelandeCherry, 1986;
Pratt, 1976) have alluded to the presence of a health behavior family
characteristic, the term family health behavior was not specified nor
included in the stated purposes of their investigations. However,
the study of three generations by Blaxter & Paterson (1982)
identified the potential for "a family health culture" and Schor,
Starfield, Stidley & Hankin (1987) investigated a perceived family
variable which they described as family health care utilization. No
evidence could be located demonstrating that either intrinsic
motivation or health status has been conceptualized as a potential
family system characteristic.
Methodological Issues

The discussion of methodological issues focuses on the
:measurement of health behavior, health outcomes, and family
characteristics.

Health Behavior Measurement

The measurement of health behavior as perceived from a
developmental perspective is in its infancy. As a result, current
health behavior measurement instruments differ widely in regard to
the questions asked and the format used. Contributing factors
include: (a) the differing purposes of the studies, (b) the skills
and philosophy of the researchers, (c) the lack of a common
conceptualization about what specific practices should be included in

health behavior, (d) differing viewpoints about how health behavior

45

should be measured, i.e., direct/indirect observation, selfereport,
other report, and (e) the absence of developmentally appropriate
reliable and valid instruments. These factors have contributed to
inadequate information about the development and psychometric
properties of’many potentially valuable and useful measurement
instruments.
Health Outcome Measurement

The approach selected to measure health outcome has been
dependent upon the purpose of the investigation, the proposed
conceptual framework, and the availability of appropriate measurement
instruments. Several measurement approaches of health outcome are
available. One approach used to measure health status was described
by Harnly and Williams (1987). Healthy functioning was measured by
accumulating data in six different areas: (a) medical and health
events, (b) perceived health, (c) psychosocial well-being, (d)
physical activity, (e) nutrition, and (f) the use of substances,
i.e., alcohol and drugs. Belloc & Breslow (1972) and Breslow &
Enstrom (1980) examined current physical health status in relation to
the presence of seven positive health practices (seven to eight hours
of sleep, controlling one’s weight, regular exercise, limiting
alcohol consumption to fewer than five drinks per sitting, never
having smoked cigarettes, eating breakfast almost daily, and seldom
eating snacks). Berkman & Breslow (1983) examined five of these
practices, i.e., cigarette smoking, physical activity, alcohol
consumption, obesity and sleeping patterns, and mortality. Although
Ware (1976) measured health status by evaluating perceived global

personal well-being, Berkely, Israel and Stokes III (1987, p.17lS)

46

proposed that the evaluation of health outcome include: (a) current
health status, i.e., physical, mental and social functioning, and (b)
risk status, i.e., "the best estimate of the probability that an
individual in a given health state at one time period will move to a
state of dysfunction over time."

The development of health outcome measures involves problems
with definitions, reliability, validity, sensitivity and
applicability. Some established instruments developed to measure
health are: (a) the Cornell Medical Index (Board-an, 1951), (b) the
measure of health perceptions (Ware, 1976), (c) the Sickness Impact
Profile (Bergner, 1978), and (d) the Nottingham Health Profile
(McEwen, Hunt and McKenna, 1987). According to McEwen et al. (1987),
the instruments are frequently too long or complicated, have scoring
systems which reflect the investigator’s view not the respondent’s,
have measures which are focused too narrowly, and derive a single
score from multiple item responses.

Family Characteristic Measurement

Although obtaining data from multiple family members, also
referred to as relational data (Fisher et al., 1985), has been
recognized as highly desirable in the assessment of a family
characteristic, the approach to be taken in the analysis of the data
has been a point of discussion among researchers. While there is
general consensus that the investigation of congruence of relational
data can reveal a structural measure of the variable, there are
significant differences of opinion regarding the method to be used to

examine congruence.

47

Fisher et a1. (1985) proposed that four assumptions provide the
basis for relational data: (a) individual data have been collected,
(h) each scale or dimension comprises more than one item of response,
(c) the data suggest continuous or ordinal level measurement, and (d)
two or more individual scores in each family are available.

Klein (1984) suggested that five models or techniques to assess
the presence of a family characteristic exist. He further suggested
that any or all of the models may be applicable and that the
application of more than one model may reveal which is most
appropriate for the data. The five models are:

l. The Additive Model: the sum or average of the scores is viewed as
the best estimate of the family characteristic. If the number of
informants varies from family to family, then the average score of
all respondents is used to standardize the measure across families.
This model lends itself to those behaviors and characteristics which
have an objective and concrete quality.

2. The Discrepancy Score Model: the discrepancy (the property of
dispersion) between reports is taken as the measure of the family
characteristic. If there are only two respondents, the difference or
ratio reflects this model. A standard deviation or mean difference
score would be appropriate for multiple respondents. Since the model
assumes that no systematic measurement error exists in the
distribution of reports, it would be possible to discover different
types of families.

3. The Disjunctive Model: although this model treats reports as
equally reliable and valid, it assumes there are multiple discrete

realities in the family. The reports from multiple repondents are

48

analyzed independently but theoretically linked to antecedent and
consequent variables. This model is most appropriate when data are
selfereported attitudes of members or perceptions of the attitudes of
other members.

4. The Weighted Model: since it can be argued that biased and
measurement errors can vary across respondent groups, this model

. utilizes specified criteria to weigh some reports as more reflective
of reality. This is similar to the Additive Model, but reports do
not have equal value. This model appears to be most appropriate when
differences in the credibility of responses is suspected.

5. The Conjunctive Model: This model combines reports according to
their convergence. It assumes that reports are equally valid and
reliable and that multiple discrete family realities exist. One or
several covariance-based grouping techniques can be utilized, e.g.,
analysis of variance with repeated measures, factor analysis, and
cluster analysis. This model is utilized when families are
conceptualized as integrated groups or when information is collected
from multiple family members.

Klein (1984, p. 10) reViewed 1,268 articles published in the
gppppal of Marriage and the Family during the 24 year period from
1959 through 1982 to assess which of these models had been used with
data collected from multiple family members. These articles were
assumed to be representative of family focused investigations.
Content analysis of these articles revealed the following: (a) few of
the articles reported collecting data from two or more family
members, (b) approximately half of these studies attempted to

manipulate the data in a meaningful way, (c) the Discrepancy Score

49

and Disjunctive Models were most frequently selected, (d) the
variables of interest and social units of analysis were limited, and
(e) ”the systematic comparison of alternative strategies was
uncommon."
The appropriateness of these models has been the focus of
considerable debate. According to Fisher et a1. (1985), there are
several problems in the use of the arithmetic mean (Additive Model)
as a relational score. They contend that: (a) it may not be
conceptually meaningful, e.g., mean scores do not reflect differences
of age among family members or differences in power or influence of
family members, (b) the distribution of mean scores of several
families may be less than that of the individual scores, (c) the
differences between or among the contributing scores are not taken
into account, and (d) the order of the scores is not considered.
They concluded that while mean scores may severely reduce the
information contained in the original scores and potentially distort
the data, the mean may be appropriate when the discrepancy between
the scores is small and a similar range of scores is reflected within
other families in the sample.

The use of the simple mean discrepant score (Discrepancy Score
Model) can be conceptually meaningful but also presents
methodological problems (Fisher et al., 1985). For example: (a) the
discrepancy could occur anywhere along a scale, i.e., between 2 and
12 or 48 to 58 in a 60 point range, (b) the score tends to be less
reliable and therefore reduces the attainment of statistical levels
of significance, and (c) no additional information beyond

correlations of family member scores and the dependent variable is

50

available. A different statistical approach is suggested in the
analysis of data from three or more family members when using the
Discrepancy Score Model, i.e., the calculation of a family mean score
by summing each.member’s mean minus the family mean (Klein, 1984).

The Disjunctive Model has been used extensively to assess dyad
rather than triad data. Klein (1984) suggested that researchers have
used aggregate scores to describe samples rather than using theory to
link the aggregated scores from multiple family members to antecedent
or consequent variables. Although there has been been some support
for the use of the Weighted Model, concern arises from its assumption
that the weighting of the members’ data may be inaccurate. Since the
weighting is determined by the investigator, it may represent
investigator bias and therefore not present a realistic picture of
the variable of interest within the family context.

Although there is growing support for the use of the Conjunctive
Model, the approach has been viewed as complex due to the degree of
sophistication needed to conduct the statistical analyses. It is
possible that this has been a deterent in its use. One approach
which has been strongly supported by Ball, McKenry, and Price-Bonham
(1983) is the use of analysis of variance with repeated measures.
This approach investigates the covariance between family dyads
simultaneously and calculates a within-family factor which indicates
support or non-support of congruency of multiple family member data.

This brief review of the models (Klein, 1984) suggests that the
investigator’s selection of the conceptual model and statistical

approach is dependent upon the theoretical framework, the scales

51

employed, the availability of appropriate statistical packages, and
the analytic expertise of the researcher.

Multi-member family data, i.e., relational data, can be used to
examine the perceptions of family members of a common event or the
behavior of family members. This data approach has been primarily
used to examine marital satisfaction, and minimally used to examine
family functioning and health behavior; no evidence was found to
suggest that relational data have been used to examine either
perceived health status or intrinsic motivation. The investigations
which examined health practices of parents and their children
(Loveland-Cherry, 1986; Pratt, 1976) will be considered further in

the discussion of this study’s findings.

CHAPTER III

METHODOLOGY

This chapter presents a description of: (a) the research design
including the sampling procedure and data collection, (b) the human
subject protection procedures, (c) the measurement instruments, (d)
the scoring procedures, (e) the statistical methods, and (f) the data
analysis procedures.

Research Design

The purposes of this cross-sectional design of multiple members
in dual and single parent families were threefold: (a) to explore the
presence of family characteristics by comparing family members’
perceptions of personal health status, personal motivation, personal
positive health behavior and family functioning, (b) to investigate
selected relationships between the identified family characteristics,
and (c) to describe selected relationships between the family
characteristics and the family realm variables of family type and
family income. The family, defined as dual parent and single parent
families each with a young adolescent between the ages of 13 and 16,
was the unit of analysis. Each parent and adolescent completed an
individual questionnaire which incorporated the same measures of the
variables of interest. A diagrammatic representation of the study

52

53

design is provided in Figure 4.

Measures Cppparison Groups
Dual parent Single-parent
family family
Father Mother Child Mother Child
I. Health status
1. Were Index x x x x x
II. Intrinsic motivation
l. HSDI x x x x x

III. Psychosocial and
behavioral health
behavior
1. The SelfeCare x x x x x
Inventory (SCI)
2. Health Protective x x x x x
Behaviors
Questionnaire-Part 2
(HPBO-Part 2)
3. Health Practices
Survey (HPS) x x x x x
IV. Family functioning
1. General Functioning x x x x x
Subscale of the
McMaster Family
Assessment Device
V. Demographic data x x x x x

Figure 4: Diagrammatic Representation of the Study Design

The guiding theoretical definitions and their corresponding
operational definitions are provided in Chapter I (pp. 18~22).
Sgppling Procedure
Due to the specificity of the family characteristics, the
limitations of time and resources, and the difficulties inherent in
locating families through sources in the Greater Lansing Area
Community, e.g., clinics, health department, and schools, a

stratified random sample of religious congregations in a specific

54

geographic portion of this same area was selected as the vehicle for
family identification. The fact that religious affiliation and
frequent church attendance have been associated with positive health
behavior (Berkman & Breslow, 1983) was not viewed as a deterrent
factor since all participating families had a religious affiliation.
While a control group of families with no religious affiliation was
initially considered, the identification of such families who met the
proposed criteria was viewed as problematic and therefore not
pursued.

Initially this investigator contacted the past president of the
Lansing Area Congregations Together in Service (Lansing A.C.T.S.) and
received the organization’s 1987—88 DIRECTORY OF CHURCHES. One
hundred and fifty-one churches representing 63 different
denominations were listed within the following selected geographical
boundaries: (a) the Ingham County, MI line on the North, (b) Waverly
Road on the West, (c) Jolly Road on the South, and (d) the city of
East Lansing’s line on the East. These boundaries were selected
because they coincided with the 1980 census tract boundaries (U. S.
Bureau of the Census, 1980).

In order to secure a stratified random sample of these churches,
the following process was used. A map with the census tract
boundaries and census tract numbers was used to determine the
location of each church by census tract (U. 8. Bureau of the Census,
1980). Once the census tract number was identified, the selected
items used for stratification, i.e., per cent of high school
graduates and median household income were secured for each census

tract (U. S. Bureau of the Census, 1980). Then each church’s name,

55

address, telephone number, census tract number, the percent of high
school graduates, and the median household income of the census tract
were noted on a reference card.

Since there was wide variation in the percent of high school
graduates and the median household income, categories which reflected
the range of the census data were constructed. The categories for
the percent of high school graduates were: (a) low (49.4 X through
63.7 X), medium (66.0 X through 77.3 X), and high (81.0 X through
100.0 X). The categories for the median household income were: low
($4,212 through $13,494), medium ($14,163 through $19,245), and high
($20,077 through $37,238). A comparison of the percent of high
school graduates to median household income by census tract revealed
that low, medium, and high categories of the percent of high school
graduates were usually accompanied by the corresponding median income
category. Therefore, only the percent of high school graduates by
census tract was used for the selection process. Of the 151
stratified churches, 45 churches were identified in the low category,
62 churches were identified in the medium category and 44 churches
were identified in the high category.

Fifteen churches, five per category, were then randomly selected
by the investigator. This was followed by a telephone call by the
investigator to each church to obtain the name and title of the
religious leader and to verify the church’s address. An introductory
letter (Appendix A) was then forwarded to each religious leader. The
content of the letter included: (a) the intent of the study, (b) who
was eligible to participate, (c) what participation meant, (d) the

incentives for participation, and (e) the investigator’s name and

56

phone number. The letter was followed several days later by a
telephone call to the religious leader by the investigator to review
the information in the letter, answer questions, and determine if the
leader would assist in the identification of the families.

These 15 contacts revealed that a greater number of churches
would need to be contacted. Although these religious leaders
indicated that they would be interested in assisting with the
identification of families, they also indicated that there were few
families with the desired characteristics in their congregations.
They stated that the parents of first born children, now between the
ages of 13 and 16, were part of the generation who had left the
church during the late 1960s and early 1970s. Since the goal of
approximately 30 to 35 dual parent families and 30 to 35 single
parent families appeared impossible to achieve with this group of 15
randomly selected churches, this investigator decided to continue to
select churches randomly in the same configuration, five churches in
each category for a total of 15, until the desired number of families
was identified. The investigator also decided to discontinue
contacting a church after four unsuccessful telephone attempts and to
select another church representing the same category. It was assumed
that this selection process would yield the number of families needed
and that they would be representative of families in other churches
in the specified geographical area.

As a result of this selection process a total of 120 of the 151
stratified churches were randomly selected. Of this number 81
leaders indicated support and interest in the study, 11 leaders

indicated they did not wish to participate, 23 leaders could not be

57

contacted by telephone, and five churches were found to have no
currently listed telephone number. Table 1 reflects this
distribution, and the distribution of all the stratified churches by
the low, medium, and high categories of percent of high school
graduates. Of the 81 churches where the leaders indicated support
and interest, only 39 yielded names of families with the
characteristics of interest. The distribution of these churches by

the low, medium, and high categories was 8, 18, and 13 respectively.

Table 1.
Summary of Church Participation by Low, Medium, and High Categories
of Percent of High School Graduates.

 

Church Participation Categories of High School Graduates

 

Low Medium High Total

 

 

Interested 22 30 29 81
Not interested 4 3 4 11
No reponse l3 7 3 23
Other 1 3 1 5
Not selected 5 19 7 31
Total 45 62 44 151

 

Qagaggollectiop
The criteria for family participation were influenced by: (a)
the lack of available information about health behavior in dual
parent and single parent families, (b) the perceptions of educators

and researchers that there are age related health behaviors which are

58

relatively orderly, predictable and measureable (Bruhn & Parcel,
1982a, 1982b; Vogel, 1987), and (c) the investigator’s perception of
the availability of families with the characteristics of interest in
the specified geographic area.

(The 39 churches yielded a potential pool of 120 families, i.e.,
20 single parent families and 100 dual parent familes. While 16 of
the single parent families agreed to participate, data were collected
from only 13. Although 65 dual parent families agreed to
participate, data were collected frem only 60. Reasons given by the
families who agreed but did not participate were: "We can’t work it
into our schedule after all," "We’ve changed our minds," and "My
child has gone to stay with a relative." The total number of
participating families, 73, represented 35 different churches. The
distribution of these churches by the low, medium, and high
categories of percent of high school graduates was 8, 16, and 11
respectively. The data were collected in Ingham County, Michigan
between June 1, 1989 and September 3, 1989.

It was noted early in the family identification phase (June,
1989) that the number of single parent families identified was
considerably lower than anticipated. After consultation with
committee members, the investigator decided that churches should
continue to be sampled, and that families should continue to be
contacted through August, 1989. It was anticipated that this would
assure an adequate number of dual parent families while providing
increased opportunity for the identification of the single parent

families.

59

The actual mechanism of family identification was determined by
either the religious leader, or the leader in consultation with an
advisory board. Usually the leader would contact the family to
explain the study and determine the family’s interest in
participating and then provide the investigator with the family’s
name, address, and phone number. However, several leaders provided
this information directly to the investigator without prior contact
with the family. Upon receipt of this information, the investigator
contacted each family by telephone and talked with a parent. These
conversations included an explanation of the study’s intent, a review
of the criteria for selection, and an assessment of the family’s
interest in participating in the study. If the parents expressed
interest, they usually stated that they wished to discuss it further
with the other members who would be involved. A date and time for a
follow—up telephone call regarding their participation was then
arranged.

When a family indicated it would participate, a confirmatory
letter (Appendix B) was forwarded to the family. A consent form
(Appendix C) which each parent and adolescent participating in the
study was to complete prior to data collection, a list of dates and
locations where all the family members could complete their
individual study questionnaires on the same selected date (Appendix
D), and a stamped self-addressed return envelope were also enclosed.
The family members were asked to complete the consent form, check the
selected date and location on the list and return both to the
investigator at least one week prior to the selected date. This

allowed the investigator to know the number of families to expect at

60

each location and make last minute arrangements for appropriate
accommodations, i.e., room size, tables and chairs. Each family was
also contacted by telephone by the investigator three to five days
before the day selected to insure its participation. In return for a
family’s participation and as agreed upon with the family in the
consent form, the investigator forwarded a $10.00 contribution on
behalf of the family to the church of their choice following their
participation.

The locations, the data collection sites, were the physical
facilities of churches which had identified families for the study.
Since families indicated during the telephone conversations that
evenings during the week were best, most of the sites were scheduled
Monday through Thursday at 7:30 p.m.. A total of 28 dates involving
17 different locations throughout the geographical area were
scheduled and utilized. Seven families requested a different
location; therefore, the investigator met five families in their
homes and two families at the Lansing Public Library, Lansing, MI.
This approach allowed the investigator to confirm the participation
of each family by identifying it by name upon arrival, to disseminate
and collect data, to standardize the introductory remarks about the
completion of the study questionnaire, and to answer questions. It
also controlled discussion and comparison of responses between family
members. Although five family members completed the questionnaire in
approximately 30 minutes and two family members completed the
questionnaire in one hour and 15 minutes, 199 family members

completed the questionnaire in approximately 45-50 minutes.

61

Human Subjects Protection Procedures

There were no obvious physical or psychological risks to the

individuals and families participating in this study. Although a

potential legal risk did exist due to questions about alcohol and

drug use, the following procedures were devised to assure anonymity

to the individuals and families. The procedures were reviewed by the

University Committee on Research Involving Human Subjects (UCRIHS),

Michigan State University, March, 1989.

1. Only the investigator had contact with the families to
confirm their participation and mail summary study findings.

2. The investigator was present at each data collection site to
assure proper data collection approaches were followed.

3. Although the names of the participating families were known,
the information the individuals provided remained anonymous.

4. The family name was used only to confirm that the family
present was the family which agreed to participate and that a
signed consent form had been received. Additional consent forms
were available as needed.

5. The adolescent in each family selected a family packet from a
box of packets prepared for the dual and single parent families.
A packet was composed of three envelopes for the dual parent
family and two envelopes for the single parent family.’ A
questionnaire was in each envelope.

6. Each participating family member received an envelope with a
questionnaire booklet. The booklet was coded so that the
information could be later related to others in the same family,

i.e., by family type, by number, and by individual status. Dual

62

parent families were identified as D1, D2, D3,...060 and single

parent families were identified as 81, $2, S3,...Sl3. The

status of the individual was identified as F father, M =
mother, and A = adolescent. Examples of the complete code are
DlF, DlM, and DlA for a dual parent family and 81M and 81A for a
single parent family.
7. The initial page in the questionnaire requested that the
respondent refrain from placing his or her name on the
questionnaire and stated that the respondent was free not to
answer any item.
8. Each participant was asked to insert the completed
questionnaire into its envelope and place the envelope in a
collection receptacle upon exiting the site. This receptacle
was a sealed box with a slot on the top. Envelopes were
collected in the box until no additional envelopes could be
inserted by the participants.
The Measurement Instruments
Each participating parent and adolescent completed a study
questionnaire which was composed of several measurement sections.
The sections are presented in the following order: (a) The Health
SelfeDeterminism Index (HSDI), (b) The SelfeCare Inventory (SCI), (0)
The Health Protective Behaviors Questionnaire - Part 2 (HPBQ-Part 2),
(d) The Health Practices Survey (HPS), (e) The General Functioning
Subscale of the McMaster Family Assessment Device (FAD-7), and (f)
Demographic Information which includes Health Status (WI). The
criteria used in the selection of the measures were: (a) the method

for data collection matched the purpose and nature of the study, (b)

63

the measure was applicable regarding age and educational
appropriateness, (c) the measure exhibited reasonable psychometric
properties of reliability and validity, and (d) the measure
demonstrated economy of measurement regarding dollar cost,
preparation time, and ease to take and score (Nunnally, 1978;
Windsor, Baranowski, Clark & Cutter, 1984).

The Health Self-Detegpipism Index (HSDI)

The HSDI (Appendix E), designed to measure intrinsic motivation
in health behavior, was developed by Cox (1985). It was derived from
the author’s clinical nursing practice, from current theoretical
perspectives on motivation and health and specifically from Deci’s
cognitive evaluation theory (Cox, 1985). Motivation is viewed as
"an important antecedent variable and correlate of the client’s
cognitive and affective responses to a health concern, the type of
health care intervention and interaction expected by the client, and
the client’s health outcomes subsequent to this intervention" (Cox,
1985, p. 177).

The HSDI demonstrates the multidimensionality of motivation
through its four subscales: self-determined health judgments,
selfedetermined health behavior, perceived competency in health
matters, and internal-external cue responsiveness. It is comprised
of 17 Likert format items, alternately worded between an intrinsic
and extrinsic orientation, and is scored and analyzed in the
following manner:

Each intrinsically worded item is scored on a scale from

1 to 5; a score of 5 indicates the maximum intrinsic

response and a score of 1 indicates the maximum extrinsic

response. Similarly, each extrinsically worded item is

on a 5-point scale where 5 indicates the maximum
extrinsic response and 1 indicates the maximum intrinsic

64

response. For analysis, all extrinsically worded items

are reverse scored (e.g., 1:5, 5:1) so that all items are

scored in the same direction (e.g., 5=intrinsic,

1=extrinsic) (Cox, 1985, p.179).

A total HSDI score is obtained by adding all item scores. As a
result the range of scores is from 17 (most extrinsic) to 85 (most
intrinsic). .

This selfeadministered paper-and-pencil measure can be easily
completed in seven to ten minutes, was designed for an eighth grade
reading level and has demonstrated highly acceptable alpha
coefficients across six studies: .84 (N=199), .80 (N=68), .87 (N=54),
.83 (N=55), .81 (N=72), and .78 (N=379) (Cox et al., 1987). A
two-week test-retest correlation of .86 was also obtained (Cox &
Wachs, 1985). Factorial validity, the presence of the subscales, was
confirmed in two studies (Cox, 1985; Cox et al., 1987).

The examination for covariance of the HSDI and its subscales and
sociodemographic variables revealed that the respondent who scores as
intrinsically motivated on the subscales of health matters, judgment,
and internal-external cue responsiveness, is more likely to be
younger, female, and better educated. On the behavior subscale, age,
gender, and income have been identified as the strongest predictors,
i.e., older people, males, and respondents with less income score
more extrinsically. In addition, it was found that women and
individuals with higher incomes and education project competence in
their health decisions and behaviors (Cox et al., 1987).

One study (Macius, 1985) demonstrated that the HSDI total score
and the competency subscale were significant predictors of the amount

and frequency of exercise, nutrition and sleep and that they

explained 27X of the variance in the positive behaviors practiced.

65

Cox et al. (1987) reported that as the functional status and level of
well-being of at-risk elders increased, the elders achieved higher
scores on the HSDI which indicated they were more intrinsically
motivated.

Thp_§elfegggeAInventogy (SCI)

The Self-Care Inventory used in this study is a modification of
the original SCI. Information on the development of the SCI and
subsequent modifications by this investigator follow.

The SCI as reported by Hager (1984) is a 77 item selfereport
paper and pencil questionnaire which assesses the frequency with
which a person uses health behavior of a psychosocial nature. The
items focus on habits which can be learned, are positive and are
health oriented. The respondent selects and circles a response from
a range of responses for each item. The responses on a five point
Likert scale are never, rarely, sometimes, most ofthe time, and
always. They correspond to the awarding of 0, l, 2, 3, and 4 points
respectively. This means that a score of zero indicates the least
positive use of a psychosocial health practice while a score of four
connotes the most positive use of a psychosocial health practice. A
total SCI score is secured by adding the points of all items circled
by the respondent and dividing the total number by the number of
items for which a response is given.

Item selection was based upon: (a) a review of literature which
included social support, stress and coping, family health and
prevention, and (b) the input from a panel of experts. The SelfeCare
Theory of nursing proposed by Oren and a model of prevention proposed

by the 1978 Task Panel on Prevention, an adjunct panel to the

66

Commission on Mental Health, were the SCI’s major theoretical bases
(Hagar, 1984). The SCI items were initially organized into self-care
categories reflected in clinical situations, i.e., cognitive,
structuring, self-nourishment, assertiveness, expression, physical
stress reducers, receptivity and sources of support.

A pilot study with a sample of 34 college senior nursing
students indicated that the SCI could be completed in approximately
15 minutes and identified items requiring word alteration to increase
clarity (Hagar, 1984). Test-retest reliability was conducted to
determine instrument stability with another sample of 76 college
students; the overall result was an r = 0.68 while the reliability
for each item ranged from 0.47 to 0.92. A full inter-item
correlation matrix was constructed to evaluate the validity of item
placement by theoretical categories. While the alpha coefficients
for each category ranged from 0.66 to 0.85 indicating moderately high
internal consistency within categories, the categories were not found
to be distinct from each other. This suggested that the organizing
categorical approach was not valid. The author suggested focusing on
the individual health habits or all the habits as a unidimensional
construct in future studies utilizing the SCI. Internal consistency
measures, via Cronbach’s alpha, secured twice on a third sample of 64
college senior nursing students at a year and a half interval were
0.888 and 0.915 respectively.

Although the psychometric properties of this measurement
instrument have not been fully explored, it was the only instrument
located which assessed an array of psychosocial positive health

practices. Since the unit of interest in this investigation was the

67

family, Hagar (personal communication, November, 1988) suggested that
all items assessing sexual relationships be deleted and that the
language be assessed for its appropriateness for the young
adolescent. The items assessing sexual relationships were deleted
and the remaining items were reviewed for language appropriateness by
the investigator and an eighth grade English teacher in a local
school district. Words deemed inappropriate for this grade were
replaced as needed. The point system was also changed from 0 through
4 to 1 through 5 to reflect the point system used in the other
selected health instruments. The revised SCI composed of 69 items
was then completed by nine seventh and eighth grade girl scouts in
communities adjacent to the specified geographic area. Suggestions
from these scouts resulted in additional language modifications. The
resulting 69 item SCI (Appendix F) was used in this study.

Ihg_§ggl§p Protective Behaviors Quastionpgira (HPBQ-Part 2)

The Health Protective Behaviors Questionnaire (HPBQ-Part 2) is
the second part of a two part questionnaire developed by Turk, Rudy,
& Salovey (1984) and modified by Salovey, at al. (1987). It is based
upon a large exploratory survey study conducted by Harris and Guten

(1979) designed to explore the concept of health protective behavior
and its empirical dimensions. Using a stratified-cluster-sampling
design of households in a large metropolitan area, 842 randomly
selected repondents were interviewed and asked to identify activities
they performed which protected their health. Health protective
behavior was defined as "...any behavior performed by a person,
regardless of his or her perceived or actual health status, in order

to protect, promote, or maintain his or her health, whether or not

68

such behavior is objectively effective toward that end" (Harris &
Guten, 1979, p. 18). Open ended questions and a card sort consisting
of 30 protective health behaviors representing a broad sample of
behaviors were used to gather the data. Social desirability effects
were reduced by having the respondents perform a double card sort.
Based upon an open ended question regarding practices performed,
three types of behavior concerning everyday health habits were most
often reported: nutrition, sleep/rest and exercise. The result of
the card sort was similar. The mean number of behaviors perfOrmed
was 12. As a result of cluster analysis of the card sort data, five
factors emerged that included 18 of the 30 items. The factors were
labeled health practices, safety practices, preventive health care,
environmental hazard avoidance, and harmful substance abuse. Harris
and Guten (1979) concluded that while the clusters were not composed
of highly related items "...they are the ’best’ clusters identifiable
from among the 30 card-sort items" (p.23).

The card sort of the 30 health protective behaviors was
converted into a self-report questionnaire by Brown & MCCreedy
(1986). They asked 386 respondents to check one of these responses
for each health behavior: "always performed", "almost always
performed", "sometimes perfOrmed", and "never performed". Responses
of "always performed" and "almost always performed" were awarded one
point while responses of "sometimes perfbrmed” and "never perfOrmed”
were awarded zero points. Therefore, the range was from 0 to 30. No
additional information regarding the combining of responses was
given. A Cronbach’s alpha of .76 for the study indicated reasonable

reliability of the measure. The mean number of behaviors performed

69

was 17.2; the behaviors of nutrition, sleep, and exercise were
practiced most frequently. Factor analysis was not repeated.

The Health Protective Behavior Questionnaire (HPBQ) as discussed
by Salovey et al. (1987) represents another modification of the
original 30 item card sort developed by Harris and Guten (1979).

Part 1 is designed to assess general attitudes about health
protective behaviors while Part 2 is designed to assess the frequency
of the health protective behaviors. Each contains a list of the
original 30 health protective behaviors. However, a 7-point scale
with end point designations of "very frequently = 7" and "never = 1"
was adopted. When the points for all items in each part are summed
and divided by the number of items with a response, two total scores
result.

Following the administration of the HPBQ to 180 respondents,
their attitude ratings of the behaviors were factor analyzed to
determine item clusters. Four factors, labeled as General Safety
Practices, Weight Control, Rest and Relaxation, and Medical

Avoidance, were identified. Items were reviewed for "

...convergent
validity (defined as a correlation with the factor 0.45) and
discriminant validity (considered satisfactory if the magnitude
between the highest and second highest factor loading was 0.15)."
(Salovey et al., 1987, pp.198). Ten items did not display adequate
discriminant validity and were not incorporated into further
analysis. Due to inadequate item to scale correlation, another item
was deleted.

Internal consistency of the individual scales was evaluted by

Cronback’s alpha. The alpha coefficient for Safety Practices (.85)

70

was favorable, while those for Weight Control (.68), Rest and
Relaxation (.53), and Medical Avoidance (.57), were less
satisfactory. The authors concluded that these scales appear to be
measuring distinct health-protective constructs because
intercorrelations among the scales were substantially lower than the
reliability estimates. An alpha coefficient for the total scale was
not stated.

Although the HPBQ-Part I focused on attitude, the HPBQ-Part 2
focused on frequency of health behavior; the latter was viewed by
this investigator as the most relevant of the two measures for this
study. No information detailing its psychometric properties was
cited by Salovey et a1. (1987). The HPBQ-Part 2 (Appendix G) was
used in this study because: (a) it incorporated the majority of the
areas and practices identified by Bruhn & Parcel (1982a), (b) it
included practices appropriate for the young adolescent and the
parent(s), and (c) it could be completed quickly and easily.

The Health Practices Survey (RPS)

Although the HPBQ-Part 2 included 30 positive health practices,
all the practices of interest as identified by the researchers
concerned with children’s health behavior (Bruhn & Parcel, 1982a)
were not included. Therefore, a second instrument focusing on the
measurement of behavioral health practices appropriate for the young
adolescent and the adolescent’s parent(s) was sought.

This investigator concluded that previously developed measures
were not appropriate for this study. This conclusion was the result
of an extensive review of references describing health behavior

questionnaires for adults and/or the young adolescent (Michigan

71

School Health Association & Michigan Health Council, 1983; Office of
Disease Prevention and Health Promotion, Health Information Center,
1987; Rodale, 1984; Solleder, 1986; Vogel, 1987) and a review of
selected questionnaires: An Inventory of Health Knowledge, Attitudes,
and Pragticgggfor Students in Grades 4-6 and 7-9 (Abt Associates,
1984), InnerView Health Assessment (National Computer Systems, 1986),
the Michigan Educational Assessment Program Student Assessment
Booklets, Grades 4L 7, and 10 (Michigan State Board of Education,
1985), Personal Lifestyle Questionnaire (Muhlenkamp & Brown, 1983),
lggtflell:4A¥Self ScoringgWellness Assessment Questionnaire (National
wellness Institute, 1983), and The HealthrPromoting Lifestyle Profile
(walker, Sechrist & Pender, 1987).

The questionnaires: (a) did not closely correspond to the nine
areas and 23 health behaviors identified by Bruhn & Parcel (1982a),
(b) were not developmentally appropriate for the young adolescent and
parent(s), (c) usually included too few health practices, and (d)
included items of a general rather than a specific nature.

Therefore, the development of a new instrument, the Health Practices
Survey (HPS) was undertaken. A description of its developmental
process follows; the process was guided by the suggestions of
Nunnally (1978) and Windsor et a1. (1984).

Although the nine areas of health behavior and their
accompanying 23 health practices provided the theoretical framework
fer the HPS, the HPS’s format and items resulted from multiple
sources: (a) the review of the objectives and lesson plans of the
Michigan Model for Comprehensive Health Education (Department of

Education, 1988; Vogel, 1987), (b) the investigation of previously

72

developed health risk appraisals and health behavior questionnaires
as previously noted, and (c) consultation with health professionals,
educators, and evaluators.
The Michigan Model for Cogpreheggiye Health Education

This Model is viewed as the most progressive health education
program in the united States (Jubb, personal communication, January,
1989). It is a developmentally based health curriculum for grades K
through 12 which took 10 years to develop and utilized the expertise
of many individuals and professional groups related to health,
education, fire, and police. The objectives, lesson plans and
associated materials for grades K through 6 have been developed and
are at various stages of implementation in 75X of Michigan school
‘ districts (Jubb, personal communication, January, 1989). The
objectives, lesson plans and associated materials for grades 7 and 8
became available late in Spring 1989 so that implementation can occur
during the 1989-1990 academic year. Those for the higher grades will
be forthcoming. Evaluation of the impact of the Model will begin in
the 1989-1990 academic year. Measures of knowledge, attitudes, and
behaviors for grades K thrOugh 6 are in the process of being
developed (Vogel, personal communication, January, 1989).
Consultation from Health Professionals, Educators, and Evaluators

Based upon the review of items included in other health risk
appraisals and health behavior questionnaires for children,
adolescents, and adults, and the review of the objectives and lesson
plans of the Michigan Model for Comprehensive Health Education, a
preliminary draft of items for the PHS was developed by the

investigator. Content validity of the items selected, item clarity,

73

and appropriateness of survey techniques were confirmed by
individuals with expertise in health, education and evaluation in the
Greater Lansing Area. Their positions and organizational affiliation
follow: (a) Director of Health Education, Ingham County Health
Department, (b) 8th grade English educator, East Lansing Public
Schools, (c) Program Coordinator of Health, Safety and Youth
Services, American Red Cross, Lansing, (d) Nutritionist, College of
Nursing, Michigan State University, (e) Professor, Measurement and
Evaluation, College of Education, Michigan State University, (f)
Supervisor, Sleep Disorder Program, Ingham Medical Center, Lansing,
(g) President, Michigan Public Health Association, (h) Project
Officer for the Michigan Model of Comprehensive Health Education,
Michigan Department of Public Health, (i) Chief Health Surveilance
Section, Michigan Department of Public Health, and (j) Project
Director, Measurement and Evaluation, Michigan Model for
Comprehensive Health Education.

The initial Health Practices Survey (BPS) was composed of
approximately 100 items grouped into 10 focal areas, the nine areas
identified by Bruhn & Parcel (1982a) and a tenth area derived from
the investigator’s review. These areas were: (a) Food Related
Practices, (b) Dental Practices, (c) Sleep and Rest Practices, (d)
Exercise and Fitness Practices, (e) Recreational Practices, (f)
Automobile and Pedestrian Practices, (g) HOusehold and Related
Practices, (h) Stress Related Practices, (1) Substance Use Practices,
and (j) Personal Care Practices. Normative statements based upon a
standard baseline recommendation in the health literature or by the

consultants were used. The majority of the statements were stated

74

positively; some negative statements were used to avoid a response
set. Since the respondent would be asked to describe how often each
health practice was performed, a variety of response categories were
initially considered, e.g., using percent of time, using examples for
comparison, or categories of almost always, frequently, sometimes,
and never. A Likert scale was selected: ”Never”, ”Rarely",
”Sometimes", "Usually", and "Always" with "Never = l" and "Always =
5". This approach was consistent with other health behavior
questionnaires.

Based upon the reviews and comments of the consultants and the
members of the investigator’s doctoral committee, several items were
both added and omitted and minor language changes were implemented.
The resulting HPS was comprised of 128 items.

Pilgt Testing

In order to assure that the items were clearly written for the
young adolescent and to obtain initial reliability and validity
measures on the HPS, the investigator contacted the Girl Scouts,
Lansing, Michigan (Harte, personal communication, February, 1989).
While this organization indicated that they had several junior girl
scout troops composed of 12 through 15 year olds, they were reluctant
to release the names and phone numbers of the troop leaders until
their staff had reviewed the HPS as well as the revised Self-Care
Inventory (SCI). The investigator decided that the latter also
needed to be reviewed for clarity by several girl scouts in this age
range.

After the staff reviewed the HPS and SCI, the investigator

received the names and telephone numbers of six troops in the Greater

75

Lansing Area which were outside the specified geographic area used in
the sampling of churches. Each leader was contacted and information
about the intent and content of the HPS, and the previous review by
the staff of Girl Scouts, Lansing was provided by the investigator.
All leaders agreed to have the scouts in their troops complete the
HPS. The investigator requested that the largest troop of ten scouts
complete both the HPS and SCI; the leader and scouts agreed.

The investigator met with each troop and their leaders at their
regularly scheduled meeting times. The meetings were conducted
during March and April, 1989 and each was approximately two hours in
length. The seven scouts in the first troop completing the HPS were
encouraged to ask questions and.make comments so that greater item
clarity could be obtained. Based upon their input, additional minor
changes in the phrasing of items and a few word substitutions were
made. The revised HPS was then completed by 29 girl scouts between
13 and 16 years of age representing five different troops.

Data from the 29 completed HPS questionnaires were entered by a
computer consultant and analysis was conducted using SPSS-X via the
Michigan State University Computer Laboratory. Item analysis and
reliability analysis utilizing the covariance matrix were performed.
Cronbach’s alpha was used to assess the internal consistency of each
of the ten health practice areas or subscales. The results were
reviewed by two members of the investigator’s doctoral committee, the
consultant and the investigator.

According to Nunnally (1978), a reliability standard (alpha
coefficient) of .70 for new scales is acceptable. Four of the 10

subscales in the HPS met this standard: Food Related Practices (.83),

76

Stress Reduction Practices (.78), Household and Related Practices
(.74), and Sleep Practices (.72). The alpha coefficients of the
other subscales follow in descending order: Exercise and Fitness
Practices subscale (.60), Dental Practices (.50), Recreational
Practices (.46), Personal Care Practices (.38), Substance use
Practices (.20), and Automotive and Pedestrian Practices (-.29).

Upon review of the alpha coefficients and item analysis, the
following changes in the HPS were made. It was noted that one item
was repeated in the Food Related Practices Subscale, therefore it was
deleted. Two items were rewarded in the Sleep and Rest Practices
Subscale while three additional items were added to the Exercise and
Fitness Practices Subscale to enhance its alpha coefficient of .60 to
the .70 minimum. One item was deleted from the Household and Related
Practices Subscale, and two items were deleted from the Stress
Reduction Practices Subscale. No changes were made in the remaining
subscales. Since the pilot testing was restricted to young
adolescent girls in primarily rural or small town settings, the
introduction of adults and young male adolescents could alter the
subscales with alpha coefficients of .50 and less. It was speculated
that the poorest subscale performers, Substance Use Practices (.20)
and Automotive and Pedestrian Practices (-.29) might need to be
assessed for separate young adolescent and adult subscales. This
review resulted in a 127 item revised HPS (Appendix H), which was
used in this investigation. The reliability coefficients resulting
from this study are discussed in "Data Analysis Procedures" and are

presented in Table 3.

77

The General Functioning Subscale of the

McMaster Family Assessment Device (FAD-Z)

 

The McMaster Family Assessment Device (FAD) is a self-report 60
item measure of family functioning, i.e., how the family works
together, and is based upon the McMaster Model of Family
Functioning which utilizes systems theory (Epstein et al., 1983).
The respondent is asked to select one of the following responses for
each item: "strongly agree", "agree”, "disagree", or "strongly
disagree". The FAD is composed of seven subscales and has been
recognized as a reliable and valid measure in assessing family
functioning (Byles, Bryne, Boyle, & Offord, 1988; Epstein et al.,
1983; Forman & Hagan, 1983; Kabacoff, Miller, Bishop, Epstein &
Keitner, in press; McCubbin & Thompson, 1987; Miller, Epstein, Bishop
& Keitner, 1985).

Byles et a1. (1988) utilized the 12 item General Functioning
Subscale of the FAD as a global measure of family functioning in a
large random sample of 1,839 respondents. Positive features of the
subscale which facilitated its use were: (a) its ease of
administration and scoring, (b) its brevity, and (c) its high
correlation with the 60 item FAD. Because it was highly correlated
with the other six dimensions, it was considered to be a measure of
the overall health/pathology of the family and therefore included as
a self‘report measure for parents in the Ontario Child Health Study.
Byles et a1. (1988) state that: (a) construct validity of the
subscale was supported since it correlated as predicted with other
family variables in the study, (b) reliability was consistent with

previous findings, and (c) internal consistency was .86. The General

78

Functioning Subscale "...is a good measure for use in surveys where
cost and ease of administration are important considerations, and
where only scores of general family functioning rather than specific
dimensions of family functioning are required" (Byles et al., 1988,
p.103).

Kabacoff et al. (in press) indicate further support for the use
of the General Subscale of the FAD as a global measure of family
functioning. Their investigation of the correlation of this subscale
and the other six subscales revealed that it was highly correlated
with the first principal component of the other 48 items for
nonclinical, psychiatric, and medical samples, i.e., .85, .87, and
.88 respectively.

Due to copyright restrictions on the use of the FAD, the
investigator followed the guidelines provided by Epstein et a1.
(1983) and obtained the FAD packet from the Director, Brown
University Family Research Program, Butler Hospital, Providence, RI.
The packet included the measurement instrument and scoring format
(Epstein, Baldwin, & Bishop, 1982), and articles describing the FAD’s
psychometric properties. A review of the instrument revealed the 12
items of the General Functioning Subscale, (Appendix I), six items
(2, 4, 6, 8, 10, and 12) describe healthy functioning and six items
(1, 3, 5, 7, 9, and 11) describe unhealthy functioning. To score the
Subscale, the response selected by the respondent for each of the
items is coded as follows: Strongly Agree = 1, Agree = 2, Disagree =
3, and Strongly Disagree = 4. The score for each item describing
unhealthy functioning is transformed by subtracting it from 5.

Therefore one represents a healthy response and four represents an

79

unhealthy response. The resulting 12 numerical responses are then
averaged to provide a subscale score with a possible range from 1.00
(healthy) to 4.00 (unhealthy) (Epstein et al., 1982). According to
Miller et a1. (1985) a mean of greater than 2.0 endorses an unhealthy
direction suggesting that the family is having difficulties. In
addition, they note that, based upon clinical experience, scores in
the unhealthy range may occur in 19 to 363 of non-clinical or
supposedly healthy families.
Demoggaphic Information

The demographic measures (Appendix J) selected were the result
of the investigator’s review of items in other health behavior
questionnaires, THE MICHIGAN EARLY ADOLESCENT SURVEY (Keith &
Hoopfer, 1987), other dissertations, and the literature review
related to health behavior and family research. They were reviewed
for clarity and appropriateness by the members of the investigator’s
doctoral committee; their operational definitions are provided on
pp.18-22.

Scoring Procedures

According to the conceptual framework adopted for this study,
the variables of interest include: (a) personal characteristics,
i.e., age, gender, religiosity, ethnicity, marital status,
educational level, occupation, employment status, health status,
height and weight, perceived family functioning, health behavior
(psychosocial and behavioral), intrinsic motivation, (b) the family
realm, i.e., family type and family income, and (c) family

characteristics, i.e., family functioning, family intrinsic

80

motivation, family health behavior (psychosocial and behavioral), and
family health status. The measurement of each of these variables was
guided by their corresponding operational definitions, pp. 20-23, and
the scoring procedures described for the measurement instruments, pp.
62-79. Each of the personal characteristics and the family realm
variables of family type and family income were coded for data entry
by the investigator. These codes were then utilized for computer
entry of the data as well as the statistical analyses. The latter
included the formulation of measures for each of the family
characteristic variables.
The Statistical Methods

The statistical methods utilized in the analysis of the study
data are reviewed in this section. Table 2 presents an overview of
the hypotheses tested, the conceptual approach selected for their
analysis, the data used, and the statistical approach. All
statistical analyses were conducted by SPSS-X through the Computer
Laboratory at Michigan State University, East Lansing, MI.

Descriptive statistics, i.e., measures of central tendency and
measures of variability, were used to characterize the sample where
applicable. Cronbach’s alpha, a summary measure of reliability, was
utilized to indicate the degree to which each of the underlying items
in a scale or subscale consistently measured the same underlying
concept. According to Nunnally (1978, p. 207) ”the precision with
which the reliability is estimated for any test is a direct function
of the precision with which the average correlation of items in a
test estimates the average correlation of all items in the domain."

A satisfactory coefficient alpha depends upon how a measure is used;

Table 2

81

Summary of Statistical Procedures.

 

 

Purposes of Conceptual Data Used Statistical
Analysis Approach Used in Analysis Approach
to Determine
Family
Characteristics
Test of Discrepancy Individual & Family
hypotheses Score Model triad data for Discrepancy Mean
#1 - #4 both models on: Score and
1. Health status t-test.
2. Intrinsic
Motivation
3. Health
behavior
(psychosocial
and behavioral)
3. Family
functioning
4. Health status
Conjunctive Within-family
Model factor from
ANOVA
with repeated
measures &
F—test.
Test of Individual data Family means,
hypotheses on 1. - 4. Pearson
#5 -#13 above correlations on
pairs of family
characteristics
&
frequencies of
correlations by
family type.
Test of Same as for ANOVA/F-test.
hypotheses hypotheses
#14 - #17 #1 - #4
Test of Same as for Descriptive
hypothesis hypotheses Pearson
#18 #1 -#4 plus correlations

family income
& family type

between family
means on family
characteristics.

 

82

in basic research .80 is adequate while .70 is reasonable in the
early stages of instrument development (Nunnally, 1978, p.245).

The Pearson product moment correlation coefficient (Pearson
coefficient) was utilized to reflect the magnitude and direction of
the linear relationship between two variables or between two
observations of the same variable. Variance in the sample can be a
major influence of r; "the greater the variability among the
observations, the greater the value of r" (Glass & Hopkins, 1984, p.
92).

The Pearson coefficient was also used to investigate the
psychometric property of reliability of scales. Items which were
correlated most highly with the total score were retained. Items
below .30 were excluded (Nunnally, 1978). When a concept has
multiple dimensions with multiple indicators, items with low
correlations may occur due to missing key conceptual aspects (Green &
Lewis, 1986; Nunnally, 1978).

Analysis of variance (ANOVA), an inferential statistical
technique, was used to determine whether the difference among two or
more means was greater than that expected from sampling error
(chance). The statistical significance of the F-test was used to
determine whether a significant difference existed between the means.
Analysis of variance with repeated measures was utilized to
investigate family congruency (Ball et al., 1983). When two or more
family members provide data on a variable of interest, e.g., health
behavior, the family has been "repeatedly measured". This repeated
measure can be referred to as a within-family factor. According to

Ball et al. (1983) the main advantage of the within-family factor is

83

that it increases the power for the test of family-member main
effect, i.e., the F test is more powerful. This occurs since the
error, with one observation per cell, is only measurement error and
not the differences between subjects in the cell. Assuming that the
measurement error is small and individual differences are quite
large, the denominator of the family—members’ F test should be much
smaller. A small denominator suggests "a larger value of F and a
greater probability of a significant result" (Ball et al., 1983, p.
889).

The t-test was utilized to draw inferences about the central
tendency of populations. Specifically the t-test was used to examine
two alternative hypotheses: (a) that the difference between the means
of two populations or samples was equal to zero, and (b) that they
were different from zero.

Data Analysis Procedures

Prior to any attempt to analyze the data, the accuracy of the
scoring formats, i.e., reverse coding of items, and the accuracy of
the data as entered were confirmed. The former occurred by reviewing
the individual measurement guidelines and the commands on the data
printout. The accuracy of the entered data was confirmed by the
investigator and a second reviewer who together compared the original
data set for each individual with a printout of the individual’s
entered data. Following the correction of the errors, descriptive
statistics were obtained.

Cross tabulations in conjunction with reliability analyses were
then conducted on the selected measurement instruments, i.e., the

SCI, the HPBQ-Part 2, the FAD-7, the HPS and its subscales, and the

84

HSDI. Cross tabulations of each item by family type and family
subsample, the adolescents, the mothers, and the fathers, were
conducted. Reliability analyses were also completed separately for
the subsamples. They included Cronbach’s alpha coefficient (alpha
coefficient), the mean and standard deviation for each item, and the
item-total statistics for subscales, i.e., the scale mean, variance,
and alpha if item deleted, the corrected item—total correlation, and
the squared multiple correlation.

The results from the SCI, the HPBQ—Part 2, the FAD-7, the HPS
and its subscales, and the HSDI were reviewed by the investigator and
the statistical consultant. The review resulted in the deletion of
several items with inadequate variance and low item-total
correlation. The initial alpha coefficients for each of the
subsamples on the SCI were above .80; therefore, no items were
deleted. The initial alpha coefficients for the HPBQ-Part 2 were .68
for the fathers, .73 for the mothers, and .68 for the adolescents.
The review of the item analysis resulted in the deletion of three
items: #12, "Avoid contact with doctors when feeling okay”, #23,
”Don’t smoke", and #30, "Ignore health advice from friends,
neighbors, and relatives". Their deletion resulted in the
enhancement of the alpha coefficients to .70 for the three
subsamples.

The initial alpha coefficients for the total HPS were also above
.80, but those accompanying some of the subscales were still below
.60, e.g., Dental Practices, Automotive and Pedestrian Practices,
Substance Use Practices, and Personal Care Practices. The review of

the item analysis of all subscales resulted in the deletion of

85

several items to enhance the alpha coefficients of the subscales: (a).
Item #36, "I use a soft toothbrush", was deleted from the Dental
Practices Subscale; (b) items #49, "I walk up stairs rather than use
the elevator" and #50, "I maintain good posture when sitting or
standing", were deleted from the Exercise and Fitness Subscale, (c)
items #54, "I operate a three wheeler, all terrain vehicle (ATV),
snowmobile, or motorboat”, #55, "I ride with drivers of all terrain
vehicles (ATV), snowmobiles or motorboats who have been using drugs
or alcohol", and #58, "I walk or jog on a road or street with the
flow of traffic", were deleted from the Recreational Practices
Subscale, (d) item #91, "I am able to get to and handle guns in my
home" was deleted from the Household Practices Subscale, and (e) item
#96, "I eat more food when I feel anxious or tense" was deleted from
the Stress Reduction Practices Subscale. The other subscales
remained intact as originally developed.

The alpha coefficients for the HSDI for each of the subsamples
were less than those reported in other investigations. The scoring
mechanism was reviewed with Cox (personal communication, October 27,
1989); it was consistent with the scoring approach used prior to the
reliability analysis for this study. An enhancement of the
coefficients resulted when two items were deleted: #10, "I do things
to help my health even though a doctor or nurse has not suggested
these things to me" and #17, "I know what I’m doing when it comes to
taking care of my health". Since sample homogeneity was revealed
through the descriptive statistics of the respondents’ individual
characteristics, the lowered reliabilities of the subsamples on the

HSDI could also be reflecting sample homogeneity. The FAD-7 was not

86

altered due to the high alpha coefficients across the subsamples.
The resulting reliabilities for these multiple item measures by the
subsamples are presented in Table 3.

Once the reliabilities of the measurement instruments were
secured, a score or simple mean for each respondent on each
measurement instrument was obtained according to the authors’
instructions. These results and the one item score for personal
health status for each respondent were utilized in the testing of the
hypotheses. The first four hypotheses explored the presence of
family characteristics related to health behavior by investigating
the degree of congruency on the data reported by multiple family
members. The hypothesized family characteristics were: family
functioning, family intrinsic motivation, family health behavior, i.
e., family psychosocial health behavior and family behavioral health
behavior, and family health status. Two conceptual approaches, the
Discrepancy Score Model and the Conjunctive Model (Klein, 1984) were
selected as being most appropriate for their investigation.
According to Klein (1984), Fisher et al. (1985) and Gillis (1983),
different statistical methods can be used with each of these models.
The statistical methods were determined by the investigator’s review
of the family and health behavior literature, consultation with
members of the investigator’s doctoral committee, and the
recommendations of the statistical consultant.

It must be acknowledged that confusion can occur when evaluating

the results; this is due to the conceptual and statistical

87

Table.3
Reliability Analysis* of the HSDI, the FAD-7, the SCI, the HPBQ-Part
2 and the HPS and its Subscales by Subsamples.

 

 

Measures Subsamples
Fathers Mothers Adolescents
HSDI .66 .62 .72
FAD-7 .88 .88 .90
SCI .92 .94 .92
HPBQ—Part 2 .71 .77 .70
HPS .90 .91 .91
Automotive &

Pedestrian Practices .71 .60 .46
Dental Practices .44 .52 .65
Exercise and

Fitness Practices .78 .83 .71
Food Related Practices .80 .78 .74
Household & Related

Practices .63 .77 .79
Personal Care ,

Practices .41 .44 .50
Recreational Practices .64 .74 .65
Sleep & Rest Practices .59 .71 .65
Stress Reduction

Practices .75 .79 .69
Substance Use

Practices .37 .50 .38

 

*Cronbach’s Alpha Coefficient

approaches. Although the conceptual hypotheses and the statistical
null hypotheses are consistent in that they both indicate the
existence of congruency, rejection of the statistical null hypotheses
would imply that congruency does not exist. In this analysis then,
non-significant results are desired and would support the conceptual
hypotheses of congruency. The conceptual formula utilized for the

Discrepancy Score Model (Klein, 1984) was:

88
i} (:11 - :22) i.
m‘

This formula creates a family measure (score); an example for a
family HSDI score with three family members follows:

lAMNHSDI-FAMNHSDI' + lFMNHSDI-FAMNHSDII + 'MMNHSDI-FAMNHSDII =

FADISMNHSDI.
In this equation AMNHSDI = the adolescent’s mean score on the
HSDI,
FMNHSDI = the father’s mean score on the
HSDI,
MMNHSDI = the mother’s mean score on the
HSDI.
FAMNHSDI = the family measure mean score on
the HSDI.
FADISMNHSDI = the family discrepancy mean score on
the HSDI

This same equation was utilized in creating the other family
measures, i.e., family functioning, family health behavior
(psychosocial and behavioral), and family health status . The
creation of the family measures was accomplished through the program
SPSS-X and the evaluation of the presence of a family characteristic
was determined by a hand computed t via the t-test for each potential
family characteristic.

If congruency existed, the t scores were not significant, i.e.,
they were near zero and the t was less than 3.46 (p=.01) for dual
parent families and less than 4.32 (p=.01) for single parent
families, and the statistical null hypothesis was retained. However,

if congruency did not exist, the t scores were significant, i.e.,

89

they were distant from zero and the t was greater than 3.46 (p=.01)
for dual parent families and greater than 4.32 (p=.01) for single
parent families, and the statistical null hypothesis was rejected.
The analysis of variance with repeated measures was selected as
the statistical method to investigate the presence of a family
characteristic as conceptualized in the Conjunctive Mbdel (Klein,
1984). Unlike the statistical method for the Discrepancy Score
Mbdel, this method does not yield a derived family score, but does
reveal the presence of a family characteristic through the F
statistic. The F statistic was determined through the use of SPSS-X

and the following formula:

Xij = u + aj + Eij (Glass & Hopkins, 1984)

In this fermula Xij represents the simple mean scores of the
adolescent, mother and father or only the adolescent and mother, the
aj represents the with-in family effect and Eij represents the error.
J is the number of individuals in the family and i is the observation
or family. This same equation was utilized in creating the other
family characteristics, i.e., family functioning, family health
behavior (psychosocial and behavioral), and family health status,.

If congruency existed, the F statistic was not significant,
i.e., the F value was less than 3.15 (p=.05) for the dual parent
family and less than 4.75 (p=.05) for the single parent family, and
the statistical null hypothesis was retained. If congruency did not
exist, the F statistic was significant, i.e., the F value was greater

than 3.15 (p=.05) for the dual parent family and greater than 4.75

90

(p=.05) for the single parent family, and the statistical null
hypothesis was rejected.

Once these statistical analyses were completed and the results
indicated support for the proposed hypotheses, i.e., that congruence
was present, further investigation of these family characteristics
was conducted. This included the investigation of hypotheses
focusing on the relationships between these family characteristics
and within the individual family. Pearson product moment
correlations were calculated for pairs of the family measures (mean
scores). Therefore, for every proposed relationship between two
family characteristics there were 73 correlations, 60 correlations
for the dual parent families and 13 correlations for the single
parent families. The frequencies of these correlations were then
assessed by the following decision rule. If more than 503 of the
families exceeded a Pearson coefficient of .40, the hypothesis about
the relationship was supported.

The final series of statistical analyses focused on
relationships between family characteristics and the family realm
variables of family type, i. e., dual and single parent families, and
family income. The Pearson product moment correlations were based
upon the families’ means of the family characteristics. If a Pearson

of .30 was exceeded, the relationship was supported.

CHAPTER IV

RESULTS OF DATA ANALYSIS

This chapter includes the description of the study sample, a
summary of the major characteristics of the sample, the testing of the
hypotheses, and a summary of the hypotheses testing. The description
of the study sample includes the demographic and personal
characteristics, i.e., the respondents’ reports of marital status,
family composition, family income, age, ethnicity, educational
attainment, occupation, religiosity, perceived health status, and long
term health impairments or health problems.

Description of the Study Sample

The sample was composed of 60 two parent families and 13 single
parent families (Table 4). Data were collected from three members of
the two parent family, the father, the mother and the adolescent, and
from two members in the single parent family, the mother and the
adolescent. The adolescent in both families was the oldest child
between the ages of 13 and 15. Therefore, 60 fathers, 60 mothers and
60 adolescents in dual parent families, and 13 m,thers and 13
adolescents in single parent families participated. The sample total
was 206 family members representing 73 families.

Mothers in dual parent families reported that their families were

91

92

Table 4
Marital Status of Parents in Dual and Single Parent Families.

 

 

 

 

Marital Status Dual Parent Single Parent
n X n X n X
Married 60 100.0 60 100.0 - -
Divorced - - - - 11 84.6
Separated - - - - 2 15.4
Total 60 100.0 60 100.0 13 100.0

 

composed of three to seven family members. The majority of these
families, 51.7% (n=3l), had four members, the parents and two
children. Mothers in the single parent families reported that their
families were composed of two to five family members. Table 5
reflects the number of family members in both family types; the
distribution of members is the result of the investigator’s review of
uncoded data.

Table 5

Number of Family Members as Reported by Mothers in Dual and Single
Parent Families.

 

 

 

 

 

Number of Family Members Dual Parent Single Parent
n X n X
Two - - 3 23.0
Three 5 8.4 4 31.0
Four 31 51.7 2 15.0
Five 17 28.3 4 31.0
Six 5 8.3 - -
Seven 2 3.3 - -
Total 60 100.0 13 100.0

 

Family income was determined by the highest response category
selected by an adult in the dual parent family and by the mother in

the single parent family. Twenty income categories beginning with

93

"Less than $10,000", increasing at increments of $5,000, and
concluding with "More than $100,000" were used. Variance of family
income was most noticeable in the dual parent families, i.e., from
$15,000 to more than $100,000. Income categories of the

single parent families clustered from ”Less than $10,000" to $30,001 -
$35,000 (Table 6). In the dual parent families, 24* (n=l4) of the
families indicated their family income was between $15,000 and
$40,000, 362 (n=21) indicated their family income was between $40,001
and $55,000, and 403 (n=24) indicated their family income was between
$55,001 to more than $100,000. In the single parent families, 46%
(n=6) indicated a family income of between $15,000 and $25,000.

Age was calculated by the investigator from the respondent’s
birthdate and viewed within the context of categories of years. The
majority of the fathers, 78% (n=47), were between the ages of 41 and
50; the majority of the mothers in the dual parent families, 73%
(n=44), were between the ages of 36 and 45. The majority of the
mothers in the single parent families, 69* (n=9), were between 31 and
40. All adolescents were between 13 and 15 years of age. The
adolescents in the dual parent families were overall slightly older,
age 14, than those in the single parent families, age 13 (Table 7).

Although a sample diverse in ethnicity and educational attainment
had been sought, the repondents were predominantly "White not
Hispanic" and had either attended college or completed four years of
college (Tables 8 & 9). Each of the other ethnic groups, "Black not
Hispanic”, "Hispanic", "Asian", and "American Indian", was represented
by at least two people (Table 8). A majority of the fathers (66%,

n=33) and 422 (n=22) of the mothers in the dual parent families, and

94

Table 6
Family Income by Dual and Single Parent Families.

 

 

 

Income Value Dual Parent Single Parent
n % n %
Less than $10,000 1 - - l 7.7
$10,001 to $15,000 2 - - 2 15.4
$15,001 to $20,000 3 2 3.4 3 23.1
$20,001 to $25,000 4 3 5.1 3 23.1
$25,001 to $30,000 5 l . 1.7 2 15.4
$30,001 to $35,000 6 3 5.1 l 7.7
$35,001 to $40,000 7 5 8.5 - -
$40,001 to $45,000 8 7 11.9 - -
$45,001 to $50,000 9 7 11.9 - -
$50,001 to $55,000 10 7 11.9 1 7.7
$55,001 to $60,000 11 4 6.8 -
$60,001 to $65,000 12 5 8.5 - -
$65,001 to $70,000 13 3 5.1 - -
$70,001 to $75,000 14 3 ‘5.1 - -
$75,001 to $80,000 15 l 1.7 - -
$80,001 to $85,000 16 3 5.1 - -
$85,001 to $90,000 17 - - - -
$90,001 to $95,000 18 - - - -
$95,001 to $100,000 19 2 3.4 - -
More than $100,000 20 3 5.1 - -
Missing Data 21 1 7.7 - -
Total 60 100.0 13 100.0
Mean Value 10.356 4.000
Median Value 10.000 4.000
Mode Value 8.000 3.000
Standard
Deviation Value 4.246 2.273

 

58% (n=7) of the mothers in the single parent families had completed
at least four years of college with the majority of these completing
five or more years of college (Table 9).

A review of the occupational categories by fathers and mothers
revealed that the majority of both hold professional or managerial
positions, i.e., 76% (n=44) of the fathers, 43% (n=26) of the mothers
in dual parent families, and 69% (n=9) of the mothers in the single

parent families (Table 10). Only ten (16.7%) of the mothers in the

95

Table 7

Age in Years Reported by Respondents in Dual and Single Parent
Families.

 

 

 

 

Age in Years Value Dual Parent Single Parent
n % n %
Fathers
36-40 2 6 10.0 - -
41-45 3 24 40.0 - -
46-50 4 23 38.3 - -
51-55 5 5 8.3 - -
Missing data 6 2 3.3 - -
Total 60 100.0
Mean Value 3.550 -
Median Value 3.500 -
Mode Value 3.000 -
Standard
Deviation Value .910 -
Mothers
25-30 1 2 3.3 - -
31-35 2 9 15.0 4 30.8
36-40 3 26 43.3 5 38.5
41-45 4 18 30.0 3 23.1
46—50 5 4 6.7 l 7.7
Missing data 6 1 1.7 - -
Total 60 100.0 13 100.
Mean Value 3.267 3.077
Median Value 3.000 3.000
Mode Value 3.000 3.000
Standard
Deviation Value .972 .954
Adolescents
13 13 22 36.6 7 53.8
14 14 24 40.0 4 30.8
l5 15 14 23.3 2 15.4
Total 60 100.0 13 100.0
Mean Value 13.850 13.615
Median Value 14.000 13.000
Mode Value 14.000 13.000
Standard

Deviation Value .799 .768

 

96

Table 8
Ethnicity of Respondents in Dual and Single Parent Families.

 

 

 

Ethnicity Dual Parent Single Parent
£3 £2.- AA; 1Q; 11$
n % n % n % n % n %
White not
Hispanic 56 93.3 57 95.0 51 86.4 12 92.3 9 75.0
Black not
Hispanic 2 3.3 l 1.7 1 1.7 l 7 7 l 8.3
Hispanic 1 1.7 l 1.7 l 1.7 - - 1 8.3
Asian 1 1 7 - - 1 1.7 — - — -
American Indian - - l 1.7 2 3.4 - - 1 8.3
Other - - - - 3 5.1 - - - -
Missing Data - - - - l 1.7 - - l 8.3
Total 60 100.0 60 100.0 60 100.7 13 100.0 13 100.3

 

dual parent families indicated they were full time homemakers.

Although families representing diverse religious preference,
e.g., Catholicism, Protestantism, and Judaism, were sought, the sample
was primarily Protestant, i. e., 75% (n=45) of the fathers, 82% (n=49)
of the mothers, and 61.7% (n=37) of the adolescents in the dual parent
families, and 69.2% (n=9) of the mothers and 69% of the adolescents
(n=9) in the single parent families (Table 11). Two other categories
were also selected by respondents: Catholic and Other.

A review of the frequency of attendance of the repondents at
religious services or functions during the five weeks prior to
completing the questionnaire revealed that in general these families
were active church participants. Approximately two-thirds of
fathers, mothers, and adolescents had attended at least three church
related activities (Table 12). Approximately 25.7% (n=16) of these
fathers, 36.75% (n=22) of these mothers, and 26.7% (n=16) of these

adolescents in the dual parent families, and 38.5% (n=5) of these

97

Table 9

Years of Education As Reported by Parents in Dual and Single
Parent Families.

 

 

 

Education Value Dual Parent Single Parent
n % n % n %

1-3 years

of high

school 2 l 2.0 - — - —
Completed

4 years of

high school

(diploma or

G.E.D.) 3 4 8.0 8 15.4 4 33.3
Less than

4 years

of college 4 12 24.0 22 42.3 1 8.3
4 years of

college 5 11 22.0 6 11.5 4 33.3
5 or more

years of

college 6 22 44.0 16 30.8 3 25.0
Missing Data 7 10 16.6 8 13.3 1 8.3
Total 60 100.0 60 100.0 13 100.0
Mean Value 3.980 3.577 3.500
Median Value 4.000 3.000 4.000
Mode Value 5.000 3.000 2.000
Standard

Deviation Value 1.097 1.091 1.232

 

mothers in the single parent families participated in more than five
church related activities.

In general the respondents in this sample viewed themselves as
being in good health (Table 13). The categories of "Excellent", "Very
Good" and "Good" were selected by 96% (n=56) of the fathers, 92%
(n=55) of the mothers, and 92% (n=55) of the adolescents in the dual
dual parent families, and 100% (n=l3) of the mothers and adolescents

in the single parent families. However 41 respondents, 13 fathers, l3

98

mothers, and nine adolescents in the dual families, and five mothers
and one adolescent in the single parent families, collectively
indicated they had 49 long term impairments or health problems (Table
14). These were categorized into frequently used disease and disorder
groupings. The top four categories were: neurological (n=16),
endocrine/nutritional/metabolic (n=10), respiratory (n=8), and

circulatory (n=6).

Table 10
Current Occupation of Respondents in Dual and Single Parent Families.

 

Current Dual Parent Single Parent
_‘_1L._°C° at ion Ea £4.2- Ii<.i.-. M_°-. a.
n % n % n % n % n %

 

Professional 28 48.3 22 3 .
Management 16 27.6 4
Proprietor (small

business) 3 5.2 7
Clerical worker
Sales worker 1 1.7 1
Skilled craftsman

or foreman 5 8.6 - — - - - - - -
Operative

unskilled

laborer
Homemaker
Student
Other
Missing Data
Total 6

(DO)
4‘!

I

I
(A)
N
(A
H

I

I

I
I
H
w
NI—I
HHH
QQQ
I
I
o:
N
00
H
I
I

7
.7 60 100.0 - 13 100.0
3

<3wa I H

60 100.0 60 100.0 1 100.0 13 100.0

 

99

 

 

 

Table 11
Religious Preference of Respondents in Dual and Single Parent
Families.
Religious Dual Parent Single Parent
P_____reference E2; Ma. Q 1% AA:
n % % n % n % n %

Catholic 6 10 0 5 8 3 6 10.0 1 7.7 - -
Protestant 45 75 0 49 81 7 37 61.7 9 69.2 9 69.2
Other 9 l5 0 6 10 0 17 28.3 2 15.4 3 23.0
Missing

Data - - - - - - 1 7.7 1 7.6
Total 60 100.0 60 100.0 60 100.0 13 100.0 13 100.0
Table 12

Frequency of Attendance at Religious Services or Functions During
Past Five Weeks by Respondents in Dual and Single Parent Families.

 

 

Frequency Value Dual Parent Single Parent
2! F8 M2; A._d_- E91 Ad.
Attendance n % n % n % n % n %
None 0 9 15.0 8 13.3 9 15.0 2 15 4 3 23.1
Once 1 3 5.0 2 3.3 3 5.0 - - — -
Twice 2 5 8.3 5 8.3 5 8.3 - - 3 23.1
Three times 3 11 18.3 7 11.7 11 18.3 - - — -
Four times 4 8 13.3 8 13.3 8 13.3 4 30.8 1 7.7
Five times 5 8 13.3 8 13.3 8 13.3 2 15.4 4 30.8
More than

five times 6 16 25.7 22 36.7 16 26.7 5 38.5 2 15.4
Total 60 100.0 60 100.0 60 100.0 13 100.0 13 100.0
Mean Value 2.567 4.950 3.567 5.308 3.231
Median Value 3.000 5.500 4.000 6.000 4.000
Mode Value 5.000 7.000 6.000 7.000 5.000
Standard

Deviation

Value 2.100 2.135 2.110 2.097 2.315

 

100

Table 13

Perceived Overall Health Status of Respondents in Dual and Single
Parent Families.

 

 

 

Perceived Value Dual Parent Single Parent
Lam: 119.: 119L- Acl: Ha; ASL.
Status n % n % n % n % n %
Excellent 1 13 22.4 11 18.3 11 18.3 1 8.3 l 8.3
Very Good 2 24 41.4 24 40.0 24 40.0 7 53.8 7 53.8
Good 3 19 32.8 20 33.3 20 33.3 5 38 5 5 38.8
Fair 4 2 3.4 4 6.7 4 6.7 - - - —
Poor 5 - - 1 1.7 1 1.7 - — - -
Missing

Data 6 2 3.4 - - - - - - - -
Total 60 100.0 60 100.0 60 100.0 13 100.0 13 100.0
Mean Value 2.172 2.033 2.333 2.000 2.308
Median Value 2.000 2.000 2.000 2.000 2.000
Mode Value 2.000 2.000 2.000 1.000 2.000
Standard

Deviation

Value .819 .920 .914 1.000 .630

 

A comparison of the reported heights and weights of fathers and
mothers to the suggested weights for given heights (Stanhope &
Lancaster, 1984) revealed that the mothers in both groups, dual and
single parent, were more inclined to be in the appropriate weight
range for their height than the fathers in the dual parent families
(Table 15). There were 46.6% (n=28) of the mothers in the dual parent
families, 38.5% (n=5) of the mothers in single parent families, and
30.1% (n=19) fathers within the suggested weight range for their
respective heights. Of those not within the suggested range, 63.2%
(n=38) of the fathers, 61.5% (n=3) of the mothers in the single parent
families, and 48.3% (n=29) of the mothers in the dual parent families
exceeded the suggested weight range. Only three parents, one father
and two mothers in the dual parent families, were below the suggested

weights for their heights.

Table 14

101

Long Term Impairment or Health Problem as Reported by 41* Respondents

in Dual and Single Parent Families.

 

Longéterm Impair-
ment or Health
Problem

n(13)

so.-

n(l3

V

Dual Parent

n(9)—%

Single Parent

 

Diseases or dis-
orders of the
neurological

system 7 11.

Endocrine,
nutritional,
and meta-
bolic

disorders 4 6.

Diseases or dis-
orders of the
respiratory

system 3 5.

Diseases or dis-
orders of the
circulatory

system 4 6.

Diseases or dis-
orders of the
muscula-
skeletal

system 1 1.

Psychological
disorders - -
Diseases of the
skin - -
Diseases of the
digestive
tract - -

Total 19

15

1 1.6

2 15.3 -

Total

7.5 2

49

 

*The total number of respondents who identified from one to
health conditions, therefore the total number of conditions
all the conditions identified by these respondents.

four
reflects

102

Table 15

Comparison of Suggested Heights and Weights to Reported Heights and
Weights of Parents in Dual and Single Parent Families.

 

 

 

 

 

 

Suggested Reported Body Weighgg

Heights and

Weights*

Hei t Weight Dual Parent Single Parent
Ft.In. Lbs. leg N9 Egg N9

Above Below Above Below
n(%) n(%) n(%) n(%) n(%) n(%)

§_ther_ (n=60)
5 1 109-138 - l( 1.7) - - - -
5 7 128-161 1( 1.7) - - - - -
5 8 132-166 2( 3.3) 4( 6.7) - - - -
5 9 136—170 1( 1.7) 3( 5.0) - - - -
5 10 140-174 6(10.0) 9(15.0) - - - -
5 11 144-179 2( 3.3) 8(13.3) - - - -
6 0 148-184 3( 5.0) 5( 8.3) - - - -
6 1 152-189 4( 6.7) 1( 1.7) - - - -
6 2 156-194 - 4( 6.7) 1( 1.7) - - —
6 3 160-199 - 2( 3.3) - - - -
6 4 164-204 - 1( 1.7) - - - -
6 6 172-214 - l( 1.7) - - - -

Total 19(30.l) 38(63.2) 1( 1.7) - - —

Missing Data=2( 3.3%)

Mothers (n=60) (n=l3)
5 0 96-125 - l( 1.7) - - 2( 3.3) -
5 1 99-128 - - 1( 1.7) - - -
5 2 102-131 4( 6.7) 3( 5.0) - 1( 7.7) l( 7.7) -
5 3 105-134 3( 5.0) 5( 8.3) - 1( 7.7) 1( 7.7) -
5 4 108-138 5( 8.3) 3( 5.0) - - l( 7.7) -
5 5 111-142 4( 6.7) 6(10.0) - 2( 15.4) - -
5 6 114-146 7(1l.6) 2( 3.3) l( 1.7) - 2(15.4) -
5 7 118-150 1( 1.7) 5( 8.3) - l( 7.7) 1( 7.7) -
5 8 122-154 2( 3.3) 4( 6.7) - - - -
5 9 126-158 1( 1.7) - - - -
5 11 134-168 1( 1.7) - l( 1.7) - - -

Total 28(46.6) 29(48.3) 3( 5.0) 5(38.5) 8(61.5) -

 

*From the Department of Health, Education and Welfare Conference on
Obesity, Washington, 0.0., 1973 as cited by Stanhope & Lancaster
(1984, p.470).

103

The 25% to the 75% range of the physical growth percentiles for
height, weight, age, and gender published by Ross Laboratories (1982)
was used to analyze the reported heights and weights of the 73
adolescents. Forty-four adolescents were males; 36 were from dual
parent families and eight were from single parent families.
Twenty-nine adolescents were females; 24 were from dual parent
families and five were from single parent families. Data from two
adolescents, one male and one female were not available. In general,
the sample of adolescents reflects an appropriate distribution of
height and weight by age and gender as might be anticipated (Tables 16
and 17). However, it must be noted that a disproportionate number of
the 15 year old males in this sample are taller and heavier than that
anticipated for 15 year olds in general.

Analysis revealed that 11 of the 16 thirteen year old males, and six
of the 12 thirteen year old females were within their percentile range
for height while the five remaining males and five remaining females
were taller. Only one female 13 year old was below the 25% for
height. This distribution was identical for their weights. Twelve of
the 18 fourteen year old males and eight of the nine 14 year old
females were within their percentile range for height while three
males and one female were taller. Three males were below the 25% for
height. Nine of the males and seven of the females were within their
range for weight while seven males and one female weighed more and two
males and one female weighed less. Two of the nine 15 year old males
and five of the seven 15 year old females were within their percentile
range for height while six males and two females were taller and only

one male was below the 25% for height. Only one male was within the

104

percentile range for weight; the other eight exceeded the weight
range. Six of the females were within their range while one also

exceeded the weight range.

Table 16
Comparison of Suggested Heights and Weights by Age to the Heights

Weights and Ages Reported by Female Adolescents in Dual and Single
Parent Families.

 

Suggested Body Heights and Weights bygAggt

 

 

Female
Adolescents
(n=28)i%
13(n=12) 14 n=9 l5§n=72
Ht. WT. Ht. Wt. Ht. Wt.

an?" 8851—12 5'1"——5"5" 923ng 5’2'"—-'S’6" 10m
Yes No Yes No Yes No Yes No Yes No Yes No
A“ BO A“ B6 A“ 86 A“ B6 A“ BC A“ BC

 

Dual Parent
Families
(n=22) 5 3 1 5 3 1 6 2 - 6 1 l 4 2 - 5 l -

Single

Parent

Families

(n=6) 1 2 - 2 1 - - l - l - - 1 - - l - -

Total 6 5 l 7 4 1 6 3 - 7 1 1 5 2 - 6 l -

 

*From Girls: 2;to 18 YeargiPhysiggl Growth NCHS Percentileg, Ross
Laboratories, 1982.

*XMissing Data=l in a dual parent family.

“=Above

O=Below

105

Table 17

Comparison of Suggested Body Heights and Weights by Age to the
Heights, Weights and Ages Reported by Male Adolescents in Dual and
Single Parent Families.

 

Sgggested Bogy,Heights and Weights by Age*

 

 

 

 

 

 

Male
Adolescents
(B=43)**
13(n=16) l4(n=18) 15(n=9)
Ht. WT. Ht. Wt. Ht. Wt.
sum" 88-11T 5'2"-5”‘7" loo—15:? 5'5"-5“‘9" 112313;
Yes No Yes No Yes No Yes No Yes No Yes No
A“ Be A“ 36 A“ 36 A“ Be A“ B6 A“ B6
Dual
Parent
Families
(n=36) 7 5 - 4 6 2 10 3 3 8 6 2 2 6 - 1 7 -
Single
Parent
Families
(n=7) 4 - - 2 l l 2 - - 1 1 - - 1 - - l -
Total 11 5 - 6 8 3 12 3 3 9 7 2 2 7 - 1 8 -

 

*From Boys: 2 to 18 Years Phygical Growth NCHS Percentileg, Ross
Laboratories, 1982.

##Missing Data=1 in single parent family.
“=Above
6=Below
Summary of Major Characteristics of the Sample

This sample of 206 respondents representing 73 families, 60 two
parent families and 13 single parent families, was generally
homogeneous. Respondents were predominantly white Protestants who
actively participated in church related activities, and perceived
themselves to be in good health although long term health conditions
were present. The reported weights exceeded the suggested weight for

height for approximately half of the respondents. The fathers tended

to be between 40 and 50 years of age, the mothers in dual parent

106

families between 35 and 45, and the mothers in single parent families
between 31 and 40. A very high proportion of the all adults had
attended or completed four years of college and were currently
employed in managerial or professional positions. The adolescents,
age 13 to 15, were students. The dual parent families ranged in size
from three to seven members with 51.7% composed of four members while
the single parent families were almost equally distributed in size
from two to five members. Dual parent families represented middle to
upper middle income categories while the single parent families
represented the low to middle income range.

Testing of the Hypotheses

The major variables in this investigation were health status,
intrinsic motivation, family functioning, and health behavior
(psychosocial and behavioral). Data on each of these variables and
other personal characteristics were collected on a total of 206
individuals representing 73 families. The individuals selected were
the father, the mother, and a young adolescent from 60 dual parent
families and the mother and a young adolescent from 13 single parent
families.

There were three primary purposes of this investigation: (a) to
determine if there was congruency of family members’ reports of the
major variables, (b) to investigate relationships between variables
which had support for congruency within the context of the family, and
(c) to describe relationships between the variables which had support
for congruency and the family realm variables, i.e., family type and
family income, from the perspective of all family members. The

discussion of the testing of hypotheses reflects these purposes.

107

Hypotheses 1 through 4: Comgruence of meily Members’ Measures

The influence of the family on the health behavior of its
members was investigated through the examination of the congruence of
the family members’ reports on family and personal factors related to
individual health and health behavior. It was assumed that
congruence reflects the presence of previously unknown or minimally
investigated family phenomena referred to as family characteristics.
The hypothesized family characteristics investigated included: family
health status, family intrinsic motivation, family functioning, and
family health behavior, psychosocial and behavioral.

Two statistical approaches were used to examine congruency. The
first approach, the Discrepancy Score Model (Klein, 1984), initially
required the calculation of a family mean for each variable. This
mean was then used to derive a family discrepancy mean (score) for
each variable. The use of the t-test on all family discrepancy means
(scores) determined congruency. Analysis of variance with repeated
measures and the F-test were utilized in the second approach, the
Conjunctive Model (Klein, 1984). Since the Health Practice Survey
(HPS) was specifically designed to measure ten subscales of
behavioral health behavior, both statistical approaches were utilized
to examine congruency of the total measure as well as each of the
subscales. The results of the data analyses using the t-test are
presented in Table 18 and those using the F-test are presented in
Table 19. Both will be discussed in relation to each of the

hypotheses (Ho).

108

Hg_l

Health status scores of the parent(s) and young adolescent
within the same family will be congruent.

The analyses indicated that congruency between health status
scores (WI) of the parent(s) and young adolescent existed for the
dual and single parent families. However, this result was supported
only by the Conjunctive Model (Klein, 1984) and examination of the
within-family factor available through the analysis of variance with
repeated measures. The F-statistic was 1.32 (p=.272) for dual parent
families and 1.16 (p=.303) for the single parent families.
Congruency was not supported by the Discrepancy Score Model (Klein,
1984) which utilized calculated t-scores. The t score for the dual
parent families was 16.71 (p=.01) and 6.74 (p=.01) for the single
parent families.

Hg_g

Intrinsic motivation measures of the parents and young
adolescent within the same family will be congruent.

Analyses of intrinsic motivation as measured by the Health
Self-Determinism Index (HSDI) indicate that congruency between
personal intrinsic motivation scores of the parents and young
adolescent within the same family was not present. This result was
supported by both the Discrepancy Score and Conjunctive Models
(Klein, 1984). The t-scores for both the dual and single parent
families exceeded their t statistic values, the t-scores were:

12.30 (p=.01) for the dual parent families and 6.48 (p=.01) for the
single parent families. The F statistics were: 19.51 (p=.000) for

the dual parent families and 17.84 (p=.001) for single parent

109

Table 18
The t-test for Determining Congruence (Aggregate Descriptive Statistics)
on the Major Variables by Dual and Single Parent Families.

Variables Dual Parent Families Single Parent Families

n Mean SD t Sig. n Mean SD t Sig.

 

 

Health .
Status (WI) 60 1.59 .74 16.71 .01: 13 .92 .49 , 6.74 .011
Intrinsic

Motivation

(HSDI) 60 .67 .42 12.30 .01! 13 .67 .37 6.48 .01:

Health

Behavior(p)

(SCI) 60 .74 .37 15.46 .013 13 .47 .39 4.29 .01

Health

Behavior(b)

(HPBQ-

Part 2 ) 59 1.03 .55 14.29 .013 13 .55 .55 4.24 .01

Health

Behavior(b)

(HPS) Total 60 .48 .26 14.21 .01! 13 .28 .28 4.90 .013
Auto/Fed 60 65 .37 13.75 .013 13 .39 .31 4.51 .01:
Dental 60 .90 .42 16.71 .013 13 .46 .37 4.82 .01#
Exercise 60 1.80 .92 15.11 .013 12 .62 .60 3.75 .01
Food 60 .62 .34 14.24 .01: 13 .28 .17 5.98 .01:
Household 60 .78 .46 13.13 .011 13 .53 .46 4.16 .01
Personal

Care 60 .85 .45 14.70 .018 13 .70 .56 4.46 .01!
Recreation 60 .89 .55 12.55 .013 13 .79 .46 6.04 .01!
Sleep 60 .98 .52 14.40 .01! 13 .55 .44 4.55 .01:
Stress 60 .79 .41 14.92 .013 13 .35 .22 5.73 .013
Substances 60 .46 .32 9.89 .01! 13 .41 .39 3.75 .01

Family

Functioning

(FAD-7) 60 .79 .43 14.24 .013 13 .37 .10 3.60 .01

p = psychosocial b = behavioral

#Significant: the statistical null hypothesis is rejected and therefore
congruence does not exist.

110

Table 19
Analysis of Variance with Repeated Measures and F-test for Determining
Congruence (Within-Family Factor) by Dual and Single Parent Families.

 

22519212; Qmml ngent Familigg Single Pgrent Families

SS 0? MS F Sig. SS DF MS F Sig.

 

Health

Status (WI) 1.70 2 .85 1.32 .272 .62 1 .62 1.16 .303

Intrinsic ’

Motivation -

(HSDI) 3.87 2 1.93 19.51 .0008 2.22 2 2.22 17.84 .0013

Health

Behavior(p)

(SCI) 1.08 2 .54 4.00 .0211 1.37 1 1.37 16.71 .0023

Health

Behavior

(HPBQ-

Part II) 2.10 2 1.05 3.81 .025! .22 l .22 .84 .378

Health

Behavior

(HPS) Total .69 2 .34 6.07 .003# .25 l .25 6.09 .030!
Auto/Pad. 2.62 2 1.31 11.94 .000t .45 l .45 4.87 .0488
Dental 2.28 2 1.14 6.24 .0033 .08 1 .08 .47 .508
Exercise 24.03 2 12.02 18.94 .000: .08 1 .08 .22 .650
Food 1.57 2 .78 8.83 .0003 .00 l .00 .06 .811
Household 2.28 2 1.14 7.47 .0018 1.56 1 1.56 12.53 .0043
Personal

Care 3.18 2 1.59 8 85 .0033 2.80 1 2.80 15.08 .0023

Recreation 1.02 2 .51 2.35 .100 .24 1 .24 .58 .460
Sleep 1.12 2 .56 2.25 .110 .80 1 .80 4.20 .0633
Stress .55 2 .28 1.71 .186 .32 l .32 5.24 .0413
Substances 1.43 2 .72 13.34 .000! .24 1 .24 1.63 .225

Family

Functioning

(FAD-7) 1.60 2 .80 5.30 .006# .16 l .16 1.24 .287

 

#Significant- the statistical null hypothesis is rejected and therefore
congruence does not exist.

111

families. All supported the rejection of the hypothesis.
Hm_§

Health behavior scores of the parent(s) and young adolescent
within the same family will be congruent.

Since there were three measures of health behavior, psychosocial
measured by the Self-Care Inventory (SCI), and behavioral measured by
the Health Protective Behavior Questionnaire - Part 2 (HPBQ-Part 2)
and the Health Practices Survey (HPS), each was analyzed separately
in evaluating this hypothesis. The results reveal that congruency
between health behavior scores of the parent(s) and young adolescent
within the same family system were not congruent as measured by the
SCI when using the Conjuctive Model (Klein, 1984). The results for
the SCI using the Discrepancy Score Model were mixed. The t-scores
were: 15.46 (p=.01) for the dual parent families and therefore
significant and 4.29 (p=.01) for the single parent families and
therefore not significant. The F statistics were: 4.00 (p=.021) for
the dual parent families and 16.72 (p=.002) for the single parent
families.

On the HPS (Total), the t-scores were: 14.21 (p=.01) for the
dual parent families and 4.90 (p=.01) for the single parent families.
The F statistics were: 6.07 (p=.003) for the dual parent families and
6.09 (p=.030) for the single parent families. Again, both rejected
the hypothesis.

The results were very mixed for the HPBQ-Part 2. The t-scores
were: 14.29 (p=.01) for the dual parent families and 4.24 (p=.01) for
the single parent families. Therefore, congruency was not supported

for the dual parent families, but was supported for the single parent

112

families when assessed by the Discrepancy Score Model (Klein, 1984).
Mixed findings also resulted when the analysis of variance with
repeated measures and the F-test (Conjunctive Approach, Klein, 1984)
were used. The F statistic, 3.81 (p=.025) was significant for the
dual parent families and thus indicated congruency was not supported,
but the F statistic, .84 (p=.378) was not significant for the single
parent families and thus supported congruency.

Although the subscales of the HPS were not included in the
hypotheses, they were investigated for congruency. This decision was
supported by evidence that health behavior is multidimensional
(Kulbok, 1985) and that the categories of behaviors conceptualized in
the HPS (Bruhn & Parcel, 1982a) reflected multidimensionality. Once
again the analyses revealed mixed results. The t-tests (Discrepancy
Score Model, Klein, 1984) for each of the subscales for the dual
parent families were significant and thus did not support congruency.
However, congruence was supported for three subscales for the single
parent families. The t statistics (p=.01) were: exercise (3.75),
househould (4.16), and substance use (3.75). The F-tests
(Conjunctive Model, Klein, 1984) indicated support of congruence on
some subscales for both the dual and single parent families. The F
statistics for dual parent families were:. (a) recreation--2.35
(p=.100), (b) sleep-2.25 (p=.110), and (c) stress-1.71 (p=.186).
There was also support for congruence for five subscales for the
single parent families: (a) denta1-.47 (p=.508), (b) exercise-.22
(p=.650), (c) food--.06 (p=.811), (d) recreation--.58 (p=.460), and
(e) substance use--l.63 (p=.225). The only subscale demonstrating

overlap between family types when the F-test was used was recreation.

113

A comparison of overlap of subscales for single parent families with
the conceptual models and analytical approaches revealed that the
subscales of exercise and substance use were consistently supported
for congruency.

Hg_g

Family functioning scores of the parent(s) and young adolescent
within the same family system will be congruent.

The analyses of the family functioning scores as measured by The
General Functioning Subscale of the McMaster Family Device (FAD-7)
indicated mixed support of this hypothesis. The family functioning
scores of the single parent families were supported for congruency
when using both the Conjunctive and Discrepancy Score models (Klein,
1984). The F-statistic was 1.24 (p=.287) and t-statistic was 3.60
(p=.01). However, neither model supported congruency for family
functioning for the dual parent families, i.e., the F statistic was
5.30 (p=.006) and the t-statistic was 14.24 (p=.01).

Hypotheges 5 through 13: Relationships Between meimbles whigm;are
Congruent Within the Context of the Family System
According to the F statistics resulting from the analysis of
variance with repeated measures, congruency was supported on three
major variables: health behavior in single parent families as
measured by the SCI and the HPBQ-Part 2, health status in both single
parent and dual parent families, and family functioning in single
parent families. Although congruency was not supported for health
behavior as measured by the HPS as a total measure, congruency was
supported for several of the subscales in the HPS. The investigation

of the relationships between these subscales and the subscales with

114

other variables with support for congruence were not included in the
hypotheses and therefore were not assessed at this time. This
section will focus on the relationships between the major variables
with support for congruency as presented in the hypotheses. Only
those hypotheses having both variables with support for congruency
were examined; therefore, hypotheses which included intrinsic
motivation, i.e., #7, #8, #9, and #10, were not examined. These
were:
Hypothesis 7: The intrinsic motivation mean score
of each family system will correlate positively with the
family’s psychosocial health behavior mean score.
Hypothesis 8: The intrinsic motivation mean score
of each family will correlate positively with each
of the family’s two behavioral health behavior scores.
Hypothesis 9: The intrinsic motivation mean score
of each family will correlate positively with the
family’s health status mean score.
Hypothesis 10: The intrinsic motivation score
of each family will correlate positively with the
family’s composite family functioning mean score.
Hypotheses #5, #11, #13 and one component of hypotheses #6 and
#12 were examined. The partial hypotheses resulted due to the
examination of only one of the two measures of behavioral health
behavior, i.e., the HPBQ-Part 2. The hypotheses as examined were:
9999
The family health status mean score of each family will

correlate positively with the family’s psychosocial health behavior

115

mean score.
fl9_§

The family status mean score of each family will correlate
positively with the family’s health behavior mean score as measured
by the HPBQ-Part 2.

.30—11

The family functioning mean score of each family will correlate
positively with the family’s psychosocial health behavior mean score.
112.12

The family functioning mean score of each family will correlate
positively with the family’s behavioral health behavior mean score as
measured by the HPBQ-Part 2.

E°_13

The family functioning score of each family will correlate
positively with the family’s health status mean score.

The process used in the examination of these hypotheses follows.
Initially a Pearson coefficient was obtained for the pairs of the
variables for each family as identified in the hypotheses; then
frequencies of these correlations by family type were established.

It was determined a priori that each hypothesis would be supported if
the majority (50%) of the families revealed a positive correlation of
r = .40. Since the number of participating single parent families
was small (n=13) and since Pearson correlations can not be computed
with an n=2, correlations were calculated only for the dual parent
families (Table 20). The maximum potential range of correlations,
i.e., -l.00 to +1.00, with numerous negative correlations resulted

for each hypothesis. The correlations were confirmed on a second

116

independent program suggesting the algebra was correct. The
presentation of the results, i.e., as negative Pearson coefficients,
means that four of the five hypotheses were rejected since 50% of the
correlations did not exceed r=.40. However, Ho 13 regarding the
FAD-7 and W1, i.e., family functioning and health status, was
supported since 50% (n=27) of the correlations exceeded r=.40.
Hypotheseg 14 thromgh 18: Relgtionmhip9:Between Selected Major
Variables and the Familngealm Variables
of Family Type and Family Incgmg
The investigation of Hypotheses 14 through 18 was based upon the
previous finding of support for congruency of the variables.
Therefore hypothesis 15, and one of the three components of
hypothesis 17 were not investigated. These hypotheses were:
Hypothesis 15: Dual parent families will have higher
family mean scores on intrinsic motivation than
.female-headed single parent families, and
Hypothesis 17: Dual parent families will have higher
family mean scores on health behavior as measured by
the SCI and PHS than female-headed single parent families.
Hypotheses 14, 16, two components of 17, and 18 were investigated.
M
Dual parent families will have higher family mean scores on
health status than female-headed single parent families.
ELLE
Dual parent families will have higher family mean scores on

family functioning than female-headed single parent families.

117

Table 20
Frequency of Pearson Correlations for Dual Parent Families on Combinations
of Major Variables with Support for Congruence.

 

Frequengy of Pearson Correlations

 

Value Range FAD-7 & HPBQ:Part 2 FAD-7 & Fad-7 & fll_9
_L__m Part 2 LE! £1 £1. .52!
(n=59) (n=51) (n=52) (n=60) (n=52)
-1.00 to .91 10 13 4 16 12
.90 to .81 4 5 1* 10 6
.80 to .71 4 2 l 9 3
.70 to .61 5 4 l 3 -
.60 to .51 3 - 3 l l
.50 to .41 3 2 2 2 2
.40 to .31 2 l 1 2 4
.30 to .21 l 1 3 2 3
.20 to .11 4 1 2 l l
.10 to -.01 l 1 2 - -
.00 l 3 l - 1
+.01 to .10 1 2 O - 1
.11 to .20 3 2 2 - 1
.21 to .30 l - 1 - 1
.31 to .40 l - 2 - l
.41 to .50 1 4 6 - 2
.51 to .60 - 1 3 2 2
.61 to .70 2 1 0 - 1
.71 to .80 2 2 2 2 2
.81 to .90 2 - 4 4 -
.91 to +1.00 8 6 12 6 8
Mean - .175 - .224 .218 - .399 - .192
Median - .339 - .436 .383 - .767 - .380
SD .692 .705 .653 .727 .102
Kurtosis -1.236 -1.257 -1.l36 - .533 -1.344
Skewness .484 489 - .415 1.093 .442

Ranfie 1.995 1.996 1.998 1.997 1.999

 

118

B_°_1_7

Dual parent families will have higher family mean scores on
health behavior as measured by the SCI and HPBQ-Part 2 than
female-headed single parent families.

The results for these hypotheses are presented in Table 21.

Table 21

Family Mean Scores on Family Functioning, Health Behavior, and
Health Status by Family Type

 

Family Memp Scorgg by Family Type

 

Measures Dual Parent Single Parent
Mn SD Mn SD

 

Family Functioning
(FAD-7) 1.93 .33 1.89 .37

Health Behavior
(HPBQ-Part 2) 4.80 .40 4.87 .58

Health Behavior
(SCI) 3.50 .21 3.62 .44

Health Status
(WI) 2.18 .59 2.17 .66

 

The family mean scores for dual and single parent families on
all three variables, i.e., family functioning, health behavior and
health status are essentially the same. This suggests that these two
samples represent similar populations of families on these variables.
Therefore, there is no support of the hypotheses as stated. Review
of the standard deviation of the means reveals that the variance of
the family mean scores for both family types is once again
essentially the same for family functioning and health status.
However, an appreciable difference is noted for health behavior as

measured by both the SCI and HPBQ-Part 2. The single parent families

119

 

in this sample are less homogeneous in respect to health behavior
than are the dual parent families.

The last hypothesis investigates the relationship between the
major variables with support for congruency and family income. Since
intrinsic motivation, and one of the measures of health behavior,
i.e. behavioral as measured by the PHS received no support for
congruency, they were deleted from the hypothesis for the analysis.
The hypothesis tested was:

.H£_1_§

The family mean scores of the family characteristics of family
health status, family functioning, and family health behavior as
measured by the SCI and HPBQ-Part 2 will correlate positively with
family income.

The results are presented in Table 22. Pearson product moment
correlations were calculated to assess these relationships. Although
not specified in the hypothesis, Pearson coefficients for both family
types are given. Very low positive correlations between health
behavior and income are revealed for the two parent families {the
HPBQ-Part 2 (r=.13) and the SCI (r=.02)} and between the SCI and
income for the single parent families (r=.20). A very low positive
correlation (r=.06) between health status and income for single
parent families also resulted. All other hypothesized relationships
had negative correlations. Specifically, these relationships in
descending order were: family functioning and income for dual parent
families (r=-.ll), health status and income for dual parent families
(r=-.17), health behavior (as measured by the HPBQ-Part 2) and income

for single parent families (r=-.22), and family functioning and

120

income for single parent families (r=-.39). Although, there was
partial support for this hypothesis, the Pearson coefficients did not
exceed the predetermined significance level (r = .30), and therefore
the hypothesis was rejected.

Table 22

Pearson Product Moment Correlation Coefficients Between the Major
Variables with Support for Congruency and Income by Family Type.

 

 

 

 

 

Major Variables Incomg by FamilyiType
Dual Parent Single Parent
(n=59) (n=l3)

 

Family Functioning
(FAD-7) - .11 - .39

Health Behavior
(HPBQ-Part 2) .13 — .22

Health Behavior
(SCI) .02 .20

Health Status
(WI) - .17 .06

 

Summary of the Hypotheses Testing

The results of the hypotheses testing are presented in Table 23.
The first four hypotheses explored the presence of family
characteristics as determined by the congruency of multiple family
perceptions on the variables of family functioning, intrinsic
motivation, health behavior, i.e., psychosocial and behavioral,
andhealth status. Congruency was examined through the application of
two conceptual and statistical approaches: (a) the Discrepancy Score
Model and the t-test, and (b) the Conjunctive Model and analysis of
variance with repeated measures and its associated F-test. There was

support for some hypotheses but the support differed by model and

121

Table 23
Result of Hypotheses Testing

 

Dual Parent Family Single Parent Family

 

 

 

 

 

 

 

Hypotheses
Conj.* Disc.0 Conj.* Disc.@
S“ R+ S“ R+ S“ R+ S“ R+
5911,“ Ho 4
(Congruent measures)
Ho 1: Health Status X X X X
Ho 2: Intrinsic
Motivation X X X X
Ho 3: Health Behavior:
a) Psychosocial X X X X
b) Behavioral
l) HPBQ-
Part 2 X X X X
2) HPS X X X X
Aud/Ped. X X X X
Dental X X X X
Exercise X X X X
Food X X X X
Household X X X X
Personal
care X X X X
Recreation X X X X
Sleep X X X X
Stress X X X X
Substances X X X X
Ho 4: Family
Functioning X X X X
Ho 5 - Ho 13
(Relationships between supported family characteristics within family)
8“ R+
Ho 5
(Health status & psychosocial health) X
Ho 6 (Revised)
(Health status & HPBQ-Part 2) X
Ho 7 (Not examined)
Ho 8 (Not examined)
Ho 9 (Not examined)
Ho 10 (Not examined)
Ho 11
(Family functioning & psychosocial behavior) X

Ho 12 (Revised)
(Family functioning & HPBQ-Part 2) X

122

Table 23
Result of Hypotheses Testing (ContinUed)

 

 

 

Ho 5 - Ho 13

(Relationships between supported family characteristics within family)
S“ R+

Ho 13

(Family functioning & health status) X

 

Ho 14 - Ho 18

(Relationships between supported family characteristics, family
type and income)

 

8" 12+

Ho 14 _
(Health status mean scores of dual parent families

higher than single parent families’) X
Ho 15 (Not examined)
Ho 16
(Family functioning mean scores of dual parent families

higher than single parent families’) X

Ho 17 (Revised)
(Health behavior mean scores {SCI and HPBQ-Part 2} of
dual parent families higher than single parent
families) X
Ho 18 (Revised)
(Family mean scores of supported characteristics related
to family income) X

 

* Conjunctive Model

9 Discrepancy Score Model
“ Supported

+ Rejected

123

family type. The Discrepancy Score Model and the t-test failed to
support congruency for all major variables for the dual parent
families. However, there was support of family functioning and two
of the health behavior measures, i.e., SCI and HPBQ-Part 2 for single
parent families. This model and statistical approach also supported
congruency for three of the ten subscales (exercise, household, and
substances) of the third health behavior measure, the HPS, but again
only for the single parent families.

There was also support of congruency for several of the
variables when the Conjunctive Model and its allied statistical
approach was used. Health status, behavioral health behavior as
measured by the HPBQ-Part 2, and five subscales of the PHS (dental,
exercise, food, recreation, and substance use) were supported for
congruency for single parent families. Health status and three
subscales of the HPS (sleep, recreation, and stress) were supported
for congruency for dual parent families.

Hypotheses #7, #8, #9, and #10 were not investigated since there
was no support for congruency of their respective variables.
Hypotheses #5, #11, #13, and modifications of #6 and #12 were
investigated for only dual parent families. Positive correlations
between and among health status, behavioral health behavior as
measured by the SCI and HPBQ-Part 2, and family functioning within
the context of the individual family were proposed. The analysis
yielded a large number of negative correlations and rejection of four
of the hypotheses. Only the relationship regarding the family

functioning and health status was supported.

 

124

Hypotheses #15 and one component of #17 were not investigated.
Hypotheses #14, 16 and two components of #17 were investigated.
These hypotheses focused on the relationship between and among the
families’ mean scores on health status, family functioning, and
health behavior as measured by the SCI and HPBQ-Part 2 and the family
realm variable of family type. Although higher mean scores for dual
parent families were hypothesized, no support could be found and the
hypotheses were rejected.

The final hypothesis, #18 focused on the relationship between
health status, family functioning, and behavioral health behavior as
measured by the SCI and HPBQ-Part 2 and the family realm variable of
family income. Pearson product moment correlations gave minimal
support for this hypothesis. Although low positive Pearson
correlations resulted, they did not exceed the apriori r=.30, and the

hypothesis was rejected.

 

CHAPTER V

OVERVIEW, DISCUSSION, CONCLUSIONS AND IMPLICATIONS

This chapter presents an overview of this study, a discussion of
the findings, study conclusions, and implications for future
investigations.

Overview of the Study

Numerous investigations aimed at the identification of
individual factors which influence health behavior have been
conducted by researchers representing a variety of disciplines. It
is assumed that factor identification will enable health and related
professionals to develop better intervention approaches which would
facilitate better practices and promote personal well-being.
Although many factors have been identified, personal, social, and
environmental influences have been the primary components of the
health behavior frameworks and models used in these investigations.

The primary purpose of this study was to investigate the
possible influence of the family, an important social factor, on the
health behavior of its members. Family researchers suggest that
phenomena unique to the family, i.e., family characteristics, may
exist and if present could reveal additional insights about the
family’s influence. Therefore, selected variables identified in

125

126

health behavior frameworks and models were evaluated from the
perspective of a family characteristic. The assessment of the family
characteristic was dependent upon the congruency among multiple
family members’ perceptions and behavior. The family characteristics
investigated were: family health status, family intrinsic motivation,
family health behavior (psychosocial and behavioral), and family
functioning. Two approaches were used: (a) the Discrepancy Score
Model (Klein, 1984) which utilized the statistical approach of the
t-test on a derived discrepancy family score, and (b) the Conjunctive
Model (Klein, 1984) which utilized the statistical approach of
analysis of variance with repeated measures and the F-test.

Relationships were explored between the confirmed family
characteristics within the context of the individual family and
between the same characteristics and the family realm (Beutler et
al., 1989) variables of family type and family income. Individual
family member data including age, gender, ethnicity, religiosity,
marital status, educational level, occupational status, and height
and weight were also obtained for use in the description of the
sample.

Sixty dual parent families composed of the father, the mother,
and the young adolescent between the ages of 13 and 15, and 13
female-headed single parent families composed of the mother and the
young adolescent between the ages of 13 and 15 comprised the sample.
Families were identified through contacts with the religious leaders
from a stratified random sample of religious congregations in a
specified geographical area in the Greater Lansing Area Community.

Each family member completed the same paper and pencil questionnaire

127

which represented a composite of measurement instruments appropriate
to the selected major variables, i.e., health status [Ware Index
(WI)], intrinsic motivation [Health Self-Determinism Index (HDSI)],
health behavior [psychosocial as measured by the Self-Care Inventory
(SCI) and behavioral as measured by the Health Protective Behaviors
Questionnaire-Part 2 (HPBQ-Part 2) and the Health Practices Survey
(HPS)], and family functioning [The General Subscale of the McMaster
Family Functioning Device (FAD-7)]. Although each instrument had
documented psychometric properties, the reliabilities of these same
instruments were also established in this investigation for the
subsamples of fathers, mothers, and adolescents.

Discussion of the Findings

The discussion of the findings has been organized from the,
perspective of the three primary research objectives of this
investigation.

Objective 1

The first objective was to determine if there was congruence of
family members’ reports of health status, intrinsic motivation,
health behavior, and family functioning.

The results of the analyses indicated that congruence was
supported for some of the proposed family characteristics for both
the dual and single parent families. They also demonstrated that
differences occurred in relation to the conceptual and statistical
approaches utilized and that support for congruency was present for a
greater number of variables for single parent families than for dual
parent families. Although Klein (1984) recommended researchers

utilize multiple conceptual and statistical approaches, he also

128

indicated that the results may be difficult to interpret. The
current outcome is a good example. Two questions were asked: Which
of the two conceptual models and associated statistical approaches
reflect congruency? and Why are there differences in the results?

The first question was considered initially from the
perspective of number of hypotheses confirmed and applicability
across groups. In this study the application of the Conjunctive
Model (Klein, 1984) and its associated statistical approach, analysis
of variance with repeated measures and the F-test demonstrated both.
This model supported the presence of family characteristics for both
family types on the greatest number of variables, i.e., 12 of the 32
tests (37.5%.) In contrast, the application of the Discrepancy Score
Model (Klein, 1984) and its associated statistical approach, the
t-test, supported the presence of the family characteristics only for
single parent families and on fewer variables , i.e., 6 of the 32
tests (18.7%). Although this suggests that the former approach may
be superior, other factors need to be considered. In general,
congruency was supported much more frequently for the single parent
families. Potential explanations include the following: the family
identification process influenced this outcome or the single families
who participated are different from the participating dual parent
families. Some questions resulting from these results are: (a) Would
the same result occur if the single parent family sample size had
been equivalent to that of the dual parent families? (b) Is there
something special about the dyad relationship? (c) Is there really

less variance in the single parent family?

129

The analyses were limited to two of the five possible approaches
identified by Klein (1984). Analyses of the same data using the
other models, i.e., the Additive Model, the Disjunctive Model and the
Weighted Model, could assist in determining which variables are
stronger contenders for being true family characteristics. It may be
that one model and/or analytical approach is more appropriate when
investigating health and health behavior variables. It is also
important to note that there are other statistical approaches, e.g.,
factor analysis, cluster analysis, canonical correlation, and
discriminant function analysis, which are appropriate for use with
the Conjunctive Model (Klein, 1984). Their application to data from
this study could further support or possibly refute the current
findings.

An investigative concern regarding the comparison of the results
of this study to other investigations was present. These analyses
were exploratory in nature and therefore could not be directly
compared to those of other investigations. There was no evidence
that these selected conceptual and statistical approaches had been
previously used on similar data. In the current study, health status
was the only variable to be supported as a family characteristic for
both family types. However, no previous investigations were located
which viewed health status as a family characteristic and therefore
no comparison could be made. Also, health status was a one item
variable; it was the only variable which did not meet the criteria
for a family characteristic as proposed by Fisher et a1. (1985).

Family functioning and health behavior, as measured by the

HPBQ-Part 2 and SCI, were also supported as family characteristics

 

130

for the single parent families. Although behavioral health behavior,
as measured by the total HPS, did not support congruency, several of
its subscales did and therefore each could be considered as a family
characteristic.

The investigation of both family functioning and health behavior
by other researchers has been limited. According to Epstein et a1.
(1983) minimal investigation of the FAD (the total instrument) and
the FAD-7 as family measures has occurred. However, exploration of
both as family measures has been recommended (Epstein et al., 1983).
In regards to health behavior, there was no evidence that
psychosocial health behavior as measured by the SCI had been
investigated as a family measure. However, composite measures from
the family perspective for a range of health practices were utilized
by Loveland-Cherry (1986) and Pratt (1976). Although both have been
cited earlier in the text, they are reviewed here because each
utilized health behavior as a family measure. An in-depth review of
these studies demonstrates the current status of the use of family
measures (scores) in regards to health and health behavior.

Pratt (1976) developed individual member indexes as composite
measures on 15 different variables. The individual’s score on each
index resulted by applying a formula that assigned arbitrary weights
to the answer categories and summed the answers. These sums were
used in the calculation of family index scores. Two family indexes
were developed: (a) a family structure index, which included the
individual indexes related to interaction among members, autonomy,
and extramural participation, and (b) a family health/health behavior

index, which included the individual indexes related to personal

 

131

health practices, e.g., sleep, exercise, dental hygiene, nutrition,
substance use, use of preventive medical services, health problems
and health status. After experimenting with different formats,
individual family member scores on each index were dichotomized into
high or low scores. High scores were coded as number one and low
scores were coded as number two. These codes were then summed for
the family members, i.e., mother, father, and child, and a family
index score with a range from 0 to three was created. These family
index scores were used in the analyses, correlation coefficients and
regression analysis, for the postulated hypotheses. No additional
information about the individual or family index scores was
available. However, Pratt (1976) did state that these global
measures of the variables were research limitations. This research
assumed that the family characteristics of family structure, which is
considered consistent with family functioning, and family
health/health behavior existed and analysis proceeded based upon this
assumption.

Loveland-Cherry (1986) modified selected indices developed by
Pratt (1976) to measure health practices and selected variables of
social networks and socialization patterns between a convenience
sample of 20 dual parent and 21 female-headed single parent families.
The variable of personal health practices for each individual was
measured by combining the scores for the indices of sleep, exercise,
dental hygiene, smoking, alcohol use, and nutrition. Social networks
were measured by Pratt’s (1976) indices for community participation,
number of towns used for any purpose, child’s activities, and

combined family extramural participation. The variable socialization

 

132

patterns was measured by the indices of control of child by parent,
supportiveness of child by parents, and child’s autonomy. A
composite family score was derived for each index by summing the
individual family members’ scores and dividing this sum by the number
of family members. The investigation did not hypothesize the
presence of the family variable, but utilized the family scores to
detect similarities between the dual parent and single parent
families. There was no statistical evidence, via the t-test, that
differences existed between the two family types.

Some comparisons between the study by Loveland-Cherry (1986) and
the current study can be made. Although Loveland-Cherry (1986)
utilized the Additive Model (Klein, 1984), rather than the
Discrepancy Score or Conjunctive Models (Klein, 1984), the practices
included in the personal health practices index were similar to those
in both the HPBQ-Part 2 and the HPS. Also, dual and single parent
families participated in both investigations. One comparison can be
made regarding the respondents’ personal characteristics. The age
and educational level of the mothers and fathers participating in
these two studies were similar. In addition, the mean age of the
children in Loveland-Cherry’s sample was 11 and the mean age of the
children in the current study sample was approximately 13 to 14.
Family income of both family types in these two studies was also
similar.

Loveland-Cherry (1986) concluded that there were no differences
between families. When the simple family mean is used for family
comparisons in the current study, Loveland-Cherry’s conclusion was

confirmed. Hewever, if the family discrepancy mean scores (Table 18)

 

133

are used for the comparisons a different result occurs. A comparison
of means on the HPBQ-2 and the HPS for both dual and single parent
families suggests differences in practices by family type. On the
HPBQ-Part 2, the dual parent family mean was 1.03 while the single
parent family mean was .55. On the HPS the dual parent family mean
was .48 while the single parent family mean was .28. Since both
instruments attempt to assess the frequency of the respondents’
health practices, these findings suggest that the dual parent
families engage in more health enhancing practices than the single
parent families. However, these findings are subject to error and
cannot be considered accurate due to the process utilized to develop
the family discrepancy mean score.

The lack of congruence for the variables in this study may be
due to several factors. The first to be considered is the fact that
congruence truly does not exist. This may indeed be the case for
intrinsic motivation since neither conceptual approach supported
congruency for dual or single parent families. Another factor to
consider is the impact of the conceptual model and the selected
statistical approach. A third factor to consider is the adequacy of
the measurement instruments; they may enhance or diminish the
assessment of congruence. In this study composite measures were used
for intrinsic motivation, health behavior, and family functioning.
The use of single item measures, e.g. health status, or 4-6 item
measures, such as those in the HPS subscale, sleep, may be more
appropriate and facilitate the identification of congruence. A
fourth factor, sample size and characteristics, may also be

important. In this study the number of single parent families (n

 

134

=13) was small and not adequate. It is also possible that the number
of dual parent families (n = 60) while sizable was not adequate. The
investigation of congruence on these variables may require an even
larger sample. Family configuration may also be a factor, this
investigation focused on three members for the dual parent family and
two members for the single parent family. The relationships between
the pairs in the three member families were not assessed. The
investigation of the proposed family characteristics from the
perspective of pairs of family members may yield additional
infbrmation about congruence and the family. Another factor to
consider is that children of this age range, 13 to 15, although
capable of incorporating the health practices, which was an
assumption of this investigation, may not reflect the incorporation
of the health practices at this point in their development.

Objective 2

The second objective was to investigate the relationships
between the family characteristics of family health status, family
intrinsic motivation, family health behavior, i.e., psychosocial and
behavioral, and family functioning.

The investigation of this objective was dependent upon the
support for congruence of the proposed family characteristics. The
variables with support for congruence were health status, health
behavior, as measured by the SCI and HPBQ-Part 2, and family
functioning. Therefore, only the relationships between these family
characteristics were assessed. The approach selected to assess these
relationships follows: individual member mean scores on these

variables within each family were compared and Pearson coefficients

 

135

for the relationships determined. Since the number of single parent
families was only 13 and since Pearson correlations can not be
computed with an n = 2, correlations were calculated only for the
dual parent families. Although each of the hypotheses indicated that
a positive correlation would result, only the relationship between
the family functioning (FAD-7) and health status (WI) was supported.
Although this suggests that an increase in both occurs, it is not
known which variable facilitates the positive relationship. Pratt
(1976) demonstrated that ”family elements", i.e., variables which are
congruent with the items of the FAD-7, are related to health
behavior. Although the relationship between health behavior and
health status was not specifically investigated, it can be speculated
that health behavior may be a mediating variable or that a direct
relationship exists between family functioning and health status.
Many negative correlations occurred when investigating these
hypotheses. These may occur due to sample size. According to Glass
& Hopkins (1984), correlation cannot be relied upon when the n is
small. Perhaps 60 families were inadequate for this statistical
approach and therefore the results are not reflecting an accurate
picture of the hypothesized relationships. Another possibility
requiring consideration is that the negative correlations may be true
negative correlations. If so, the analyses suggest that families
cluster at the two ends of the possible range of correlations and
that there are differences between families on the combinations of
these variables. Another possible explanation of the negative
correlations is that one member’s score may be considerably higher

than the other two members’ scores, thus facilitating the negative

 

136

correlation. This could be examined by looking at the mean scores
for the compared variables for each family. Another possibility is
that the Pearson product moment correlation is not the appropriate
test. Further analysis using a non-parametric approach, e.g., the
Spearman Rank Correlation (Glass & Hopkins, 1984), or Kappa
coefficient (Uphold & Harper, 1986) could provide different results.

The intent of these hypotheses was to investigate relationships
between the major variables from the within-family perspective.
Although other investigations focusing on similar variables from a
between family perspective were located (Loveland-Cherry, 1986;
Pratt, 1976), no other investigation could be located which focused
on these variables from the within-family perspective.

Objective 3

The third objective was to describe selected relationships
between the family characteristics, i.e., family health status,
family intrinsic motivation, family health behavior, and family
functioning, and the family realm (Beutler et al., 1989) variables of
family type and family income.

The relationships investigated were once again dependent upon
the variables supported for congruency. Therefore, the relationships
between the family characteristics with support, i.e., health status,
health behavior, as measured by the SCI and HPBQ-Part 2, and family
functioning, and the family realm variables of family type and family
income were investigated. Pearson product moment correlation
coefficients between and among these variables revealed low positive
and low negative relationships. Although three low positive

correlations occurred between health behavior and income, they were

 

137

considered to be non-significant since they failed to reach the
criteria of r=.30. According to Glass and Hopkins (1984, p.115),
"the lower the correlation, the greater the margin of error". Two of
these correlations require specific comment: (a) behavioral health
behavior as measured by the HPBQ-Part 2 (r=.13 for dual parent
families), and psychosocial health behavior as measured by the SCI
(r=.20 for single parent families). A similar finding to the first
was reported by Loveland-Cherry (1986); however, neither finding was
suggestive of the positive relationship described in other
investigations (Berkman & Breslow, 1983; Coburn & Pope, 1974; Kasl &
Cobb, 1966). The high educational level of the adult respondents may
be a significant influencing factor.

The second finding suggested that psychosocial health practices
and income are related for single parent families. Investigation of
which variable influences the other requires further study. Two low
positive relationships, i.e., between health status and income
(r=.06) for the single parent families and between psychosocial
health behavior and income for dual parents, were also
non-significant.

The other correlations were negative; the largest negative
correlations, r=-.22 and r=-.39, occurred respectively between health
behavior and income and between family functioning and income in the
single parent families. Again possible explanations for these
results include the small size of the two samples, the homogeneity of
the samples, and the utilization of inadequate and/or inappropriate

measurement instruments.

138

Conclusions
The major conclusions of this study as organized by the three
major objectives are presented:
Objective 1

1. Personal health status was supported as a family
characteristic for both dual and single parent families
as assessed by the Conjunctive Mbdel.

2. Personal intrinsic motivation was not supported as a
family characteristic for either the dual or single
parent families as assessed by the Conjunctive or
Discrepancy Score Models.

3. Psychosocial health behavior was supported as a family
characteristic for single parent families as assessed by
the Discrepancy Score Model.

4. Behavioral health behavior (as measured by the
HPBQ-Part 2) was supported as a family characteristic
for single parent families as assessed by the Conjunctive
Model.

5. Three personal health behavior subscales of the HPS,
i.e., sleep, stress, and recreation, were supported as
family characteristics for dual parent families as
assessed by the Conjunctive Model.

6. Five health behavior subscales on the HPS, i.e.,
dental, exercise, food, recreation, and substance use,
were supported as family characteristics for single
parent families as assessed by the Conjunctive Model.

7. Three health behavior subscales on the HPS, i.e.,

10.

11.

12.

13.

14.

139

exercise, household, and substance use, were supported
as family characteristics for single parent families as
assessed by the Discrepancy Score Model.
Family functioning was supported as a family
characteristic for single parent families as assessed by
the Discrepancy Score and Conjunctive Models.

Objective 2
The health status mean score of each family
did not correlate positively with the family’s
psychosocial health behavior mean score as measured by
the SCI.
The health status mean score of each family did not
correlate positively with the the family’s behavioral
health behavior score as measured by the HPBQ-Part 2.
The family functioning mean score of each family did not
correlate positively with the family’s psychosocial
health behavior mean score as measured by the SCI.
The family functioning mean score of each family did
not correlate positively with the family’s behavioral
health behavior mean score as measured by the HPBQ-Part 2.
The family functioning mean score of each family did
correlate positively with the family’s health status mean
score.

Objective 3
Dual parent families and single parent families had
similar family mean scores on family functioning, health

behavior (as measured by the HPBQ-Part 2. and the SCI),

140

and health status.

15. There were non-significant low positive Pearson
correlations between health behavior (as measured by the
HPBQ-Part 2) and income for dual and single parent
families and between health status and income for single
parent families.

16. There were non-significant negative Pearson correlations
between family functioning mean scores and income, and
between family health status mean scores and income
for the dual parent families.

17. There were non-significant negative Pearson correlations
between family functioning mean scores and income and
between health behavior mean scores (as measured by the
HPBQ-Part 2) and income for the single parent families.

Implications of the Study

The implications of this study can be viewed from two
perspectives: the continued investigation of this study’s data set,
and future studies.

The ContinmgdiInvegpigmtion of this Data Set

This data set is large and therefore offers many opportunities
for continued analysis. The investigator plans to pursue the
investigation of the data and presents here some of the analyses
currently being considered.

Congruency was supported for single parent families on nine of
the 16 possible variables which were evaluated as potential family
characteristics. A question previously raised concerned the effect

family size and composition had on this outcome. The investigation

141

of the conceptual models using only mothers and adolescents from the
dual parent families is planned. This could provide additional
information regarding the impact of a third family member, the
father, on congruency. The investigation of the father-adolescent
dyad as representative of the family using the same approaches is
also anticipated. Another potential approach for the examination of
congruency would be to create dual and single parent families
artificially. This could be accomplished by randomly selecting non
related family members who participated in the study. A created dual
parent family could be composed of the mother from D1, the father
from D25, and the young adolescent from D45. A created single parent
family could be composed of the mother from 82 and the young
adolescent from 812. These created family groups, i.e., dual and
single parent families, could then be used as comparison groups of
unrelated individuals. Congruency should be least likely to exist
under these circumstances.

The investigator plans to assess these variables through the
application of alternative models and statistical approaches (Klein,
1984). This is particularly important since this study’s results
showed there was inconsistency in confirmation of a variable when the
two approaches, Discrepancy Score and Conjunctive Models, were used.

Although it was not the intent of this investigation to
investigate the variables by subsamples, the examination of the data
by gender and family position, i.e., father, mother, and adolescent
are just two of the potential areas to explore. Data were not
examined from the individual perspective or by a specific health

practice. The investigation of smoking practices or use of

 

142

automobile safety belts are just two of the 127 practices included in
the Health Practices Survey (HPS) which can be analyzed. Preliminary
data of the correlations between the major variables and the
subscales of the HPS by subsample dyads, i.e., the father and the
mother, the mother and the adolescent, and the father and the
adolescent, have been included as Appendix H and will be investigated
in the future.

The HPS has demonstrated potential as a measure of health
practices, but requires a great deal of refinement and verification
of its psychometric properties. Only the initial steps were taken in
this study. Evaluation of the subscales as dimensions of health
behavior as conceptualized needs to occur. According to Nunnally
(1978) factor analysis or cluster analysis may be used. However,
cluster analysis may be more appropriate since it can be used when
the sample size is reasonably small.

Several unplanned outcomes related to the HPS have occurred. A
copy of the HPS has been forwarded to each contributing party or
organization participating in its development. It is viewed as a
potential pool of items from which specific items will be selected
and tested in the 1989—1990 health practices component of different
tests developed for the Michigan Model for Comprehensive Health
Education (Vogel, personal communication, February, 1989). The
representative at the Sleep Disorder Clinic, Ingham Medical Center,
Lansing, Michigan, has also indicated an interest in future
cooperative investigative studies which could utilize the HPS as part
of the health evaluation of clients using the clinic services. Two

of the religious leaders requested copies; they thought the items

143

could be used to generate discussion about health practices with
adolescents in their youth organizations. At the various data
collection sites, many parents had questions about the items included
in the HPS; they indicated that it facilitated self-assessment of
their personal health practices, raised questions about the influence
of their behavior on their children, and encouraged them to review
which specific practices were emphasized in their family.

Two additional specific planned activities are: (a) the
communication of information about the alpha reliabilities and sample
characteristics to the authors of the measurement instruments, and
(b) the communication of the use of the General Functioning Subscale
(FAD-7) as a composite family measure with the authors of the
McMaster Family Functioning Device.

[mture Studies

This investigation revealed that the application of multiple
conceptual models to relational data needs to be pursued. It is
possible that different models are appropriate for different types of
data. Therefore, the experimentation with a variety of models and
statistical approaches could narrow the possibilities to a few. The
comparison of the models’ results using different configurations of
family data, i.e., father-mother, father-child, mother-child,
father-mother-child, father-mother-chi1d-chi1d, and child-child, may
also assist in the exploration of congruence.

This study utilized a newly created conceptualization of the
family and individual factors which influence health and health

behavior. The adoption of this framework by other investigators may

 

144

further clarify the proposed conceptualization and the frameworks
from which this model was constructed.

This study assumed that family characteristics were the result
of the family’s social atmosphere and members’ interactions. Future
investigations need to consider the process involved in the
development of the family characteristics in addition to the
assessment of the presence or absence of family characteristics.
This suggests the examination of individual and family factors
related to health and health behavior over time, e.g., longitudinal
or cross-sectional investigations by family developmental levels. In
addition, each of the measurement instruments used in this study
should be utilized in other similar studies so that their
psychometric properties can be further assessed. It was apparent
from the literature review that few psychometrically supported
appropriate instruments were available.

The assumption was made that the information from the
respondents was accurate. This assumption needs confirmation which
may occur through a variety of methods, e.g., observation of the
behaviors by the investigator, reports of the respondent’s practices
by another family member, and/or the use of an individual measure of
social desirability (Nunnally, 1978).

Another important consideration for further research is sample
size and composition. Certainly sample size for the single parent
families was a limiting factor in this study. Perhaps, samples
larger than 60 are essential when conducting research on family
characteristics. The homogeneity of the sample in this study, i.e.,

income, education, ethnicity, may have contributed to the lack of

145

significant findings. This result is primarily attributed to the
sampling approach. Although a stratified random sampling of churches
by census tract data, i.e., level of education and median income, was
conducted in order to locate families with different socioeconomic
characteristics, the families’ characteristics were highly
homogeneous. They did not reflect the characteristics of the
individuals in the census tract data; however, the 1980 census data
are nearly ten years old and likely to be inaccurate. Another
complication of this sampling approach was that many of the small
churches located in the low income and low educational level census
tracts could not be contacted and, when a contact was made, the
clergy declined to participate in the family identification process.
Family identification through other sources, e.g., public schools,
clinics, could yield a more heterogeneous sample of families.

This study represents an initial step in the investigation of
the influence of the family on the health and health behavior of its
members. Although the exploration of the selected family
characteristics related to health and health behavior presented
numerous conceptual and methodological challenges, the results are
encouraging and suggest an array of potential future investigations.
The results of these investigations could provide guidance to
educators, health care and related professionals in their quest to

promote health enhancing practices of individuals and families.

APPENDIX A

146

July 11, 1989

Dear :

During the past several weeks I have been contacting the religious
leaders in the Greater Lansing Area. The purpose of these contacts
has been to secure their assistance in identifying families,
specifically two parent and single parent families who have an oldest
child between 13 and 16 years of age, who would be interested in
participating in a study regarding the influence of the family on the
development of health practices of its members. This letter provides
introductory information about this study. Your review of the
following information and assistance in identifying families in your
congregation would be most appreciated.

During the past decade individuals and families have been encouraged
by their health professionals, the media, and educational programs to
get adequate nutrition, sleep and exercise, to use sound dental
practices, to avoid smoking and drug use, to reduce stress, and to
avoid or drink alcohol in moderation. This is based on the belief
that the regular use of these health practices will enhance personal
health and well being. While the family has been identified as an
important influence on the development of these health practices by
its members, only a few studies have attempted to examine the role of
the family in the adoption of these practices by family members.
Family and health promotion researchers believe that more detailed
information about the family, its characteristics and health
practices, is essential to the identification of more effective
intervention stategies aimed at altering poor health practices and
improving the health of families and their members.

While each family would certainly contribute to our current
understanding, past research suggests a selection of families with
certain characteristics is highly desirable. Therefore, two parent
famiIies and female-headed single parent familes having a 13 to 16
year old who is the oldest child are being sought.

All information given by the family members, the adults and the
adolescent, will be treated with strict confidence. No one,
including myself, will be able to associate the responses with the
individual or family name.

147

Each participating family will be asked to select a common date, time
and location when the participating members can complete individual
paper and pencil questionnaires which can be completed in
approximately one hour. The dates and locations are being scheduled
during July and August, Monday through Thursdays at 7:30 p.m., at
various churches throughout the Lansing-East Lansing area. Such an
arrangement could be made with you so that families could complete
the questionnaire at your church.

The questionnaire is composed of questions in four areas: (a)
personal characteristics, such as age, sex, education, occupation,
(b) family relationships, (c) attitudes toward health and health
care, and (d) health practices, such as those related to nutrition,
sleep, exercise, smoking, stress reduction, dental care, and the use
of alcohol and drugs. No information of a sexual nature is
requested.

A composite of the study’s findings which reflects the information
from all participating families will be available at the family’s
request. A $10.00 contribution will be forwarded to the religious
congregation specified by each participating family. In addition a
composite of the study’s findings would be forwarded to you.

Thank you for reviewing this information and considering this
request. I will contact you by phone in a couple of days to answer
questions you may have and discuss your interest in assisting with
the identification of families for this study.

Sincerely,

gm €1.de

Joan E. Wood

Doctoral Candidate

Department of Family and Child Ecology
College of Human Ecology

Michigan State University

(Ph.) 332-0985

APPENDIX B

148

August 8, 1989

Dear Ms.:

Thank you for consenting to participate in my study about the
influence of the family on the health practices of children.
Enclosed for your review is the "Consent Form" and a list of the
dates and locations scheduled where you can complete the study
questionnaire. Please sign the "Consent Form" and select a
date/location by placing an "x" next to it on the enclosed form.
Return both to me in the stamped addressed envelope provided so that
I will know a couple days in advance of the date selected. I will
contact you prior to the date to confirm your selection. If these
dates are not convenient, please contact me by phone at 332-0985 to
schedule another date.

Thank you again.
Sincerely,

9W 5.. (Jot—4.11;

Joan E. Wood

Doctoral Student

College of Human Ecology
Michigan State University

APPENDIX C

149

Consent Form

We, the undersigned, voluntarily consent to participate in a
scientific and educational study conducted by Joan E. Wood, a doctoral
student in Family and Child Ecology, College of Human Ecology, Michigan
State University. We understand that this study is being conducted
under the guidance of Ms. Wood’s program committee chaired by Dr. Linda
Nelson.

We understand that the main purpose of this study is to examine the
role of the family in the adoption of health practices by its members.
Information about our practices, such as nutrition, sleep, exercise,
smoking, stress, dental care, and the use of alcohol and drugs, will be

sought. -

We understand that each of us will complete a questionnaire which
asks for information about our health practices, attitudes toward health
and health care, family relationships and ourselves, such as age, sex,
and religion.

We understand that a $10.00 donation will be forewarded to the
religious congregation specified below upon completion of our
participation.

We understand that we or any one of us may discontinue our
participation at any time without penalty, are free not to answer
certain questions, and may contact Ms. Wood at 332-0985 if we should
have any questions or concern about the study.

We understand that our responses will not be identified
individually but will be incorporated into a composite of the study’s
findings which will include information from all participating families,
that a copy of this composite will be made available upon our request,
that information given by each of us is not available to other family
members who complete the survey, and that our individual responses will
remain anonymous, so that even Ms. Wood is unable to associate our
responses with our names.

We desire to participate in this study and consent and agree.
We/I, as legal parent(s) of the young adolescent named below, give
our/my permission for him/her to participate in the study.

 

 

 

 

 

 

 

 

Adult Female Signature Date
Adult Male Signature Date
Young Adolescent Signature Date
Address City, Town, State ZIP

 

Religious Congregation

PLEASE MAIL THE Consent Form IN THE STNMPED ENVELOPE DY

 

APPENDIX D

150

List of Dates and Locations

Thursday, June 1, 1989
(7:30 p.m.)

Tuesday, June 6, 1989
(7:30 p.m.)

Wednesday, June 7, 1989
(7:30 p.m.)

Thursday, June 8, 1989
(7:30 p.m.)

Tuesday, June 13, 1989
(7:30 p.m.)

Wednesday, June 14, 1989
(7:30 p.m.)

Monday, June 19, 1989
(7:30 p.m.)

Wednesday, June 21, 1989
(7:30 p.m.)

Monday, June 26, 1989
(7:30 p.m.)

Tuesday, June 27, 1989
(7:30 p.m.)

Wednesday, June 28, 1989
(7:30 p.m.)

Tuesday, July 11, 1989
(7:30 p.m.)

Thursday, July 12, 1989
(7:30 p.m.)

Tuesday, July 18, 1989
(7:30 p.m.)

Wednesday, July 19, 1989
(7:30 p.m.)

Edgewood United Church
469 N. Hagadorn E. Lansing

Pilgrim Congregational Church
Church of Christ
125 8. Pennsylvania Lansing

Edgewood United Church
469 N. Hagadorn E. Lansing

Faith United Methodist Church
4301 S. Waverly Rd. Lansing

Pilgrim Congregational Church
Church of Christ

125 S. Pennsylvania Lansing

Faith United Methodist Church
4301 S. Waverly Rd. Lansing

St. Andrew Orthodox
Catholic Church
1216 Greencrest Ave. E. Lansing

First Church of the Brethren
3020 S. Washington Lansing

Faith United Methodist Church
4301 S. Waverly Rd. Lansing

Grace United Methodist Church
1900 Boston Blvd. Lansing

St. Andrew Orthodox
Catholic Church
1216 Greencrest Ave. E. Lansing

River Terrace Christian
Reformed Church
1509 River Terrace Dr. E. Lansing

North Presbyterian Church
108 W. Grand River Ave. Lansing

River Terrace Christian
Reformed Church
1509 River Terrace Dr. E. Lansing

St. Casimir Catholic Church
815 Sparrow Ave. Lansing

Thursday, July 20,
(7:30 p.m.)

1989

Sunday, July 23, 1989
(7:15 p.m.)

Tuesday, July 25, 1989
(7:30 p.m.)

Thursday, July 27, 1989
(7:30 p.m.)

Sunday, July 30, 1989
(7:15 p.m.)

Wednesday, August 2, 1989
(7:30 p.m.)

Thursday, August 3, 1989
(7:30 p.m.)

Tuesday, August 8, 1989
(7:30 p.m.)

Thursday, August 10, 1989
(7:30 p.m.)

Wednesday, August 16, 1989
(7:30 p.m.)

Thursday, August 17, 1989
(7:30 p.m.)

Sunday, August 20, 1989
(7:15 p.m.)

151

Bethlehem Lutheran Church

549 E. Mt. Hope Ave. Lansing

Holmes Road Church of Christ

321 E. Holmes Rd. Lansing
Maranatha Baptist Church

2300 N. Waverly Lansing
Trinity Lutheran Church

501 W. Saginaw Lansing

Holmes Road Church of Christ
321 E. Holmes Rd. Lansing

Emanuel First Lutheran Church
1001 N. Capitol Lansing

South Lansing Church of Christ
4002 S. Pennsylvania Lansing

River Terrace Christian
Reformed Church
1509 River Terrace Dr. E. Lansing
Unitarian Universalist Church of
Greater Lansing
855 Grove Street E. Lansing
University United Methodist Church
1120 S. Harrison E. Lansing

St. Casimir Catholic Church
815 Sparrow Lansing

Holmes Road Church of Christ
321 E. Holmes Rd. Lansing

APPENDIX E

152

The following statements are about health and health-related issues.
Please circle one number to indicate how much you disagree or agree
with each statement.

Strongly Strongly
Disagree Disagree Undecided Agree Agree
a. For me, it takes more
willpower than I have
to do the things that
I know are good for -
_ my health ................... l ......... 2 ......... 3 ......... 4 ......... 5

b. Most of the thee I
know what to do for
my health without
needing to contact
a doctor .................... 1 ......... 2 ........ 3 ......... 4 ......... 5

c. Only a doctor really
knows whether or not
I am in good health ......... 1 ......... 2 ......... 3 ......... 4 ......... 5

d. Some people think
that a doctor should
decide about what to
do about their
health care, but I
feel that I should
decide ...................... 1 ......... 2 ......... 3 ......... 4 ......... 5

e. I worry about my
health ...................... l ......... 2 ......... 3 ......... 4 ......... 5

f. Whatever a doctor
suggests about my
health is OK for
me to do .................... 1 ......... 2 ......... 3 ......... 4 ......... 5

g. I know, without
someone else telling
me, when I am in
good health ................. 1 ......... 2 ......... 3 ......... 4 ......... 5

h. I more often agree
with what doctors
and nurses think
instead of my own
opinion ..................... l ......... 2 ......... 3 ......... 4 ......... 5

153

Strongly Strongly
Disagree Disagree Undecided Agree Agree
1. I feel good about
how I take care of
my health ................... 1 ......... 2 ......... 3 ......... 4 ......... 5

j. I do things to
help my health
even though a
doctor or nurse
has not suggested
these things to me.. ........ .1 ......... 2 ......... 3.. ....... 4 ......... 5

k. I’m really never
sure that I’m
doing the right
things for my
health until I’ve
checked it out
with a doctor ............... l ......... 2 ......... 3 ......... 4 ......... 5

1. My own ideas about
taking care of my
health are often
better than the
ideas which doctors
and nurses have ............. l ......... 2 ......... 3 ......... 4 ......... 5

m. I don’t do as well
at taking care of
my health as other
people I know ............... l ......... 2 ......... 3 ......... 4 ......... 5

n. I prefer that
doctors and nurses
help me plan my
health practices ............ l ......... 2 ......... 3 ......... 4 ......... 5

o. I know, without a
doctor telling me,
that I’m doing the
right thing for my
health ...................... l ......... 2 ......... 3 ......... 4 ......... 5

p. What a doctor thinks
about my health is
more important than
what I think..... ........... l ......... 2 ......... 3 ......... 4 ......... 5

q. I know what I’m
doing when it comes

to taking care of my
health ...................... 1 ......... 2 ......... 3 ......... 4 ......... 5

APPENDIX E

154

In addition to physical health habits, there are many other things a
person may do to influence his or her health and sense of well-being.
Some of these are specific to certain co-only-expsrienced situations.
Sue are more general.

For each statement about these situations, circle the m response which
most accurately reflects how ogtg you carry out the practice.

[Never

[Barely

M I am ffl with 9 problg:

1. I find out more about the probl- so
Immdl.1tutterOOOOOOCCCOQOOOIOOOOIOII1 2 3 4 5

 

2.Idealwithitassoonaspossible...........1 2 3 4 5
3.Ihendleitonestepatetime...............1 2 3 4 5

4. I think through different ways to
“I with itol0.0.0....OIIOOOIIOIIOOIOOOOOIOII 2 3 4 5

5. I make positive stataemts to myself
Inches:"Icenhandlethis."................l 2 3 4 5

8. I prepare myself by imagining doing
and saying exactly what I would want
todoandsaytohandletheproblcm..........l 2 3 4 5

7.Ilookuponproble-aschallenges...........1 2 3 4 5

8. When faced withaproblem that‘I
cannot solve, I share my feelings about
theproblemwithsomeoneltrust.............l 2 3 4 5

9.Iexpressmyfeelingsfreely.................1 2 3 4 5

10.Iseeksupportfromotherpersons............l 2 3 4 5
When I am depressg o: feeling down:

11. I treat myself to some thing or some
activity which usually makes me feel good... .1 2 3 4 5

12. I participate in strenuous physical
exercise such as jogging, or biking..........1 2 3 4 5

l3.Iexpressmyfeelings........................l 2 3 4 5

 

 

14. I talk over my feelings and my
situation with someone I trust...............l 2 3 .4 5

 

 

 

155

M I a feeling tags or anxious:

15.
18.
17.
18.

19.
20.

21.

 

I consciously allow my uncles to relax... . .

I participate inphysical exercise...........

I take a deep breath and let go of

tb tmion....'.l'...'..DIUUD'OIOIOOOOOOIICl
I find a place to be that is quiet.........1.

I WIY at .IWIYaaaassaaeesessasseel-do

Itakeshowersorbathsorsitina

Mt t” to "lueeaessassasaasasesesssaese-a.

I speak openly about my feelings if

the situation is appropriate.................

I picture myself in a relaxing

sitmtion to help myself relaxmflw.

I take the time to get may or do
something relaxing, such as listening
to music. reading a book. meditating.

praying. playing with a pet, etc.............

m 1 - winning a anlict with someone:

24.

a.

26.

28

I try to resolve the conflict by

tslkingwith that person............ ........ .

I select a time to talk with that

person. when he/she is not too busy..........

I keep to the issue at hand

I direct the conversation to the
behavior and not the personality
characteristics of the person.

(n..- «male, "I want you to do the
dirty dishes that you agreed to do"
rather than "You sure are irresponsible

to let these dirty dishes pile up.").. ...... l

I listen to the person's side of the
story u well as express my View of

Never
Karel
Somet imes
[Usua_l_ly
PALM:

. 1 2 3 4 5
l 2 3 4 5
1 2 3 i1 5
1 2 3 4 5
1 2 3 4 5
l 2 3 4 5
l 2 3 4 5
l 2 3 4 5
l 2 3 4 5

 

th .1tmt1Meeessaaesssseeseaseeaaseeessssns

 

 

 

 

2 3 4 5
2 3 4 5
2 3 4 5
2 3 4 5

156

 

 

 

 

 

 

 

Sometimes
Usually
A1125.

am ri in con ict wit someone:
29. If I a extremely angry, I let off

stenbeforedealingwiththatperson........l 2 3 4 5
.Genegi Self-Care Habits:
30. I keep track of what I have todo

bykeepingacalendar........................l 2 3 4 5
31.Isetrealisticgoalsformyself.............1 2 3 4 5
32.Ifindsuetimeeachdaytobealone........l 3 4 5
33. I express wora of affection to

personsIcareaboutl 2 3 4 5
34. When someone expresses anger towards me.

ItellthemthatIknowtheyareangry.......1 2 3 4 5
35. Iorganizemytime so thatIcan

dowhatIneedtodo............ ..... ........1 2 3 4 5
36. I change my goals if they prove to be

too high. too low, or not good for the

current timel 3 4 5
37. I listen to people without finishing

theirsentencesforthem.....................l 2 3 4 5
38. I express affection physically to

personsIcareflout...”...................1 3 4 5
39. I express my feelings of anger without

hurtingmyselforothers.....................1 3 4 5
40.Itrytolearnfrommymistakes........ ..... .1 2 3 4 5
41. I accept physical expressions of

affectionfromthosewhoeareaboutse.......1 3 4 5
42. When I thi__n_k someone is angry. I

check with the person to see if he/she

rally1.MCCOIODIIIOOOOOODODOOOOIOI003I01 3 4 5
43. I have regular conversations with the

people I live with to solve house-hold

issues,suchaschores,food.money,etc.....l 2 3 4 5

 

157

 

 

 

 

 

Never
' Rarely
Sometimes
Usual
4.14212
Whonnnmd)
“.Icqlimentmyselfwhenldoagoodjob.....l 2 3 4 5
45. I arrange my living environment (for _
equehome. apartment. or room) so
tut it1.PlumtonOOOOOOOOOOOOO:OOOOI.1 2 3 4 5
48. I accept verbal expressions of
affection fromthosewhocareaboutme.....1.l 2 3 4 5
47. When I feel like crying. I allow myself
to cry if the situation is appropriate.......l 2 3 4 5
48. I do something nice for myself at
I“:m.NO...OOOOOOOOOOOOOOOOOOOOO.‘.1 2 3 4 5
49.'IaskforwhatIneedratherthan
waiting for others to meet my needs
withoutmyuking .................. . ....... ..1 2 3 4 5
50. then someone expresses anger toward
me, I respond by trying to understand
what thatperson isangryabout ..... . ...... ..l 2 3 4 5
51.7lhsnIreallywant to refuses
rmt {I'm M. I d0”..............¢ol 2 3 4 5
52. I a comfortable accepting onliments
(“Ml.eeeeeaseeeeeseee eeeee eseee eeeeee del 2 3 4 5
53. The direction I a presently taking
inmylife ismsaningful tome. ....... .......l 2 3 4 5
54.Italkwithsueoneatleastonceaweek
aboutthingsthatareiqortanttome ...... ..l 2 3 4 5
55. I receive hugs from someonewhocares
aboutmeetleastonceaday ........ . ...... 4.1 2 3 4 5
56.Iplantimeeachweektospendwitha
personorpersonsIcareabout...............l 2 3 4 5

 

158

 

  
    

 

     

 

 

 

 

 

 

 

 

Never
Rarely
Sometimes
Usuam
Alwafl

992"]: Eff-Care 391g; (continued)
Inthepatmonth.Mhaveyou

haderelatiVe (suchuaparent,

brother. sister. spouse or in-lmv. etc.)

eto b honeori

57.Whorespectsandvaluesyou?............... .1 2 3 4 5
58.Whousuallyunderstandshowyoufeel?...... .1 2 3 4 5
59."blovuandetym7............. .1 2 3 4 5
60. Nithwhomyou feel comfortable crying? ..... .1 2 3 4 5
In the past month, how often have you

had a friend (not a relative) available

t o b he 0 i on?
61.Whorespectsandvaluesyou?............... .1 2 3 4 5
62. Whousually understanchhowyou feel? ...... .l 2 3 4 5
63.WhoIovesandcaresd>outyou?... ..... ......1 2 3 4 5
64. With when you feel comfortable

Mimi.0.0000........IOOIOOOOCCOOCIOIa.Ia1 2 3 4 5
In the past month. how often have you

usedeachofthe followingas asource

ofsgmrt?
65.Agroupofwhidiyouareamed>er? ......... .1 2 3 4 5
“OA”t?‘IO0.0.0.0000.........IIOOOOIIIICOO-I I1 2 3 4 5
6'7.Natureortheout-of-doors?........... ..... .1 2 3 4 5
68. One or more of the fine arts (for

exuple,micordance)?............ ...... .1 2 3 4 5
69. Spiritual resources (such as talk with rabbL.

priest,pastor,orreadscriptures?........1 2 3 4 5

 

 

 

 

 

 

M
m

159

This section asks you to indicate how frequently you personally engage in
selected health behaviors. For each item place an ”X" in the most
appropriate blank under each behavior to indicate how often you actually
engage in that particular behavior.

1. Discuss health with friends, neighbors, relatives.
Very frequently : : : : : : Never

2. See a doctor for a regular check-up.
Very frequently : : : : : : Never

3. Get enough sleep. 5 —
Very frequently : -: : : : : Never

4. Get enough exercise.
Very frequently : : : : : : Never

5. Avoid parts of the city with a lot of crine.
Very frequently : : : : : : Never

6. Check condition of electrical appliances, the car, etc.
Very frequently : . : : : : Never

7. Do things in moderation.
Very frequently : ' : : : : : Never

8. Use dental floss.
Very frequently : z : : : : Never

9. Avoid over-the-counter (non-prescription) medicines.
Very frequently : : : : z : Never

10. Spend free tine out of doors.
Very frequently : : : : : : Never

11. Take vitamins.
Very frequently : z : : : : Never

12. Avoid contact with doctors when feeling okay.
Very frequently : : : : : : Never

13. Watch my weight.
Very frequently : : : : : : Never

14. Don’ t let things "get me down."
Very frequently : : : : : : Never

15. Avoid getting chilled.
Very frequently : : : : : : Never

16.

17.

18.

19.

20.-

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

160

Don’ t drink alcoholic beverages.
Very frequently '

 

Bat sensibly.
Very frequently

 

Keep eaergency phone numbers near the phone.
Very frequently '

Fix broken things around the home right away.
Very frequently ° :

Avoid overworking.
Very frequently

Destroy old or unused medicines.
Very frequently °

Never

Never

Never

Never

Never

Never

 

Have a first aid kit in bone.
Very frequently

Don’t smoke.
Very frequently

Never

Never

 

Pray or live by the principles of religion.
Very frequently

Never

 

Get enough relaxation.
Very frequently '

Never

 

Limit foods like sugar, coffee, fats, etc.
Very frequently ' :

 

See a dentist for a regular checkup.
Very frequently

Never

Never

 

Avoid parts of the city with a lot of pollution.

Very frequently

 

Near a seat belt when in the car.
Very frequently

Never

Never

Ignore health advice from friends, neighbors, and relatives.

Very frequently

Never

 

APPENDIX H

161

This section is divided into 10 personal health practice area. Each area has a
series of state-ants about related health practices.

lead each statement and circle M next to the stataent that b_eg_t describes
hgg_gf§gn you do the health practice.

-If you 3333; do the health practice, circle the number 1.

-If you [3:311 do the health practice, circle the number 2.

-If you gglgfflgg do the health practice, circle the number 3.

-If you 3:93111 do the health practice, circle the number 4.

-If you 3131; do the health practice, circle the other 5.

 

 

g a 2g 3 2
In 5 a a
. z a: m :3
MW
1. Ihave at leasttwomfgn (if overage18)

or g; 122': four segingg (if age is 13 to 18)
of dairy products even day.

{One serving 8 one 8 ounce glue of milk

1 cup of yogurt

2 cups of cottage cheese

2 one inch square cheese cubes
2 slices of processed cheese)......l....2....3..“4.4.5

2- I hav- W of Int.
poultry, fish, beans or a codination of
then 93:01.43!-

{One serving 8 2 ounces of meat, fish or
poultry
8 1 up of cooked dry beans or

peas
3 1,4 m 0! m“.)..............a...l..r..2..b.3..p.4..”5

3- I hm W of I
codination of whole grain, enriched or
fortified breeds and cereals m.
{One serving : 1 slice of bread

a 1/2 to 3/4 cup of pasta, rice 0
a cooked cereal such as oatmeal
8 4 two inch square crackers}........1....2....3....4....5

 

 

 

 

 

 

 

1(52

lead each statuent and circle the odor next to the statement that M
describes w you do the health practice.

4.

10.

t
I law-Wot cook-401'
rut fruits and vegetables 93331.!!!-
{Oneserving8l/2cupofcookedorrmv
vegetables

8 l/2 cup of canned fruit

8 l/2clyof Juice

8 1 piece of run fruit)...........

I hm W of carrots.
hrk green leafy vegetables, bok choy, sweet
potato, winter squash, apricots, or broccoli

W- .
{One serving8 l/2 up} ...... .

I luv- W of Ire-n m.
broccoli, potato, tuto, spinach, cabbage,
green peas, strmuberries or orange Juice mg
day;
{One serving 8 l/2 cup of vegetables or
str-berries
8 4 «races of orange Juice}.......

I M'- W
of black-eyed peas, pinto beans, chick peas,
asparagus, mustard or collard green, li-
besns, broccoli, g; peanuts.
{One serving 8 U2 cup of vegetfiles

8 1/4 cup peanuts}................

I eat snacks such a cheese, crackers

 

NEVER

RARELY

soun'huas

USU ALLY

WAYS

 

peels

)eels

peelse

lad/or breed, fruit or vegetables .............. . .

I eat candy bars, cookies, doughnuts,

patri“. ’1“. dcauseeeeeeeeseeeeeeeeeees es s

I tria all visible fat from east and
remove the skin and fat from chicken and

turkey before cooking or eating.......... ...... ..

 

00020.

.b.2..

00.0200

L0200

000200

 

Deane

Deane

..3.. . .. .

pesos

.040.

,4” .

.0400

..3.. . .

 

020.

L030.

 

.0400 0

L5

..6

 

 

 

163

Read each statement \and circle the m next to the statement that MS
describes w you do the health practice.

_, a .. m

WM '
11. I eat butter, highly marbled seats (fat seen

as lines throughout the seat). fried foods and

salad dressingssuchasmayonnaise ............. ...l....2....3..
12. I eat foofi that are baked, broiled

stir-fried, boiled, steaed or roasted

”. ”0000000000000000000000000000000 0000000 ‘00100P0200p030.
13. I snack in front of the television........ ..... ...l....2....3..
14. I eat low-fat dairy products such as

2!, 1/23 or ski-ed silk, low-fat

cottage cheese, low-fat cheeses,

low-fat yogurt, ad low-fat ice crea. ......... +...l.r......2r3
15. I eat fish such as tuna, sole, perch,

catfish ”MW. ........ ,..l....2....3..
18. I eat poultry, such as chicken ad

turkey g§_1gg.t_§gigg_3_gggk ................... ...l....2..r.3..
17. I eat smoked fish, smoked cheese, or salt-cured

ha, bacon, or saueage............ ......... ...l. ...2L.3..
18. I eat hot dogs, lmcheon aeats such as

bologna. dill pickles, potato chips, corn

$1”. t” at” or p"t:.l.eeseee 000000 000000‘00100L020000300
19. I drink regular cola, tea, or coffee ....... .......l....2....3..
20- I drink that WW-

{1 ‘1“. 8 8 ma} eeeeeeeeeeeseseeeeessessee1eelseae20¢pe3ee
21. Ieat eggs, hardcheeses, suchacheddar,

“#10-tn eeeeeeeeeeee eeesseeeeeee eeeee ee e4eeleebe2esp03e
22. I eat pa fried and deep fried food, such a

fried chicken, fish, ."rench fries or haaurgers...l....2..,.3..
23. I eat 3t_1ggIg_2_g:_§_plgggg§_lggll_g_g§y......4..l....2....3..
24. Iaddsalt toeyfood atthatable ......... .......l....2....3.

 

USUALLY

ALWAYS

 

 

 

 

”4..

F'4°°

.0‘00

 

be4ee e

 

...5

 

 

16li

Dead each statasnt ad circle the mael- next to the stat-ant that has
describes w you do the health practice.

Winn!

W

Wm

 

ILARUILY

S£fl0©rllufl3

noun!

_ mus

 

I eat pre-sweetaed or sugar coated
«ml. .m or m0000000000000000000000004

I eat m1 (at least fruit or Juice ad
cereal or toast) m.
I read food laels for clues about the sugar.

fat ad edit. or salt contat ..

I wash vegetables ad fruits M95
“tm m00000000000000000000000000000 ..... 004

I wash ay hads m hadling food............

I bnnh my teeth m meals and snacks.........

IbrmhayteethW.... ....... 4

Ihaveadentalcheck-upormyteeth
cleaed W...................

I floss ay teeth m.....................

I use a stinldent, tooth pick, water
pick or dental stialator mm. . . . .

I use a fluoride toothpaste.....................
Iuseasoft tootbruh .......... ...... .4

I rinse ay mouth out with water after eating
ifIaaale to brusheyteeth... ...... .....q

I sleep continuously for m (if age
13 to 18) or m (if over age 18)
in mm

0000000000000000000000000004

sssess1e

 

s4

I have a quiet time for ayself ”$4131....”

0100
00100

0 0 10.0

..l.. .

00100

..2..

000200

szee
pe2ee

. ..2..
....[2

r.2..
..2..

pe2es
bezee
..2..

Deane

 

bs2ss

posse

“3..
y.3..

De3es
..3..

5.3..
'0300
tease

hoses

4.3..

 

00300

V3" .

be‘se e

Is4se

he‘s
a s4e s

,.4.. .

00‘00
D0400

p.4.. .

.L.5

.05

L.4.. .

 

'0400J 0

 

 

165

Read each statamat and circle the other next to the statement that best
describes w you do the health practice.

WW

“0

41. I get a aout the sas time mm”
42. I follow a bedtias routine, such as washing,

brushing teeth and reading, ”1913

mm t0 .l~00000000000000000000000000000000040010000200
43. I follow a was-w routine, such as fishing,

dressing ad having breakfat, WHWIZ
44- 1mm“ quiet

activitiesmgoing to sleeplz
W
46- ! participcto in I We:

2m. and: 8- Jossins. bicyclins. brisk

walking or swi-ing, at my home, office,

ml am000000000000000000000000000000000000010000200

49.
50.
51.

 

NEVER
RARELY

ALUAYS

SOIETIIES

USU ALLY

 

Igotoslesp at “twee-stile“...

I do warm-up exercises, including stretching,
for W dons th- excreta-u

I do non-stop light to aoderate aerobic
exercise. such a Jogging, bicycling, brisk
mum. at out-ins for W

00100002000030000400005

00100002000030000400006

0030000400005

0030000400005

eslessZessusssu0005

m ”0000000000000000000000000000000000001000020000300.0400005
I do cool-down exercises similar to the

urn-up mil-8 for W

doing the non-stop aerobic exercise...............l....2....3....4....5
I walk up stairs rather than use the elevator.....l....2....3....4....5
I naintain good posture when sitting or stadium..l....2....3....4....5

I participate in group or individual sports

 

regularly... .......... .................. ...1....2....3....4....5

 

1656

Read each statement and circle the number next to the stateaent that best
describes hgg_gztgg you do the health practice.

W

52.

53.

57.

580

59.

I follow the suggested safety rules
when doing a sport..........

I wear/use the suggested or required safety
equipment for sport activities, such as a

halaet when riding a bicycle or shin guards
when playing soccer............................
I operate a three wheeler, all terrain vehicle
(ATV), snowmobile, or aotorboat................

I ride with drivers of all terrain
vehicles (ATV), snowsobiles or aotorboats .
who have been using'druge or alcohol ...........

I have reflectors on ay bicycle or use
light-colored or reflective clothing when

"811:1“. Joainl or Midas Mwm

I check recreational vehicles I use, such as
a bicycle, three wheeler or all terrain
vehicles, to sake sake sure they’re safe
and running properly.................

I walk or Jog on a road or street with the
a“ or tnfflc0000000000000000000

I can swia or stay afloat for gt_1gggt_zg
mumm- thatisoverayhead .......

I use the "buddy systea" when swiaming.........

I know the depth of the water and location
of obstacles before entering pools, rivers

M lau00000000000000000000000000000000000000 00

I ride a bicycle during high traffic tiaes.....

I follow traffic rules when riding a bicycle
9; using other recreational vehicles, such as
a three wheeler or all terrain vehicle, a

 

IHEVER

IRAfiUEDY

SIIUETIIUES

[UNUAIJAI

AUJIAYSB

 

 

snot-obile or aotorboat ..... . . ..................

0010

00100

00100

100

0.0200

“..2..

.5.2..

 

5.2..

0000200 0 0

.0300

00300

.0300

5.3..

00300

5.3..

000300

 

D0400

be‘ee

..5

5.5

00400 0

.0300 0 00
00400 0

00400
00400

 

 

..3.. .

 

 

167

Read each statement and circle the maer next to the statement that m
describes Mg you do the health practice.

WM

70.

710
72.

73.

74.

75 .

 

NEVER

RARELY

SOIETINES

USUALLY

ALWAYS

 

I operate a three wheeler, a all terrain
vehicle (ATV), a snot-obile or a aotorboat

while under the influence of alcohol or drugs... .

I check autuotive vehicles I use, such as the
car, truck, aotorcycle, aoped, to sake
sure they’re safe ad running properly........

‘ 1

I hitchhiks 2: pick up a hitchhiker.............

I ride with drivers of cars, trucks, aopeds or
aotorcycles who have been using drugs or
.lml000000000

000000000000000'00000000000000001

I follow traffic rules when driving a car,
truck, aoped, or sotorcycle.........

000000000001

I use seatbelts when riding in 9; driving a
car.

I drive a car, truck, eotorcycle, or soped

when under the influence of drugs or alcohol...... .

I accept a ride only from someone I know..........

I cross streets with the “walk" sin, green

light, adat cross walks .......... ......... ...

I cross between parked cars.

d t t
I practice a fire drill procedure
.t “00000000000000000000000000000 00000000000 W

I review «an amergacy procedures,
such as those for cute, burns, bleeding

0000000000000000000000000000000000000000000t00

 

“WWW .......... ...

00100 0

..l..

0010

...2...

”5.2.. .

 

00300 0

000300
...3..

spszss s as s

000200 0

 

she‘s

.5.2.. e se e es s

 

OI

0005
0005

 

00L05

 

Bead each statement and circle the under next to the state-ant that best-

168

describes w you do the health practice.

 

 

 

 

 

 

 

3 ~ .,
a 3
E 3 3 a S
- z a: m :3
WM
78. Ihaveaergacyphonemaers,such
as police, fire, poison control
center and doctor, posted by the phone............l....2..5.3..5.4....5
77. I answer thephonewith the faily nae,
suchas "ihis is thaJones's residence.".......4..l..,.2..5.3....4. ..5
78. I read the labels of housebld products,
suchascleansers,paintsadpoisons
mtmtu 000000000 000000000000000000000100100.0200'030000400D05
79. I follow the safety instructions when
u1utmm10pmu.......u. sssssss 1ee1esbs2es|s3esbe4e0b05
so. Iput deadly substaces, suchaspainte,
medicines ad drugs, ad cleaners on
ufimlmwu.lwc¢mteeeee eeeeeee (eelseae2eepe3ssbe4eepe5
81. Ikeepssdicinesaddrugsinaares
separate fromthekitchaadfood..... ........ ...l....2. ..3....4....5
82. I give infor-tion about people who are
Mt 0thtomt0lm c0110r0....o.....l....2....3..5.4..5.5
83. I read the directions for using electrical
appliacesanddevices beforeusing tha..........l....2..,.3....4....5
84. I dry ay hands before using electrical
mlim00000000000000000000000000000000000001001000020 ’0300I040 D05
fl. Ichsckfor homehold safety hazards, suchas
M“ or {m lqmmlim “meeea10slssee2e De3es+e4esbss
fl. Icut towards ayaelfwhenuingeknife...........l....2..5.3..L.4. ..5
87. I keep hallwus ad stairs free from
clutter, such on toys or clothing ....... ..............4l2 ..3..5.4..5.5
88. I look for sins of bulging or rounded swelling
of Jarandcmnenbbefore opening them.. ..... ....l....2. ..8....4....5
89. Ikeepthedoorsofsybouseorapartaat
lfldmm‘10“.0000000000000000000 000000 400100£0200L030000400b05

 

169

Read each statesent and circle the mar next to the state-ant that best
describes Mg you do the health practice.

 

 

 

 

 

 

 

 

§ ..
(4 >0
_ g g 3 3 .
an 53 a
2 G 0) :3
90. I discard Jars ad cans which have bulging
ends or rounding swelling, spurting, leaks,
mt“. 0' ”Off. ”assesseseesseeseeeeesss ..loo5o2..5.3..5.4....5
91. Iaabletoget toadhadlegunsinay
“0000000000000000000000000000000000000000000 00100P0200’0300I0400005
92. Ithrowoutdrugsvdiicherepastthedete
0! 0‘4“.“ messsssessasseesssossessoseessss>ss1ssPs2ee|e3eebs4sabs5
93. I store food in airtight containers in the .
"a“.ntorsseass eeeee sseeseassessssaaasssseae eslssIOZOslsaaeps4ssnss
Wm
94. I eeditate or pray when I have a probla, feel
twat--m1” eeeeeeeeee s eeeeeeee eessssees eslesbeZeeleaeess4esss5
%. I try to relax by doing an activity I like.
such 'as playing a eusical instrI-ent, sewing,
playing cards, reading, going to a aovie,
01' “tm “a ssssssss sseeeesesseseeseessssees4eslsebazssbsaseaeQeeass
Q. I eat sore food wha I feel axious or tense... ..l....2....3....4....5
97. I tighten ad relax the Insoles in Iy
body, such as ay toes, feet, legs,
shoulders ad/orneckwhenIfeel stressed ..... ..l....2....3....4....5
98. I take several slow deep breath in
adoutwhenIfeal stressed..... .............. ..l....2..,.3....4....5
99. I talk things over with a person I ca
tmtm I h".Pml-0000000000000000000010010000200P030000400006
100. I think aout different ways to hadle
upmlu000000 0000000000000 00000 0000000000000 100100002000030000400005
101. I exercise when I feel tense or anxiom...........l..5.2....3....4..5.5
102. I take drugs 2: alcohol to forget aout
.90-*1- eeeeeeeeeeeee eseeoeseesee eeeeeeeeeeeee ...l....2..5.3..5.4..bo5

 

 

170

lead each statasnt and circle the other next to the statement that m
describes w you do the health practice.

 

 

scum-ms
USUALLY

NEVER
RARELY

WW

ALWAYS

 

108. I set goals for ayeelf which I ca achieve...”...l..5.2..5.3....4.

104. I ake workale plane to reach my goals...........l..5.2....3....4.. .

105. I coaliant eyself when I go a good Job..........l....2....3..5.4.. .

108. I think aout suicide as a way to deal with

. prObl..bseeseeeeeseeeeeseeseeessssee eeeee ......ol..5.2..p.3....0.. e

107. I think aout how I will respond to a situation
before it occurs, such as accepting or refining
aridsfraapersonwhoisunder the influence
ofdrueoralcohol ...... ....l..5.2....3....4..

108. I do scathing nice for ayself m .

M0000000000000000000000000000000000000‘00100002000030000400 0

109. I pla tia m to spend with people
Icareabout..... .............................. T..l....2..5.3..5.4.. .

W93
110. I drink alcoholic beverages, such as beer,

”100, Of liqmrseesesseee ssssss sees eeeeeeeeeeee desleeesZesIs3eePe4se e

111. I drink mm 2 ounces of liquor, 93;
10 ounces of beer 2; 5 ounces of table wine

m”... ....... ....... . ................ ..l..r.2....3..5.4.. .

112. I ak for inforation about drugs the doctor
tells a to take, such as the

nae, purpose, when ad how to m. it ...... ......l....2..5.3....4.. .

113. I use ”street drugs" such a heshieh, cocaine,

".1 “tarot“: 111ml m00000000000000000100002000030000400 0

114. I sniff glue or solvent, such as airplaine
‘1“. M11". or “11 ”1m0000000 00000 00000000100002000030000400

115. I follow the instructions provided with
mmItmoessseaeseeeeesssese eeeeeeeeeeee esleebeZeees3eeLe4ee e

 

 

 

 

116. Isnoke cigarettes. ......................... ......l....2....3....4...

 

 

171

need each etataeat ad circle the odor next to the state-ant that best

describes w you do the health practice.

 

 

 

 

 

 

 

_ E a 3. a z:
‘ h! g D <
.. 5 m 3
z a: U) _ D
W
117. I ueesmokelese or chewing tobacco. ......... ...l. ..2. ..3....4. ..5
1180 I “k. “H00000000000 00000 000000000000000000001000020 .0300P0400005
119. I use drugs aich the doctor has prescribed
for otheraaers ofu faily ........ ..............l.2. ..3....4....5
MW
120. I apply a sunscreen or sunblocking product
tomyekinwhsnworking or playing outdoors.......l....2..5.3....-4....5
1210 Imum.ft.r min‘ t“ “thru00000000100100002000030000400005
1&0 I”. Mm. or .W ”000000000000 00000 J00100L0z0000300‘0400005
123. I share personal itea with others, such as ay
cod: or bmh, hat or scarf, or cosmetics ....... ..l. 5.2....3....4....5
124. I aar clothing that is right for the
weather, suchaehatorecarftocover
"M in “Id ”thr0000000000000000000000000100100‘0200?030000400005
125. I exaine ey breasts (if a feeale) or ay
tuti¢1u (t! . -1.) Wessssessesdoelsspe2eebe3esee4esos5
128. Ikespacurrentrecordofths
iaaisatione, such as aales, tetaue, ad
”110““. I mm0000000000 00000000000 000000100002000030000400005
127. Ihaveayblood pressure checked”
W ..... . ......................... ..l....2....3....4....5

 

APPFTND I X T

172

This section contains a number of statements about families. Please
read each statement carefully, and decide how well it describes your
own family. You should answer according to how you see your family.

For each statement there are four (4) possible responses:

Strongly Agree

Agree

Disagree

Strongly Disagree
that

family

Check ”Strongly Agree ”if you feel that
the'statement describes your family very
accurately.

Check ”Agree" if you feel that the
state-eat describes your family for the
most part.

Check "Disagree” if you feel that the
statement does not describe your family
for the seat part.

Check "Strongly Disagree” if you feel

the statement does not describe your

at all.

1. Planning family activities is difficalt because we aisunderstaad

each other.

___ Strongly Agree

Disagree

-.. Agree ___ Disagree ___ Strongly

2. In tiaas of criaia we can turn to each otaer for suaaart.

_._ Strongly Agree

Disagree

___ Agree ___ Disagree ___ Strongly

3. We caanot talk to each other about the sadness we feel.

___ Strongly Agree

Disagree

 

___ Agree ___ Disagree ___ Strongly
4. Individuals araaaccepted for what they are.
_ Agree _ Disagree ___ Strongly

Disagree

Strongly Agr

10.

11.

12.

173

. He_avoid disgassing our fears and concerns.

___ Strongly Agree _-_ Agree ___ Disagree
Disagree

. We cap expreaaifeelinga to each other.

___ Strongly Agree ___ Agree _._ Disagree
Disagree '

___ Strongly Agree __; Agree ___ Disagree
Disagree
We feel acgapted for what we‘are.

___ Strongly Agree
Disagree

__ Agree _._ Disagree

. Making deciaiana is avgroblaa for our family.

___ Strongly Agree ___ Agree ___ Disagree
Disagree

___ Strongly

. Theregare lotayof bad feeliaga in the family.

Strongly

___ Strongly

_ Strongly

___ Strongly

We are able to aake decisions about how to solve

___ Strongly Agree ___ Agree ___ Disagree
Disagree

 

We doa’t get alongLEell together.
___ Strongly Agree ___ Agree ___ Disagree
Disagree

We confide in each other.

___ Strongly Agree ___ Agree ___ Disagree
Disagree

probleaay

___ Strongly

___ Strongly

Strongly

APPEND IX J

174

This section asks for information about yourself and your family. Circle the

number next to the word which best describes you 9; insert the information
requested.

1. What is your birthdate?

 

 

month day year

2. What is your sex?

Male .............. l -
- Female ............ 2

3. What is your current marital status? ‘

Married ........... l
Divorced ....... ...2
Widowed ........... 3
Separated ......... 4
Never married ..... 5
Other ............. 6
(Specify)

 

4. What is your ethnic origin or descent?

White not Hispanic....1
Black not Hispanic....2

 

Hispanic... ........... 3

Asian ................. 4

Pacific Islander ...... 5

American Indian ....... 6

Other..... ............ 7
(Specify)

5. Which of the following best describes your religious preference?
Catholic .......... 1
Protestant ........ 2
Jewish ............ 3
Other ............. 4

(Specify)

 

6. How often have you attended religious services or functions during
the past five weeks?

None ................. 1
Once ............. ....2
Twice ................ 3
Three tines.... ...... 4
Four times ........... 5
Five times ........... 6

More than five tines.7

 

7.

Kindergarten..
First grade...
Second grade..
Third grade...
Fourth grade..
Sixth grade...
Seventh grade.
Eighth grade..

8.

10.

11.

12.

175

What is the last year of school you completed?

Tenth grade.... .........
Eleventh grade ..........
Twelth grade ....... .....
First year of college...
Second year of college..
Third year of college...
Fourth year of college..

mqmubwwp

Ninth grade .............. 9

10
ll
12
13
14
15
16

Fifth year of college...17
Sixth year of college...18
Seventh year of college.19

 

Technical school........20
(Specify years
completed)

Other ................... 21
(Specify) '

 

 

What is your current employment status? (Circle all that apply.)-

Self-employed full-time ......... l
Self-employed part-time ......... 2
Employed by other full-time ..... 3
Employed by other part-time ..... 4
Unemployed ...................... 5
Student full-time ............... 6
Student part-time ............ ...7
Other ........................... 8
(Specify)

 

What is your current occupation? Please give your current position (for

example dentist, typist, housewife, or student) and a brief description

of your activities.

 

 

 

 

In general, how would you describe your overall health?

Excellent .......... 1
Very good .......... 2
Good ............... 3
Fair ............... 4
Poor..... .......... 5

Do you have a long-term impairment or health problem?

No ................. 1
Yes...... .......... 2

If yes, please describe briefly:

 

 

 

About how much do you currently weigh without shoes?

 

Number of Pounds

 

176

13. About how tall are you without shoes?

 

 

 

 

 

 

 

 

 

 

 

Feet Inches
14. Including yourself, how many people are living with you at your current
address?
Two ............ 1
Three .......... 2
Four ........... 3
Five ........... 4
Six ............ 5 ” -
Seven .......... 6
Eight .......... 7
Nine or more...8
15. Describe each person by giving their age and their relationship to
you, for example, mother - 34, sister - 5.
Relationship Age
16. Which of the following best describes your family’s income from all
sources before taxes for 1988?
Less than $10,000....1 $55,001 to $60,000...11
$10,001 to $15,000....2 $60,001 to $65,000...12
$15,001 to $20,000....3 $65,001 to $70,000...l3
$20,001 to $25,000....4 $70,001 to $75,000...14
$25,001 to $30,000....5 $75,001 to $80,000...15
$30,001 to $35,000....6 $80,001 to $85,000...16
$35,001 to $40,000....7 $85,001 to $90,000...17
$40,001 to $45,000....8 $90,001 to $95,000...18
$45,001 to $50,000....9 $95,001 to $100,000...19
$50,001 to $55,000...10 More than $100,001...20

APPEND IX K

 

177

Pearson Product Moment Correlation Coefficients on the Major
Variables between the Subsamples of Adolescents, Fathers, and

 

 

 

 

 

 

 

 

 

Mothers.
Pearson Product Moment Correlation Coefficients
Major by Subsamples
Variables AdolescentsG Adolescents# Fathersé
& Fatherse & Mothers# & Mothersé
HSDIt .18 .23 .42
(p=.082) (p=.026) (p=.000)
FAD-7* .20 .31 .42
(p=.059) (p=.004) (p=.000)
SCI* .04 .00 .12
(p=.387) (p=.486) (p=.185)
HPBQ-Part 2* .20 .11 .27
(p=.069) (p=.188) (p=.017)
HPS* .27 .15 .22
(p=.018) (p=.097) (p=.049)
WIS .21 .14 .17
(p=.060) (p=.112) (p=.100)
9 n=60
# n=73
* Pearson coefficients of total scale means
$ Pearson coefficients of the one item responses

 

 

178

Pearson Product Moment Correlation Coefficients of Subscale Means on
the HPS between the Subsamples of Adolescents, Fathers, and Mothers.

 

Pearson Product Moment Correlation Coefficients

 

 

 

 

 

 

 

 

 

§£§ of Subscale Means between Subsamples
Subscales Adolescentsé Adolescents# Fathersé &
& Fatherse & Mothers# & Mothersé
Automotive & .03 .11 .33
Pedestrian (p=.398) (p=.176) (p=.005)
Practices
Dental Practices .34 .29 .44
(p=.004) (p=.007) (p=.000)
Exercise & .06 .18 .45
Fitness (p=.327) (p=.071) (p=.000)
Practices
Food Related .44 .30 .46
Practices (p=.000) (p=.005) (p=.000)
Household & .32 .19 .33
Related (p=.006) (p=.051) (p=.005)
Practices
Personal Care .23 .22 .24
Practices (p=.040) (p=.031) (p=.032)
Recreational .26 .09 —.17
Practices (p=.024) (p=.215) (p=.449)
Sleep & Rest .17 .07 .27
Practices (p=.096) (p=.275) (p=.019)
Stress -.16 .23 .03
Reduction (p=.115) (p=.024) (p=.397)
Practices
Substance Use .09 -.07 .38
Practices (p=.253) (p=.272) (p=.001)
G n-60

 

LIST 01" REFERENCES

 

LIST OF REFERENCES

Abt Associates Inc. (1984). An inventory of health knowledge;
attitudes, and practices for students in grades 4—6, 7-9.
Washington, D. 0.: Centers for Disease Control and Office of
Assistant Secretary for Health under contract #200-81-0600.

Andrews, M. P., Bubolz, M. M., & Paolucci, B. (1980). An ecological

approach to study of the family. Marriage and Family Review, 3(1
& 2), 29—49.

Ball, D., McKenry, P. C., & Price-Bonham, S. (1983). Use of

repeated-measures designs in family research. Journal of Marriage
and the Family, 11, 885-896.

Bausell, R. B. (1985). The prevention index. Evaluation and the Health
Professions, 3, 3-6.

Becker, M. R., & Maiman, L. A. (1983). Models of health-related
behavior. In D. Mechanic (Ed.), Handbook of health, health care;
and the healthgprofessions (pp. 539-568). New York: Free Press.

Belloc, N. 3., & Breslow, L. (1972). Relationship of physical health
status and health practices. Preventive Medicine, 1, 401-415.

Bergner, M. (1978). The sickness impact profile: Conceptual
formulation and methodology for the development of a health

measure. International Journal of Health Services, 6 (3),
393-415.

Berkely, J. L., Israel, I., & Stokes III, J. (1987). Health assessment
in the Framingham offspring study: A research proposal. Journal
of Chronic Disease, 49 (1), 1698-1768.

Berkman, L. F., & Breslow, L. (1983). Health and ways of living: The
Alameda county study. New York: Oxford University Press.

Berman, W. H., & Turk, D. C. (1981). Adaptation to divorde: Problems

and coping strategies. Journal of Marriage and the Family, g§(1),
179-189.

Beutler, I. F., Burr, W. R., Bahr, K. S., & Herrin, D. A. (1989). The
family realm: Theoretical contributions for understanding its
uniqueness. Journal of Marriage and the Family, 51(8), 805-816.

179

 

 

180

Blaxter, M., & Paterson, E. (1982). Mothers and daughters: A three
generational study of health attitudes and behaviors.
London: Heinemann Educational Books Ltd..

 

Boardman, K. (1951). The Cornell medical index and health
questionnaire as a diagnostic tool. Journal of the American
Medical Association, 145, 152-157.

Breslow, L. & Enstrom, J. (1980). Persistence of health habits and
their relationship to mortality. Preventive Medicine, 9, 469-483.

Brown, J. S., & McCreedy, M. (1986). The hale elderly: Health behavior
and its correlates. Research in Nursiagi& Health, g, 317-329.

Brown, L. H., & Kidwell, J. S. (1982). Methodology in family studies:

The other side of caring. Journal of Marriage and the Family, 11,
833-839.

Bruhn, J. G. (1988). Life-style and health behavior. In D. S. Gochman
(Ed.), Health behavior: Emergiagyresearch perspectives (pp.
71-86). New York: Plenum Press.

Bruhn, J. G., & Cordova, F. D. (1977). A developmental approach to
learning wellness behavior, I. Infancy to early adolescence.
Health Values, 1, 246-254.

Bruhn, J. G., & Parcel, G. S. (1982a). Current knowledge about the
health behavior of young children: a conference summary. Health
Education Quarterly, 9(2 & 3), 238-261.

Bruhn, J. G., & Parcel, G. S. (1982b). Preschool health education
program (PREP): An analysis of baseline data. Health Education
Quarterly, g(2&3), 116-129.

Bryan, M. S., & Lowenberg, M. E. (1958). The father’s influence on
young children’s food preferences. Journal of the American
Dietetic Association, 34, 30-34.

Butcher, J. (1983). Socialization of adolescent girls into physical
activity. Adolescence, 18, 753—766.

Byles, D. S. W., Bryne, C., Boyle, M. H., & Offord, D. R. (1988).
Ontario child health study: Reliability and validity of the
General Functioning Subscale of the McMaster Family Assessment
Device. Family Process, 21(3), 97-107.

Campbell, T. L. (1987). Eagilyfs impaet on health: A critical review
and annotated bibliography (Series DN No.6 - National Institute
of Mental Health). Washington, D. C.: Superintendent of
Documents, United States Government Printing Office.

Child Trends, Inc. (1980). Special tabulationaifrompthe 1976 National
Sgrveypof Children. Washington, D. C.: Child Trends, Inc..

 

181

Coburn, D., & Pope, C. R. (1974). Socioeconomic status and preventive
health behavior. Journal of Health & Social Behavior, §, 67-78.

Coletta, N. D. (1983). Stressful lives: The situation of divorced
mothers and their children. Journal of Divorce. l§(3), 19—31.

Cox, C. (1982). The interaction model of client health behavior:

Theoretical prescription for nursing. Advances in Nursing
Science, 4, 41-56.

Cox, C. (1985). The selfedeterminism index. Nursing Research, 34,
177-183.

Cox, C. (1988). Untitled grant proposal in process of being developed
for the U.S. Department of Health and Human Services, Public
Health Service.

Cox, C., Miller, E., & Mull, C. (1987). Motivation: Measurement,

antecedents and correlates. Advances in Nursing?Sciences, 9(4),
1-15.

Cox, C., & Wachs, J. (1985). Motivation: Vehicle for public health
nursing interventions? Public Health Nursing, 2(4), 202-212.

Crawford, C. O. (1971). Health and the familyi- A medical-sociologaeal
analysis. New York: The MacMillan Company.

Deci, E. (1980). The psychology of self-detegginism. Lexington, Mass.:
Lexington Books.

Department of Education, Department of Mental Health, Department of
Public Health, Department of Social Services, Office of Health
and Medical Affairs, Office of Highway Safety Planning, & Office
of Substance Abuse Services (1988). Michigaanodel for
Comprehensive School Health Education (HP 103). Lansing, MI.

Dielman, T. E., Leech, 8., Becker, M. H., Rosenstock, I. M., Horvath,
W. J., & Radius, S. M. (1982). Parental and child health beliefs
and behaviors. Health Edaaation Quarterly, 9(2 & 3), 156-173.

Doherty, W. J., & McCubbin, H. I. (1985a). Families and health care:

An emerging area of theory, research and clinical intervention.
Family Relations. 34, 5-11.

Doherty, W. J., & McCubbin, H. I. (1985b). The family_in health care.
Lake Mills, Iowa: National Council on Families.

Duffy, M. E. (1986). Primary prevention behaviors: the female-headed,
one parent family. Research in Nursing and Health, 9, 115-122.

Duffy, M. E. (1988). Health promotion in the family: Current findings

and directives for nursing research. Journal of Advanced Nursing,
13, 109-117.

 

182

Epstein, N. B., Baldwin, L. M., & Bishop, D. S. (1982). Family
assessment device (Version 3). Providence, R.I.: The Brown
University/Butler Hospital Family Research Program.

Epstein, N. B., Baldwin, L. M., & Bishop, D. S. (1983). The McMaster

family assessment device. Journal of Marital and Family Therapy,
2, 171-180.

Ezell, M. P. (1982). Family member’s perceptions of household
prodaction in relationship toggaality of life. Unpublished
doctoral dissertation, Michigan State University, East Lansing.

Fabrega, H. (1976). A behavioral framework for the study of human
disease. Annals of Internal Medicine, 24, 200-208.

Fisher, L., Kokes, R. F., Ranson, D. C., Phillips, S. L., & Rudd, P.
(1985). Alternative strategies for creating "relational" data.
Family Process, 24(6), 213-224.

Forman, B. D., & Hagan, B. J. (1983). A comparative review of total
family functioning measures. The American Journal of Family
Thera , 22(4), 25-40.

Fuchs, V. R. (1974). Who shall live? Health econogics and choice. New
York: Basic Books.

Gillis, C. (1983). The family as a unit of analysis: Strategies for
the nurse researcher. Advances in Nursing Science, 5, 50—59.

Glass, G. V. & Hopkins, K. D. (1984). Statistical methods in education
and psychology (2nd ed.). Englewood Cliffs, N. J.: Prentice-Hall,
Inc.

Gottlieb, N. H. & Chen, M. S. (1985). Sociocultural correlates of
childhood sporting activities: Their implication for heart
health. Social Science and Medicine, 5, 533-539.

Green, L. (1970). Status identity and preventive health behavior.
Pacific Health Education Reports, 2.

Green, L. W., & Lewis, F. M. (1986). Measurement and evaluation in
health education and health promotion. Palo Alto, CA: Mayfield
Publishing Company.

Hager, R. (1984). Development of the self-care inventory: A
psychosocial health habits assessment tool. Unpublished
manuscript, Grand Valley State College, Michigan.

Harnly, M. E., & Williams, G. 2. (1987). Positive health status
measurement and clinical application. In T. Abelin, Z. J.
Brzezinski, & V. D. L. Carstairs (Eds.), Measurement in health
promption and protection (pp. 555-563). Copenhagen: World Health
Organization Regional Publications.

 

 

183

Harris, D. M., & Guten, S. (1979). Health-protective behavior: An

exploratory study. Journal of Health and Social Behavior, 29(3),
l7-29.

Haydon, D. E. (1987). The family and health fitness. Health Values,
22(2), 36-39.

 

Herman, S. J. (1977). Women, divorce, and suicide. Journal of Divorce,
2(2), 107-117.

Hertzler, A. A. (1988). Food dynamics and children’s food and
nutrition behavior. Home Econgmics Fopum, 2(1), 22-23.

Holman, A. M. (1983). Family assessment. Beverly Hills: Sage.

Horowitz, J. A., & Purdue, B. J. (1977). Single-parent families.
Nursing Clinics of North america, 22(3), 503-511.

Hubbard, P., Muhlenkamp, A. F., & Brown, N. (1984). The relationship

between social support and self-care practices. NursiagiResearch,
22(5), 266-269.

Hunt, S. M., McEwen, J., & McKenna, S. P. ( 1986). Measmringyhealth
status. London: Croom Helm.

Janz, N. K., & Becker. M. H. (1984). The health belief model. Health
Education Qmarterly, 22(1), 1-47.

Kabacoff, K. I., Miller, I. W., Bishop, D. S., Epstein, N. B., &
Keitner, G. I. (in press). A psychometric study of the McMaster
Family Assessment Device in psychiatric, medical, and nonclinical
samples. Journal of Family Psychology2

Kandzari, J. R., Howard, H. R., & Rock, M. S. (1981). The well family:

A developmental approach to assesament. Boston: Little, Brown and
Company.

Kasl, S., & Cobb, S. (1966). Health behavior, illness behavior and

sick role behavior. Archives of Environmental Health, 22,
246-266.

Keith, J., & Hoopfer, L. (1987). Michigan early adolescent survey:
Parent Questionnaire. East Lansing, MI: Michigan State
University.

Kintner, M., Boss, P. G., & Johnson, N. (1981). The relationship
between dysfunctional family environments and family member food
intake. Journal of Marriage and the Family, 2, 633-641.

Kirscht, J. P. (1983). Preventive health behavior: A review of
research and issues. Health Psychology, 2(3), 277-301.

Klein, D. (1984, October). The problem of multiple perceptigma in
family research. Paper presented at the meeting of the National
Council on Family Relations, Milwaukee, Wisconsin.

 

184

Kulbok, P. P. (1983). A concept analysis of preventive health
behavior. In P. L. Chinn (Ed.), Advances in nursingytheory
development. Rockville, MD: Aspen Systems.

Kulbok, P. P. (1985). Social resources, health resources, and
preventive health behavior: Patterns and predictions. Public
Health Nursiag, 2(2), 67-81.

Lairson, D., & Swint, J. A. (1978). A multivariate analysis of the
likelihood and volume of preventive demands in a prepaid group
practice. Medical Care, 22, 730-739.

Langlie, J. (1977). Social networks, health beliefs, and preventive
health behavior. Journal of Healthaand Social Behayior, 22,
244-266.

Langlie, J. (1979). Interrelationships among preventive health

behaviors: A test of competing hypotheses. Emblic Hea2th Reports,
24, 216-225.

Laskey, P. A., & Eichelberger, K. M. (1985). Health related views and

self-care behaviors in young children. Eamily Relation§,, 24,
13-18.

Lansing area congregations together in service: 1987 ~1988 Directory
of chmpches. Royal Oak, MI: Cathedral Publishers.

Larzelere, R. E., & Klein, D. M. (1987). Methodology. In M. B.
Sussman, & S. K. Steinmetz (Eds.) Handbook of marriage and the
family (pp. 125-155). New York: Plenum Press.

Lauer, R. M., Akers, R. L., Massey, J., & Clarke, W. R. (1982).
Evaluation of cigarette smoking among adolescents. Preventive
Medicine, 22, 417-428.

 

Litman, T. J. (1974). The family as a unit in health and medical care:

A social-behavioral view. Social Science and Medicine, 2,
495-519.

Loveland-Cherry, C. J. (1986). Personal health practices in single
parent and two parent families. Family Relationa, 22, 133-139.

Macius, P. (1985). An examination of the predictive validity of the
HDSI. Thesis. College of Nursing, University of Illinois at
Chicago.

McCubbin, H. I., & Thompson, A. I. (1987). Familyiassessment
inventories for research and practice. Madison: University of
Wisconsin.

 

185

McEwen, J., Hunt, 8. M., & McKenna, S. P. (1987). A measure of
perceived health: The Nottingham health profile. In T. Abelin, Z.
J. Brxexinski, & V. D. L. Carstairs (Eds.) Measurement in health
prqmption and protection (pp. 591~603. Copenhagen: World Health
Organization Regional Publications.

McLanahan, S. (1983). Family structure and stress: A longitudinal
comparison of two-parent and female-headed families. Journal of
Marriage and the Family, é, 347-357.

McLanahan, S., & Booth, K. (1989). Mother-only famlies: Problems,

prospects and politics. Journal of Marriage and the Family, 2,
557-580.

Mechanic, D. (1964). The influence of mothers on their children’s
health attitudes and behavior. Pediatrics, 22, 444-453.

Mechanic, D. (1979). The stability of health and illness behavior:
results from a 16-year follow-up. Journal of Health and Social
Behavior, 22, 244-260.

Michigan School Health Association & Michigan Health Council (1983).
Individual wellness profile - level III. Lansing, MI: Michigan
School Health Association.

Michigan State Board of Education (1985). Michigan edaeational
assessment program stmdent assessment booklet -;gpades 4, 7, and
29. MI: Michigan State Board of Education.

 

Miller, E.W., Epstein, N. B., Bishop, D. S., & Keitner, G. I. (1985).
The McMaster family assessment device: Reliability and validity.
Journal of Marital and Family Therapy, 22, 345-356.

Muhlenkamp, A. F., & Brown, N. J. (1983, September). The development
of an instrument to measure health practices. Presented at the
American Nurses Association Council of Nurse Researchers
Conference, Minneapolis, MN.

Mullen, P. D. (1983). Promoting child health: Channels of
socialization. Eamily and Community Health, 2, 52-88.

Nathanson, C. (1977). Sex, illness, and medical care: A review of
data, theory, and method. Social Science and Medicine, 22, 13-25.

National Computer Systems (1986). Innerview health aasessment. Iowa
City, IA: National Computer Systems Information Services.

National Wellness Institute (1983). Testwell: A self-scoring wellness
assessment geestionnaire. Stevens Point, WI: National Wellness
Institute.

Nunnally, J. (1978). Psychqmetric theory (2nd ed.). New York:
McGraw—Hill Book Company.

 

186

Office of Disease Prevention and Health Promotion (ODPHP) Health
Information Center (1987). Healthfinder: Health risk appraisals.
Washington, D. C.: U. S. Government Printing Office.

Palmore, E. (1970). Health practices and illness among the aged. 22a
Gerontologiag, 22(4), 313-316.

Parcel, G. S., Bruhn, J. G., & Murray, M. A. (1984). Preschool health
educational program (PHEP): Analysis of educational and
behavioral outcome. Health Education Qaarter2y, 22(3/4), 149-171.

Philips, J. W., & Kolasa, K. K. (1980). Vegetable preferences of

preschoolers in daycare. Journal of Nutrition Education, 22,
192-196.

Pratt, L. (1971). The relationship of socioeconomic status to health.
American Journal of Public Health, 22, 281-291.

Pratt, L. (1973). Child rearing methods and children’s health
behavior. Journal of Health & Social Behavior, 24(3), 61-69.

Pratt, L. (1976). Family structure and effective health behavior - The
energized family. Boston: Houghton Mifflin Company.

Rodale, R. (Ed.). (1984). The prevention index - a report card on the
nation’s health. Emmaus, Pa.: The Prevention Research Center.

Rogers, S. S. (1984). Community as client - A Multivariate model for
analysis of community and aggregate health risk. Emplic Health
Nursing, 2(4), 210-222.

 

Rosenstock, I. M. (1974). Historical origins of the health belief
model. Healthegducation Monograph, 2, 328.

Ross Laboratories (1982). Boys: 2 - 18 yeama physical growth NCHS
percentiles. Columbus, OH: Abbott Laboratories.

Ross Laboratories (1982). Girls: 2 -18 years physical growth NCHS
percentiles. Columbus, OH: Abbott Laboratories.

Ross, J. G., Pate, R. R., Caspersen, C. J., Damberg, C. L., & Svilar,
M. (1987). Home and community in children’s exercise habits (The
National Children’s and Youth Fitness Study II). Journal of
Physical Education, Reaearch and Development, NovZDec, 85-92.

Rothman, R. (1978). Inequality and statification in the United States.
Englewood Cliffs, N. J.: Prentice-Hall.

Sallis, J. F., & Nader, P. R. (1988). Family determinants of health
behaviors. In D. S. Gochman (Ed.) Health behavior: Emerging
research perspectives (pp. 107-124). New York: Plenum Press.

 

187

Salovey, P., Rudy, T. E., & Turk, D. C. (1987). Preaching and
practicing: The structure and consistency of health-protective
attitudes and behaviors. Health Education Research, 2(3),
195-205.

Schor, E., Starfield, B., Stidley, C., & Hankin, J. (1987). Family
health: Utilization and efforts of family membership. Medical
Care, 22(7), 616-626.

Schumm, W. R., Barnes, H. L., Bollman, S. R., Jurich, A. P., &
Milliken, G. A. (1985). Approaches to the statistical analysis of
family data. Home Economics Research Jounnal, 24(1), 112-122.

Select Panel for the Promotion of Child Health (1981). Better health
for our children: A national strategy - a statistical analysis.
2(#017-002-00142-9). Washington, D. C.: United Stated Government
Printing Office.

 

Smith, M. J. (1980). The social consequences of single parenthood: A
longitudinal perspective. Family Relations, 22, 75-81.

Solleder, M. K. (1986). Evaluation instruments in health education
(4th edition). Reston, VA: Association for the Advancement of
Health Education.

Stanhope, M. & Lancaster, J. (1984). Community health nursing -
Procesapand practice for prampting health. St.Louis: The C. V.
Mosby Company.

Suchman, E. (1975). Preventive health behavior: A model for research

on community health campaigns. Journal of Health and Social
Behavior, 2, 197-209.

Thompson, J. J., & Chin, C-H, Y. (1985). The use of an ecological
framework in graduate research in the College of Haman Ecology:
1975-1984. Unpublished manuscript.

Turk, D. C., Rudy, T. E, & Salovey, P. (1984). Health protection:
Attitudes and behaviors of LPNs, teachers, and college students.
Health Psychology, 2, 189-210.

Uphold, C. R., & Harper, D. C. (1986). Methodological issues in
intergenerational family nursing research. Advances in Nursing
Research, 4, 38—49.

U. S. Bureau of the Census (1980). 1980 census tracts. Lansing - East
LansingprI, State and Metropolitan Statistical Area.

U. S. Department of Health and Human Services, National Center for
Health Statistics (1980). 22ghlights from7the wave I of the
national survey of personal health practices and consequences,
United States (#017-022-00751-7). Washington, D. C.: United
States Government Printing Office.

188

Vogel, P. G. (1987). A plan for evaluating the Michigan model for
comprehensive school health education (Final Rep. Contract
#3433). East Lansing: Michigan State University.

Walker, S. N., Sechrist, K. R. & Pender, N. J. (1987). The
health-promoting lifestyle profile: Development and psychometric
characteristics. Nursing Research, 22(2), 76-81.

Ware, J. E. (1976). Scales for measuring general health perceptions.
Health Servicea Reaearch, Winter, 396-415.

Ware, J. E., Brook, R. H., Davies, A. R., & Lohr, K. N. (1981).
Choosing measures of health status for individuals in the general
population. American Journal of Public Health, 12, 620-625.

Windsor, R. A., Baranowski, T., Clark, N., & Cutter, G. (1984).
Evaluation of health promotion and educational programa. Palo
Alto, CA.: Mayfield Publishing Co.

Wood, J. E. (1989). Health practices survey. Unpublished measurement
instrument.

 

HICHIGRN smr: UNIV. LIBRARIES
\IHIWIIWIIINI”WIN“llll”Will"||1H|||HN||H|
31293007762242