Support network resources of single adolescent mothers and development of their high-risk infants

MISSING IMAGE

Material Information

Title:
Support network resources of single adolescent mothers and development of their high-risk infants
Physical Description:
viii, 131 leaves : ; 28 cm.
Language:
English
Creator:
Stegelin, Dolores Hoffman, 1947-
Publication Date:

Subjects

Subjects / Keywords:
Teenage mothers   ( lcsh )
Unmarried mothers   ( lcsh )
Child development   ( lcsh )
Mother and child   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1983.
Bibliography:
Includes bibliographical references (leaves 126-130).
Statement of Responsibility:
by Dolores Hoffman Stegelin.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000365763
notis - ACA4580
oclc - 09882429
System ID:
AA00003850:00001

Full Text















SUPPORT NETWORK RESOURCES OF SINGLE ADOLESCENT
MOTHERS AND DEVELOPMENT OF THEIR HIGH-RISK INFANTS




BY


DOLORES HOFFMAN STEGELIN


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1983













ACKNOWLEDGMENTS


I wish to express gratitude to the following people who helped

to make this study possible:

To Dr. Athol Packer and Dr. Michael Resnick who made this study

possible by encouraging the use of medical data to study inter-

disciplinary research questions and to Dr. Fonda Eyler for her sug-

gestions in the organization of the study and in the collection of

the data;

To Dr. Linda Lamme for her encouragement and responsiveness to

my needs as a doctoral student;

To Dr. Steve Olejnik, Dr. Pat Ashton, and Dr. Suzanne Krogh

for their constructive feedback during the organizational stages and

the writing process;

To my husband, Forrest, and my two children, Stephen and Amber,

for their patience and understanding during three years of graduate

work;

To my fellow graduate students, Tess, Tish, Dwight, Beth, and

Bruce, for their support and encouragement.













TABLE OF CONTENTS

PAGE
ACKNOWLEDGMENTS. . . .. ...... ii

LIST OF TABLES . .. . v

ABSTRACT . . .. .. vii

CHAPTER

I INTRODUCTION. . .. .. 1

Overview of Adolescent Pregnancy .. 1
Questions Raised by the Research . 5
The Purpose of the Study . 6
Hypotheses to be Studied . 6
Definition of Terms. . 7
Limitations of the Study . 8

II REVIEW OF THE LITERATURE. . .. 11

Adolescent Pregnancy and Outcomes. .. 11 .
The Theoretical Basis for Social Support Networks. 16
Research Related to Social Support Networks. 26
Conceptual Framework for Research on Networking
and Child Development. ............. 34
Research Related to Child Development and
Networking . .... .35
Summary. . . ... .. 44

III METHODOLOGY . .. .. .46

Design of the Study. . ... 46
Variables in this Study. . ... 47
Hypotheses to be Tested. .. .. 50
Description of Subjects. . ... 50
Instrumentation. . ... 53
Data Collection. .. ....... 57
Data Analysis--Regression Analysis ........ 58
Regression Models for this Study ... 61
Analysis Procedure . ... 62
Limitations of Methodology . .... .63
Summary . . .. 64








IV RESULTS . .


Preliminary Anaysis. . .... 65
Infant Scores. . . 66
Hospital Visits . . 67
Availability of Support Resources. .... ... 68
Actual Support Resources . .... .68
Organizations. . . .. 72
Perceptions of Support . ..... .73
Secondary Independent Variables. . 74
Regression Analysis. . .. 81
Hypotheses Tested. . ... .86
Stepwise Regression Analyses .. . 92

V DISCUSSION, SUMMARY, AND CONCLUSIONS. .. 96

Discussion of Results. . ... 96
Regression Analysis. . ... 108
Stepwise Regression Analyses . .. 113
Summary. . . 115
Findings As Related to the Literature. .. 116
Recommendations for Future Research ... 119
Implications for Parent Education. .. 121
Conclusions. . . ... 122

APPENDIX QUESTIONNAIRE. . . 123

REFERENCES . . .. . 126

BIOGRAPHICAL SKETCH.. . . .. .. 131













LIST OF TABLES


TABLE PAGE

1 Out-of-wedlock births per 1,000 women 15 to 19 years
of age, 1950-1978. . . ... 12

2 Independent, dependent and control variables in this
study. . . ... ...... .49

3 Birthweights of infants born to single adolescent
mothers . . .. ... 51

4 Ages of single adolescent mother subjects. .. 53

5 Years of education completed by mothers. .. 52

6 Race of single adolescent mother subjects. ... 53

7 Distribution of the Bayley MDI and PDI scores of
infants at 12 months of age. . .. 66

8 Number of visits made by infant's immediate family
during infant's hospitalization . .. 67

9 Support resources available to the adolescent mother 69

10 Actual measures of support resources: Relatives in
the area. . . 70

11 Number of relatives and/or friends mother can count
on in times of crisis. . ... 71

12 Does mother get together with relatives/friends often. 71

13 Times per week mother gets together with friends/
relatives. . ... ..... .72

14 Types of organizations mothers participated in 73

15 Mother's perception of her support network ...... 75

16 Times per month to get away by yourself. . 76

17 Is this enough time to get away by yourself. 76







18 Mother's perception of her own health. ... 77

19 Mother's perception of her physical well-being .... 78

20 Mother's perception of her own body. . .. 78

21 Mother's perception of her own weight. .. .. 79

22 Mother's perception of her life in general ...... 80

23 Stress scores of adolescent mothers. . .. 81

24 Correlational matrix . .. 84

25 Correlations of control and dependent variables ... 85

26 Correlation coefficients of four control variables in
quadratic form . . ... 86

27 Linear regression of MDI and PDI by number of visits
by support network family members during the
infant's hospitalization . .... .88

28 Linear regression of MDI and PDI by measures of the
mother's accessibility to her support network
resources. . ... ..... .89

29 Linear regression of MDI and PDI by measures of the
adolescent mother's actual support network resources 89

30 Linear regression of MDI and PDI by measures of
adolescent mother's perception of her support network. 91

31 Stepwise regression models for MDI--Max R procedure. .. 93

32 Stepwise regression models for PDI--Max R procedure. 94













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

SUPPORT NETWORK RESOURCES OF SINGLE ADOLESCENT
MOTHERS AND DEVELOPMENT OF THEIR HIGH-RISK INFANTS

By
Dolores Hoffman Stegelin

April 1983
Chairman: Linda L. Lamme
Cochairman: Athol Packer
Major Department: Curriculum and Instruction

The purpose of this study was to examine the association between

measures of support network resources of single adolescent mothers

and the Bayley scores of Mental and Psychomotor development attained

by their infants at 12 months of age.

The theoretical framework for this research originated from

Bronfenbrenner's theory of ecological human development. Bronfen-

brenner, along with recent theorists Cochran and Brassard, in 1979

proposed that social support might be associated with qualitative

child development. This research was an exploratory study of this

proposed relationship.

Four hypotheses were tested, utilizing data from the Children's

Developmental Clinic at Shands Teaching Hospital, University of

Florida. Subjects were 35 mother-infant dyads. The factors of

marital status, socioeconomic status, infant's birth other, and

neonatal condition of the infant were controlled for. All infants







were classified as high-risk due to low birthweights. Measures of

social support were obtained through a sociodemographic questionnaire

completed by the mothers. The infants' scores on the Bayley Scales of

Infant Development, Cognitive and Psychomotor Index, were used as

the dependent variables in the study.

Four measures of social support of the single adolescent mothers

were studied in relationship to scores attained by their infants on

the Bayley Scales of Mental and Psychomotor Development. Measures of

maternal social support were not associated significantly with measures

of infant development. Measures of the mother's perception of her

social support were associated significantly with the infant's scores

on the psychomotor index.

In summary, this exploratory study showed little association

between measures of social support of single adolescent mothers and

measures of their infants' cognitive and psychomotor development at

12 months of age. The results suggested there is an association

between the mother's perception of her support and the psychomotor

development of her infant. These results indicated a need for further

research as related to factors that influence the mother's perception

of her social support.


viii












CHAPTER I
INTRODUCTION

Overview of Adolescent Pregnancy

Considerable impetus on the study of the family network as a

factor in child development is encouraged by the research of

Bronfenbrenner (1974, 1979). More recently, Powell (1979) and

Cochran and Brassard (1979) have proposed theoretical frameworks for

networking as an important contributor to the qualitative development

in young children. This study addresses the relationship between

measures of support network resources of single adolescent mothers

and the qualitative development of their infants at 12 months of age,

as measured by the Bayley Scales of Mental and Psychomotor Development.

The occurrence of adolescent pregnancy and consequent parent-

hood has risen dramatically in the past decade (Guttmacher, 1976,

1981). The proportion of births to adolescent mothers as compared to

adult mothers is currently a major national concern (Scott, 1981).

While births to adult women continue to decline, the rate of adolescent

pregnancy and childbirth continues to remain at a high level. In the

case of very young mothers (10 to 14 years of age) the rate of

adolescent pregnancy has increased noticeably in the past decade.

Compared to white females, black females maintained a high rate of

adolescent pregnancy in the past decade while the rate of adolescent

pregnancy for white females doubled during the 1970's (Guttmacher,







1981). Clearly, adolescent pregnancy presents a complex and chal-

lenging research topic.

The consequences of adolescent pregnancy are far-reaching. Not

only does the young female experience accelerated role transition

(Bacon, 1974), but the consequences for the infant born to a very

young mother are of major concern to medical researchers. Numerous

studies document that pregnant adolescent females drop out of school,

have higher divorce rates, lower earning potential, higher rates of

child abuse, and, in general, more limited and long-term goal setting

opportunities (Furstenberg, 1976; Guttmacher, 1976, 1981).

The outcomes for the infants born to these very young mothers

are also well-documented. Adolescent mothers have higher incidences

of premature delivery, increased neonatal complications, and higher

infant mortality rates (Field, 1981; Monkus, 1981; Oppel and Royston,

1971; and Quay, 1981). This is especially true for the younger

adolescent mother, from 10 to 14 years of age (Field, 1981). In-

creasingly, researchers are focusing on the medical, social, psycho-

logical, and physical consequences of adolescent pregnancy. The

adolescent mother population presents many unanswered research

questions and problems. This study addresses this unique population

and the possible relationship between support networks of adolescent

mothers and the qualitative development of their infants.

Keniston (1977) indicates that the perception that adequate

families are independent, autonomous, self-sufficient units that

insulate their members from external pressures has been challenged.

The concept of the extended family network is hardly a new one,

especially to sociologists. However, the qualitative and quantitative







investigation of the social support system or network as it relates

to the development of young children who transact within the framework

of that support network is a new and fertile area for research.

Bronfenbrenner's (1974) ecological theory and model of child

development characterizes the family as a system, influenced by

various neighborhood groups, community agencies, voluntary associa-

tions, government offices and policies, as well as by broad cultural,

ideological, and political systems. While the concept of support

networks is utilized as a part of the system's theory, relatively few

investigators have linked systems of support (or networking) to the

family's qualitative growth and development, especially in relation to

the specific measures of infant cognitive, social, and physical

development.

According to Powell (1979), social networks may enhance a

family's childrearing processes by providing material, psychological,

and emotional support and by sharing information and reinforcing

accepted beliefs and behaviors that socialize adults into the parent-

ing role. Cochran and Brassard (1979) argue that the members of the

parents' social support network influence the child's development

indirectly by their more direct influence on the parent. Transactions

and interchanges that occur frequently between members of the same

network are more likely to enhance parent functioning and children's

development.

Cannon-Bonventre and Kahn (1979) found that teenage parents rely

on informal social networks to meet a range of material needs, includ-

ing money, housing, furniture, food, clothing, and medical assistance.

Similar studies on the use of support network systems by young







parents have been conducted by Epstein (1980) and Zitner and Miller

(1980). Stolz (1967) found that parents used a variety of resources

for child development information, though the principal sources were

interpersonal in nature. Epstein (1980) found that teenage mothers

who are coping well with parenting roles attribute their success to

someone who taught them to "negotiate the system" of available in-

formation and support systems. Egeland, Breitenbucher, and Deinard

(1980) found that low-income teenage mothers rated as giving excellent

care to their infants obtained most of their child development infor-

mation from relatives and their own childhood experiences. In a

nationwide survey of parents, including high-risk and teenage mothers,

Sparling, Lowman, Lewis, and Bartel (1978) found that there was a

strong preference of parents for interpersonal sources of information,

followed by print and audiovisual media sources. McKinlay (1973)

reported that networks appear to exert restrictive as well as enabling

influences, depending upon their characteristics.

The literature has supported the idea that parents do not rear

children in a social vacuum. Social relationships may be particularly

critical in the learning of parenting or acquisition of parenting

skills across the individual's lifespan. However, there have been

little data to support this speculation. Connectedness to others has

seemed to be associated with healthy family functioning. Parents

who have exhibited more problematic patterns of interaction, such as

abuse, have been more likely to be socially isolated. The study of

support networks utilized by parents has furthered our understanding

of forces that enhance parenting skills.





5

Questions Raised by the Research


The research literature supports the notion that support networks

are utilized by parents. Since the adolescent mother represents a

unique population, there is a need to study ways in which social sup-

port networks affect the lives of adolescent mothers and their off-

spring. Specifically, there is a need to investigate possible rela-

tionships between the nature and utilization of the adolescent

mother's social support network and quantitative measures of her

infant's cognitive and psychomotor development. Several research

questions are posed.

1) What types of social support resources constitute the

adolescent mother's support network?

2) With what frequency does the adolescent mother interact

with and rely upon her support network?

3) What role does the social support network play in times of

crisis for the adolescent mother and her offspring?

4) How do adolescent mothers perceive their social support

networks?

5) How accessible are adolescent mothers to their support

networks?

6) What social, religious, educational, and political organi-

zation play a role in the adolescent mother's support network?

7) Are social networks of adolescent mothers related to the

quality of their infants' development?







The Purpose of the Study

The purpose of this study was to investigate 1) the types of

support resources that make up the adolescent mother's support network

and the frequency with which she utilizes the support resources,

2) the correlation of four specific indicators of support in relation

to her infant's mental and psychomotor development scores on the

Bayley Scales of Infant Development at 12 months of age,and 3)measures

a self-concept and life stress of the adolescent mothers.


Hypotheses to be Studied

Hypothesis I. There will be no significant linear relationship

between measures of in-hospital communication efforts by members of

the support network and the Bayley scores of Mental and Psychomotor

Development of the infants at 12 months of age.

Hypothesis II. There will be no significant linear relationship

between measures of the adolescent mother's accessibility to her

support network resources and the Bayley scores of Mental and Psycho-

motor Development of the infants at 12 months of age.

Hypothesis III. There will be no significant linear relation-

ship between measures of actual support network resources of the

adolescent mother and the Bayley scores of Mental and Psychomotor

Development of the infants at 12 months of age.

Hypothesis IV. There will be no significant linear relationship

between measures of the mother's perception of her support network

and the Bayley scores of Mental and Psychomotor Development of the

infants at 12 months of age.








Definition of Terms

Single Adolescent Mother--an adolescent mother, under the age of

20, who has never been married or is separated or divorced. None of

the subjects are married at the time of the data collection and all

are lower socioeconomic status subjects.

High-Risk Infant--an infant who was placed in the Neonatal

Intensive Care Unit (NICU) at birth, due to neonatal low birthweight

or prematurity. All infants in this study were classified as "at

risk" at birth. Babies ranged in weight from 900 to 2240 grams.

Mental Development Index score (MDI)--referred to as the MDI,

this is a score obtained from the administration of the Bayley

Scales of Infant Development, Mental Development Index. The MDI scores

obtained for this study were taken when the infants were 12 months of

age. All testing was conducted at the Children's Developmental Clinic,

Shands Teaching Hospital, University of Florida, between January 1,

1979, and December 30, 1980.

Psychomotor Development Index score (PDI)--a score obtained

from administration of the Bayley Scales of Infant Development,

Psychomotor Development Index. The PDI scores obtained for this

study were taken when the infants were 12 months of age. All testing

was conducted at the Children's Developmental Clinic, Shands Teach-

ing Hospital, University of Florida, between January 1, 1979, and

December 30, 1980.

Control Variables--from previous neonatal research on infants

born to very young mothers, four variables will be used as control

variables in the initial analysis. The four variables are infant's








birthweight, number of days infant spent in the hospital after his/her

birth, mother's age, and mother's level of education. If these four

control variables are significantly correlated to the two dependent

variables--MDI and PDI scores--they will be used as controls in order

to reduce the error variance when investigating the independent

variables--measures of support network resources--in relation to the

two dependent variables, MDI and PDI scores for the infants taken at

12 months of age.


Limitations of the Study

This study includes the following limitations:

Selection of sample. The sample was selected through ex post

facto medical data. All adolescent mother subjects were single, lower

SES, and 19 years old or younger. The infant subjects were all first

born. All infants were born between January 1, 1979, and December 30,

1980, and utilized the NICU at Shands Teaching Hospital. The mother-

infant dyads were selected from a total sample of 139 dyads, based

on completeness of medical data, which may have biased the data.

However, the reasons for incompleteness of the data varied widely

from one dyad to another.

A local population. The sample was local in nature. These

mother-infant dyads were representative of a southern lower SES

population. Therefore, generalizability of the findings to a wider

or national population is not possible.

Data collection procedures. The data were collected by highly

trained medical personnel. While strict routines were followed in

this hospital setting, some variation in the collection of data must







be acknowledged, due to individual differences among hospital staff.

Also, the data on the adolescent mothers were collected by means of

a questionnaire. Human factors impacted this data collection process,

creating additional limitations to the study. Mothers provided

self-reported data, sometimes under conditions of fatigue and environ-

mental interference. Self-reported data always have limitations.

The interpretation of each question may vary by the subjects. In

this particular population--lower SES and adolescent in age--the

ability to read and comprehend the questions as intended by the

researcher may have been limited.

Incomplete medical data. A final limitation to this study was

the occurrence of some incomplete data. Again, what factors caused

this are difficult to ascertain. Since many of these mothers com-

pleted the questionnaire after several hours of waiting, the factor

of fatigue must be considered. No pressure was put on the subjects

to complete the data although hospital staff in the Children's Devel-

opmental Clinic were always present and available for clarification

of the questions. This is a common limitation to the use of medical

data since many factors intervene in the data collection process.

Limits of correlation studies. This study was correlational.

An attempt was made to investigate possible associations between

the measures of social support of adolescent mothers and the mental

and psychomotor development of their infants. Cause and effect

relationships are not implied in this study. The researcher is at-

tempting to ascertain possible associations between a specific kind of

variable, measures of social support, and measures of infant develop-

ment. Other variables which may contribute to the infant's mental





10

and psychomotor development are acknowledged by the researcher. Given

the newness of the theoretical framework for this study, the researcher

chose to conduct an exploratory study, utilizing appropriate correla-

tional procedures.













CHAPTER II
REVIEW OF THE LITERATURE

The concept of social support systems or networking has developed

primarily through the efforts of investigators in the field of

sociology. The role of social support in the qualitative develop-

ment of young children has been identified and studied only in the

past decade. More specifically, the relationship between social

support system and specific, measurable indices of cognitive, social,

and physical development of young children has been proposed only in

recent years.

This chapter reviews the literature on the following topics:

adolescent pregnancy and outcomes for the neonate and the young

mother; the theoretical basis for social support systems and networks;

and specific research that relates to the relationship of social

support systems to such dependent factors as mental health, child

abuse, and child development.


Adolescent Pregnancy and Outcomes


The proportion of births to adolescent mothers in the United

States has increased dramatically since the 1960's. The Bureau of

the United States census in 1980 summarized the occurrence of births

to 15 to 19 year olds since 1950. In 1950, 12.6 infants were born to

every 1,000 15 to 19 year olds, out-of-wedlock. By 1960, this figure

had increased to 15.3. An escalation of births born to adolescent







mothers occurred between 1960 and 1970. By 1970, 22.4 infants were

born to every 1,000 15 to 19 year olds out-of-wedlock. The increase

continued but did level off. By 1978, the latest year for which data

were available, 25.4 births occurred out-of-wedlock to young women

between the ages of 15 to 19 years of age. This information is sum-

marized in the table below.


Table 1. Out-of-wedlock births per 1,000 women 15 to 19 years of
age, 1950-1978.*


Year 15 to 19 20 to 24
years old years old


1978 25.4

1975 24.2

1970 22.4

1960 15.3

1950 12.6

*Bureau of the Census, January 1982, p. 7


36.1

31.6

38.4

39.7

21.3


The ratio of the number of out-of-wedlock births per 1,000 live

births has more than tripled since 1950 for women 15 to 19 years old

and for women 20 to 24 years old. By 1978, the latest year data are

available, 44 percent of all births to women 15 to 19 years old and

16 percent of all births to women 20 to 24 years old were classified

as occurring out-of-wedlock. The proportions are much higher for

blacks than for white women; in 1978, approximately 8 out of every

10 births to blacks 15 to 19 years old were born out-of-wedlock,







compared with 3 out of 10 for white women this age. These differences

between the two races can be traced to different patterns of contra-

ceptive effectiveness and marriage (Bureau of the U.S. Census, 1982,

p. 8).

The absolute number of live births to teenagers reached a maximum

in 1972 and 1973 but has since shown some decline (Scott, 1981).

Lincoln, Jaffe, and Ambrose (1976), writing in the influential

Guttmacher report, used the term "epidemic" to describe the occurrence

of adolescent pregnancy; articles in newspapers picked up the term,

thus reinforcing the general notion that the rate of adolescent

pregnancy had escalated rapidly. Currently, over one million female

adolescents give birth annually (Guttmacher, 1981).

Zelnick and Kantner (1979), in a national probability sample of

young women, found that there have been changes in sexual activity as

indexed by mean age of first coitus. Between 1971 and 1976 the median

age of first intercourse shifted among white females from 18.9 years

to 18.1 years, and for black females from 17.1 years to 16.6 years.

Moore, Hofferth, Wertheimer, Waite, and Caldwell (1981) reported that

during the past decade, the United States witnessed a dramatic decline

in fertility rates. Despite this overall trend, fertility rates have

not fallen as rapidly among teenagers as among older women. Thus

the relatively slow decline in fertility rates among teenagers rela-

tive to older women, combined with large teenage cohorts, has resulted

in another phenomenon--the proportion of all babies that are born to

teenage mothers has risen significantly.

In 1950, females under the age of 20 bore 12 percent of all

children and 20 percent of all first children. In 1978, they bore







17 percent of all children and 31 percent of all first children

(Moore, Hofferth, Wertheimer, Waite, & Caldwell, 1981). An accurate

assessment of occurrence of adolescent pregnancy might be summarized

in the following way: the actual rate of adolescent pregnancy has

not increased dramatically as some sources have indicated. Instead,

the rate of adolescent pregnancy has been sustained in a relatively

stable fashion. However, when the population as a whole has been

studied, the birth rate has definitely declined. In light of the

slowed rate of childbirth among adult women in the United States,

the rate of adolescent pregnancy must be viewed with concern. In

addition, outcomes for adolescent mothers and their infants are less

positive than for adult mothers and their offspring, as described in

the following section.


Outcomes of Adolescent Pregnancy

The adverse personal and social consequences of teenage pregnancy,

particularly in the earlier (13 to 15) years, have been well documented.

Included have been school dropout (Bacon, 1974; Furstenberg, 1976),

marital problems and divorce (Furstenberg, 1976), less psychologically

competent offspring (Furstenberg, 1976; Oppel and Royston, 1971; Field,

1981), and higher rates of abusiveness (DeLissovoy, 1973).

Early childbearing and lower educational attainment have been

documented in the research (Furstenberg, 1976; Moore and Waite,

1977; Guttmacher, 1976, 1981). The importance of schooling to other

life outcomes has been found repeatedly in the research. Income,

occupation, fertility, sex role orientation, unemployment, and even

the probability of divorce are affected by education (Moore et-al., 1981).







Scott (1981) indicated that when a wide range of medical, social, and

maternal criteria are applied to the prediction of intelligence levels

in childhood, maternal education appears as a major predictor of out-

come. Thus the consequences of early childbearing for the young

mother have been multifaceted--educational, social, psychological, and

heal th-related.

Neonatal Outcomes


Studies of the neonatal outcomes of infants born to teenage

mothers suggest that these infants are a high-risk group (Field,

1981; Holstrum, 1979; Hofheimer, 1979). The teenage population is

characterized by low socioeconomic status, poor nutrition (not only i

for financial reasons but also due to cultural idiosyncracies), in-

creased illegitimacy, and lack of early prenatal care, as well as a

higher rate of pregnancy and delivery complications (Monkus, 1981).

Field (1981) reported that previous and current research suggest that

the teenage mother and her offspring are at risk primarily due to

social, educational, and economic factors. Quay (1981) and Oppel

and Royston (1971) report that teenage mothers are more likely to

give birth to less psychologically competent offspring. 6

In summary, infants born to teenage mothers are at a greater

risk of being lower birth weight than infants born to mature women.

The increased risk is largely due to the frequent association between

teenage pregnancy and other factors, such as poor socioeconomic status,

poor nutrition, or lack of prenatal care (Monkus, 1981).

White (1979), writing in The Origins of Human Competence,
states







After 20 years of research on the origins of human
competence, we are convinced that much that shapes
the final human product takes place during the
first years of life. We are also convinced that
the traditional failure of society to offer train-
ing and assistance to new parents has several
harmful consequences put simply, people
could grow up to be more able and more secure if
their first teachers did not have to be "self-
taught and unsupported." (p. 183)


With this in mind, the concept of social support systems or networks

is now presented.


The Theoretical Basis for Social Support Networks

The concept of "social support network" or system is defined by

several investigators and theorists.

Sociologists have studied the family network or system extensively.

Buckley (1967) described a system's most fundamental property as

being the interdependence of parts or variables. Adams (1970), an-

other family sociologist, indicated that proximity, not separation,

is the rule for American families, with geographical isolation from

kin being characteristic of only a small portion of the population.

Adams (1970) also stated that it is the "professional and managerial

families of the upper-middle class who are most likely to be separated

from their kin" (p. 578). According to Sussman and Burchinal (1962)

and Adams (1970), the working classes are more kin oriented. In

terms of daily living, then, kin seem more salient to working class

people.

Keniston (1977) described the Carnegie Council on Children in

1972 as having addressed a central issue--is child development itself

heavily influenced by its social context? (p. 1). Keniston (1977)







went on to say that "until policy makers and planners shift their

focus to the broad ecological pressures on children and their parents,

our public policies will be unable to do much more then help indi-

viduals repair damage that the environment is constantly reinflict-

ing" (p. 2). Sussman and Burchinal (1962) agreed with Keniston's

assessment of family networks and challenged the concept of the

healthy family being autonomous and independent.

Several current theorists in human development support the re-

search of such sociologists as Adams, Buckley, Sussman, and Keniston.

Among human development researchers at the forefront are Bronfenbrenner,

Powell, Cochran, and Brassard. Bronfenbrenner (1974) presents an

"ecological model" of human development which states


At the heart of the system is the human organism,
which through its biology and "personality" establishes
many of the parameters of human behavior. Surrounding
and containing the organism is the immediate setting
which may be defined in terms of three dimensions:
design of physical space and materials, the roles and
relationships of other people to the organism, and
human activities surrounding the organism.
(Bronfenbrenner, 1974, p. 4)

Bronfenbrenner's ecological model provides a mechanism for analyzing

the development of the child and the personal social networks of

immediate family members (Cochran and Brassard, 1979). Families are

embedded in networks of relatives, neighbors, and friends. Those

network members influence the rearing of children, sometimes directly
and often indirectly. Yet such social influences have gone virtually

unrecognized by those studying child development (Powell, 1979).

Cochran and Brassard (1979) define a network as "those people

outside the household who engage in activities and exchange of an







affective and/or material nature with the members of the immediate

family. Membership is confined to persons outside the household

and usually consists of other kin, neighbors, friends, schoolmates,

and workmates" (Cochran and Brassard, 1979, p. 602).

Cochran and Brassard (1979) used the framework of network

analysis to assess the social ecology of the parent and child in

relation to its possible effects upon child development. They state

that families have always been embedded in networks of relatives,

neighbors, and friends. Those network members have undoubtedly in-

fluenced the rearing of children, sometimes directly and often

indirectly.

Cochran and Brassard (1979) made a strong theoretical case for

the impact of the support system on early child development. They

stated that there are four ways in which the parent's social network

has influenced the child:

1) Co-workers can compete with the child for the parent's time.

2) The social network can link the parent to new life possi-

bilities.

3) The social network can influence the child's development

through the parental role itself.

4) Provision of the opportunity for active participation in

the networks of his or her parents can lay the foundation for the

development of the child's own personal social network or "network

building skills" (Cochran and Brassard, 1979, p. 4).

In reference to three and four above, the following theoretical

framework was elaborated by Cochran and Brassard. The social network

can influence the child's development through the parental role itself.







More specifically, the parent's role is influenced by the 1) exchange

of emotional material assistance between parents and network members

(Stack, 1974; Cochran and Brassard, 1979; Furstenberg, 1976);

2) provision of childbearing controls, that is, friends, relatives,

or neighbors of parents encourage or discourage participation in pat-

terns of parent-child interactions; and 3) availability of role models.

Parents adopt or modify some childrearing practices as a consequence

of watching the behavior of network members. Bandura (1977) pointed

out that participation is not a necessary condition for the impact of

the modeling to be manifested. Bandura also found that to have an

impact, the modeling person must be able to demonstrate the ability to

nurture or to control valuable resources.

Provision of the opportunity for active participation in the

networks of his or her parents can lay the foundation for the develop-

ment of the child's own personal social network or network building

skills. Furthermore, Cochran and Brassard (1979) propose that there

is general acceptance of the fact that the normal child progresses

developmentally from a more to a less egocentric view of physical and

social events and structures and that this progress also involves a

cognitive shift from a concrete, stimulus-bound perception of the

world to a representation of reality which is more abstract and

symbolic. These developmental changes have significance for the

child in relation to both the personal social networks of the parents

and the development of the young child's own network-building skills.

According to Cochran and Brassard (1979), reciprocal exchange is the

major cognitive skill which is attained through the maintenance of

network relationships.







The developmental benefits of reciprocal exchange are cited by

Cochran and Brassard as contributing to successful participation in

the peer group, preadolescence, and adulthood. It is through ex-

changes between parents and members of their social networks that the

child has the first opportunity to observe these reciprocal processes

at work. In contrast to Piaget's (1962) concept of reciprocity de-

veloping due to peer exchange, Cochran and Brassard suggested that

adult members of the parent's social network provide another significant

vehicle for the development of reciprocal skills.

Finally, Cochran and Brassard (1979) emphasized in their theore-

tical model of networking that the relationships between the child

and the network members are bidirectional in influence. The child

stimulates and reinforces the adult members of the support network to

behave in certain ways.

Very few investigators, according to Cochran and Brassard, have

examined systematically the impact of the parents' social networks

upon child development. They proposed that research be conducted in

four areas related to cognitive child development.

1) In the Piagetian sense the members of the social network

may provide the moderately conflicting and contrasting experience

which upsets the cognitive equilibrium enough to permit movement to

a higher stage of cognitive development.

2) Task completion--the parents' social network may facilitate

the development of skills required by children in goal-oriented

tasks.

3) Representation thinking--members of a social network are

sometimes present and often absent. In their absence, reference







to these individuals mayfacilitate the development of cognitive

memory capacities.

4) Cognitive receptivity--social networks may affect the

development of cognitive receptivity or openness to new intellectual

stimuli. Hess, Shipman, Brophy and Bear (1968) found that mothers

who were involved in many rather than few out-of-home activities made

greater use of home resources in interacting with their preschool

children.

Cochran and Brassard (1979) stressed the direct influence of the

support system on the child. They cited cognitive and social stimula-

tion as being the result of interaction with new people, varied activi-

ties, and varied settings. Piaget (1962) described the modeling of

temper tantrums by a friend's child, which stimulated similar behavior

in his daughter. Finally, Cochran and Brassard stressed the provision

of opportunities for active participation in the networks of his/her

parents. This, according to Cochran and Brassard, contributed to the

child's own personal social network and encouraged the "network-

building skills."

How personal, social, support networks influence the child's

development can be summarized as follows: network influences are

both direct and indirect. They include the sanctioning of parental

behaviors and the provision of material and emotional support for

both parent and child. Network members also serve as models for

parent and child. These processes (stimulation, modeling, involvement)

interact with the developmental age of the child to "stimulate the

basic trust, empathy, and mastery of the reciprocal exchange skills

essential to network functioning" (Cochran and Brassard, 1979, p. 605).







McKinlay (1973) states that although of considerable potential,

the concept of a social network is one of the most underdeveloped and

underemployed concepts in present-day sociology (p. 275). Sussman and

Burchinal (1962) state that before 1940 most Americans rejected the

notion that receiving aid from their kin is a good thing. Then, a

new empiricism emerged in the late 1940's that questioned the per-

sistence of the isolated nuclear family notion and presented.evidence

to support the viability of kin family network in industrial society.

Today, empirical evidence from studies by investigators in a variety

of disciplines substantiates the notion that the extended kin family

carries on multitudinous activities that have implications for the

functioning of other social systems of the society. "The major

activities linking the network are mutual aid and social activities

among kin related families" (Sussman and Burchinal, 1962, p. 235).

Simm and Belz (1980) describe the network in terms of learning

theory. "It is clear that for many types of learning problems and

situations, the learner's impulse is to seek out another individual

within his or her own social network so that assistance can be acquired

about the learning project in question" (p. 22). Essentially, Simm

and Belz conceptualize the social support network into many possible

learning situations or environments, whereby the learners adapt their

own resources within the framework available to them. This process

has been described as substantially one of information processing or

the exchange of information. Furthermore, Simm and Belz have stated

that networking enhances learning rather than inhibits it, as has

been the case in more traditional learning modes. Mitchell (1969)







also described networks as "communication links" which act as vehicles

for the flow of information.

Luikart (1977) utilized a learning approach to his definition

of networking. Luikart described self-directed learners as those who

are most likely to seek out and receive help from other individuals

with whom they have a common bond.

Finally, Caplan (1974) defined a support system or network to

be an "enduring pattern of continuous or intermittent ties that play

a significant part in maintaining the psychological and physical

integrity of the individual over time" (p. 9).

Thus, in recent years, the significance of support systems,

often called networking, has been emphasized by many people in many

professions, but especially the helping professions. Waters (1981)

suggested that while supports are important to people throughout their

lives, they are crucial at times of transition.

Caplan (1974) noted that the concept of support systems has

focused on the health-promoting forces at the person-to-person and

social levels which enable people to master the challenges and strains

in their lives. Therefore, Caplan urges investigators to utilize the

network concept from the standpoint of being a "wellness" rather than

a sickness model, an approach that examines methods people use to

acknowledge and mobilize their strengths. In Caplan's view, the major

feature of support networks is that the person with a problem is

"dealt with as a unique individual. Significant others help

the individual to mobilize his psychological resources and master his

emotional burdens; they share his tasks and they provide him with

extra supplies of money, materials, tools, skills, and cognitive







guidance to improve his handling of his situation" (p. 5-6). Caplan

conceptualizes the support network in a positive fashion. To Caplan,

the notion of support represents an enrichment of existing strengths,

rather than a propping up of someone who .is weak.

Gourash (1978) reviewed the literature on help-seeking behavior

and concludes that help-seeking is a critical aspect in the nuclear

family's ability to cope and survive in time of crisis. Help-seeking

is defined by Gourash as "any communication about a problem or trouble-

some event that is directed toward obtaining support, advice, or

assistance in times of distress" (p. 414). This conceptualization of

support systems is defined, then, in a help-seeking fashion. Gourash

describes help-seeking as encompassing requests for assistance from

friends, relatives, and neighbors as well as professional helping

agents.

Zimbardo and Formica (1963) described people who solicit help as

usually looking for comfort, reassurance, and advice. Booth and

Babchuck (1972), Litman (1974), and Quarantelli (1960) all found that

individuals tend to turn initially to family and friends and, as a

last resort, to professional services and agencies. Rosenblatt and

Mayer (1972) found that the sole use of professional services occurs

much less freuqntly than either exclusive reliance on family and

friends or help-seeking from both the social network and professional

sources.

Gourash (1978) posited that members of the support network can

affect help-seeking in a number of ways: 1) by buffering the experi-

ence of stress which obviates the need for help; 2) by precluding

the necessity for professional assistance through the provision of







instrumental and affective support; 3) by acting as screening and

referral agents to professional services; and 4) by transmitting attitudes

values, and norms about help-seeking (p. 416).

Speck and Rueveni (1969) utilized the social network concept

in developing a therapy network for dysfunctional family units.

Speck and Rueveni (1969) defined social network as "a group of per-

sons who maintain an ongoing significance in each other's lives by

fulfilling specific human needs" (p. 184). Like Sussman, Speck and

Rueveni have rediscovered that even in our own culture of nuclear

families, the extended family kinship system still plays a significant

role in the adaptation of nuclear families. Utilizing this approach,

Speck and Rueveni (1969) successfully treated schizophrenic individuals

by assembling family therapy networks, some numbering as many as 40

people.

In summary, the notion of support systems or networking is

defined and described by several investigators and theorists, from

sociology to the helping professions. Support systems contribute not

only to positive mental health, but there is some evidence that the

absence of support networks is associated with negative outcomes,

as described in the subsequent section on research related to sup-

port systems. It is suggested that help-seeking is especially

crucial in times of transition or crisis (Waters, 1981). Adolescent

pregnancy thrusts the young female into accelerated role transition,

thus placing her in a crisis situation (Bacon, 1974). Thus, the

need for social support for the adolescent mother appears to be

critical at this time in her life.







Research Related to Social Support Networks

In the previous section, the relatively recent emphasis on the

role of support systems or networks is established. Research utiliz-

ing support systems as a predictor variable has been conducted. The

implications from this research are that social support systems are

important to persons through all stages of the lifespan. Cobb (1976)

summarized the research on social support and concludes that it plays

a critical role in moderating crisis situations. While most of the

research identifies social support as a positive force in human

lives, McKinlay's (1973) study found that social support systems may

discourage the use of medical facilities. Tolsdorf (1976) concluded

from his study of nonpsychiatric and psychiatric subjects that mental

health may be correlated with the quality and quantity of one's

social support system and the ability to draw upon its many re-

sources. Berkman and Syme (1979) reported that social circumstances

such as social isolation may have pervasive health consequences to

the individual, such as vulnerability to disease in general and even

death. Swick, Brown and Watson's (1980) study finds that neighbor-

hoods that have a higher degree of social interaction, less transience,

and more child-child interaction are perceived by residents to be

more supportive.

McKinlay (1973) conducted a study of social networks, lay con-

sultation, and help-seeking behavior. McKinlay addressed the gen-

eral question, "What is the apparent role of the family, and its kin

and friendship networks, in the utilization of services?" Utilizing

two approaches--sociodemographic data (such as geographic proximity







of friends and relatives and frequency of contact with them) and

open-ended responses to hypothetical questions/situations--McKinlay

studied 87 unskilled working-class families, consisting of two sub-

samples of "utilizing" and underutilizingg" respondents.

The subjects were drawn from a centralized hospital-based

maternity clinic in Aberdeen City, Scotland, and interviewed four

times over a period of about one and a half years. Women who attended

the central hospital clinic at least once after the birth of their

infants were included in the study. A woman was regarded as a

"utilizer" if she a) attended the clinic for the first antenatal

visit before the end of the 17th week of gestation and b) having

had the pregnancy confirmed by the obstetric staff, attended regularly

for antenatal care.

A woman was regarded as an "underutilizer" if she a) had no ante-

natal preparation at all, b) attended for some form of care only

after the 28th week of gestation, or c) was an emergency admission

during labor without any previous prenatal care, or d) defaulted

from her clinic appointments more than three times consecutively

without offering an excuse.

The subjects were a random sample consisting of 48 underutilizers

(40 multiparae and 8 primigravida) and 39 utilizers (24 multiparae

and 15 primigravida). The variables studied were components of kin-

ship and friendship network.

The findings indicate that more underutilizers had relatives

living in the same house and, when their presence was controlled for,

still had more relatives living close geographically, despite com-

parable numbers of relatives seen by members of both study groups.







The utilizers--whether multiparous or primiparous--visited with

relatives less frequently than did underutilizers. Utilizers made

greater use of friends and husbands and less use of mothers or other

relatives and tended to consult a narrower range of lay consultants.

The findings suggest that support by family members may serve as an

aid in medical situations. Further study is needed to determine the

effect of this intervention by the family on the degree of medical

care sought by the patient.

Cobb (1976), in a review of social support as a moderator of

life stress, identifies three classes of social support: 1) informa-

tion leading the subject to believe that he/she is cared for and

loved; 2) information leading the subject to believe that he/she is

esteemed and valued; and 3) information leading the subject to

believe that he/she belongs to a network of communication and mutual

obligation.

Cobb summarized the research on social support as a moderator of

life stress. He suggests that there is strong and often quite hard

evidence, repeated over a variety of transitions in the life cycle

from birth to death, that social support has been protective. Con-

vincing evidence exists that adequate social support can protect

people in crisis from a wide variety of pathological states: from

low birth weight to death, and from arthritis through tuberculosis

to depression, alcoholism, and other psychiatric illnesses. Cobb

encourages investigators to study the effect of a support system as a

moderator of life situations so that a more thorough understanding of

the role and function of support systems might be obtained.







Tolsdorf (1976) conducted an exploratory study of social networks,

support, and coping. Using the social network model, borrowed from

sociology and anthropology, Tolsdorf described and quantified an

individual's immediate family and those with whom he/she has regular

contact.

The purpose of the Tolsdorf (1976) study was to determine the

usefulness of the social network model in the study of stress, support,

and coping. The main interest was to expand the network model to

coping and psychopathology using a system rather than an individual-

istic approach. Extensive interviews were conducted at the Northampton,

Massachusetts, and Newington, Connecticut, Veterans Administration

Hospitals with ten recently hospitalized, first-admission males,

hospitalized as medical (nonpsychiatric) and ten psychiatric subjects.

In the psychiatric sample, seven were diagnosed as paranoid schizo-

phrenic, two as acute schizophrenic episode, and one simply as

schizophrenic. All subjects were male veterans matched for age,

marital status, education, and socioeconomic status.

Data were collected on the size and membership of the subjects'

social networks, the qualities of the relationships the subjects had

with their network members, the subjects' attitudes, beliefs, and

expectations toward their networks as to the role network members

could play in helping them cope with life stresses, the presence and

type of recent life stresses, and the extent and nature of their usual

coping styles.

The methodology yielded two types of data: quantitative data

that summarized the network variables and qualitative data gathered

during the interview process. Three classes of variables were found







to discriminate between the two groups: 1) relationship density and

multiplex relationships; 2) functional indegree and outdegree;

3) kinship members and linkages. Medical subjects tended to 1) give

and receive equal numbers of functions but psychiatric subjects tended

to receive many more functions than they provided to others; 2) the

psychiatric subjects reported fewer intimate relationships with their

network members in a network that was more heavily dominated by

family members, where functional people were in a more controlling

and dominant position and where overall there were relatively few but

relatively more powerful functional people in the network.

The medical subjects, on the other hand, reported more intimate

relationships with more people in a network that was less dominated

by family members and where functional people were on equal standing

with subjects in the exchange of support, advice, and feedback.

The qualitative data indicated that the psychiatric subjects

demonstrated negative network orientation--a set of expectations or

beliefs that it is inadvisable, impossible, useless, or potentially

dangerous to draw on network resources (p. 413). They had a history

of having not drawn on the advice, support, or feedback of their

network members, and they explicitly stated that they intended not to

draw on such supports in the future. When compared to the psychiatric

group, the medical group, by and large, held positive network ori-

entations; the majority stated that they did seek out support and

resources of their network members, especially if they could not

handle a problem themselves.

The above data suggest that the medical subjects had more contact

with, and drew more heavily on, a broader and stronger base of network







resources, compared to the psychiatric subjects. Consequently, the

medical subjects received much more network support compared with the

psychiatric group. "Network mobilization" was cited frequently by the

medical subjects as a way of coping with life stresses, while psychi-

atric subjects avoided it. In summary, medical subjects utilized

their social support networks more frequently than did the psychiatric

subjects. In this study, network utilization was associated with

healthier coping skills.

Berkman and Syme (1979) presented findings on a study of social

networks, host resistance, and mortality. The relationship between

social and community ties and mortality was assessed using the 1965

Human Population Laboratory survey of a random sample of 6928 adults

in Alameda County, California, and a subsequent nine-year follow-up

study of mortality. The findings showed that people who lacked social

and community ties were more likely to die in the follow-up period

than those with more extensive contacts. The age-adjusted relative

risks for those most isolated when compared to those with the most

social contacts were 2.3 for men and 2.8 for women. The association

between social ties and mortality was found to be independent of self-

reported physical healthy status at the time of the 1965 survey, year

of death, socioeconomic status, and health practices such as smoking,

alcoholic beverage consumption, obesity, physical activity, and

utilization of preventive health services as well as a cumulative

index.

A Chi square statistic was used in this study, modified to

include more than two comparison groups. Four sources of social

context were examined: 1) marriage, 2) contacts with close friends







and relatives, 3) church membership, and 4) informal and formal group

associations. With few exceptions, respondents with each type of

social tie had lower mortality rates than did respondents lacking

such connections. In every sex and age category, people who reported

having few friends and relatives and/or who saw them infrequently had

higher mortality rates than those people who had many friends and

relatives and saw them often. A Social Network Index was constructed

to summarize the effects on mortality of increasing social isolation.

The findings from this study suggest that social circumstances such

as social isolation may have pervasive health consequences, and they

support the hypothesis that social factors may influence host re-

sistance and affect vulnerability to disease in general.

Swick, Brown, and Watson (1980) hypothesized that at least four

factors within the neighborhood determined whether it was a supportive

social network or nonsupportive network. Three issues were addressed.

1) The family's perception of itself and the outer world as

positive or negative will affect the willingness to form networks

within the neighborhood.

2) The family's perception of its needs and the needs of neigh-

borhood members and the probability of meeting their needs within the

neighborhood will affect the willingness to form networks within the

neighborhood.

3) The degree of stress the family and neighbors are under as

well as their ability to deal effectively with stress will affect

the development of supportive networks. Swick and colleagues de-

veloped the PNSS (Percention of Neighborhood Supportiveness Scale)

which contained questions that related to perceived neighborhood







support and the demographic makeup of the neighborhood. Subjects were

early childhood graduate students at the University of South Carolina

during the fall semester of 1979. There were 164 subjects, pre-

dominantly Caucasian, married, professional educators, with middle

socioeconomic status. The subjects were administered the PNSS; Part A

consisted of 27 questions that provided demographic information about

the respondent and the respondent's neighborhood. Part B was a Likert

Scale of behaviors which might be found in supportive neighborhoods.

A Crombach Alpha internal consistency reliability coefficient of .93

was computed in a pretest.

The ANOVA revealed significant relationships for scores obtained

in Part A (Demography) and Part B (Perception of Supportiveness) for

Amount of Playtime, Age of Respondent, Type of Dwelling, Longevity in

Neighborhood, Income, Frequency of Social Activity, and Frequency of

Social Activity Involving Children. Neighborhoods were perceived as

more supportive when 1) more children played with other children;

2) slightly older adults resided in the neighborhood; 3) home owners

were more numerous; 4) higher incomes existed; and 5) longevity in

the neighborhood was present. In summary, this study evaluated a

neighborhood as to its degree of supportiveness and how the subject's

perception of the neighborhood related to the subject's attitude

about his neighborhood. Findings show that neighborhoods that had a

higher degree of social interaction, less transience, and more in-

volvement with children were perceived as being more supportive

neighborhoods.







Conceptual Framework for Research on Networking
and Child Development

The current resurgence of interest in familial influences on child

behavior and development is accompanied by a strong concern for the

social and economic conditions under which families carry out their

childrearing functions (Powell, 1979). The interplay between the

social environment and the family childrearing process is relatively

uncharted terrain,and theoretical guidelines for such research are

relatively few in number. Powell (1979) makes a strong argument that

the child's socialization experience is related to the nature of the

family's interaction with its immediate social environment. The net-

working approach views the family as an open, adaptive system whose

exchanges with the environment provided emotional and material support

for the family's childrearing functions, socialize the family into

certain childrearing beliefs, practices, and family-child relationships

(Powell, 1979). Bronfenbrenner (1974) suggests that the progressive

fragmentation and isolation of the family in its childrearing role

requires support systems that undergird parents in rearing their

children.

While research is limited, there is evidence of a strong rela-

tionship between family life ecology and child behavior from research.

on the correlates of child abuse (Garbarino, 1976; Parke, 1977) and

on the development of children in single-parent families (Hetherington

Cox, & Cox, 1977).

Parke (1977) posited a social-psychological theory of child abuse,

based on research data, in which cultural sanctioning of violence and

lack of community support systems for families are viewed as supporting







the development of abusive parents. Garbarino (1976) found that

the degree to which mothers have been subjected to social-emotional

stress without adequate support systems accounted for a substantial

proportion (36 percent) of the variance in rates of child abuse and

maltreatment. Hetherington, Cox, and Cox (1976) found that the lack

of support systems played a significant role in disruptions in the

development of children of mother-headed households.

As Keniston (1977) pointed out, since the early 1800's the concept

of families as autonomous units has been an integral part of the moral,

economic, and political fabric of the United States. Thus, the new

paradigm of a socioecological perspective has emerged slowly and with

some reluctance.

How personal, social, support systems influence the child's

development can be summarized as follows: network influences are both

direct and indirect. They include the sanctioning of parental be-

haviors and the provision of material and emotional support for both

parent and child. Network members also serve as models for parent

and child. These processes (stimulation, modeling, involvement)

interact with the developmental age of the child to "stimulate the

basic trust, empathy, and mastery of the reciprocal exchange skills

essential to network functioning" (Cochran and Brassard, 1979).


Research Related to Child Development and Networking

The paradigm of the socioecological relationship to child devel-

opment has emerged only in the last decade. Research that connects

networking with child development and behavior has been limited.

Several of these studies are now presented.







Stevens (1981) and Hough and Stevens (1981) stated that we have

been preoccupied with constructing and designing formal support sys-

tems like parent education programs, Home State programs,and parent-

child center programs that have enabled parents to rear competent

children. However laudable that goal, Stevens cautions that most
parent intervention programs have been developed from ignorance.

Most researchers, according to Stevens, have not attempted to uncover

existing informal networks that enable black parents, especially,

to rear children competently. Stack (1974) found that the black

parent's social network played a significant role in the rearing

of children. Childrearing was perceived not as an isolated task of

the mothers, but rather a task to be shared by network members.

Stevens studied 300 low-income black families in a large southern

metropolitan area to see if a number of aspects of the mother's

behavior and characteristics of her network were correlated with her
infant's development. Stevens utilized a multiple correlation ap-

proach and found that mothers who were more emotionally and verbally

responsive and who had networks composed of more females had toddlers
who were developing better, as measured by the Bayley Scales of

Infant Development. Stevens concluded from his study that "natural

networks" of black families were best left intact and undisturbed.

Stevens also concluded that intervention programs have not adequately
taken into account the role of support networks in the development of

young children.

Garbarino (1976) studied ecological correlates of child abuse.

Garbarino utilized Bronfenbrenner's ecological construct to look at
the possible correlates between child abuse and parent support systems.







The mediation of immediate family settings by socioeconomic forces was

hypothesized to be related to the degree to which children experienced

abuse and maltreatment. Counties of New York were chosen as units of

analysis. It was assumed that the 1973 reports recorded by the state-

wide Central Registry were a reasonably valid index of child abuse

and maltreatment. Data reported for the period September 1, 1973, to

January 31, 1974, were used. The United States census was chosen as

the source for indices of the independent variables support systems

for parenting. The variable indices were 1) transience, 2) economic

development, 3) utilization of educational resources, 4) rurality-

urbanism, and 5) SES of the mother.

The analysis was designed to test the hypothesis that the socio-

economic support system in the family in each county was directly

associated with the rate of child abuse and maltreatment. Stepwise

multiple regression was utilized to generate the best predictor

equation based on the independent variables.

The results indicated that the ecological context generated by

economic and educational resources was an important factor in the

etiology of child abuse and maltreatment. Overall economic distress

appeared to be important through its impact on mothers as well as

directly on the community and neighborhood. Garbarino concluded that

abuse and maltreatment was, to a large extent, related to economically

depressed mothers, often alone in the role of parent, attempting to

cope in isolation without adequate facilities and resources for their

children. The results of this study suggest that one promising way

to deal with the social problem of child abuse and maltreatment is by

dealing with the support systems of mothers.







Several studies have correlated infant development with socio-

environmental factors. In a study of the three-year developmental

status of high risk infants, Holstrum (1979) investigated the develop-

ment of premature and term infants. A multiphasic battery of tests

was used to assess the developmental status of each child and the

socioemotional environment in which she/he lived. Data were gathered

to answer the following research questions: 1) Do those infants who

were sick and/or premature at birth catch up in their developmental

status to term infants by three years of age? 2) Which of the vari-

ables under investigation contribute the most to the prediction of

developmental status.

The study involved 102 three-year-olds born at or transferred to

Shands Hospital, University of Florida. Half of the children had

been placed in the NICU (Neonatal Intensive Care Unit) due to pre-

maturity or illness. The others were term infants placed in the

Newborn Nursery. The children were evaluated with the Stanford

Binet, the Carrow Test for Language Comprehension, and the Beery

Developmental Test of Visual-Motor Integration. Demographic and

Socioenvironmental data were obtained through the Personal Interview,

the Stressful Life Events Checklist, and the Behavioral Interview.

Birthweights and amount of neonatal complications were obtained from

medical records.

The results showed no statistically significant differences

between the NICU and the term infants. The data suggest, however,

that the small-infants are still small at three years of age, and

the infants who had the greatest amount of complications were found

to be the most significant predictors of three-year developmental








status. Socioeconomic status contributed significantly to the pre-

diction of IQ and visual-motor integration. It was found that mothers

with more stress experiences in their lives were more likely to per-

ceive their children as having behavioral problems.

Bender (1980), in a study of adolescent, early childbearing,

and preventive health services, hypothesized that the high rate of

adolescent pregnancy and poor use of preventive health services by

adolescents before, during, and after pregnancy are conditioned by

factors in the adolescent's private culture. This cultural analysis

included a study of the events of pregnancy and the attitudes, be-

liefs, and values which surround the pregnancy for 35 adolescent

mothers. The adolescents were residents of a town in the southeastern

United States called Farmville. Viewpoints of members of the community

and of health providers were used also.

Instrumentation included a schedule, a questionnaire, open-

ended interviews, and direct observation. Data were collected from

35 of 76 adolescents who had given birth in Farmville in 1976.

Materials from ten of these interviews, when combined with data from

reinterviews, were developed into case studies. Data from the re-

maining 35 interviews were incorporated into the analysis which is

organized in six themes: contraceptive knowledge, use, and responsi-

bility; initial reactions to pregnancy; family relationships, sup-

portive networks, acceptability of health care, and adolescent

motherhood.

The author concluded that adolescents and adults alike perceive

early, unplanned pregnancy as a mistake. The high rate of adolescent

pregnancy in Farmville was attributed to differences between the







private culture of adolescents and the community's public culture.

Specifically, adolescents lacked information related to contra-

ceptives, and they were often unable or unwilling to admit that their

sexual activity violated certain cultural norms. Recommendations

for creating culturally appropriate services for adolescents were

made in keeping with the conclusions named above.

Hofheimer (1979) studied the developmental outcomes for infants

born to adolescent mothers. The primary purpose of the study was to

assess the contributions of the mother's age, perinatal risk, and

socioenvironmental and medical resources to the prediction of dimen-

sions of the mother-infant transaction process and the developmental

status of the infant.

Data were collected in a clinical setting on an age-specific

sample of 77 mothers and their six-month old infants. Infant devel-

opment was assessed using the Bayley Scales of Infant Development.

Mother-infant transaction was analyzed using the Adapted Beckwith

Behavior Scale (ABBS). Demographic and socioenvironmental data were

obtained from the Child and Family Development Interview.

Based upon the results of the multivariate multiple regression

analysis, the variables which were identified as predictors of mental

developments were 1) the age of the mother, 2) the type of prenatal

care of the mother, and 3) the presence of prenatal complications.

Psychomotor development was found to vary as a function of 1) respon-

sive vocalization of the mother-infant transaction process and

2) the type of prenatal care received by the mother.

The results suggested that the infants of young mothers are at

risk for problematic mental development. The data supported the







idea that the mother-infant relationship is important to the infant's

development of competence. The findings suggested that more compre-

hensive interdisciplinary models of perinatal and pediatric support

are associated with enhanced development of the infant. Also indi-

cated was a need for the design of parent and infant-centered inter-

ventions for the young mother in order to enrich the quality of care

and stimulation provided by the mother and thus enhance the development

of the infant.

Colletta, Hadler and Gregg (1981) studied how adolescents cope

with the problems of early motherhood. The purpose of this study

was to measure the coping responses used by adolescent mothers (N = 64),

to determine variables related to their choice of responses, and to

examine the relationship between coping responses and emotional stress.

A nonprobability sample of adolescent mothers was obtained

through the cooperation of a public school system in a large metro-

politan area. Mothers were accepted for participation in the study

if they were between 14 and 19 years of age, had one child two years

of age or under, and were not currently married. The sample con-

sisted of 64 black adolescents who averaged 16.27 years of age when

their children were born and 17.47 years of age at the time of the

interview.

An interview schedule was designed to elicit the mother's

perceptions of the major problems in their lives, the ways they

responded to the problem situations, and the emotional stress they

experienced in each area. Support was measured by asking the mothers

which individuals and/or institutions assisted them in each problem







area. Support was first coded 1 nonsupportt) to 4 (extensive support).

Intercoder reliability ranged from .84 to .91 with a mean of .87.

The results showed that the adolescent mothers' major coping

response was to ask others for assistance. This response was most

common when the young mothers were faced with task-oriented problems.

Interpersonal problems tended to elicit avoidance as a coping re-

sponse, while conflicts with institutions elicited a range of

responses. Direct action coping responses were related to higher

self-esteem, more active support systems, and lower levels of emo-

tional stress. The results are interpreted to indicate the impor-

tance of support systems which help young mothers deal with their

daily problems.

These young mothers also cited isolation from peers and finances

as being their major concerns. Fifty-four percent of the mothers

were satisfied or very satisfied with the amount and quality of help

available from community services. Twenty-nine percent reported

problems with community services, and 15 percent were acutely unhappy

with the quality or availability of services. The style of coping

was related to other variables. Young mothers with high self-

esteem and active support systems were significantly more likely to

choose an active coping strategy while there was a trend for those

with an internal locus of control to use active coping responses.

Finally, adolescent mothers with active coping styles reported lower

levels of emotional stress across problem categories.

A second major finding was that when faced with interpersonal

problems, with their parents, peers, or childcare, the mothers chose

to cope by avoiding the situation.







Held (1981) studied the self-esteem and social network of 62

women, 17 years of age and younger, who were in their third trimester

of pregnancy. Self-esteem was assessed by administration of the

Coopersmith Self-Esteem Inventory. Social network information was

sought by asking the adolescent to rate her perception of reactions

to the pregnancy by significant others. She was then asked to rank

these people in order of importance to her. The study population

included women from Houston's three major ethnic groups--white, black,

and Mexican-American--who were in five different program settings.

All women received medical care from the obstetrical staff of the

University of Texas Medical School.

Tests of significance were done using the Chi Square method of

correlation. In order to study self-esteem, the data were artifi-

cially dichotomized into scores above and below 70; 70 and above are

normal scores reported for most population groups. It was statis-

tically significant at the .05 level of confidence that almost 60

percent of black women keeping their babies scored 70 or above on the

Coopersmith Self-Esteem Test. Fewer than 30 percent of the whites and

Mexican-Americans scored 70 or above. Self-esteem was higher for

those women attending a day school for pregnant teenagers. Scores

were higher than for women at clinic sites and at the Home for Unwed

Mothers. Social Network data indicated that the mother of the ado-

lescent was most disapproving. The father of the baby was most

approving. Ninety-two percent of the adolescents placed their mothers

as being more important to them than they themselves were. In

summary, self-esteem scores were highest among black women keeping

their babies who attended the day school for pregnant women. Held







suggests that future research should include longitudinal research

especially with regard to enactment of future plans.


Summary

Social support systems or networks have been defined and studied

primarily in the fields of sociology and anthropology. More recent

research has viewed the social support system in relation to.family

development and functioning. Only in recent years has the possible

relationship between the social support system and specific measures

of child development been proposed. Indeed, very few studies have

focused on the mother's social support system and the qualitative

development of her young child.

The widespread occurrence of adolescent pregnancy in this society

has caused concern for the young mother, the development of her in-

fant, and for the consequences for society in general. There is a

need to study this population of adolescent mothers in regards to

the amount of social support that they have available to them and

the consequent development of their infants. Since adolescent

mothers have experienced higher rates of prenatal and obstetric

complications and since infants born to adolescent mothers frequently

have been high-risk, low-birth-weight, due to neonatal complications,

there is a specific need to investigate possible correlations between

the support system of the adolescent mother and particular develop-

mental measures of the infant.

There are limitations to the cited research by Held (1981),

Colletta, Hadler, and Gregg (1981), and Bender (1980). These studies

utilized questionnaires to obtain their data. The limitations of the







use of questionnaires included 1) the data were self-reported data;

2) incomplete data collection was more common; 3) interpretation of

questions by the subjects may have varied; and 4) they were correc-

tional rather than experimental. Additional limitations of these

studies included the fact that the populations were local and not

national in makeup. Thus, the generalizability of these findings was

limited to such local populations and not to samples of a national

scope. However, the findings of the above studies reinforced one

another.

The implications of such research have significance for parent

education, infant development, and mother-infant interaction. Inter-

vention for adolescent mothers has been traditionally on a formalized

basis. The possibility exists that social support systems may be a

vehicle to further means of informal intervention that might encourage

a young mother in her demanding role as a parent. Support systems

may be related to quality child development.













CHAPTER III
METHODOLOGY

Design of the Study

This study investigated the relationship between measures of

support network resources of single adolescent mothers and the cog-

nitive and psychomotor development of their infants.

The purpose of this study was to investigate 1) the types of

support network resources utilized by single adolescent mothers and

the frequency with which these support resources are used; 2) the

correlation of four indicators of support network resources to cogni-

tive and psychomotor measures of development of the high-risk infants

at 12 months of age (both linear and stepwise regression procedures

were utilized to study the nature of the relationship between the two

dependent variables and the four independent variables, each of which

consisted of several measures of support); and 3) measures of self-

concept and life stress of the adolescent mothers.

The research design used in this study was descriptive and ex

post facto. Due to the nature of the variables--measures of existing

support network resources and measures of cognitive and psychomotor

development taken at one year of age--an ex post facto design was

required. Therefore, the interpretation of the results from this

study design was associational or relational and not causal. The

strength of the design is that it allows the investigator to study







existing variables. Also, the support network was an attribute vari-

able and was not manipulative.


Variables in this Study

Measures of four independent variables were studied. Three of

these independent variables consist of more than one measure. The

four independent measures include

1) Measures of communication efforts made by the support network

during the infant's hospitalization in the Neonatal Intensive Care

Unit at Shands Hospital, University of Florida. A composite or

overall measure of this independent variable was obtained by totaling

the actual number of visits made by members of the support network

during the infant's hospitalization. (The length of stay in the

hospital, in number of days, will be used as a control variable in the

initial analysis.)

2) Measures of the adolescent mother's accessibility to her

support network were obtained from the sociodemographic data. Three

separate measures were grouped together to obtain this single measure.

These three measures of accessibility include 1) accessibility to a

car, 2) accessibility to a telephone, and 3) the mother's own per-

ception of her accessibility to her support network (ease or difficulty).

See Appendix for examples of this measure.

3) Measures of the adolescent mother's actual support network

were quantified. Five questions from the sociodemographic data were

used as measures of support; these measures were grouped to obtain one

multiple correlation coefficient which was used as an overall measure

of the relationships between adolescent mother's actual support and








the two dependent variables. These five measures include the number

of adult relatives/friends living in the immediate area, the number

of relatives/friends that can be counted on in times of real need,

the frequency of get together with relatives/friends, in general and

per week, and the total number of organizations the adolescent mother

participates in. See Appendix for examples of this measure.

4) Measures of the adolescent mother's perception of her support

network were grouped to obtain an overall measure of perception of

support. These measures include the mother's perceptions of enough

support, a perception of being overwhelmed, a perception of time to be

by herself, a perception of the adequacy of that time to be by her-

self, and a perception of the overall level of happiness with her

present living situation. See Appendix for examples of these measures.

Dependent Variables


The two dependent variables are scores attained by the admin-

istration of the Bayley Scales of Infant Development, Mental and

Psychomotor Development Index. These scores were obtained when the

infant was 12 months old, adjusted for prematurity.

Control Variables


Four variables used as control variables in the initial analysis

were measures of 1) the mother's age, in years; 2) the mother's level

of education, in years; 3) length of the infant's hospitalization, in

days; and 4) the infant's birthweight, in grams.











Table 2. Independent, dependent and control variables in this study.


Independent Variables

1. Measures of communication efforts by the support network
during the infant's hospitalization.

2. Measures of the mother's accessibility to her support
network.

3. Measures of actual support network resources of the
adolescent mother.

4. Measures of the mother's perception of her support network.

Secondary Variables

1. Measures of the mother's self-concept.

2. Measure of the mother's life stress.

Dependent Variables

1. The score attained by the infant when assessed with the
Bayley Scales of Infant Development, Mental Index (MDI).
2. The score attained by the infant when assessed with the
Bayley Scales of Infant Development, Psychomotor Index
(PDI).

Control Variables (Used only in primary analysis)

1. Mother's age at birth of infant, in years.

2. Mother's level of education completed, in years.

3. Infant's birthweight, in grams.

4. Number of days infant was hospitalized.







Hypotheses to be Tested

Hypothesis I: There will be no significant linear relationship

between measures of in-hospital communication efforts by the support

network and Bayley scores of cognitive and psychomotor development of

the infants at 12 months of age.

Hypothesis II: There will be no significant linear relationship

between measures of the mother's accessibility to her support network

resources and her infant's scores of development on the Bayley Mental

and Psychomotor Index.

Hypothesis III: There will be no significant linear relationship

between measures of the adolescent mother's actual support network

resources and her infant's scores on the Bayley Mental and Psychomotor

Index.

Hypothesis IV: There will be no significant linear relationship

between measures of the adolescent mother's perception of her support

network and her infant's scores on the Bayley Mental and Psychomotor

Index.


Description of Subjects

The following tables present descriptive data on the single

adolescent mother subjects and their 12-month old infants. The

infants in this sample ranged in weight from 800 to 2240 grams, with

the mean weight being 1415 grams. Only one infant weighed more than

1800 grams, the weight which is considered to be "at risk." All

infants were classified as premature/high risk at birth due to lower

birth weights or other medical complications. See Table 3 for a

summary of birthweights of the infant subjects.







Table 3. Birthweights of infants born to single adolescent mothers
(in grams).


Birthweight

800-1000
1001-1200
1201-1400
1401-1600
1601-1800
1801-2000
2001-2200
more than 2200
Total


Frequency

3
7
9
5
9
1
0
1
35


Percentage

9.00
20.00
26.00
14.00
26.00
3.00
00.00
3.00
100.00


Maternal Age


Table 4 presents data on maternal age which were available for

34 of the 35 subjects. Twelve of the mothers were 19 years of age

and 11 were 18 years of age. Only three mothers were age 15. There-

fore, two-thirds of the single adolescent mothers fell into the older

category. Approximately 20 percent of the adolescent mothers were

16 years of age or under. The mean age for the mother subjects was

17.7 years of age.


Maternal Level of Education


Data on education completed by adolescent mother subjects were

available for 34 of the 35 subjects, as shown in Table 5. Over half

(18) of the mothers had completed 12 years or more of education.

The mean level of maternal education was 10.9 years. Only three

mothers had completed nine years or less.







Table 4. Ages of single adolescent mother subjects (by years).


Age Frequency Percentage

15 3 9.00

16 4 12.00

17 4 12.00

18 11 32.00

19 12 35.00

Total 34 100.00



Table 5. Years of education completed by mothers (in years).


Years of education Frequency Percentage

9 or less 3 9.00
10 6 18.00
11 7 21.00
12 16 47.00
13 or more 2 6.00
Total 34 100.00



Mother's Race


Of the 35 single adolescent mothers, 33 were black and 2 were

white. Table 6 presents the data on the race of the mother subjects.

This population sample reflected a rather homogeneous group of single

adolescent mother subjects, both for race and socioeconomic status.

All mothers were lower SES, as indicated by their responses on the

personal data part of the questionnaire; all subjects indicated lower

incomes.








Table 6. Race of single adolescent mother subjects.


Race Number of Subjects Percentage

Black 33 94.00
White 2 6.00
Other 0 00.00
Total 35 100.00



In summary, the subjects for this study were single adolescent

mothers whose high-risk infants were placed in the Neonatal Intensive

Care Unit at Shands Teaching Hospital, University of Florida, during

the period from January 1, 1979, through December 31, 1980. All

mothers were less than 20 years of age and were classified as 1) never

married, 2) divorced, or 3) separated. Only medical files which were

complete were included in this study.


Instrumentation

The instrument used to collect data on the adolescent mother's

support system was a personal questionnaire administered to the mother

when she brought her child for assessment. Sociodemographic informa-

tion was coded from the personal interview and categorized into sub-

headings of measures of support. The following categories were

obtained: 1) measures of communication efforts by the support network

during the infant's hospitalization, 2) measures of the adolescent

mother's accessibility to her support network, 3) measures of actual

support network resources available to the mothers, and 4) measures

of the mother's perceived support.







Listed below are the specific questions used in this study.

I. Measures of Indicators of Communication between the Support

System of the Mother and the Infant during the Infant's Hos-

pitalization. The total number of visits made by members of

the support network was obtained from hospital records.

II. Measures of Mother's Accessibility to Support Network..

a. Do you own or have access to a car? YES NO

b. Do you have a telephone? YES NO

c. How difficult is it for you to get out and do what

you need to do?

YES = very difficult; somewhat difficult; depends

NO = somewhat easy; very easy

III. Measures of Actual Support Available to the Adolescent Mother.

a. How many relatives/friends live around your area (adults)?

total number
b. How many of these relative/friends can you count on in

times of real need? total number

c. Do you get together often with friends and relatives?

YES NO

d. How many times a week?

NOT OFTEN = never; rarely (up to once every few months)

OFTEN = sometimes (a few times a month to once a week);

often (a few times a week); very frequently (every day)

e. Do you belong to any organizations such as social,

religious, educational, or political groups?








social religious educational __political =

total number of organizations

IV. Measures of Mother's Perception of Support Network Resources.

a. Does your present situation provide you with enough or not

enough support? YES = enough, sometimes

NO = not enough; not at all
b. Do you feel you're overwhelmed with household tasks and

children? NO = never; infrequently

YES = sometimes; often; all the time
c. How many times a month do you get away by yourself to do

something you would like to do? (total)

d. Do you think this is enough time? NO = not enough

YES = enough or too much
e. On the whole, would you describe your present living

situation as happy? NO = very unhappy; somewhat unhappy;

neutral

YES = somewhat happy; very happy


Responses that were YES or NO were quantified by assigning a

value of 0 to No and a value of +1 to Yes. A point biserial correla-

tion coefficient was determined for each dichotomous response. The

responses were categorized into four independent general support

variables. Three of the support variables consisted of three or more

measures. These measures were grouped by their correlation coefficients

to form multiple correlation coefficients. These were 1) accessibility

to support, 2) actual measures of support, and 3) the mother's per-

ception of her support. The fourth independent variable, support by







the network during the infant's hospitalization, was quantified by

totaling the actual number of visits made by the family during the

infant's hospitalization. A numerical total was obtained, thus

quantifying this independent variable for the purpose of analysis.

The instrument used to collect data on the infants' development

was the Bayley Scales of Infant Development, Mental and Psychomotor

Indexes. This instrument has been used in child development settings

for over thirty years. Bayley (1969) reported a correlation of .60

between the Mental and Motor Scales. The split-half reliability

coefficient for the Mental Scales was .88 and for the Motor Scale, .84.

The coefficient of correlation between scores obtained on the Bayley

and the Binet was .57, with considerable restriction on the range of

scores on both tests, especially the Binet.

The mental scale is designed to assess sensory-perceptual acuities,

discrimination, and the ability to respond to these; the early acqui-

sition of "object constancy" and memory, learning and problem-solving

ability; vocalizations and the beginnings of verbal communication; and

early evidence of the ability to form generalizations and classifica-

tions. Results of the Mental Scale are expressed as a standard score,

the MDI, or Mental Developmental Index.

The Motor Scale is designed to provide a measure of the degree

of control of the body, coordination of the large muscles and finer

manipulatory skills of the hands and fingers. The Motor Scale is

specifically directed toward behaviors reflecting motor coordination

and skills and is not concerned with "mental" or "intelligent" func-

tions. Results of administration of the Motor Scale are expressed








as a standard score, the PDI, or Psychomotor Development Index

(Bayley, 1969, p. 3).


Data Collection

The data for this study were obtained from medical files at the

Children's Developmental Services Project at Shands Teaching Hospital,

Gainesville, Florida. Sociodemographic data on the adolescent mothers

were taken from the personal questionnaire administered to the mother

when her infant was six months old. These data consist of 1) measures

of communication efforts made by the support network during the in-

fant's hospitalization, 2) measures of the mother's accessibility to

her support network, 3) measures of the mother's actual support re-

sources, and 4) measures of the mother's perception of her support

network. In addition, two other secondary variables were studied

descriptively. They were 1) measures of the mother's self-concept

and 2) the mother's life stress score for the previous year. Data on

the infants were obtained from medical files at Shands Teaching

Hospital. Bayley Scales for Physical and Psychomotor Development

were obtained, and these scores were taken from administration of the

Bayley on the infants at 12 months of age. The data were obtained

from complete medical folders from January 1, 1979, through December 31,

1980. The researcher signed a confidentiality oath in conjunction

with the Children's Research Project. All subjects gave approval for

the use of these data, with confidentiality assured by the researchers.








Data Analysis--Regression Analysis

Regression analysis is a statistical tool for evaluating the

relationship between one or more independent variables, X1, X2,

X to a single continuous dependent variable Y. Practical

applications of regression analysis include the following:

1) When the investigator wishes to characterize the relationship

between the dependent and the independent variables in the sense of

determining the extent, direction, and strength of the association

among these variables.

2) The researcher desires to describe quantitatively or quali-

tatively the relationship between X1, X2, Xn and Y, while con-

trolling for the effects of other variables C1, C2, C which

may have an important relationship with the dependent variables.

3) The researcher wants to determine which of several independent

variables are important and which are not for describing or predicting

a dependent variable. Also, the researcher may desire to rank in-

dependent variables in their order of importance (Kleinbaum and

Kupper, 1978, p. 34-35).

Independent Variables in this Study


As indicated earlier in this chapter, the four independent

variables used in this study were made up of several measures, each

with the exception of the first independent variable, which was a

measure of support during the hospitalization of the infant; this

measure consisted of one measure--the total number of visits by the

family during the infant's hospitalization. Each of the three other







independent variables consisted of three or more measures. These

four variables and the measures included in each one are now presented.


Hypothesis I: Communication Efforts by the Support Network During

the Infant's Hospitalization.

X1 = total number of visits made by the support network during

the infant's hospitalization in the NICU.

Hypothesis II: Measures of the Adolescent Mother's Accessibility

to her Support Network.

X2 = adolescent mother's access to a car.

X3 = adolescent mother's access to a telephone.

X4 = adolescent mother's perception of difficulty to get out

and make contact with her support network.
Hypothesis III: Measures of Adolescent Mother's Actual Support

Network.

X5 = number of relatives/friends living in the immediate area

(adults).

X6 = number of relatives/friends living in the immediate area

that can be counted on in times of real need.
X7 = measures of frequency with which adolescent mother gets

together with relatives/friends.

X8 = frequency, per week, of mother's contact with support network.

X9 = number of organizations to which adolescent mother belongs.

Hypothesis IV: Measures of Adolescent Mother's Perception of

Adequacy of Support Network.

X10 = measure of mother's perception of adequacy of support network.

Xll = mother's perception of being overwhelmed with household

tasks and childcare responsibilities.








X12 = mother's perception of frequency of time to be by herself.

X13 = mother's estimation of adequacy of time to be by herself.

X14 = mother's perception of present living situation as happy.

Secondary Variables for Descriptive Study


Five measures of self-concept were studied for descriptive pur-

poses. These included the mother's perception of her health, physical

fitness, body shape, weight and life in general. The second variable,

life stress, was studied by obtaining a life stress score from the

questionnaire completed by the mothers. This score indicated the

types of stressful events the mother had experienced during the past

12 months prior to completing the questionnaire. A total stress

score was also obtained. See Appendix for example of questions for

these two secondary measures.

Dependent Variables


Y1 = Mental Development Index from the Bayley Scales of
Infant Development, administered to the infant at 12 months

of age.

Y2 = Physical Development Index from the Bayley Scales of
Infant Development, administered to the infant at 12 months

Control Variables

C1 = mother's age, in years.

C2 = mother's level of education, in years.

C3 = birthweight of infant, in grams.

C4 = infant's length of stay in hospital, in days.







These four control variables were included in the initial inter-

correlational matrix.


Regression Models for this Study

After an initial analysis which included the four control vari-

ables cited above and which included performing reduced, full, and

partial analyses, the researcher decided to eliminate the four con-

trol variables. None of the four variables were significantly related

to the dependent variables; therefore, it was not helpful to include

them in the final regression analysis. The final regression analysis

consisted of performing regression analysis on simple models and doing

a Max-R and a stepwise regression analysis which included all of the

independent variables.

Listed below are the regression models used in this study.

Hypothesis I: There will be no significant linear relationship be-

tween measures of communication efforts made by the support

network during the infant's hospitalization and the MDI and

PDI scores attained by the infant at 12 months of age.

Regression Model Utilized: Y = B0 + B1X1 + E


Hypothesis II: There will be no significant linear relationship

between measures of the mother's accessibility to her support

network and the MDI and PDI scores attained by the infant at

12 months of age.

Regression Model Utilized: Y = B0 B2X2 + B3X3 + B4X4 + E








Hypothesis III: There will be no significant linear relationship

between measures of the mother's actual support network re-

sources and the MID and PDI scores attained by the infant at

12 months of age.

Regression Model Utilized: Y = B0 + B5X5 + B6X6 + B7X7 + B8X8


+ B2X9 + E

Hypothesis IV: There will be no significant linear relationships be-

tween measures of the mother's perception of her support and

the MDI and PDI scores attained by the infant at 12 months of

age.

Regression Model Utilized: Y = B0 + B10X10 + B11X11 + B12X12


+ B13X13 + B14X14 = E

Analysis Procedure

The initial intercorrelational study indicated that none of the

four control variables were significantly related to either of the

two dependent variables in the study--the MDI and the PDI scores of

the infants at 12 months of age. A decision was made to delete the

four control variables from the remaining analysis. Therefore, the

use of simple linear regression models was indicated.

A simple linear regression model was developed in order to test

each of the four major hypotheses. Three of the four regression

equations consisted of three or more measures, grouped together to

form a multiple correlation coefficient. Therefore, multiple








regression was utilized in testing Hypotheses II, III, and IV. The

F-values and levels of significance were determined in order to test

the four hypotheses.


Limitation of Methodology

Correlational study has been utilized extensively in educational

research. It is especially helpful because it allows the researcher

to study the extent of relationship existing between variables.

Correlational studies enable one to determine the extent to which

variations in one variable are associated with variations in another

(Ary, 1979).

There are limitations to correlational studies. Correlational

findings indicate relationships between or in association with vari-

ables but do not imply a causal relationship. One cannot interpret

findings in correlational studies to be "cause and effect." Sec-

ondly, the coefficient of correlation is not to be interpreted as an

absolute fact. Values of r are found for sample cases,and the extent

of relationship found in one sample need not necessarily be the same

found in another sample from the same population. Additional limita-

tions to this study include those associated with the measurement

procedure. Data were collected from the adolescent mothers in a

self-report format. Dichotomization of responses may have limited'

the accuracy of the data. The sample size is also a limitation. Re-

strictions in the range of data and the reliability of correlation

coefficients are limitations due to small sample size.








Summary

Bronfenbrenner (1974) and Cochran and Brassard (1979) have

argued that the social support network of the family has a major

influence on the development of the child. Relatively few studies

have focused on measures of the family's support system and specific

measures of child development. This study focused on indicators of

social support of single adolescent mothers and their relationships

to specific measures of cognitive and physical development of their

infants. Data on the social support system of the mothers were taken

from personal interviews of the mothers which were administered to

the mother when her infant was six months old. Data on the infants

are measures of cognitive and psychomotor development as determined

by the Bayley Scales administered when the infants were 12 months old.

Linear regression and stepwise regression procedures were

utilized to study the relationship between measures of infant cog-

nitive and psychomotor development and specific measures of social

support networks of the mothers.













CHAPTER IV
RESULTS

Preliminary Analysis

The purpose of this study was to investigate 1) the types of

support networks utilized by single adolescent mothers and the fre-

quency with which these support networks are used, 2) the relation-

ship of four primary measures of support network resources to cognitive

and psychomotor measures of infants of the adolescent mothers at 12

months of age, and 3) measures of self-concept and life stress of

the adolescent mother subjects.

This chapter includes the following: 1) a description of the

types of support network resources utilized by single adolescent

mothers and the frequency with which these support network resources

were used, 2) a presentation of four major hypotheses which studied

the relationship between the use of support network resources of 35

single adolescent mothers and the Bayley scores of Mental and Physical

Development of the infants at 12 months of age, and 3) a description

of the measures of self-concept and life stress of the adolescent

mother subjects.

The Bayley Scales of Infant Development were administered to

each infant at his/her 12 month evaluation. Data were available on

34 of the 35 infant subjects. Table 7 shows that scores ranged from

50 to 150 on the Mental Development Index (MDI) and from 29 to 138 on







the Physical Development Index (PDI). Nearly one-third of the MDI

scores fell into the 91-110 range, which is considered a "normal" or

average range. Seventeen infants had MDI scores above 110. This

represented one half of the infant population. The mean MDI'score

was 109.5


Infant Scores


Table 7. Distribution of the Bayley MDI and PDI scores
at 12 months of age.


of infants


MDI PDI
Scores Frequency Percentage Frequency Percentage

Data
unavailable 1 3.00 1 3.00
50 or less 1 3.00 2 6.00
51-70 2 6.00 1 3.00
71-90 2 6.00 4 11.00
91-110 12 34.00 15 43.00
110-130 13 37.00 10 29.00
131 or more 4 11.00 2 6.00
Total 35 100.00 35 100.00



Infants scored more toward the average on the Physical Development

Index. Twenty-five scored between 91 and 130. Only two scored above

the 130 point. This indicated that infant scores fell into more of an

"average" range on PDI scores and more "above average" on MDI scores.

The mean PDI score was 99.4, 10 points below the mean of the MDI scores.





67

Hospital Visits


Data on hospital visits made by the family network during the

infant's hospitalization were available on 30 of the 35 infant sub-

jects. Fourteen infants, nearly one-half, received six to ten visits

during their hospitalization. Eight received five or less visits

while eight received more than ten visits. A visit was recorded each

time a member of the immediate family or combination of members of

the immediate family made a visit to the infant. Data were collected

by the Neonatal Intensive Care Unit staff. See Table 8.


Table 8. Number of visits made by infant's immediate family during
infant's hospitalization.

Number of visits Frequency Percentage

No response 5 14.00
0 to 5 8 23.00
6 to 10 14 40.00
11 to 15 4 11.00
16 to 20 1 3.00
21 to 25 3 9.00
Total 35 100.00



The mean number of visits made by support network members was 10 with

a range of 1 to 22. The number of days in the hospital ranged from

10 to 82 days with an average of 31 days, as compared with an average

of 5 days for normal infants.







Availability of Support Resources

Several questions sought to determine the availability of support

resources to the single adolescent mother. Three questions asked were

1) do you have accessibility to a car? 2) do you have accessibility

to a telephone? and 3) do you think it is difficult for you to get

out and do the things you need to do?

Of the 35 respondents, 21 said they did not have accessibility

to a car. Twenty-four of the 34 subjects responded that they did

have accessibility to a telephone. This represents 70 percent of the

total number of subjects, as shown in Table 9.

When asked if it were difficult to get out to do the things they

needed to do, 20 of the 34 respondents answered "yes." This rep-

resents 59 percent of the subjects.


Actual Support Resources

Several questions sought to determine the extent of actual support

resources available to single adolescent mothers in regards to rela-

tives and friends. All 35 subjects responded to the question, "How

many relatives and/or friends live in the immediate area (adult)?"

Fifteen subjects responded that they had five or fewer relatives and/or

friends in the immediate area. Of these, four reported having no

adult friends and/or relatives in the immediate area. Six responded

that they had from 6 to 10 relatives and/or friends living in the

area, and 9 answered that they had 15 or less relatives and/or friends

in the immediate area. Two responded that they had between 26 and 30,

and only one subject reported having more than 35 relatives in the

area. See Table 10 for these data.



















LD LC)








00
0 0
0 CO


LO








0
0


0 r-


o 0
o*
0 0Y


0
F-


4-.)
C









c"
C


0-



S.,



C
cv
(U
L


>.-












ct
0-
(u







(,





c,




0-

LL.


0 0
0 r-
I-








r- C
b. h



F- cb


O

01 4-) 4-
C ,-3
o o
0 0
0. *i-*J
) 4- U
S- 4 0
r 0 *e-
0. I- 0


r- O -
C r- r-







Table 10. Actual measures of support resources: Relatives in the
area (adult).


Number of adult
relatives and/or
friends in the area Frequency Percentage

0 4 11.00
1 to 5 11 31.00
6 to 10 6 17.00
11 to 15 9 26.00
16 to 20 2 6.00
21 to 25 0 0.00
26 to 30 2 6.00
31 to 35 0 0.00
more than 35 1 3.00
Total 35 100.00


Support During Crisis


When asked how many of these relatives and/or friends could

be counted on in times of crisis, 24 of the 33 respondents responded

with "0-5." This represents 73 percent of the total group.


Frequency of Interaction


When the adolescent mothers were asked if they got together

often with relatives and/or friends, 30 of the 34 subjects who re-

sponded said "yes," only 4 said "no." This represents 88 percent,

a high percentage of the total group, as shown in Table 12.







Table 11. Number of relatives and/or friends mother can count on
in times of crisis.

Number of adults
mother can count
on in crisis Frequency Percentage

0 to 5 24 69.00
6 to 10 6 17.00
11 to 15 2 6.00
16 to 20 0 0.00
21 to 25 0 0.00
26 to 30 0 0.00
31 to 35 1 3.00
No response 2 6.00
Total 35 100.00



Table 12. Does mother get together with relatives/friends often.


Response Frequency Percentage

Yes 30 86.00

No 4 11.00

No response 1 3.00

Total 35 100.00


Frequency of Interaction Per Week


The adolescent mothers were asked how many times per week they

got together with relatives or friends. Responses were collapsed into

the dichotomous responses of "not often" and "often." Those who re-

sponded with "never, rarely, or sometimes" were categorized into "not







often." Those who responded with a "few times per week" and "very

frequently/every day" were categorized into "often."

Of the 35 mothers who responded, 23 said "not often" (never,

rarely, or sometimes). This represents nearly 66 percent of the

responses; these responses seem to contradict the answer given in the

previous question, which was a general question as to whether the

adolescent got together with relatives and/or friends often.- It seems

that when the questions became more specific or focused, the mothers

responded that, in fact, they did not get together with the relatives

and/or friends very many times per week. Table 13 illustrates these

findings.


Table 13. Times per week mother gets together with friends/relatives.


Response Frequency Percentage

Not often 23 66.00

Often 12 34.00

Total 35 100.00


Organizations

Organizations are thought to be positive forms of support for

most people. This question sought to identify how many different

organizations the adolescent mothers participated in. The most common

organization identified by this population was the church. Twelve

adolescent mothers responded that they participated in a religious

organization or church. Only two reported being active in social







organizations, one in an educational organization, and none reported

participating in a political organization. Twenty adolescent mothers

did not participate in any type of organization, representing over

one-half of the subjects. See Table 14.


Table 14. Types of organizations mothers participated in.


Type of organization Frequency* Percentage

No participation 20 54.00
Social organization 4 11.00
Religious organization 12 32.00
Political organization 0 00.00
Educational organization 1 3.00
Total 37 100.00

*One subject participated in more than one organization.


Perceptions of Support

Five questions sought to measure the adolescent mother's per-

ceptions of her support network. When asked if their present situa-

tion provided them with enough support, 21 out of the 32 responding

answered that they perceived their present situation to have enough

support. That represents 65 percent or nearly two-thirds of the

responses.

Thirty-one members responded to the question, "Do you feel over-

whelmed with household tasks and children?" Twenty-two said "yes"

while only nine said "no." Therefore, over 70 percent of these sub-

jects indicated that they felt overwhelmed with household tasks and

children.







Another question, "Do you have time to do things by yourself

that you would like to do?" had 28 respondents. Half of these mothers

indicated that they were able to get away by themselves two or less

times per month. Nearly half indicated that they got away from three

to five times per month.

When the adolescent mothers were asked if they thought this was

enough time (in reference to the previous question), fully two-thirds

of the respondents replied that they had enough time to get away by

themselves each month. This indicates a perception that there are

sufficient resources to allow them to do this.

The adolescent mothers were asked if their present living situa-

tions were happy. Thirty-two adolescent mothers responded to this

question. Nearly 60 percent described their present living situa-

tions as being happy. In summary, five questions related to the

mother's perception of her support were asked. The adolescent mothers

responded favorably to all questions, except for one. "Do you feel

overwhelmed with household tasks and children?" Over 70 percent

responded with "yes" to this question. See Tables 15, 16, and 17 for

these data.


Secondary Independent Variables

Data were collected on two secondary variables: adolescent

mother's self-concept and her life stress. The researcher was curious

about the young mothers' perceptions of themselves (self-concept5) and

about the degree of life stress they had experienced during the year

previous to their completion of the questionnaire. While the major

purpose of collecting these data for the study was descriptive,























4-,


S-




0)




-4_
QJ













4.

U
a,
0-















a)
0"

II-


0-


S.-
au


Q,


U
aJ


,.


Ca)
C i-


0
I.C



4'





0e E
S-U ( I

0 t-







Mc:-e S
(U03,
so T..
>- .<, s.


C

L.-

3C ..- C

0 0 -)-.

U) 4-)
S- 0 'r-
#D


o 0
o 0
nI-















o 0o
o o

r Lf


C%-


0
O a"
Q 1




0 0m
o>
0
0 a
0m CU







Table 16. Times per month to get away by yourself.


Times per month Frequency Percentage

Data unavailable 7 ;20.00
0- 2 14 40.00
3- 5 12 34.00
6- 8 1 3.00
9-11 1 3.00
Total 35 100.00



Table 17. Is this enough time to get away by yourself.


Response Frequency Percentage

Data unavailable 5 14.00
Yes 20 57.00
No 10 29.00
Total 35 100.00



further research is suggested utilizing these data in relationship

to other variables.


Mother's Self-Concept


Five measures of self-concept were obtained from the adolescent

mothers. A scale from 1-7 on five different measures was used. This

was a continuous scale ranging from 1-7. Higher scores indicated a

lower self-concept. Twenty-seven of the 35 adolescent mothers re-

sponded to the measure of self-concept regarding health. Over 40

percent rated themselves as having excellent health and 22 of the 27







ranked themselves in the top four ratings for health. Only one sub-

ject perceived herself as having poor health. In general, these

subjects rated themselves as being in good to relatively good health.

The mean health score was 2.26. See Table 18.


Table 18. Mother's perception of her own health.


Score Frequency Percentage

Data unavailable 8 23.00
1 (excellent) 12 34.00
2 4 11.00
3 6 17.00
4 4 11.00
5 0 0.00
6 0 0.00
7 (poor) 1 3.00
Total 35 100.00


When asked about her perception of her physical well-being, 28

subjects responded. Over 90 percent rated themselves from excellent

to good on perception of well-being. This reflects a relatively

positive sense of physical well-being. The mean score on physical

well-being was 2.63. See Table 19.

As shown in Table 20, 25 subjects responded to the measure of

perception of her own body. Of the 25 respondents to this question,

nearly one-third rated perception of their bodies as excellent, and

88 percent rated their body perception in the four most positive

rankings. This also reflects a relatively positive body image. The







Table 19. Mother's perception of her physical well-being.


Score Frequency Percentage

Data unavailable 8 23.00
1 (excellent) 7 20.00
2 6 17.00
3 6 17.00
4 7 20.00
5 0 0.00
6 1 3.00
7 (poor) 0 0.00
Total 35 100.00



Table 20. Mother's perception of her own body.


Score Frequency Percentage

Data unavailable 10 29.00
1 (excellent) 8 23.00
2 6 17.00
3 4 11.00
4 4 11.00
5 2 6.00
6 1 3.00
7 (poor) 0 0.00
Total 35 100.00



mean score for the adolescent mothers' perceptions of their own

bodies was 2.52.

When asked about their perceptions of their weight, 27 adolescent

mothers responded to this self-concept item. Over 25 percent said







that they perceived their weight to be excellent and nearly 70 percent

responded that they perceived their weight to be very good to excellent.

In terms of perceived weight, these adolescent mothers held very

positive perceptions, as shown in Table 21. The mean score was 3.00.


Table 21. Mother's perception of her own weight.


Score Frequency Percentage

Data unavailable 8 23.00
1 (excellent) 7 20.00
2 5 14.00
3 7 20.00
4 3 9.00
5 2 6.00
6 0 0.00
7 (poor) 3 9.00
Total 35 100.00



The adolescent mothers were asked how they perceived their lives

in general. Of the 27 respondents, over one-third said that they

felt very positive about their lives in general. While 70 percent of

the responses fell into the more positive categories, it is interest-

ing to note that three of the mothers responded that they perceived

their lives to be in a "poor" state. However, the mean score was

3.03, the highest average of any of the measures of self-concept. See

Table 22 for these results.

In summary, an attempt to measure the adolescent mother's self-

concept was made by asking them to characterize their health, body







Table 22. Mother's perception of her life in general.


Score Frequency Percentage

Data unavailable 8 23.00
1 (excellent 10 29.00
2 4 11.00
3 1 3.00
4 5 14.00
5 4 11.00
6 0 0.00
7 (poor) 3 9.00
Total 35 100.00



image, weight, physical well-being, and life in general. This was

done by having the subjects indicate on a continuum from 1-7 their

own perceptions of their self-concepts. In general, these single

adolescent mothers described themselves in very positive terms.

Three subjects marked their weight and life status in general as

being "poor." However, a much larger proportion of the subjects

marked responses that fell into the top four categories. The overall

mean score for the five measures of self-concept was 2.71.

Life Stress


Another secondary variable was a measure of life stress.

Twenty-five mothers completed the Holmes Stress Scale (1973) which

was a part of the questionnaire completed by the subjects. As shown

in Table 23, the stress scores were evenly distributed. There was a

range of 5 to 260. Twenty percent fell into the 161-180 category.

Most of the stress scores were moderate. The mean score was 119.72.








Table 23. Stress scores of adolescent mothers.


Score Frequency Percentage

Data unavailable 10 29.00
0- 20 2 6.00
21- 40 1 3.00
41- 60 1 3.00
61- 80 1 3.00
81-100 5 14.00
101-120 2 6.00
121-140 4 11.00
141-160 2 6.00
161-180 5 14.00
181-200 0 0.00
201-220 0 0.00
221-240 1 3.00
241-260 1 3.00
Total 35 100.00




Regression Analysis

Regression analysis was utilized in this study. Several prelimin-

ary steps were taken to analyze the dependent and independent vari-

ables. The initial procedure was to obtain an intercorrelational matrix,

using both dependent variables--Mental Development Index scores and

Psychomotor Development Index scores--and all independent variables,

which were measures of support network resources of the single ado-

lescent mothers. The independent variables were all included in the

initial correlational matrix. Specifically, the four primary categories

were 1) Measures of Communication Efforts by the Support Network







During the Infant's Hospitalization, 2) Measures of the Adolescent

Mother's Accessibility to her Support Newtork, 3) Measures of Actaul

Support Resources Available to the Adolescent Mothers, and 4) Measures

of the Adolescent Mother's Perception of Her Support. Two secondary

measures related to the adolescent mother's social support were in-

cluded; they were 1) Measures of the Adolescent Mother's Self-Concept

and 2) Measures of the Adolescent Mother's Life Stress.

The correlational matrix represents correlations between the

dependent variables, between the independent variables, and between

the dependent and independent variables. This preliminary correla-

tional study was done in order to better understand the relationships

that existed among all the variables, both dependent and independent.

The researcher was especially interested in learning 1) which indepen-

dent variables were most highly correlated with the two dependent

variables and 2) which independent variables were most highly cor-

related with one another. If independent variables were highly cor-

related to one another, then some independent variables would be

eliminated in order to increase power in the analysis.

The intercorrelational matrix code and the matrix are seen

below. Note that the code listing includes the two dependent vari-

ables and the fourteen independent variables which were utilized in

the hypothesis testing. The intercorrelational matrix is read by

reading down the left-hand column initially and then reading to the

right to determine the correlation between the two variables in

question. For example, MDI has a correlation of 1.00 with MDI,

162 with the PDI score of the infant, .33 with the number of

visits by the family, .17 with the availability of a car by the







adolescent mother, and so on. Correlations which were significant

at the .05 level of probability or less are noted with an asterisk.

Correlational matrices are helpful in providing an overview of the

relationships that exist among the dependent and independent

variables.


Correlational Code and Matrix


CORRELATIONAL MATRIX CODE

MDI = Infant's score on the Bayley Mental Development Index.

PDI = Infant's score on the Bayley Psychomotor Development
Index.

NVF = Number of visits by family during infant's hospitalization
at birth.

AVCAR = Availability of a car for the mother.

ATEL = Mother's access to a telephone.

DIFOUT = Mother's estimation of difficulty to get out to do the
things she needs to do.

NREL = Number of adult relatives/friends in the immediate area.

COUNT = Number of adult relatives/friends in the immediate area that
can be counted on in crisis.

TOG = Mother's estimation of general frequency of her interaction
with adult friends/relatives.

FREQ = Frequency per week of mother's interaction with adult
friends/relatives.
TORG = Total number of organizations the mother participates in.

ENSUPP = Mother's perception of adequacy of support system.

OVWH = Mother's perception of being overwhelmed by children and
household tasks.

GETMH = Mother's perception of times to get away by herself per month.

SUFTIM = Mother's perception of the adequacy of her time to be alone.

HAPGEN = Mother's perception of her happiness with her life situation.






















) o R o R O .


I I


S00
00
























II
N
aM






S01














ia





M o






NC C






I I



































LI I







Control Variables


A general linear regression analysis was completed in order to
ascertain which, if any, control variables should be included in the

final regression analysis. Four control variables were initially

identified, based on previous research on premature infants born to

adolescent mothers (Holstrum, 1979; Hofheimer, 1979). The four con-

trol variables investigated were 1) Mother's age at birth of infant

mageE); 2) Mother's level of education, in years (YED); 3) Infant's

birthweight, in grams (BWT); and 4) Number of days in the hospital

(infant) (NDH). As can be seen in Table 25, the following correlations

were obtained.


Table 25. Correlations of control and dependent variables.

Control variable MDI P-value PDI P-value

1) Mother's age -.18 .33 -.09 .60
2) Mother's level of education -.09 .62 -.31 .08
3) Infant's birthweight .20 .25 .05 .78
4) Number of days in hospital -.19 .32 -.09 .64



None of the four control variables were related significantly to the

two dependent variables. Therefore, the researcher made a decision

to eliminate the four control variables from further analysis.

In addition, a correlational matrix was obtained using quadratic
forms of the four control variables to determine if a quadratic rela-

tionship existed between the four control variables and the two depen-

dent variables, therefore accounting for more of the variance in the







MDI and PDI scores. As shown in Table 26, none of the four control

variables was significantly correlated with the two dependent vari-

ables when quadratic forms of the control variables were used. There-

fore, the researcher decided to 1) eliminate all four control variables

in the regression study of the support network measures with the MDI

and PDI scores and 2) to allow a linear relationship to explain the

relationship between the independent variables and the two dependent

variables.


Table 26. Correlation coefficients of four control variables in
quadratic form.


N R P-value

For MDI

1) Mother's age 32 .18 .33
2) Number of days in hospital 29 -.26 .18
3) Birthweight of infant 33 .23 .19
4) Maternal years of education 32 -.15 .41

For PDI

1) Mother's age 32 -.20 .26
2) Number of days in hospital 29 -.08 .66
3) Birthweight of Infants 33 .06 .75
4) Maternal years of education 32 -.33 .07


Hypotheses Tested

Four primary hypotheses were tested. The four primary hypotheses

tested related to measures of support network resources. These four

hypotheses and the results of the regression analyses are now presented.







HYPOTHESIS I: There will be no significant linear relationship
between measures of in-hospital communication ef-
forts by the support network and the Bayley scores
of Mental and Psychomotor Development at 12 months
of age.

For each mother-infant dyad, the number of actual visits made by

the immediate family during the infant's hospitalization was obtained.

These visits were limited to only the immediate family and were closely

monitored and recorded by the NICU (Neonatal Intensive Care .Unit) staff.

The total number of visits made by the support network was designated

as NVF. This independent variable was entered into the overall

regression model--


MDI PDI = Number of Visits to the Infant by Support Network
During Hospitalization.

Table 27 lists the R2 for each of the simple regression models tested

for MDI and PDI. The independent variable used in this analysis is a

measure of the adolescent mother's support during the infant's hos-

pitalization. More specifically, the total number of visits made by

members of the support system was obtained and used as a measure of

support for the adolescent mother.

As noted in Table 27, the simple regression model using Number of

Visits by the Support Network members yielded a very small variance

for both dependent variables, MDI and PDI. Neither of the dependent

variables showed a statistically significant relationship with the

independent variable, Number of Visits by the Family.

In summary, Hypothesis I was not rejected. There was no signifi-

cant and linear relationship between the number of visits made by the

family during the infant's hospitalization and the infant's scores on

the Bayley Mental and Physical Development Index at 12 months of age.







Table 27. Linear regression of MDI and PDI by number of visits by
support network family members during the infant's
hospitalization.

Regression model analyzed N R2 F P > F

MDI = Number of visits 24 .0005 .01 .91
PDI = Number of visits 24 .0045 .10 .74



HYPOTHESIS II: There will be no significant linear relationship be-
tween measures of the mother's accessibility to her
support network resources and her infant's scores of
development on the Bayley Mental and Psychomotor
Index at 12 months of age.
Hypothesis II tested the association of measures of the mother's

accessibility to her support network with her infant's scores of devel-

opment on the Bayley Mental and Psychomotor Index at 12 months of age.

Regression models were analyzed for the simple linear models of Mental

Development and Psychomotor Development with the three measures of the

mother's accessibility to her support network. (See Appendix for ex-

amples of these three measures.) As shown in Table 28, the proportions

of variance explained by the regression of both MDI and PDI were both

small, .08 and .10, respectively. The F values obtained from these

analyses were both nonsignificant. Therefore, the results of the

linear regression analysis failed to reject Hypothesis II. There was

no significant linear relationship between measures of the mother's

accessibility to her support network and her infant'sscores of

development on the Bayley Mental and Psychomotor Development Index at

12 months of age.







Table 28. Linear regression of MDI and PDI by measures of the
mother's accessibility to her support network resources.


Regression Model Analyzed N R2 F P > F

MDI = Three measures of
accessibility* 32 .08 .84 .482
PDI = Three measures of
accessibility 32 .10 1.16 .342

*See Appendix for examples of these three measures.


HYPOTHESIS III: There will be no significant linear relationship
between measures of the adolescent mother's actual
support network and her infant's scores on the
Bayley Mental and Psychomotor Index at 12 months of
age.
As shown in Table 29, the linear regression analysis of the meas-

ures of the adolescent mother's actual support network with the infant's

scores on the Bayley Scale of Infant Development, Mental and Psycho-

motor Indexes, yielded R2's of .026 and .142, respectively. The F

values for both analyses were nonsignificant. Therefore, the re-

searcher failed to reject Hypothesis III. There was no significant

linear relationship between measures of the adolescent mother's actual

support network and her infant's scores on the Bayley Scales of Infant

Development, Mental and Psychomotor Index, at 12 months of age.

Table 29. Linear regression of MDI and PDI by measures of the adoles-
cent mother's actual support network resources.

Regression Models Analyzed N R2 F P > F

MDI = Five measures of actual support* 32 .026 .15 .979
PDI = Five measures of actual support 32 .142 .90 .497

*See Appendix for examples of these measures.







HYPOTHESIS IV: There will be no significant linear relationship
between measures of the adolescent mother's percep-
tion of her support network and the infant's scores
on the Bayley Mental and Psychomotor Index at 12
months of age.
Five measures of the adolescent mother's perception of her support

were utilized in this analysis. As shown in Table 30, these five

measures included the mother's perception of the adequacy of her sup-

port, her perception of being overwhelmed with household tasks and

children, her perception of her amount of time to be by herself, the

adequacy of that time to be by herself, and her perception of her own

happiness in regards to her living situation.

The linear regression of the MDI scores with the five measures of

the mother's perception of her support yielded an R2 of .25 and an

F value of 1.17. This finding was nonsignificant. Therefore, Hypothe-

sis IV, as related to the MDI, was not rejected. There was no signifi-

cant linear relationship between the MDI scores and the five measures

of the mother's perception of her support network.

The linear regression of the PDI scores with the five measures of

the mother's perception of her support network yielded an R2 of .53

with an F value of 4.02, significant at the .012 level. Therefore,

Hypothesis IV, as related to the PDI regression analysis, was rejected.

There was a significant relationship between the PDI scores of the

infants and the five measures of the mother's perception of her support

network.

Listed in Table 30 are the Type I and Type IV Sums of Squares.

Results for special Type I and Type IV Sums of Squares provided useful

information. The Type I Sums of Squares measured incremental sums of

squares for the regression model as each of the five variables was







added. The Type IV Sums of Squares was the sum of squares due to

adding each variable last in the model.


Table 30. Linear regression of MDI and PDI by measures
mother's perception of her support network.


of adolescent


Regression models analyzed N R2 F P > F

MDI = Five measures of mother's
perception of her support 23 .25 1.17 .363
PDI = Five measures of mother's
perception of her support* 23 .53 4.02 .012**

Type I Sums of Squares


Variable Sums of Squares F-value P > F

GETMH 166.105 1.03 .322
SUFTIM 394.888 2.46 .134
ENSUPP 1144.50 7.13 .016**
OVWH 1209.29 7.53 .013**
HAPGEN 312.86 1.95 .80

Type IV Sums of Squares

Variable Sums of Squares F-value P > F

GETMH 158.46 .99 .33
SUFTIM 3.16 .02 .89
ENSUPP 175.23 1.09 .31
OVWH 1018.97 6.35 .02**
HAPGEN 312.86 1.95 .18

*See Appendix for examples of these measures.
**Significant at the .05 level.


As noted, two variables had F values of significance in the

Type I Sums of Squares. The mother's perception of the adequacy of

her support (ENSUPP) yielded an F-value of 7.13, significant at the

.01 level. The second variable, a measure of whether the mother felt







overwhelmed by her childrearing responsibilities and household tasks,

(OVWH), yielded an F-value which was 7.53, also significant at the

.01 level.

In the Type IV Sums of Squares, only one variable yielded an

F-value which was significant; the variable OVWH, the mother's percep-

tion of being overwhelmed by childrearing responsibilities and house-

hold tasks, yielded an F value of 6.35, significant at the .02 level.


Stepwise Regression Analyses

Two stepwise regression procedures were conducted. In the first

stepwise analysis, a regression model was entered that included both

dependent variables, MDI and PDI, and all of the independent variables.

A "Max-R" procedure was conducted in which all of the independent

variables were sorted according to maximum contribution to the variance

of each of the two dependent variables. In this procedure, the com-

puter sorted all of the independent variables in many different com-

binations, arranging the variables in a descending order of contribution

to the variances of MDI and PDI. Listed in Table 31 are the best

models as determined by the Max-R procedure.

The best models for independent variables in relation to PDI

(Physical Development Index) scores were also obtained. Again, the

regression model entered included the dependent variable, PDI, and

all of the independent variables. This sorting procedure yielded a

series of models which accounted for the greatest amount of variance

in PDI scores. Table 32 illustrates these results.