Citation
A comparison of successful and unsuccessful adoption seekers based on a national sample

Material Information

Title:
A comparison of successful and unsuccessful adoption seekers based on a national sample
Creator:
Holzheuer-Guinan, Melanie
Publication Date:
Language:
English
Physical Description:
ix, 46 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Adoptive parents ( lcsh )
Adoption ( lcsh )
Adoption ( fast )
Adoptive parents ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 45-46).
General Note:
Department of Sociology
Statement of Responsibility:
by Melanie Holzheuer-Guinan.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
71778686 ( OCLC )
ocm71778686
Classification:
LD1193.L66 2006m H64 ( lcc )

Full Text
A COMPARISON OF SUCCESSFUL AND UNSUCCESSFUL
ADOPTION SEEKERS BASED ON A NATIONAL SAMPLE
by
Melanie Holzheuer-Guinan
A.A., Community College of Denver, 2002
B.A., University of Colorado at Denver, 2003
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Sociology


2006 by Melanie Holzheuer-Guinan
All rights reserved.


This thesis for the Master of Arts
degree by
Melanie Holzheuer-Guinan
has been approved
by
Andrea Haar
tjliqloy
Date


Holzheuer-Guinan, Melanie (M.A., Sociology)
A Comparison of Successful and Unsuccessful Adoption Seekers Based on a
National Sample
Thesis directed by Associate Professor Candan Duran-Aydintug
ABSTRACT
Since the 1930s, researchers have conducted several investigations in order to
describe the characteristics of individuals who support, consider, seek or
sought and already experienced adoption. Important topics that have emerged
from the previous literature relate to age, marital status, marriage, parity,
education, income, race, and labor force status. This study uses data from the
1995 National Survey of Family Growth to compare the characteristics of
individuals who go through the adoption process and succeed with those who
do not, an area that has not been formerly studied. Results suggest that
women who successfully adopt a child are more likely to be older including
their age at certain stages associated with fertility values -, to have longer and
more stable marriages, to desire more children, to have a higher level of
education and income, to be white and they are less likely to work full-time.
Although the difference in marital status is not statistically significant, other
evidence suggests that this is due to the small sample size.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed
Candan Duran-Aydintu,


DEDICATION
I dedicate this thesis to my parents and my husband for their unfaltering
understanding and support during my entire academic career.


ACKNOWLEDGEMENT
My thanks to my thesis committee and department chair, Candan Duran-
Aydintug, for her guidance and support during my research. I also wish to
thank the remaining committee members, Andrea Haar and Yili Xu, for their
valuable insights. Last, I would like to acknowledge my appreciation for my
friend Diane, who has kept me levelheaded during this entire process.


TABLE OF CONTENTS
List of Tables...........................................ix
CHAPTER
1. INTRODUCTION...........................................1
2. REVIEW OF THE LITERATURE...............................3
Age.................................................3
Marital Status......................................6
Marriage............................................7
Parity..............................................8
Education...........................................9
Income.............................................11
Race...............................................13
Labor Force Status.................................15
Eligibility and Selection Criteria.................16
3. METHODS...............................................18
Data...............................................18
Sample, Sample Characteristics and Sample Size.....19
Variables and Measures
.23


Statistical Procedures
,25
4. RESULTS............................................27
Chi-Square Tests................................27
Independent Samples T-Tests................... 31
Logistic Regression.............................34
5. DISCUSSION AND LIMITATIONS.........................37
BIBLIOGRAPHY..............................................45
viii


LIST OF TABLES
Table
3.1 Descriptive statistics for the continuous variables relevant to the issues
of adoption.............................................................20
3.2 Frequency table for the categorical variables used in this investigation.21
4.1 Chi-square test results for adoption versus race.......................27
4.2 Chi-square test results for adoption versus labor force status.........28
4.3 Chi-square test results for adoption versus marital status.............29
4.4 Chi-square test results for adoption versus marital status at first
conception............................................................30
4.5 Chi-square test results for adoption versus parity.....................30
4.6 Independent samples t-test for numeric variables.......................33
4.7 Logistic regression predicting who will successfully adopt............36
IX


CHAPTER 1
INTRODUCTION
Most Americans state they have a favorable opinion of adoption and
think it serves a significant purpose in society (Benchmark Adoption
Survey, 1997). Over a quarter of all ever-married women between the ages
of 18 to 44, or 9.9 million women, have ever considered adopting at some
point during their lives (Chandra, Abma, Maza & Bachrach, 1999). About 16
percent of these women, or 1.6 million, have ever taken steps toward the
adoption process (Chandra, Abma, Maza & Bachrach, 1999). Of those, 31
percent, or 487,000 have adopted at least one child (Chandra, Abma, Maza &
Bachrach, 1999). Thus, the success of adopting, as measured by the
percentage of those who have actually adopted a child out of all the women
who have taken steps toward adoption, is 31 percent.
Previous research has looked at various characteristics of all of these
groups of women including their age, marital status, marriage, parity,
education, income, race, and labor force status. Researchers have compared
some of the groups with each other and they have also compared them to the
1


general public in order to find any differences or similarities. One
comparison, which could answer an extremely important question, has been
neglected thus far: What characteristics determine if an adoption seeker will
have a positive or negative adoption outcome? This study, therefore, extends
past research and fills an important gap in current knowledge by comparing
the characteristics of individuals who go through the adoption process and
succeed with those who do not.
2


CHAPTER 2
REVIEW OF THE LITERATURE
Since the 1930s, researchers have conducted many investigations in
order to describe the characteristics of individuals who support, consider, seek
or sought and already experienced adoption. The researchers have either
compared subgroups of these individuals with each other or they have
compared them to the general public. Important topics that have emerged
from the previous literature relate to age, marital status, marriage, parity,
education, income, race, and labor force status. Since there is no previous
research that compares the characteristics of individuals who go through the
adoption process and succeed with those who do not, the following literature
review will focus on the findings of the individuals formerly studied. Results
of those studies have been ordered by the important topics mentioned above.
Age
The Benchmark Adoption Survey conducted in 1997 found that
adoption attitudes do not vary by age. Hence, young Americans are as likely
3


to be full supporters of adoption as older ones (Benchmark Adoption
Survey, 1997). Age does appear to be a significant factor for individuals
who considered adoption though. Data from the 1995 National Survey of
Family Growth showed that women who ever considered adopting a child are
significantly more likely to be older (Chandra, Abma, Maza & Bachrach,
1999). The belief that age contributes to the propensity to adopt was also
strengthened by data from the 2002 National Adoption Attitudes Survey
(National Adoption Attitudes Survey, 2002 & Hansen, 2005). According to
the survey, the highest percentage of individuals who have considered
adopting a child can be found among those ages 35 to 44 and 45 to 54
(National Adoption Attitudes Survey, 2002).
Age has also a strong positive effect on the odds of having sought
adoption (Bachrach, London & Maza, 1991). Compared to women who ever
considered adoption, women who actually took concrete steps toward
adoption tend to be older (Chandra, Abma, Maza & Bachrach, 1999). The
same appears to be true for adoptive parents. Adoption data obtained in the
1920s in Minnesota showed that adoptive parents are generally older than
parents who have biological children of comparative age (Leahy, 1933).
Bachrach (1983) also found that age at interview is strongly associated with
4


the percentage of women who have adopted. According to her study, adoptive
parenthood is more frequent among those women ages 30 to 44 at the time of
interview (Bachrach, 1983).
In a later study, Bachrach also compared various characteristics of
adopted children and children living with their birthmothers. She discovered
that adopted children have older mothers than children who live with their
birthmothers (1986). While the average age of adopted childrens mothers is
36, the average age of birthmothers is 33 (Bachrach, 1986). In 1989,
Moorman and Hernandez also compared various family structures and found
that in purely adoptive families parents are more likely to be at least 45 years
of age and less likely to be under 35 years of age. Further, in joint biological-
adoptive families parents are also less likely to be under 35 years of age, but
they were more likely than purely adoptive families to have mothers in the 35
to 44 age bracket (Moorman & Hernandez, 1989). The finding that the
prevalence of adoption increases with age was also supported and documented
by data from the 1995 National Survey of Family Growth (Chandra, Abma,
Maza & Bachrach, 1999).
5


Marital Status
According to the Benchmark Adoption Survey (1997), married
Americans are more likely to be full supporters of adoption than those who
are not married. The belief that marital status is associated with the
propensity to adopt was also corroborated by data from the 2002 National
Adoption Attitudes Survey (National Adoption Attitudes Survey, 2002 &
Hansen, 2005). According to the National Attitudes Survey, married couples
are more likely to have considered adopting (National Adoption Attitudes
Survey, 2002). Being currently or formerly married is also positively related
with having sought to adopt (Bachrach, London & Maza, 1991). A more
recent analysis of the 1995 Survey of Family Growth also showed that one of
the factors most closely tied to taking steps among current seekers/planners is
being currently married (Chandra, Abma, Maza & Bachrach, 1999).
A similar pattern emerges from most of the research on adoptive
parents. In 1933, Leahys investigation of 2,414 adoptive parents in
Minnesota showed that the vast majority of people who adopt children are
married. Two studies using data from the 1976 and 1982 Survey of Family
Growth appear to support Leahys findings (Bachrach, 1983 & 1986). These
studies found that adoption is more common among women who are married
6


at the time of the interview (Bachrach, 1983) and that adopted children are
more likely to have a mother who is currently married than children living
with their birthmother (Bachrach, 1986). On the contrary, Bonhams analysis
of the 1973 Survey of Family Growth showed no difference in the adoption
rate of women who were currently married compared to women who were
widowed, divorced or separated (1977).
Marriage
In 1933, Leahy reported that the period of time between the wedding
of adoptive parents and the actual adoption is approximately five or six times
as long as the period between the wedding and the birth of the first child to
parents in general. Further, she speculated that the general population gets
married at a considerably younger age than adoptive parents (Leahy, 1933).
Her hypothesis is supported by Moorman and Hernandez (1989) who found
that parents in purely adoptive families are more likely to have marriages of
long duration than parents in other types of married-couple families. In
addition, parents in joint biological-adoptive families were second most likely
to have long marriages (Moorman & Hernandez, 1989). The same study also
showed that both parents have been married just once in the majority of purely
7


adoptive and joint biological-adoptive families (Moorman & Hernandez,
1989).
Parity
According to a study from 1999, women who have ever considered
adoption are significantly more likely to be nulliparous compared to all ever-
married women (Chandra, Abma, Maza & Bachrach). Having sought to adopt
appears to be more common among childless women as well. Bachrach,
London and Maza (1991) found that childless women are 2.1 times more
likely to have sought than women who have given birth by the time of the
interview. This study also showed that although the desired number of
children is not significantly associated with having sought to adopt, women
who wish to have more children than they are expecting to give birth to, are
substantially more likely to have sought than women who wish to have fewer
(Bachrach, London & Maza, 1991). Further, women who actually took
concrete steps toward the adoption process are different from those who have
ever considered adoption in that they are more likely to be nulliparous
(Chandra, Abma, Maza & Bachrach, 1999).
8


The same trend appears to apply to adoptive parents. In the 1920s,
the majority of adoptive parents in Minnesota were childless (Leahy, 1933).
Later studies have focused on the parity of women instead of the number of
children in the family household. These studies found that adoptive
parenthood is more common among nulliparous women (Bonham, 1977,
Bachrach, 1983, Chandra, Abma, Maza & Bachrach). Despite these findings,
Bonham (1977) remarked that even though women who have borne children
are less likely to adopt than women who have not, their predominance in the
U.S. population means that over half of all women who have adopted a child
have also given birth to a child (p.303).
Education
The 1997 Benchmark Adoption Survey found that Americans with a
college degree are much more likely to be full supporters of adoption than
those with a high school education. In 2002, Hansen also reported that
education level contributes to the propensity to consider adoption. Further,
she observed a statistically significant positive effect for post graduate
education (Hansen, 2002). Although a different study confirmed that
individuals with a graduate degree are slightly more likely to consider
9


adoption (National Adoption Attitudes Survey, 2002), the same study also
questioned the previous findings by claiming that education is not a main
factor in considering adoption. Thus, those with a high school diploma have
the same propensity to consider adoption than those with a college degree
(National Adoption Attitudes Survey, 2002).
Among adoption seekers, the proportion of those who have ever
sought to adopt is positively associated with educational attainment
(Bachrach, London & Maza, 1991). However, regression results show no
significant effect of educational attainment on the odds of having sought to
adopt (Bachrach, London & Maza, 1991). The relationship between
education and adoptive parents appears to be less ambiguous. In 1977,
Bonham found that women who have graduated from college are more likely
to have adopted compared to women with any other educational level. This
finding is supported by Bachrachs analysis (1986) of the 1973 National
Survey of Family Growth, which showed that the proportion of those who
have adopted is significantly higher for women with some college education
than for women who have not completed high school.
When comparing children living with adoptive parents to those living
with birthmothers, Bachrach (1986) noted that adopted children have better
10


educated mothers. The average number of school years completed for
mothers of adopted children is 13.4, it is 12.3 for birthmothers, and only 10.7
for birthmothers who have never been married (Bachrach, 1986). Moorman
and Hernandezs study also showed that mothers in purely adoptive families
are less likely to have completed less than four years of high school and more
likely to have completed at least one year of college (1989). More recent data
from the 1995 Survey of Family Growth also suggested that the prevalence of
adoption increases with education level (Chandra, Abma, Maza & Bachrach,
1999).
Income
In 1999, Chandra, Abma, Maza and Bachrach reported that women
who have ever considered adoption are significantly more likely to have a
higher level of income. The positive relationship between income and interest
in adoption has also been documented by Hansen (2002). Her research
showed that the proportion of women who report that they have ever
considered adopting increases with income; the difference in the average
proportions in the three lowest income groups is statistically different from the
average proportion in the three highest income groups (Hansen, 2002).
11


However, a different study did not support the finding that income has a
positive effect on the probability of adoption seeking. Instead, the results
revealed no significant differences between different income ranges, so the
researchers concluded that income is not an indicator on whether someone
considers adoption (National Adoption Attitudes Survey, 2002).
When examining adoption data in 1933, Leahy discovered that most of
adopted children live in homes of superior economic status. Subsequent
research has only supported this idea. In 1977, Bonham found that as family
income increases, the percentage among women who have adopted increases
as well. When studying the differences among income groups, Bachrach
(1983) also observed that they were statistically significant. Further, women
with family incomes at least twice the poverty level are more likely to have
adopted than women with incomes below the poverty level (Bachrach, 1986).
Comparing various types of family structures, Moorman and Hernandez
(1989) discovered that purely adoptive families and joint biological-adoptive
families are more likely to have greater percentages of comparatively high
incomes and higher median incomes. A more recent study has also confirmed
an increase in the prevalence of adoption with rising income (Chandra, Abma,
Maza & Bachrach, 1999).
12


Race
The Benchmark Adoption Survey conducted in 1997 showed that
regardless of their racial or ethnic background, Americans generally support
adoption (National Adoption Attitudes Survey, 2002). However, Whites are
three times as likely as African-Americans to be fully supportive of adoption
(Benchmark Adoption Survey, 1997). In contrast, a recent study from 2002
revealed that Hispanics are much more likely to consider adopting than
African-Americans and Whites (National Adoption Attitudes Survey).
Hansen (2002) also found that race contributes to the propensity to consider
adoption. According to her research, there is a statistically significant positive
effect for being of African-American or Hispanic origin (Hansen, 2002). The
same survey also showed that while African-Americans and Hispanics are
more likely to consider adoption, they are also less likely to have experience
with adoption compared to Whites (National Adoption Attitudes Survey,
2002).
Race appears to play an important role for adoption seekers as well. A
2002 study reported that the propensity to currently seek adoption is higher for
African-American women (National Adoption Attitudes Survey, 2002).
When comparing women who are currently seeking or planning to adopt with
13


women who ever considered adoption and took concrete steps, Chandra,
Abma, Maza and Bachrach (1999) discovered that women who are currently
seeking or planning to adopt are significantly less likely to be White. Their
study also revealed that being of a racial/ethnic group other than non-Hispanic
black was one of three factors most strongly associated with taking steps
among those who are currently seeking or planning to adopt (Chandra, Abma,
Maza & Bachrach, 1999).
So far, research on adoptive parents has not supported any of those
previous findings pertaining to adoption supporters and seekers. In 1977,
Bonham reported that about the same proportion of White women had
adopted a child as had African-American women. An analysis of data from
the 1976 Survey of Family Growth showed no significant differences in the
rate of adoption by race (Bachrach, 1983). The 1973 and 1982 waves of the
survey also revealed no significant differences by race in the proportion of
ever married women who had adopted a child (Bachrach, 1986). Moorman
and Hernandez (1989) similarly found that African-Americans are just as
likely to include adopted children as Whites. On the other hand, when
comparing children living with adoptive parents and children living with their
birthmothers, Bachrach (1986) revealed that unrelated adopted children are
14


much more likely to have White mothers than children living with their
birthmothers.
Labor Force Status
According to Bonham (1977), women not in the labor force and
women who work part-time are more likely to have adopted than women who
work full-time. Data from the 1976 Survey of Family Growth also showed
that adopted children are less likely to have mothers employed full-time
outside the home compared to children living with birthmothers (Bachrach,
1986). For the 1982 wave though, the difference was much smaller and not
statistically significant (Bachrach, 1986). In 1989, Moorman and Hernandez
discovered that in the majority of joint biological-adoptive families, the father
provides economically for the family through active participation in the labor
force and the mother stays at home and provides for the care of the children.
In purely adoptive families, the percentage of families with only the father in
the labor force is smaller, but it is still greater than the percentage of both
parents in the labor force (Moorman & Hernandez, 1989).
15


Eligibility and Selection Criteria
In their comparison of public and private adoption agencies, Daly and
Sobol (1994) investigated the importance of various criteria used to determine
the eligibility to adopt an infant. They found that both types of agencies rank
the criteria similarly. Marriage stability, motivation to parent, problem
solving ability, adaptability, warmth and nurturance, and understanding of
adoption are all ranked as important or very important (Daly & Sobol, 1994).
Interestingly, criteria that are not considered to be as important include
parents education, income and employment, presence of other children in the
family and marital status (Daly & Sobol, 1994).
In addition, Daly and Sobol (1994) studied the importance of factors
that both types of agencies consider when they have multiple eligible
applicants to choose from. The most remarkable difference between public
and private adoption agencies appears to be the greater emphasis that is put by
private adoption practitioners on the expectations that birthmothers have for
the kind of adoptive parents they would prefer (Daly & Sobol, 1994). The
selection factors include marital status, consensus between birth parents and
adoptive parents on the desired level of openness, physical characteristics,
16


ethnicity, temperament of the child, educational level, hobbies and interests,
and occupation (Daly & Sobol, 1994).
Previous research on the characteristics of individuals who support,
consider, seek or sought and already experienced adoption has demonstrated
some differences in age, marital status, marriage, parity, education, income,
race, and labor force status. Age appears to be the strongest determinant.
Daly and Sobols findings suggest that the majority of these factors should
also be of importance when agencies have to choose between various
applicants. Although age has not been mentioned in the findings, it is
intertwined with some of the other factors such as marital status and income.
17


CHAPTER 3
METHODS
Data
Data were drawn from the National Survey of Family Growth (NFSG),
Cycle V, conducted in 1995 by the National Center for Health Statistics, an
agency of the Department of Health and Human Services. This survey is the
fifth in a series of periodic surveys performed since 1973. The 1995 NSFG
sample was drawn from 14,000 households interviewed in the 1993 National
Health Interview Survey (NHIS), which is based on a stratified multistage
cluster sample. Of these 14,000 civilian, noninstitutionalized women of
reproductive age (ages 15 to 44), 13,795 were found to be eligible for the
1995 NSFG and 10,847 completed the interviews. Black and Hispanic
women were oversampled in the NHIS and its subsample, the 1995 NSFG.
Personal interviews were conducted in the homes of the respondents
between mid-January and October 1995 by trained female interviewers
utilizing laptop computers. The average interview lasted 105 minutes and
covered topics including fertility, family planning, and reproductive health
18


along with a variety of demographic and economic information.
Supplementary data on sensitive issues such as abortion and force intercourse
were gathered in brief self-administered interviews, in which the respondents
recorded their own answers into the laptop computer after listening to the
question over headphones. Respondents received a $20 token of appreciation
for their participation.
Sample, Sample Characteristics and Sample Size
The 1995 interviews have been divided into two files. The
Respondent File contains one record for each woman in the survey, while the
Interval File contains one record for each completed pregnancy experienced
by a woman in the survey. Since the purpose of this study is to investigate
what factors significantly contribute to the success of adoption, only the
Respondent File was used, which originally contains 5,711 variables for
10,847 cases. Further, the sample used in the study had to include only those
respondents who actually initiated, physically engaged in, and went through a
complete process of adoption with the help of an agency or lawyer, regardless
of whether it is a success or not.
19


Following this criterion, a group of twenty variables that measure how
the adoption was arranged for other child #1 through #20 (howadpOO through
howadpl9) were selected. This group of variables can unequivocally and
unambiguously identify those respondents that meet the criterion. Using this
variable as the selection criterion, a subsample consisting of 77 cases was
created by first selecting the 84 respondents who had valid values (or non-
missing values) on any of the twenty variables. Upon further review 7 cases
were excluded from the subsample since the respondents have had mixed
adoption results. The characteristics of the subsample are shown in tables 3.1
and 3.2.
Table 3.1 Descriptive statistics for the continuous variables relevant to the
issues of adoption
Variable N M SD
Rs age at interview 77 37.51 6.09
Highest school grade/year attended 77 13.68 2.38
Number of times R has been pregnant 77 1.62 1.95
Rs ideal number of children to have 77 2.82 1.18
Number of completed pregnancies 77 1.62 1.95
Age at first live birth 41 23.66 4.95
Age at most recent live birth 41 26.90 4.59
20


3.1 (Cont.)
Variable N M SD
Age at first conception 50 23.06 5.48
Age at first pregnancy outcome 50 23.70 5.22
Number of marriages 77 1.12 0.51
Months between first marriage & dissolution or interview 72 155.64 81.93
Years since first marriage 72 16.03 6.64
Rs age at first sex (even if before menarche) 77 18.49 3.30
Rs age at first voluntary sex since menarche 77 18.69 2.90
Months between first sex & first marriage or interview 77 42.60 62.71
Months between first voluntary sex after menarche & first marriage or interview 77 40.04 61.27
Poverty level of income 77 379.38 193.63
Table 3.2 Frequency table for the categorical variables used in this
investigation
Variable Frequency Percentage
Race Not White 21 27.3
White 56 72.7
Labor Force Not working full-time 37 48.1
Status Working full-time 40 51.9
21


3.2 (Cont.)
Variable Frequency Percentage
Marital Status Not married 18 23.4
Married 59 76.6
Marital Status at Not married 17 34.0
First Conception Married 33 66.0
Parity Not parous 36 46.8
Parous 41 53.2
Although the small sample size of 77 cases is not ideal, it ensures that
those respondents who have not successfully adopted a child have indeed
taken concrete steps toward legal adoption. When using a small sample there
is a chance that the power could be too low to detect even large meaningful
differences. That means that there is an increased risk of making a Type II
error. Since one method of increasing power is to raise the alpha level, all
alpha levels have been set to 0.10 for the statistical procedures. The small
sample size also requires using additional procedures such as Yates' correction
for particular Chi-square tests and recalculating independent t-tests by hand.
22


Variables and Measures
Out of the 5,711 variables of the original study, twenty-two variables
were selected to be included in this study based on prior research and their
relevance to the issues that the current study is investigating. As discussed in
the literature review, important topics that have emerged include age, marital
status, marriage, parity, education, income, race, and labor force status. Age
appears to be the strongest determinant. Hence, additional variables
pertaining to this variable have been added to the list of variables. It is
advisable to not only include the age at the time of the interview, but also to
investigate how age at first sex, at first conception, at first live birth, and at
most recent birth for instance influence adoption outcome.
The variables chosen for this study include twenty-one independent
and one dependent variable. Of the twenty-one independent variables, five
are dichotomous variables and sixteen are numeric variables. The
dichotomous variables include: respondents race (race), labor force status
(laborfor), marital status (martstat), marital status at first conception
(,marconOl) and parity (parity). All but parity had originally been nominal
variables with various categories and they have been recoded to be used in the
23


data analysis. Respondents race has been recoded not White = 0 and White =
1, labor force status has been recoded 0 = not working full-time and 1 =
working full-time, marital status has been recoded not married = 0 and
married = 1, marital status at first conception has been recoded not married =
0 and married = 1. Parity had originally been a numeric variables and has also
been recoded not parous (nulliparous) = 0 and parous = 1.
The numeric independent variables include: respondents age at
interview (age), highest school grade/year attended (higrade), number of
times respondent has been pregnant (numpregs), respondents ideal number of
children to have (intexact), number of completed pregnancies (compreg), age
at first live birth (agebldgl), age at most recent live birth (agels2dg), age at
first conception (agecldg2 ), age at first pregnancy outcome (agepldg2 ),
number of marriages (fmamo), months between first marriage and dissolution
or interview (marldiss), years since first marriage (marl_now), respondents
age at first sex even if before menarche (vrylstag), respondents age at first
voluntary sex since menarche (sexlage), months between first sex and first
marriage or interview (sexmar), months between first voluntary sex after
menarche and first marriage or interview (volsxmar) and poverty level of
income (poverty).
24


The dependent variable, did respondent legally adopt other child
(adptot), is a dichotomous variable that has been recoded with no = 0 and yes
= 1. This variable is a compilation of 20 different variables that asked the
same question regarding other child #1 through #20 (adptotOO through
adptotl9). It was originally coded with yes = 1 and no = 2. Chi-Square
analyses and independent t-tests were conducted to find out whether
respondents who did adopt a child differ from those who did not in regard to
the twenty-one independent variables mentioned above. In addition, binary
logistic regression was used to identify those factors that significantly predict
whether or not a respondent adopted a child
Statistical Procedures
Chi-Square analyses are used to assess two types of comparison: tests
of goodness of fit and tests of independence. This study used tests of
independence, which assess whether paired observations on two variables,
expressed in a contingency table, are independent of each other. Yates'
correction for continuity adjusts the formula for the Chi-Square analysis by
subtracting 0.5 from each observed value in a 2 x 2 contingency table. This
formula is mainly used when at least one cell of the table has an expected
25


frequency less than 5. Independent samples t-tests are used to assess the
association between a dependent numeric variable and two categories of
another, independent variable. Lastly, binary logistic regression is used to
predict a categorical outcome or dependent variable from a set of predictor
variables.
26


CHAPTER 4
RESULTS
Chi-Square Tests
Chi-Square analyses were conducted to determine whether
respondents who did adopt a child differ from those who did not in regard to
respondents race, labor force status, marital status, marital status at first
conception, and parity. The relation between adoption outcome and
respondents race was significant. Table 4.1 shows that respondents who are
White are more likely to successfully adopt a child than respondents who are
not. Since one of the expected frequencies was less than 5, Yates' correction
was used to avoid making a Type I error. The relation remained significant.
Table 4.1 Chi-square test results for adoption outcome versus race
Race Model Fit Index
Adoption Not White White
1 4
No (3) (8) X2 = 8.56 (p = 0.003) X2(Yates Corrected) = 8.55
14 52 ll o o
Yes (18) (48) N = 77
27


Note: Expected frequencies appear in parentheses below observed
frequencies.
The relation between adoption outcome and respondents labor force
status was also significant. Table 4.2 shows that respondents who are working
full-time are less likely to successfully adopt a child than respondents who are
not. The relation between adoption outcome and respondents marital status
was ultimately not significant. Table 4.3 shows that Pearsons Chi-square was
significant indicating that respondents who are married are more likely to
successfully adopt a child than respondents who are not. Since one of the
expected frequencies was less than 5 though, Yates' correction was also used
to avoid making a Type I error. The relation did not remain significant.
Table 4.2 Chi -square test results for adoption outcome versus labor force
status
Labor Force Status
Adoption Not Working Full-time Working Full- time Model Fit Index
No 1 10
(5) (6) t = 7.804 (p = 0.005)
Yes 36 (32) 30 (34) N = 77
28


Note: Expected frequencies appear in parentheses below observed
frequencies.
Table 4.3 Chi -square test results for adoption outcome versus marital status
Adoption Marital Status ^[ Married Married Model Fit Index
No 5 (3) 6 (8) t = 3.492 (/? = 0.062) X2(Yates Corrected) = 2.202
13 53 (p = 0.138)
Yes (15) (51) r~- t II z
Note: Expected frequencies appear in parentheses below observed
frequencies.
The relation between adoption outcome and respondents marital
status at first conception was significant. Table 4.4 shows that respondents
who are married are more likely to successfully adopt a child than respondents
who are not. Since two of the expected frequencies were less than 5, Yates'
correction was also taken into consideration to avoid making a Type I error.
The relation remained significant. The relation between adoption outcome
and respondents parity was not significant. Table 4.5 shows the results of the
Chi-square test.
29


Table 4.4 Chi -square test results for adoption outcome versus marital status
at first conception
Adoption
Marital Status at First
Conception
Not
Married
Married
Model Fit Index
No 5 (2) 0 (3) X2= 10.784 (p = 0.001) X2(Yates Corrected) = 7.764
Yes 12 (15) 33 (30) (p = 0.005) N = 50
Note: Expected frequencies appear in parentheses below observed frequencies.
Table 4.5 Chi -square test results for adoption outcome versus parity
Adoption Parity Not Parous Parous Model Fit Index
No 6 (5) 5 (6) f- = 0.313 (p = 0.576)
Yes 30 (31) 36 (35) N = 77
Note: Expected frequencies appear in parentheses below observed
frequencies.
30


Independent Samples T-Tests
Independent samples t-tests were conducted to determine whether
respondents who did adopt a child differ from those who did not in regard to
respondents age at interview, highest school grade/year attended, number of
times respondent has been pregnant, respondents ideal number of children to
have, number of completed pregnancies, age at first live birth, age at most
recent live birth, age at first conception, age at first pregnancy outcome,
number of marriages, months between first marriage and dissolution or
interview, years since first marriage, respondents age at first sex even if
before menarche, respondents age at first voluntary sex since menarche,
months between first sex and first marriage or interview, months between first
voluntary sex after menarche and first marriage or interview, and poverty
level of income.
Table 4.6 presents the independent samples t-test results as well as the
means and standard deviations for each variable by adoption outcomes. There
was a significant effect for respondents age, age at first live birth, age at most
recent live birth, age at first conception, age at first pregnancy outcome, age at
first sex even if before menarche and age at first voluntary sex since
menarche with respondents who successfully adopted a child being or having
31


been older than those who did not. There was a significant effect for highest
school grade/year attended, with respondents who successfully adopted a child
having attended more schooling than those who did not.
There was a significant effect for number of times the respondent has
been pregnant and for number of completed pregnancies, with respondents
who successfully adopted a child having been pregnant more often and
completing more pregnancies than those who did not. There was a significant
effect for respondents ideal number of children to have, with respondents
who successfully adopted a child desiring more children than those who did
not. There was a significant effect for number of marriages, with respondents
who successfully adopted a child having been married more often than those
who did not. There was a significant effect for months between first marriage
and dissolution or interview and years since first marriage, with more time
having passed for respondents who successfully adopted a child than those
who did not.
There was a significant effect for months between first sex and first
marriage or interview and for months between first voluntary sex after
menarche and first marriage or interview, with less time having passed for
respondents who successfully adopted a child than those who did not. Lastly,
32


there was a significant effect for poverty level of income, with respondents
who successfully adopted a child having a higher poverty level of income than
those who did not. Since all of the tests involve one large and one small
sample, all tests were recalculated by hand and all variables remained
significant.
Table 4.6 Independent samples t-tests results for numeric variables
Variable Adoption N M t P
Rs age at interview No 11 29.45 -3.775 0.003
Yes 66 38.85
Highest school grade/year No 11 12.36 -2.031 0.048
attended Yes 66 13.89
Number of times R has been No 11 0.73 -1.668 0.099
pregnant Yes 66 1.77
Rs ideal number of children No 11 2.27 -1.679 0.097
to have Yes 66 2.91
Number of completed No 11 0.73 -1.668 0.099
pregnancies Yes 66 1.77
Age at first live birth No 5 20.60 -2.770 0.016
Yes 36 24.08
Age at most recent live birth No 5 22.20 -2.616 0.013
Yes 36 27.56
Age at first conception No 5 18.80 -4.665 0.000
Yes 45 23.53
33


4.6 (Cont.)
Variable Adoption N M t P
Age at first pregnancy No 5 19.40 -5.095 0.000
outcome Yes 45 24.18
Number of marriages No 11 0.73 -1.872 0.088
Yes 66 1.18
Months between first marriage No 6 78.50 -2.496 0.015
& dissolution or interview Yes 66 162.65
Years since first marriage No 6 9.00 -2.837 0.006
Yes 66 16.67
Rs age at first sex (even if No 11 15.18 -3.916 0.000
before menarche) Yes 66 19.05
Rs age at first voluntary sex No 11 16.00 -3.483 0.001
since menarche Yes 66 19.11
Months between first sex & No 11 107.73 2.333 0.040
first marriage or interview Yes 66 31.74
Months between first No 11 97 09
voluntary sex after menarche Yes 66 30.86 2.205 0.069
& first marriage or interview
Poverty level of income No 11 258.82 -2.292 0.025
Yes 66 399.47
Logistic Regression
Binary logistic regression was employed to assess how well adoption
outcome can be predicted from a combination of six variables: respondents
age at interview, marital status, race, highest school grade/year attended,
34


parity, and respondents ideal number of children to have. When all six
predictor variables are considered together, they significantly predict whether
or not a respondent legally adopt a child. Table 4.7 presents the results of the
logistic regression. The odds ratios, which suggest that the odds of estimating
correctly who successfully adopts a child improve by 47% if one knows the
respondents age, by 2628% if one knows the respondents race and by 64% if
one knows the highest school grade/year the respondent attended.
The odds ratios could also be interpreted as suggesting that for each
additional year of the respondents age, the odds of successfully adopting a
child increase by a factor of 1.47, assuming that race and highest school
grade/year attended are held constant. Further, respondents who are White are
27.28 times more likely to have a successful adoption outcome, assuming that
age and highest school grade/year attended are held constant. Last, for each
additional year in highest school grade/year attended, the odds of successfully
adopting a child increase by a factor of 1.64, assuming that age and race are
held constant.
The table also shows that the Nagelkerke R2 is 0.700, which indicates
that 70% of the variance in whether or not respondents adopted a child can be
predicted from the linear combination of the six independent variables.
35


Table 4.7 Logistic regression predicting who will successfully adopt
Variable P SE Odds Ratio p
Age 0.38 0.17 1.47 0.001
Marital Status 0.81 1.49 2.26 0.585
Race 3.31 1.41 27.28 0.019
Highest school grade/year attended 0.49 0.28 1.64 0.079
Parity -0.18 1.15 0.84 0.877
Ideal number of children 1.01 0.70 2.74 0.148
Constant -22.73 7.23 0.00 0.002
X2 = 38.307, Nagelkerke R2 = 0.700, N = 77, p < 0.001
36


CHAPTER 5
DISCUSSION AND LIMITATIONS
The goal of this study was to fill one gap in current knowledge
regarding adoption by comparing the characteristics of individuals who go
through the adoption process and succeed with those who do not. Previous
research on the characteristics of individuals who support, consider, seek or
sought and already experienced adoption has demonstrated some differences
in age, marital status, marriage, parity, education, income, race, and labor
force status. Some of the findings of this study correspond to the previous
research, while other findings do not. Age appeared to be the strongest
determinant in the many of the studies and for that reason, additional variables
pertaining to this variable had been added to this study.
Previous research has shown that adoption seekers and adoptive
parents tend to be older than parents with biological children and the general
public (Leahy, 1933, Bachrach, 1983 & 1986, Moorman & Hernandez, 1989).
This study found that respondents who successfully adopted a child were not
only older at the time of the interview, but also at their first live birth, most
recent live birth, first conception, first pregnancy outcome and first sex.
37


These findings indicate that being older at certain stages associated with
fertility values could be beneficial for individuals seeking to adopt.
Interestingly, age has not been named by agencies in Daly and Sobols 1994
study as a selection factor between applicants.
The majority of prior studies have also recognized the importance of
marital status pertaining to adoption attitudes and behaviors. Those studies
found that adoption supporters, seekers and adoptive parents are more likely
to be married than parents with biological children and the general public
(Bachrach, 1986, Moorman & Hernandez, 1989, Hansen, 2002). Further,
marital status is one of the selection factors used by agencies when confronted
with multiple eligible adoption applicants (Daly & Sobol, 1994). This study
could not ultimately find a difference between respondents who successfully
adopted a child and those who did not regarding marital status. It appears that
the failure to establish statistical significance might be the result of the small
sample size. It is still noteworthy that 80% of all respondents who
successfully adopted a child are married compared to 55% of those who are
not successful.
In their study from 1989, Moorman and Hernandez found that parents
in most purely adoptive families and most joint biological-adoptive families
38


have only been married once. This study also shows a significant effect for
number of marriages. While respondents who successfully adopted a child
have been married on average 1.18 times, those who did not have only been
married 0.73 times, which also implies a preference towards marriage. This
study further found that respondents who successfully adopted a child tend to
be married at their first conception than respondents who are not. The finding
might be explained by the importance agencies attribute to marriage stability
and marital status when determining eligibility and selecting applicants for a
particular child (Daly & Sobol, 1994).
Taking into consideration that the majority of all respondents who
successfully adopted a child are married, have only been married once on
average and they have been married at their first conception, then this would
certainly point towards a stable marriage. Further, prior research indicates
that adoptive parents have been married longer than other types of parents and
the general public (Leahy, 1933, Moorman & Hernandez, 1989). This study
shows that respondents who successfully adopted a child tend to be married
longer. For them, more time has passed between their first marriage and its
dissolution or the interview, and their first marriage has also been longer in
duration. 83% of all respondents who successfully adopted a child have only
39


been married once, which would suggest that they have been indeed married
longer.
All prior studies have shown that nulliparous women are more likely
to be adoption supporters, seekers, and adoptive parents (Leahy, 1933,
Bonham, 1977, Bachrach, London & Maza, 1991, Chandra, Abma, Maza &
Bachrach, 1999). This study could not find a difference between respondents
who successfully adopted a child and those who did not with regard to parity.
45% of all respondents who successfully adopted child were nulliparous
compared to 55% of those who did not. Interestingly, Daly and Sobol (1994)
do not mention parity as one important factor agencies use in their decision
process either. Further, in 1977 Bonham stated that over half of all women
who have adopted a child have also given birth to a child, a finding that has
been confirmed by this study.
Surprisingly, there was a significant effect for number of times the
respondent has been pregnant and for the number of completed pregnancies.
Respondents who successfully adopted a child have been pregnant more often
and completed more pregnancies than those who did not. Further research is
needed to investigate these findings in detail. A study conducted in 1991 had
also suggests that women who desire more children than they expect to bear
40


are substantially more likely to have sought than women who desire fewer
(Bachrach, London & Maza, p.712). This study found that respondents who
successfully adopted a child generally desire more children than those who
did not. Maybe adoption agencies and birth mothers prefer couples who
envision large families by inferring that those individuals are more family-
oriented.
The majority of prior studies have also recognized the importance of
education pertaining to adoption attitudes and behaviors. Those studies found
that adoption supporters, seekers and adoptive parents tend to be more
educated than parents with biological children and the general public
(Bachrach, London & Maza, 1991, Moorman & Hernandez, 1989, Hansen,
2002). This study shows that respondents who successfully adopted a child
have attended more schooling than those who did not, which coincides with
Daly and Sobols findings (1994) that educational level is one of the selection
factors used by agencies when confronted with multiple eligible adoption
applicants.
Many researchers investigating adoption seekers and adoptive parents
have found that they have more income than parents with biological children
and the general public (Leahy, 1933, Bonham, 1977, Moorman & Hernandez,
41


1989, Hansen, 2002). This study also shows that respondents who
successfully adopted a child have a higher income (expressed as poverty level
of income) than those who did not. Although income has not been identified
as a selection factor for agencies to choose between applicants, occupation
has. Using occupation instead of income might make agencies appear more
welcoming, but occupation is certainly intertwined with both income and
education. Due to the nature of the secondary data and the small sample size,
this study was not able to investigate the relationship between occupation and
adoption outcome.
The relationship between race and adoption has been studied by
several researchers. While some of them have discovered that adoption
seekers are less likely to be white, others have found no racial difference
regarding adoptive parents (Bonham, 1977, Bachrach, 1983, Moorman &
Hernandez, 1989, Chandra, Abma, Maza & Bachrach, 1999). This study
found that White respondents are more likely to have successfully adopted a
child than respondents who are not White and that race appears to have the
most significant effect. Ethnicity has also been named by agencies as a
selection factor between applicants as reported by Daly and Sobol in 1994.
42


Apparently, both private and public agencies have some concerns about
placing a child with parents of a different race (Daly & Sobol, 1994).
Research on the labor force status and adoptive parents has shown that
adoptive mothers are less likely to work full-time than mothers with biological
children and the general public (Bonham, 1977, Bachrach, 1986, Moorman &
Hernandez, 1989). This study found that respondents who successfully
adopted a child are less likely to work full-time than respondents who did not.
Agencies have also voiced minor concerns about placing children in families
where parents were working full-time (Daly & Sobol, 1994). It might be that
agencies just want to ensure that the parents have enough time to integrate the
new child into their family in order to have a good transition.
The major limitations of this study are the small sample size and the
nature of the secondary data. The National Survey of Family Growth only
includes women ages 15 to 44 and their characteristics such as age and marital
status are measured as of the time of the interview. Further, many of the
individuals who did not successfully adopt a child did not specify how the
adoption was arranged. While 100% of those who successfully adopted
specified their method of adoption, only a small portion of those who did not
43


adopt did. All of those cases could not be included though, since it was not
known if the respondent actually sought to adopt the child.
This study should be replicated in the future with a larger sample and
variables that more specifically ask about the characteristics of individuals at
the time of the adoption. It would also be interesting to study male
respondents in order to achieve a more complete picture of the whole family.
Future research should also investigate the differences between individuals
who adopt successfully and those who do not more closely, since this is the
first study of this kind as of the researchers knowledge.
44


BIBLIOGRAPHY
Bachrach, Christine A. (1983). Adoption as a Means of Family Formation:
Data from the National Survey of Family Growth. Journal of Marriage
and the Family, 45 (November), 859-865.
Bachrach, Christine A. (1986). Adoption Plans, Adopted Children, and
Adoptive Mothers. Journal of Marriage and the Family, 48 (May), 243-
253.
Bachrach, Christine A., London, Kathryn A., & Maza, Penelope L. (1991).
On the Path to Adoption: Adoption Seeking in the United States, 1988.
Journal of Marriage and the Family, 53 (August), 705-718.
Bonham, Gordon Scott. (1977). Who Adopts: The Relationship of Adoption
and Social-Demographic Characteristics of Women. Journal of Marriage
and the Family, 39 (May), 295-306.
Chandra, Anjani, Abma, Joyce, Maza, Penelope, & Bachrach, Christine.
(1999). Adoption, Adoption Seeking, and Relinquishment for Adoption in
the United States. Advanced Data, 306. Retrieved February 20, 2006,
from the Department of Health and Human Services, Centers for Disease
Control and Prevention website:
http://www.cdc.gov/nchs/data/ad/ad306.pdf
Daly, Kerry J & Sobol, Michael P. (1994). Public and Private Adoption: A
Comparison of Service and Accessibility. Family Relations, 43, 86-93.
Dave Thomas Foundation For Adoption. (2002). National Adoption Attitudes
Survey. Retrieved February 20, 2006, from the Dave Thomas Foundation
For Adoption website:
http://www.davethomasfoundationforadoption.org/html/resource/Adoption
Attitudes.pdf
45


Evan B. Donaldson Adoption Institute. (1997). Benchmark Adoption Survey:
Report on the Findings. Retrieved February 20, 2006, from the Evan B.
Donaldson Adoption Institute website:
http://www.adoptioninstitute.org/survev/Benchmark Survey 1997.pdf
Hansen, Mary Eschelbach. (2002). Measuring the Relationship between
Income and Interest in Adoption. Retrieved February 20, 2006, from the
Center for Adoption Research website:
http://www.centerforadoptionresearch.org/
Leahy, Alice M. (1933). Some Characteristics Of Adoptive Parents.
American Journal of Sociolog, 38 (January), 548-563.
Moorman, Jeanne E., & Hernandez, Donald J. (1989). Married-Couple
Families With Step, Adopted, and Biological Children. Demography, 26
(2), 267-277.
46