Discussion

27 (2005) 1031–1044

www.elsevier.com/locate/childyouth

A comparison of family functioning in gay/lesbian,

heterosexual and special needs adoptions

Patrick LeungT, Stephen Erich, Heather Kanenberg

University of Houston, United States

Received 25 October 2004; received in revised form 17 December 2004; accepted 20 December 2004

Available online 9 February 2005

Abstract

The purpose of this study was to identify possible contributing factors to family functioning in

three types of adoptive families: those headed by gays/lesbians, those headed by heterosexuals, and

those involving the adoption of children with special needs. These three adoptive family types were

examined concurrently so that commonalities and differences could be identified and considered for

use in adoption practice. A multiple regression analysis was used to assess the relationship between

the dependent variable (standardized family functioning score) and independent variables (child

behavior scores, special needs adoption, gay/lesbian headed families, age at adoption and at

interview, diagnoses of disabilities, total social support score, number of previous placements,

previous abuse and co-sibling adoption). Results indicated no negative effects for the parenting of

adopted children by gay/lesbian headed families. Higher levels of family functioning were found to

be associated with special needs, younger, and non-disabled child adoptions. Gay/lesbian headed

family adoptions of older children, non-sibling group adoptions, and children with more foster

placements also experienced higher levels of family functioning. Implications include the need to (1)

place a child in an adoptive family as early as possible, (2) ensure strong support networks for

adoptive families of children with disabilities and with those who adopt sibling groups, and (3)

encourage the practice of adoption by gay/lesbian headed families, especially for older children.

D 2005 Elsevier Ltd. All rights reserved.

Keywords: Special needs; Gay/lesbian and heterosexual adoption

0190-7409/$ -

doi:10.1016/j.

T Correspon Houston, TX

E-mail add

Children and Youth Services Review

see front matter D 2005 Elsevier Ltd. All rights reserved.

childyouth.2004.12.030

ding author. Graduate School of Social Work, University of Houston, 237 Social Work Building,

77204-4013, United States.

ress: [email protected] (P. Leung).

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441032

1. Introduction

Family has been conceptualized in a variety of ways. Barker (1995) emphasizes, ba family consists of a primary group whose members assume certain obligations for each

other and generally share common residencesQ (p. 130). The National Association of Social Workers (NASW) Family Policy (1999) asserts, b[t]he family is the primary socializing agent as well as the primary economic unit in our cultureQ (p. 1). Since the Industrial Revolution, the conceptualization of family has changed and developed (Cherlin

& Furstenberg, 1994; Furstenberg, 1999; NASW, 1999). In its broadest sense, family is

two or more people who regard themselves as family and who take upon themselves

commitments and responsibilities that are commonly deemed fundamental to family life

(NASW, 1999). A child’s family, and the typical protection, socialization, security and

companionship offered by family members, is essential to the human condition. A family

and its members, whether biological or adoptive, are indispensable in the growth and

development of children. The family unit, no matter how it is defined, serves to foster

children’s view of the external world, their emotional capacities, and their individual

identities (Goldenberg & Goldenberg, 1996).

Research indicates that there were 542,000 children in foster care in the United States in

2001 (U.S. Department of Health and Human Services [DHHS], 2003). At the time of data

collection, only 27% (143,208) of these children were awaiting adoption or placement

with a family member, and 11% (62,014) had no defined case plan (DHHS, 2003). Many

children waiting to be adopted may be older, have a disability, have siblings that want to

be adopted together, or may be from a minority culture. Research indicates (National

Adoption Information Clearinghouse, 1999) that between 30% and 50% of children

awaiting adoption have a developmental disability. Special needs children, like all

children, thrive with the love and stability that come from belonging to permanent families

(NAIC, 1999).

Many gay and lesbian adults and families are interested, willing and even resolute in

their quest to adopt children (American Civil Liberties Union, 1999). Historically,

organizational policies, legislation, and social stigma have made it difficult for gay and

lesbian adults and couples to complete the adoption process (Brodzinsky, 2003). With

each state court having discretion over the laws and processes for adoption and custody,

the door is open to inconsistencies and to the creation of a patchwork system for

families in the United States. While generalizations and stereotypes have shadowed the

possibilities for adoption by homosexual couples or individuals, a review of the research

on homosexual parents stated that b[t]here is no evidence of any kind that demonstrates that living with a homosexual parent has any significant negative effects on children. In

fact, it appears that gay parents are as effective and may be even more so in some ways

than non-gay parentsQ (Bigner & Bozett, 1989, p. 160). Recent research indicates that there is an increasing number of lesbians and gay men who are becoming parents, and

that approximately two in five of all adoption agencies in the country have placed

children with adoptive parents whom they recognize to be gay or lesbian (Brodzinsky,

2003). While these are encouraging findings for homosexual individuals and couples

seeking to adopt, approximately 25% of respondents to a national adoption survey

indicated bthat birth parents have objected to placing their child with gays or lesbians, or

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–1044 1033

have specifically requested that their child not be placed with homosexualsQ (Brodzinsky, 2003, p. 4).

A lack of research designed to explore the possible effects of gay and lesbian parents on

their adoptive children has certainly contributed to these barriers to adoption. While some

of the controversy regarding adoptions by gay and lesbian adults and couples is

presumably a byproduct of homophobia, much of the debate appears to be the result of a

lack of assurance with regard to the well-being of children placed within these family units

(Brooks & Goldberg, 2001).

This study proposes to examine predictors of family functioning among a cohort which

includes three groups of adoptive families: gay/lesbian, heterosexual, and families who

have adopted a child with special needs status. The Process Model of Family Functioning

(Steinhauer, Santa-Barbara, & Skinner, 1984) is used in this study as a conceptual

framework for assessing family functioning. According to this model, the overriding goal

of the family is the successful achievement of a variety of basic, developmental, and crisis

tasks (Task Accomplishment). It is through these processes that the family attains, or fails

to attain, objectives central to its life as a group. These functions include allowing for the

continued development of all family members, providing security, ensuring sufficient

cohesion to maintain the family as a unit, and functioning effectively as part of society.

Successful Task accomplishment involves the differentiation and performance of various

roles (Role Performance). Essential to the performance of these roles is the process of

interaction (Communication), by which information essential to Task Accomplishment and

the ongoing role definition is exchanged. A vital element of the communication process is

the expression of affect (Affective Expression), which can impede or facilitate various

aspects of Task Accomplishment and successful role integration. Critical elements of

Affective Expression include the content, intensity, and timing of the feelings involved.

Similarly, family members’ involvement with one another (Involvement) can either help or

hinder Task Accomplishment. Involvement refers to both the degree and quality of family

members’ interest in one another. Control is the process by which family members

influence each other. The family should be capable of successfully maintaining ongoing

functions as well as adapting to shifting task demands. Finally, how tasks are defined and

how the family proceeds to accomplish them may be greatly influenced by the family and

culture (Values and Norms) (Steinhauer et al., 1984).

The objectives of this study are: (1) to review the literature regarding disrupted

adoptions, adoptive child’s behavior, and familial support networks; (2) to develop a

regression model indicating predictors of adoptive family functioning; (3) to identify

whether there is a significant difference in family functioning among homosexual,

heterosexual and special needs adoptions so that commonalities and differences among

these adoptive family types can be identified; and (4) to discuss the implications of these

findings in relation to further research and adoption practices.

2. Literature review

The research on families who adopt children with special needs status is generally

limited to those parents who are heterosexual or assumed to be heterosexual. However,

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441034

within this parameter, the majority of these adoptive families are considered successful

placements. Barth and Berry (1988) found that recent studies on dissolution or disruption

rates for adoptive families who adopt a child with special needs status suggest that less

than 15% of these placements fail to sustain. Rosenthal and Groze (1991) found that three

quarters of the parents of special needs children reported that the effect of adoption on the

family had been mostly positive or very positive. However, Rosenthal and Groze (1992)

also found that behavioral problems were reported to be prevalent among many of these

children. In fact, 41% scored in the abnormal range on the Achenbach Child Behavior

Checklist (Achenbach & Edelbrock, 1983), which is indicative of severe emotional or

behavioral problems. Barth and Berry (1988) reported that children from disrupted

adoptions had significantly higher levels of behavior problems on the externalizing

subscale of the child behavior checklist than those from intact adoptions. A substantial

body of research on parents who adopt children with special needs supports the premise

that incidents of adoption disruption are associated with increased age of the child at the

time of adoption (Rosenthal, 1993). Several studies indicate a strong correlation between

an adopted child’s behavioral problems, legal problems, psychological difficulties, and/or

a history of abuse or neglect with adoption disruption (Barth & Berry, 1988; McDonald,

Liberman, Partridge, & Hornby, 1991; Rosenthal & Groze, 1992). Additionally, several

studies have found the number of previous placements to be correlated with adoption

disruption (Barth & Berry, 1988; Festinger, 1986; Rosenthal & Groze, 1992).

Westhues and Cohen (1990), utilizing the Family Assessment Measure (FAM), reported

that intact adoptive families differed from disrupted adoptive families on several indices of

family functioning. Leung and Erich (2002) reported a significant inverse relationship

between family functioning and the adopted child’s behavioral status. Several studies also

suggest that familial support networks impact adoptive family functioning (Kagen & Reid,

1986; Leung & Erich, 2002; Rosenthal, 1993).

When the literature on family functioning of gay and lesbian parents and their children

is reviewed, the results strongly suggest that lesbian mothers and gay fathers are capable of

fostering warm, positive, encouraging relationships. Several studies of lesbian mothers

revealed that they are as child focused, loving, confident, nurturing, and responsive as their

heterosexual counterparts (Golombok, Spencer, & Rutter, 1983; Miller, Jacobsen, &

Bigner, 1981; Tasker & Golombok, 1995). Chan, Raboy, and Patterson (1998) found that

children’s outcomes relating to social competencies, behavior, and adjustment were

unrelated to parents’ sexual orientation. Further, this study found that when assessing

children’s social competencies and behavior problems, it was not possible for the

researchers to distinguish between children raised by lesbian mothers and their

heterosexual counterparts. In essence, both lesbian and heterosexual mothers are able to

provide home environments that support healthy development (Chan et al., 1998).

With regard to gay fathers and couples, Bigner and Jacobsen (1989) reported that gay

fathers have been found to possess parenting skills similar to heterosexual fathers. Gay

fathers exhibit healthy intimate bonds with their children, motivation of their children,

provision of recreational opportunities, encouragement of autonomy, and experience

parental satisfaction as often as heterosexual fathers. In addition, Bigner and Jacobsen

found gay fathers superior in attentiveness to children’s needs, paternal nurturing, and in

communicating reasons for appropriate behavior to their children.

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–1044 1035

In another study, Brooks and Goldberg (2001) found that children raised by gay or

lesbian parents were not impacted in a negative way. In fact, their research reported that

gay and lesbian parents may have special strengths that make them particularly well suited

for adoption and foster parenting. Many participants in the Brooks and Goldberg study

indicated the benefits of a strong system of extended family and friends that aided in the

nurturing and support of their children. It is important to note that many of these studies

have been criticized for having small samples, no comparison groups or otherwise have

weak designs.

Interestingly, research regarding homosexual adoptive families remains unavailable

(Brooks & Goldberg, 2001). Furthermore, in a content analysis of twelve prominent social

work journals covering a 12-year period, Van Voorhis and Wagner (2001) found that there

were no articles dealing with gay men or lesbians as adoptive or foster parents. However,

in a recent study in exploring adoptive family functioning in gay/lesbian families, scores

were within normal ranges on measures of family functioning, familial support networks

and adopted child’s behavior (Erich, Leung, Kindle, & Carter, 2004). When these scores

were compared to a similar cohort of heterosexual adoptive parents, no significant

differences on these measures were found (Erich, Leung, & Kindle, in press).

3. Method

3.1. Sample and design

The study sample consisted of three data sets. All three data sets were obtained by

convenience sampling to ensure an adequate sample size for this comparative study.

The first data set consisted of adopted children with special needs status and their

parents. Special needs status included children older than three years of age, children with

physical and/or mental handicapping conditions, those with psychological or emotional

problems, and children adopted as part of a sibling group, as well as those originating from

minority groups. The majority of families were recruited through four adoption programs

located in a large metropolitan area of a southern state. From this group, 86 adoptive

parents (respondents) participated in this study and reported demographic, historical, and

behavioral information on 117 of their adopted children. The majority of the adopted

children were male (40%). About 33% were Caucasian; 43% were non-Caucasian (13%

African American, 15% bi-racial, 12% Latino and 3% Asian); and 23% did not identify

their ethnicity. The mean age at adoption and at the time of interview was 4.38 and 10.74,

respectively. About 55% had some form of disability. These children had a mean of .23

previous placements prior to adoption, and over 48% were reported to have had a history

of abuse prior to adoption (see Table 1).

In terms of the characteristics of adoptive parents in the first data set, a majority were

Caucasian (81%) and had 12 years or more of education (95%). Only 8% indicated that

they were previously foster parents, and all 86 parents adopted their children via Child

Protective Services (CPS) (Table 2).

The second data set included a total of 47 gay and lesbian adoptive parents

(respondents) who also reported demographic, historical, and behavioral information on

Table 1

Characteristics of adopted children

Variable Heterosexual Gay/lesbian Special needs

N % N % N %

Gender of adopted child

Female 20 46.5 32 47.0 43 36.8

Male 23 53.5 36 53.0 47 40.2

No response – – – – 27 23.0

Race of adopted child

Caucasian 11 25.6 20 29.4 39 33.3

African-American 9 20.9 7 10.3 15 12.8

Latin/Hispanic 4 9.3 14 20.6 14 12.0

Asian 5 11.6 11 16.2 4 3.4

Bi-racial 14 32.6 14 20.6 18 15.4

No response 0 0.0 2 2.9 27 23.1

Age at adoption

0–4 41 95.3 54 79.4 55 47.0

5–9 2 4.7 4 5.9 25 21.4

10+ 0 0.0 1 1.5 10 8.5

No response 0 0.0 9 13.2 27 23.1

(X: 5.98; S.D.: 5.39) (X: 1.69; S.D.: 1.93) (X: 4.38; S.D.: 3.40)

Age now

0–4 18 41.9 28 41.1 7 6.0

5–9 13 30.2 28 41.1 26 22.2

10+ 12 27.9 12 17.6 51 43.6

No response – – – – 33 28.2

(X: 7.08; S.D.: 4.26) (X: 6.09; S.D.: 3.71 ) (X: 10.74; S.D.: 4.61)

Diagnostic characteristics at adoption

Physical handicap 2 4.7 0 0.0 10 8.5

Learning disability 2 4.7 4 5.9 8 6.8

Mental retardation – – – – 5 4.3

Psych. disorder 2 4.7 3 4.4 20 17.1

Other/comb. 5 11.6 10 14.7 21 17.9

None 32 74.3 50 73.5 26 22.2

No response 0 0.0 1 1.5 27 23.1

Number of placements prior to adoption

0 31 72.1 50 73.5 79 67.5

1 6 14.0 5 7.4 10 8.5

2–4 5 11.6 11 16.2 1 0.9

5+ 1 2.3 2 2.9 27 23.1

(X: .58; S.D.: 1.24) (X: .99; S.D.: 2.72) (X: .23; S.D.: 1.19)

Type of abuse prior to adoption

Physical abuse 0 0.0 5 7.4 10 8.5

Sexual abuse 1 2.3 7 10.3 25 21.4

Neglect 5 11.6 11 16.2 – –

Physical and sexual 1 2.3 1 1.5 – –

Others – – – – 21 17.9

None 36 83.7 44 64.7 34 29.1

No response – – – – 27 23.1

Child adopted as sibling group

Yes 7 16.3 10 14.7 28 23.9

No 36 83.7 58 85.3 62 53.0

No response – – – – 27 23.1

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441036

Table 2

Characteristics of adoptive parents

Variable Heterosexual (N=25) Gay/lesbian (N=47) Special needs (N=86)

N % N % N %

Gender of Respondent

Female 24 96.0 24 51.0 (not available)

Male 1 4.0 23 49.0 (not available)

Race of respondent

Caucasian 22 88.0 43 91.5 70 81.4

African-American 0 0.0 0 0.0 8 9.3

Latino/Hispanic 0 0.0 2 4.3 4 4.7

Asian 1 4.0 0 0.0 – –

Other/comb 2 8.0 1 2.1 2 2.3

No response 0 0.0 1 2.1 2 2.3

Respondent’s years of education

1–12 2 8.0 1 2.1 2 2.3

13–16 10 40.0 12 25.6 52 60.5

17–18 4 16.0 16 34.0 23 26.7

19+ 8 32.0 17 36.2 7 8.1

No response 1 4.0 1 2.1 2 2.3

Respondent previously a foster parent

Yes 6 24.0 6 12.8 7 8.1

No 19 76.0 41 87.2 79 91.9

Type of adoption

Private 8 18.6 21 30.9 – –

International 7 16.3 23 33.9 – –

CPS 3 6.9 13 19.1 86 100.0

Private-non profit 15 34.9 10 14.7 – –

Other 10 23.3 0 0.0 – –

No response 0 0.0 1 1.4 – –

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–1044 1037

68 of their adopted children. In order to recruit participants, gay and lesbian parenting

support groups and informational web sites were located using common Internet search

engines. Several sites were then contacted by electronic mail to solicit respondents.

Volunteers and inquiries were directed to the primary researcher, who provided informed

consent information and questionnaires, where appropriate. The primary researcher

screened potential participants for eligibility by ensuring that respondents were gay or

lesbian adoptive parents. Questionnaires were returned to the primary researcher by mail.

The majority of the adopted children were male (53%) and over 67% were minority (10%

African American, 21% Latino, 21% bi-racial and 16% Asian) children. The mean age at

adoption and at the time of the interview was 1.69 and 6.09, respectively. Over 25% had

some form of disability. On average, these children had one previous placement prior to

adoption, and over 35% were reported to have had a history of abuse prior to adoption.

In terms of the characteristics of adoptive parents in the second set, 51% were lesbian

parents. A majority of them were Caucasian (92%) and had over 12 years of education

(95%). Only 13% indicated previous foster parent experience. About 31% adopted their

children privately, 34% through the international program and 15% through private non-

profit organizations. Only 19% adopted their children through CPS.

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441038

The third data set included a total of 25 heterosexual adoptive parents (respondents)

who also reported demographic, historical, and behavioral information on 43 of their

adopted children. In a similar fashion, networks of adoption support groups and

informational web sites were used to solicit heterosexual adoptive parents. Again,

volunteers and inquiries were directed to the primary researcher, who provided informed

consent information and questionnaires, where appropriate. Preliminary eligibility

screening by the primary researcher ensured that respondents were heterosexual adoptive

parents. Questionnaires were returned to the primary researcher by mail. The majority of

the adopted children were male (54%) and over 74% were minority (21% African

American, 9% Latino, 33% bi-racial and 12% Asian) children. The mean age at adoption

and at the time of the interview was 5.98 and 7.08, respectively. Over 26% had some form

of disability. On average, these children had a mean of .58 previous placements prior to

adoption, and over 16% were reported to have had a history of abuse prior to adoption.

In terms of the characteristics of the adoptive parents in the third data set, 96% were

female. A majority of them were Caucasian (88%) and had over 12 years of education

(88%). Only 24% indicated that they had previous foster parent experience. About 19%

adopted their children privately, 16% through international programs and, 35% through

private non-profit organizations. Only 7% adopted their children through CPS.

3.2. Measures

3.2.1. Self-Report Family Functioning (SFI)

The instrument used to assess family functioning for participants in the first data set

(children with special needs) was a subscale adapted from the Self-Report Family

Functioning (SFI) scale known as Family Health. The SFI has good internal consistency

with an alpha coefficient of .80 (Beavers, Hampson, & Hulgus, 1985). Total Family Health

scores are determined by summing the scores from each scale item. High scores indicate

better family functioning. This score was reversed so that it could be combined with the

FAM-III standardized score.

3.2.2. Family Assessment Measure III, General Scale

The Family Assessment Measure III, General Scale (FAM-III) was used with the

second and third data sets (heterosexual and homosexual adoptive parents). The FAM-III

is a 50-item self-report instrument that quantifies indices of family strengths and weakness

on a Likert-type scale. Internal validity is supported by an alpha coefficient of .93. The

total scale scoring of FAM-III is used in this study as a measure of family functioning,

ranging from optimal to less than adequate, with low scores indicating better family

functioning (Skinner, Steinhauer, & Santa-Barbara, 1995; Skinner, Steinhauer, &

Sitarenios, 2000).

3.2.3. Eyberg Child Behavior Inventory (ECBI)

The instrument used to assess the adopted child’s behavior in the first data set was the

Eyberg Child Behavior Inventory (ECBI) (Psychological Assessment Resources, Inc.,

1974). Specifically, parents completed the intensity scale from the ECBI. The ECBI

intensity scale is a 36-item instrument designed to measure the intensity of conduct

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–1044 1039

problems in children and adolescents (Eyberg & Ross, 1978). The intensity scale is

presented in a Likert-type scale format. Scale items are summed to obtain a total behavior

problem score for the scale. The ECBI intensity scale has excellent internal consistency

with an alpha coefficient of .93 (Fischer & Corcoran, 1987). A high score indicates more

child behavior problems.

3.2.4. Child Behavior Checklists (CBCL)

The Child Behavior Checklists (CBCL) were used with the second and third data sets

(heterosexual and homosexual adoptive parents). The adopted child’s behavioral

functioning was assessed with two versions of the CBCL. The CBCL/4–18 (Achenbach,

1991) is a 113-item self-report instrument which was used for adoptive parents with

children between the ages of 4 and 18 years of age. The CBCL/2–3 (Achenbach, 1992) is a

100-item self-report instrument which was used for adoptive parents with children

between the ages of 2 and 3. Both parent assessments are on a Likert-type scale, which

produces a total problem score composed of internalizing and externalizing subscales

(Achenbach & Edelbrock, 1983). The CBCL/2–3 and CBCL/4–18 are highly correlated.

Internal validity is appropriate for research, with alpha coefficients ranging from .65 to .91

on all subscales (Achenbach, 1991, 1992; Achenbach & Edelbrock, 1983). Again, a high

score indicates more child behavior problems.

3.2.5. Family Support Scale (FSS)

A modified version of the Family Support Scale (FSS) was used with all three data sets.

The FSS is an 18-item self-report instrument. The FSS has an alpha coefficient of .79,

indicating good internal consistency. Scoring is on a Likert-type scale with higher scores

representing higher perceptions of helpfulness from support networks (Dunst & Trivette,

1988a, 1988b). A high score indicates more family support.

The unit of measurement for this study was the child. Thus, a parent’s perception of

family functioning score and a parent’s perception of helpfulness from support networks

score was matched with each child’s total behavior problem score for all adopted children

from each family. In order to make the data comparable, the measures to assess family

functioning (FAM-III and SFI: hereafter referred to as the standardized functioning score)

and child’s total behavior problem (EBCI and CBCL: hereafter referred to as the

standardized child behavior score) were converted to a standardized z score in all three

data sets.

4. Results

Pearson’s correlations were utilized to determine the relationship between the

standardized family functioning score and socio-economic/demographic variables to

include: standardized child behavior score, total support scale score, number of previous

placements of child, age at adoption and age at the time of the interview. The data

indicated that the standardized child behavior score (r=.258, p=.000), child’s age at

adoption (r=.275, p=.000) and child’s age at the time of the interview (r=.279, p=.001)

were positively associated with the standardized family functioning score. However, total

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441040

support scale score (r=�.136, p=.012) was negatively correlated with the standardized family functioning score. No significant relationship was found between the number of

previous child placements and standardized family functioning scores.

T-tests (see Table 3) were performed to determine the impact of the socio-economic and

demographic characteristics of the child (ethnicity, gender, disability, previous abuse and

sibling adoption) relative to the standardized family functioning score. The results

indicated that those children who were diagnosed to have a disability (t=3.06, df=198,

p=.003), had been previously abused (t=4.104, df=199, p=.000), and were adopted as a

sibling group (t=4.028, df=199, p=.000) had significantly lower family functioning (or

high standardized family functioning score) than those lacking such conditions. However,

no statistically significant relationships were found between ethnicity, gender and

standardized family functioning score.

Based on the literature and the preliminary statistical analysis, a step-wise multiple

regression procedure was performed between the dependent variable, standardized family

functioning score, and the demographic independent variables, which included the

standardized child behavior score, special needs adoption (1=special needs; 0=other),

homosexual adoptive parent (1=homosexual; 0=other), age at adoption (age at interview

was excluded, as it would present a multicollinearity problem), diagnoses of disabilities

(1=diagnosis; 0=none), total support scale score, number of previous placements, previous

abuse (1=yes; 0=no) and sibling adoption (1=non-sibling; 0=sibling). These variables had

demonstrated statistical significance in the t-tests and correlation analysis, and were

supported in the literature as having a relationship with family functioning. Multiple

regression assumptions were checked and none were violated. A power analysis was also

conducted to ensure that the sample size was adequate for the multiple regression analysis.

The results of the regression analysis indicated that six variables contributed significantly

Table 3

T-test of standardized family functioning score

Variable N Mean S.D. t

Diagnosed to have a disability

Yes 45 .2135 1.0651

No 156 �.2222 .9484 3.060TT Previous abuse

Yes 84 .3112 .9825

No 117 .2669 .9868 4.104TTT Ethnicity

Caucasian 70 �.0420 1.0201 Non-Caucasian 274 .0107 .9927 �.395

Child adopted as sibling group

Yes 45 .4965 .9794

No 156 �.1758 .9884 4.028TTT Gender of child

Male 106 �.1521 .9881 Female 95 .1162 1.0483 �1.868 TT pb.01. TTT pb.001.

Table 4

Multiple regression of standardized family functioning score

Variable R 2

B BETA t Significance

Age at adoption .078 .121 .368 4.101 .000TTT Child adopted as sibling group .108 �.492 �.199 �2.793 .006TT Diagnosed to have a disability .128 .525 .258 3.351 .001TTT Special needs adoption .168 �.824 �.405 �4.325 .000TTT Number of previous placements .198 �.059 �.168 �2.350 .020T Interact between homosexual .217 �.133 �.165 �2.157 .032T Adoptive parent and age at adoption constant

.751 2.133 .032T

F(6,184)=8.514, p=.000; R 2 =.217.

Lower standardized family functioning score indicates higher level of family functioning.

T pb.05. TT pb.01. TTT pb.001.

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–1044 1041

to the prediction of the standardized family functioning score, F(6,184)=8.514, p=.000

(see Table 4). Four other variables, including the standardized child behavior score,

homosexual adoptive parent, total support scale score and previous abuse were excluded in

the analysis as they were not statistically significant in the equation. These six variables

accounted for a total of 21.7% of the variance. The data indicated that those families who

adopted a child at a younger age would have a lower standardized family functioning score

(or better family functioning). Additionally, an adopted child who was not adopted as part

of a sibling group, was not diagnosed to have a disability, qualified for special needs

adoption, had experienced multiple placements, or had been adopted by a homosexual

adoptive parent at an older age (interaction effect created by multiplying the homosexual

adoptive parent variable with the age of the adopted child) would have a lower

standardized family functioning score (or better family functioning).

5. Discussion and implications for practice

The regression analysis indicates that a child’s age at adoption, sibling group, disability

and special needs adoption status, number of previous placements and the interaction effect

between homosexual adoptive parent and child’s age at adoption are associated with family

functioning. The results of this study are generally consistent with the findings from the

current literature. A substantial body of research supports the premise that adoption

disruption (or poorer family functioning) increases as the age of child at the time of adoption

increases (Boyne, Denby, Kettenring, & Wheeler, 1984; McDonald et al., 1991; Rosenthal,

1993). However, previous research has been inconsistent regarding the impact of sibling

group placements on adoptive family functioning when compared with single child

adoptions (Benton, 1985; Rosenthal, Schmidt, & Conner, 1988). Several studies indicate a

strong correlation between adoption disruption and an adopted child’s behavioral and legal

problems, psychological difficulties, and/or a history of abuse or neglect (Barth & Berry,

1988; Festinger, 1986; Rosenthal & Groze, 1992). Glidden’s (1991) study of long-term

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441042

outcomes of families who adopted children with special needs found that several years after

adoption, mothers reported that their families functioned quite well. The findings indicated

that these adoptive parents scored similarly to parents of children without disabilities on

family functioning. As reported previously, several studies indicate that the number of

previous placements is correlated with adoption disruption (Barth & Berry, 1988; Festinger,

1986; Rosenthal & Groze, 1992). However, the findings from this study suggest that higher

number of previous placements was related to better family functioning. Children who had

higher number of previous placements might find it easier to adapt to the new environment.

As a result, the level of family functioning might be significantly higher. Research regarding

homosexual adoptive families still remains largely unavailable (Brooks & Goldberg, 2001).

This study found no negative effects upon family functioning associated with gay/lesbian

sexual orientation of adoptive parents. Further, the results suggest family functioning was

actually enhanced when homosexual families adopted older children.

The purpose of this study was to identify factors that may contribute to family

functioning in adoptive homosexual families, heterosexual families, and families who

adopted children with special needs. The results from this study suggest that adoption

agencies should place a child in an adoptive family as early as possible, regardless of

the parent’s sexual orientation. This study also suggests that sibling adoption is

associated with lower family functioning, presumably because these families often

require more of scarce resources to support family functioning. Therefore, sibling

adoptions should only be consummated when adequate post-adoptive support services

and financial supports are available to adoptive families. The availability and provision

of post-adoptive support services can most effectively be enhanced through changes in

public policy that increase funding for these vital services. Furthermore, when a child

is diagnosed as having a disability, families are likely to require stronger support

networks. The findings from this study also indicated that special needs adoptions are

associated with higher levels of family functioning. The children with special needs in

this study were adopted through government-funded agencies specializing in placing

children with special needs status. The expertise of these agencies, and their

connections with a wide variety of post-adoption services, may represent a plausible

rationale for their clients’ high levels of reported family functioning. Finally, no

previous research supports the frequently held belief that lesbian and gay adults or

couples are less effective parents than their heterosexual counterparts. In fact, this

study found no negative effects regarding the parenting of adopted children by gay and

lesbian adults and couples. Moreover, support for the practice of adoption by gay and

lesbian adults and couples, especially with older children, was indicated by this study.

In brief, states remain bereft of data to support the continued discrimination against

lesbian and gay adults as adoptive parents. Research with gay and lesbian adoptive

parents should be continued to further explore the effects and previously reported

benefits (Brooks & Goldberg, 2001) of such adoptions.

In summary, the limitations of this study are notable. Convenience sampling and

cross-sectional studies often help to yield larger study samples, but they are limited in

terms of their effectiveness against several threats to internal and external validity.

Adopted sibling family functioning measures cannot be treated as independent scores. In

addition, the use of self-report questionnaires may lead to erroneous declarations to

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–1044 1043

researchers. Since random samples are not easily available, and therefore not used in this

study, caution should be exercised in generalizing the results of this study to a larger

population.

References

Achenbach, T. M. (1991). Manual for the child behavior checklist/4–18 and 1991 profile. Burlington, VT7

University of Vermont Department of Psychiatry.

Achenbach, T. M. (1992). Manual for the child behavior checklist/2–3 and 1992 profile. Burlington, VT7

University of Vermont Department of Psychiatry.

Achenbach, T. M., & Edelbrock, C. (1983). Manual for the child behavior checklist and revised child behavior

profile. Burlington, VT7 University of Vermont, Department of Psychiatry.

American Civil Liberties Union. (1999). Fact sheet: Overview of lesbian and gay parenting, adoption and foster

care. New York7 Author.

Barker, R. L. (1995). The social work dictionary, (3rd ed.). Washington, DC7 NASW Press.

Barth, R., & Berry, M. (1988). Adoption and disruption: Rates, risks, and responses. New York7 Aldine/

DeGruyter.

Beavers, R., Hampson, R., & Hulgus, Y. (1985). Commentary: The beavers systems approach to family

assessment. Family Process, 9(24), 398–405.

Benton, B. (1985). Evaluation of state activities with regard to adoption disruption. Washington, DC7 Urban

Systems Research and Engineering.

Bigner, J. J., & Bozett, F. W. (1989a). Parenting by gay fathers. Marriage and Family Review, 14, 155–175.

Bigner, J. J., & Jacobsen, R. B. (1989b). Parenting behaviors of homosexual and heterosexual fathers. Journal of

Homosexuality, 18, 173–186.

Boyne, J., Denby, L., Kettenring, J., & Wheeler, W. (1984). The shadow of success: A statistical analysis of

outcomes of adoptions of hard to place children. Westfield, NJ7 Spaulding for Children.

Brodzinsky, D. M. (2003). Adoption by lesbians and gays: A national survey of adoption agency policies,

practices, and attitudes. New York7 Evan B. Donaldson Adoption Institute.

Brooks, D., & Goldberg, S. (2001). Gay and lesbian adoptive and foster care placements: Can they meet the needs

of waiting children? Social Work, 46(2), 147–157.

Chan, R. W., Raboy, B., & Patterson, C. J. (1998). Psychosocial adjustment among children conceived via donor

insemination by lesbian and heterosexual mothers. Child Development, 69(2), 443–457.

Cherlin, A. J., & Furstenberg, F. F. (1994). Stepfamilies in the United States: A reconsideration. Annual Review of

Sociology, 20, 359–381.

Dunst, C., & Trivette, C. (1988a). A family systems model of early intervention with handicapped and

developmentally at-risk children. In D. Powell (Ed.), Parent education as early childhood intervention:

Emerging directions in theory, research, and practice (pp. 131–180). Norwood, NJ7 Ablex.

Dunst, C., & Trivette, C. (1988b). Determinants of caregiver styles of interaction used with developmental at-risk

children. In K. Marfo (Ed.), Parent–child interaction and developmental disabilities: Theory, research, and

intervention (pp. 3–31). New York7 Praeger Publishers.

Erich, S., Leung, P., & Kindle, P. (in press). An analysis of adoptive family functioning: Outcomes for

families with gay and lesbian adoptive parents and heterosexual adoptive parents. Journal of GLBT

Family Studies.

Erich, S., Leung, P., Kindle, P., & Carter, S., (2004). Gay and lesbian adoptive families: An exploratory

study of family functioning, adoptive child’s behavior, and familial support networks. (Submitted for

publication).

Eyberg, S., & Ross, A. (1978). Assessment of child behavior problems: The validation of a new inventory.

Journal of Clinical Child Psychology, 7(2), 113–116.

Festinger, T. (1986). Necessary risk: A study of adoptions and disrupted adoptive placements. New York7 Child

Welfare League of America.

Fischer, J., & Corcoran, K. (1987). Measures for clinical practice. New York7 The Free Press.

P. Leung et al. / Children and Youth Services Review 27 (2005) 1031–10441044

Furstenberg, F. F. (1999). Is the modern family a threat to children’s health? Society, 36, 31–37.

Glidden, L. (1991). Adopted children with developmental disabilities: Post-placement family functioning.

Children and Youth Services Review, 13(5 and 6), 363–377.

Goldenberg, I., & Goldenberg, H. (1996). Family therapy: An overview, (4th ed.). Pacific Grove, CA7 Brooks

& Cole.

Golombok, S., Spencer, A., & Rutter, M. (1983). Children in lesbian and single-parent households: Psychosexual

and psychiatric appraisal. Journal of Child Psychology and Psychiatry, 24(4), 551–572.

Kagen, R., & Reid, W. (1986). Critical factors in the adoption of emotionally disturbed youths. Child Welfare,

65(1), 63–73.

Leung, P., & Erich, S. (2002). Family functioning of adoptive children with special needs: Implications of familial

supports and child characteristics. Children and Youth Services Review, 24(11), 641–658.

McDonald, T., Liberman, A., Partridge, S., & Hornby, H. (1991). Assessing the role of agency services in

reducing adoption disruption. Children and Youth Services Review, 13(5–6), 425–438.

Miller, J. A., Jacobsen, R. B., & Bigner, J. J. (1981). The child’s home environment for lesbian vs. heterosexual

mothers: A neglected are of research. Journal of Homosexuality, 7(1), 49–56.

National Adoption Information Clearinghouse. (1999, July). Adopting children with developmental disabilities.

Washington, DC7 Author.

National Association of Social Workers. (1999, August). Statement on family policy. Washington, DC7 Author.

Psychological Assessment Resources, Inc. (1974). Eyberg child behavior inventory. Lutz, FL7 Author.

Rosenthal, J. (1993). Outcomes of adoption of children with special need. Adoption: The Future of Children, 3(1),

77–88.

Rosenthal, J., & Groze, V. (1992). Special needs adoptions: A study of intact families. Westport, CT7 Praeger.

Rosenthal, J., Schmidt, D., & Conner, J. (1988). Predictors of special needs adoption disruption: An exploratory

study. Children and Youth Services Review, 10(2), 101–117.

Rosenthal, J. A., & Groze, V. (1991). Adoption outcomes for children with handicaps. Child Welfare, 70(6),

623–636.

Skinner, H., Steinhauer, P., & Santa-Barbara, J. (1995). Family assessment measure version III. North

Tonawanda, NY7 Multi-Health Systems, Inc.

Skinner, H. A., Steinhauer, P. D., & Sitarenios, G. (2000). Family assessment measures (FAM) and process model

of family functioning. Journal of Family Therapy, 22, 190–210.

Steinhauer, P. D., Santa-Barbara, J., & Skinner, H. (1984). The process model of family functioning. Canadian

Journal of Psychiatry, 29(2), 77–88.

Tasker, F., & Golombok, S. (1995). Adults raised as children in lesbian families. American Journal of

Orthopsychiatry, 65(2), 203–215.

U.S. Department of Health and Human Services, Administration for Children and Families, Children’s Bureau.

(2003, March). Adoption and foster care analysis reporting system: Preliminary FY2001 estimates (AFCARS

Report No. 8). Washington, DC: Author.

Van Voorhis, R., & Wagner, M. (2001). Coverage of gay and lesbian subject matter in social work journals.

Journal of Social Work Education, 37(1), 147–157.

Westhues, A., & Cohen, J. (1990). Predicting disruption of special needs adoptions. Child Welfare, 69(2),

141–155.