Preschool selection considerations and experiences of school mistreatment among lesbian, gay, and heterosexual adoptive parents

Abstract

The current study is the first to investigate the school selection considerations and school-related experiences of sexual-minority parents with young children. The sample consisted of 210 parents in 105 couples, including 35 lesbian couples, 30 gay male couples, and 40 heterosexual couples, all of whom had adopted a child three years earlier. We found that parents with less income were more likely to consider cost in choosing a preschool, and parents with less education were more likely to consider location. More educated parents tended to emphasize racial diversity and the presence of adoptive families, and, among sexual-minority parents, the presence of other lesbian/gay parents. Sexual-minority parents were more likely to consider racial diversity than heterosexual parents. In reporting on their experiences with schools, heterosexual parents were more likely to perceive mistreatment due to their adoptive status than sexual-minority parents, and sexual-minority parents living in less gay-friendly communities were more likely to perceive mistreatment due to their sexual orientation than sexual-minority parents living in more gay-friendly communities. Our findings have implications for early childhood educators and administrators seeking to create an inclusive learning community for all types of families.

1. Introduction

Despite advances in securing equal rights for sexual minorities and their families, sexual-minority (i.e., lesbian, gay, and bisexual; LGB) parents and their children encounter explicit and implicit forms of marginalization, exclusion, and stigma embedded in various societal institutions, such as the legal system and the schools (Byard, Kosciw, & Bartkiewicz, 2013; Goldberg, 2010). Within the school context, sexual-minority parent families may encounter teachers, school staff, and other parents who possess ambivalent or unsupportive attitudes toward families like their own (Gartrell et al., 2000). Such attitudes may manifest in either lack of acknowledgment or explicit stigmatization of children with sexual-minority parents. On a more subtle level, marginalization of these families may be embedded in the curriculum (e.g., by focusing entirely on the experiences of heterosexual people and families) and school paperwork (e.g., by failing to allow for representation of diverse family forms; Byard et al., 2013). Sexual-minority parent families who are “different” in additional ways, beyond parents’ sexual orientation, may also be vulnerable to bias and exclusion in school settings. Many sexual-minority parent families are adoptive and/or multiracial (Gates, Badgett, Macomber, & Chambers, 2007), introducing other dimensions of difference that may meet challenges in the school setting.

Little research has examined sexual-minority parents’ experiences in school settings. The little research that exists has focused primarily on their experiences in elementary school settings (Kosciw & Diaz, 2008), as opposed to their experiences in early childhood settings, where they may have more contact with teachers and staff (Beveridge, 2005), and thus may be more attuned to insensitivities directed at them and/or their child(ren). Indeed, we know little about the family–school interface of sexual-minority parent families with young children, including their school selection process and potential experiences of exclusion and mistreatment within their children’s schools. Likewise, we know little about the school experiences of sexual-minority parents who have adopted their children, and how such experiences may be shaped by adoption- or race-related factors. The current exploratory study examines the experiences of lesbian, gay, and heterosexual adoptive parents of preschool-age children with respect to their (a) school-related selection considerations; and (b) perceived experiences of mistreatment at school.

1.1. School decision-making and selection in parents of preschool-age children

Research on the school-related concerns of sexual-minority parents of school-age children suggests that parents are often aware of the potential for homophobic bullying at school (Gartrell et al., 2000). Several studies further suggest that some sexual-minority parents purposefully seek out progressive and diverse schools and communities in an effort to decrease the stigma to which their children are exposed (Casper & Schultz, 1999; Kosciw & Diaz, 2008; Mercier & Harold, 2003). Such efforts may be particularly pronounced among sexual-minority parents of children of color (Kosciw & Diaz, 2008). Parents who adopt children of color face the possibility that their children might be mistreated (i.e., confront stereotypes and ignorance) on the basis of their race; in turn, White sexual-minority parents of children of color recognize that their children might be mistreated on the basis of both their race and family structure (Goldberg, 2009). White parents who adopt children of color also confront the possibility that the multiracial makeup of their families – the fact that children look “different” from them – may render their children vulnerable to intrusive questions about adoption (e.g., “where are you from?”; “who are your ‘real’ parents?”; Vaschenko, D’Aleo, & Pinderhughes, 2012).

Parents of preschool-age children play a greater role in selecting their children’s school environments than parents of school-age children. Whereas most children in elementary school attend their local public schools, most children in preschool attend private programs, although some public programs (which are funded by the city and/or state) are available (Magnuson & Waldfogel, 2005). In turn, parents – especially middle-class parents – typically play an active role in selecting preschool or day-care programs for their children (Cryer, Tietze, & Wessels, 2002). Research on heterosexual parents’ school selection process has found that parents tend to consider a range of factors in choosing schools and daycares for their young children, with frequently mentioned concerns including logistical/practical factors (e.g., cost, location), school quality considerations (e.g., the school curriculum or philosophy), and, more rarely, value-related considerations (i.e., the degree to which parents’ values match the schools’ values) (Galotti & Tinkelenberg, 2009; Glenn-Applegate, Pentimonti, & Justice, 2011). Working-class parents may be more likely to emphasize practical concerns such as cost and location in their selection process than middle-class parents, in part because of differing constraints on their choices (e.g., in terms of money, transportation, and time; Smrekar & Goldring, 1999); but also because of differing views of education (e.g., middle-class parents may be more likely to view education as a “calculated decision that matches the values and attributes of the family… and child… to the best-fitting school”; Goyette, 2008, p. 117; Wells & Crain, 1997). Likewise, middle-class parents may be more likely to emphasize school quality in their selection process (Peyton, Jacobs, O’Brien, & Roy, 2001), perhaps in part because, as Larner and Phillips (1994) note, parents with more education tend to view the role of early childhood education as a key context for learning and preparing for grade school, and as a starting point for children’s long-term educational success.

No known research has focused on whether and to what extent the above school selection considerations are endorsed by sexual-minority parents of young children. It is expected that, like parents of children with special needs (Glenn-Applegate et al., 2011) and parents of bilingual children (McClain, 2010), sexual-minority parents – as well as adoptive parents – may consider additional issues related to the inclusiveness of the school community. Understanding what factors sexual-minority parents consider in selecting early educational environments for their children is important, as it will provide crucial insight into the school-related concerns and values of sexual-minority parents, and can inform teaching and practice in early childhood education. Furthermore, knowledge of sexual-minority parents’ school-related considerations is of interest in that they may foreshadow what types of early childhood environments their children ultimately inhabit. Such knowledge is important, given that the early childhood educational environment impacts children’s social, emotional, and cognitive development (Burger, 2010).

1.2. Sexual-minority parents’ school selection

Research on school selection among sexual-minority parents with elementary school-age children can lend some insight into the school selection considerations of sexual-minority parents of preschoolers. Key data on this topic come from the Gay, Lesbian, and Straight Education Network (GLSEN)’s 2008 survey of 588 LGBT parents, most of whom were women and had a child in elementary school (Kosciw & Diaz, 2008). Although most parents reported that their children attended public schools (78%), this percentage was significantly lower than the national percentage (89%). Of the parents who sent their children to private schools, these schools were less likely to be religiously-affiliated schools (7%) and more likely to be independent schools (16%) than national percentages. Regarding LGBT parents’ reasons for selecting their children’s schools, parents most often reported that they chose the local or neighborhood school (59%) and that they chose the school based on academic reputation (54%). Other common reasons cited were the diversity of the school population (31%), the school’s reputation for valuing diversity (22%), they knew other families at the school (29%), the sports/arts/music reputation of the school (29%), they knew that there were other children with LGBT parents there (17%), the school’s reputation for being welcoming of LGBT families (17%), the academic approach (e.g., Montessori) (13%), special education services (12%), and language programs (10%). Parents of children of color were more likely to choose schools based on the diversity of the school population (43%) than were parents with a White student (25%), regardless of the race/ethnicity of the parents (about 16% of the families represented had White parent(s) and a child of color, and 14% of the families represented were comprised of one White and one non-White parent; Kosciw & Diaz, 2008).

In addition to the GLSEN survey, several qualitative studies have examined the school-related experiences of lesbian parents; indeed, research on gay fathers’ school-related experiences is particularly sparse. Mercier and Harold (2003) interviewed 15 lesbian-parent families with children ranging from six months to 18 years. Similar to the GLSEN sample, the parents in this study often emphasized the importance of sending their children to schools that valued diversity, because they believed that “schools that value diversity of any type are more likely to respond well to lesbian-parent families” (p. 39). Consistent with this, Gartrell et al. (1999) interviewed 84 lesbian-parent families with toddlers about their plans for child care or preschool and found that 87% of mothers said that they planned to enroll their children in programs that included children and teachers of different social classes, genders, races, ethnicities, and cultures, out of a belief that “exposure to diversity was the most effective method of fortifying their children against homo-phobia” (p. 367). When the children in the study were five, 74% of the children’s schools were described as multicultural and 33% had lesbian/gay staff members; by extension, under one-fifth (18%) of children had reportedly experienced homophobia from peers or teachers (Gartrell et al., 2000). Of note is that Gartrell et al. sample was drawn from metropolitan areas (e.g., San Francisco), which may have facilitated access to multicultural, gay-friendly school environments.

Finally, a qualitative study of 20 lesbian-parent families in Australia found that parents who could afford to send their children to private school often did so, in part because they seemed to believe that these schools would be more likely to accept and be inclusive of their families (Lindsay et al., 2006). Thus, in sum, the existing research suggests that lesbian mothers value diversity, broadly defined, as well as racial diversity, specifically (Kosciw & Diaz, 2008).

1.3. Experiences of mistreatment in early childhood settings

The early parent–school relationship is important, in that it can have positive implications for child development, and also lays the foundation for future parent involvement and school connection (Beveridge, 2005). In addition to playing an active role in the selection of preschools, parents of young children may also actively interact with their children’s schools. Unlike parents of school-age children, parents of preschool-age children are usually in direct contact with the teachers and staff who are caring for their children, providing information about their children’s personality and developmental stage, as well as recommendations about how best to support their children’s growth (Beveridge, 2005; Rimm-Kaufman & Pianta, 2005). Parents of young children also often seek out more information from schools about what their children are learning, since their children may lack the cognitive or verbal skills to explain these details to them (Glover & Bruning, 1987) – although parents with more resources tend to have more communication with teachers (Bryant, Peisner-Feinberg, & Miller-Johnson, 2000; Lareau, 1987). Parents of young children may also inquire about their children’s social experiences at school; although negative peer behavior is less of a concern in early childhood settings than elementary school (Sawyer, Mishna, Pepler, & Wiener, 2011), teasing in these settings does occur (Kirves & Sajaniemi, 2012).

The early interactions between sexual-minority parents and early educational environments are of great significance, insomuch as these experiences can set the stage for parents’ expectations about and involvement in their children’s school lives more generally (Casper & Schultz, 1999), and, in turn, parent school involvement is related to child developmental outcomes, such that children with highly involved parents demonstrate greater social and cognitive skills, and fewer behavioral problems (Powell, Son, File, & San Juan, 2010). Perceived experiences of marginalization or mistreatment in early childhood settings, then, are particularly important to attend to, as they may have profound implications for sexual-minority parents’ school connectedness and involvement (Kosciw & Diaz, 2008). Some research has examined sexual-minority parents’ perceptions of exclusion and mistreatment in elementary school settings. More than half (53%) of the LGBT parents in the GLSEN survey described various forms of exclusion related to their sexual orientation from their children’s school communities (Kosciw & Diaz, 2008). For example, parents were told that they could not be aids in their children’s classrooms; that only one parent was allowed to attend a school event; and that their offers to assist with creating a more inclusive classroom were not welcome or needed. Further, 26% of LGBT parents reported being mistreated by other parents (e.g., being stared at or ignored). Not surprisingly, parents who reported higher levels of exclusion and mistreatment were less likely to be involved in volunteering at their children’s schools (Kosciw & Diaz, 2008).

There is some evidence that sexual-minority parents with greater educational, financial, and geographic resources may perceive less exclusion in their children’s schools. Casper and Schultz (1999) interviewed 17 lesbian and gay parents with children in preschool to 5th grade and found that parents who were middle-class, living in urban or “progressive” areas, and/or sent their children to private schools, reported more ease in being “out” and advocating for their children than parents with limited financial or geographic resources. In a rare study of working-class lesbian mothers, Nixon (2011) found that participants often felt like outsiders at their children’s schools due to both their socioeconomic status (and, specifically, their own poor school histories) and also their sexual orientation. In turn, very few mothers discussed their concerns about bullying directly with their children’s teachers, choosing instead to try to prepare their children to avoid, and if necessary cope with, bullying.

1.4. Adoptive parents’ experiences in school settings

Many sexual-minority parents adopt their children (Gates et al., 2007), creating additional dimensions of difference that they may consider in choosing school environments for their children. Of note is that a range of practical resources are available for adoptive parents of young children (e.g., that provide guidance on how to talk to teachers about their children’s adoptive status as well as how to educate teachers about adoption in general; Child Welfare Information Gateway, 2012; Gilmore & Bell, 2006). However, little empirical work has examined adoptive parents’ considerations in selecting schools for their children, or navigating and responding to adoption bias (e.g., the use of adoption-insensitive language by teachers and/or in curricula). The absence of this work is concerning, given that misinformation and stigma related to adoption – particularly transracial adoption – are still pervasive in the broader society (Goldberg, Kinkler, & Hines, 2011) and may trickle down into the attitudes and practices of teachers and other school personnel. Teachers may fail to understand or attend to the multiple dimensions of difference that may impinge upon the identity or experiences of adopted children (Enge, 1999). They may also neglect to discuss racial or family diversity in the classroom, perhaps because they believe that young children are too young to understand these issues (Husband, 2012; Robinson & Ferfolja, 2002), despite evidence to the contrary (Park, 2011).

Few studies have assessed adoptive parents’ experiences related to their children’s school environments. In a study of racial socialization practices among heterosexual parents who had adopted transracially, Vonk, Lee, and Crolley-Simic (2010) found that 54% of parents reported that they had chosen child care providers, teachers, or other role models similar to their children’s race or ethnicity. Other studies (Goldberg, 2009; Mercier & Harold, 2003) have also observed that White parents who adopt transracially sometimes describe efforts to move to an area that is racially diverse, so that their children will grow up around, and attend schools with, other people who look like them. These studies suggest that for parents who adopt a child of a different race, the racial diversity of teachers and schools may take on heightened importance.

Speaking to issues of inclusion versus exclusion in schools, Nowak-Fabrykowski, Helinski, and Buchstein (2009) surveyed 23 heterosexual foster parents and found that most respondents reported their children’s teachers and classrooms did not have any materials related to adoption, and felt that teachers should make more of an effort to assign lessons about adoption (e.g., during Adoption Month). Thus, respondents demonstrated a general sense that schools could be doing more than they were to incorporate the experiences and needs of adopted individuals and their families into their materials and curricula.

The research discussed thus far is limited by the fact that it has rarely included parents of young children, is primarily qualitative, and, when it includes sexual-minority parents, has tended to focus on lesbian mothers, not gay fathers. Despite these limitations, it provides insights into (a) the type of criteria that sexual-minority adoptive parents may value in choosing schools for their children; and (b) the possibility for sexual-minority adoptive parents to perceive mistreatment by their children’s schools. First, the literature suggests that sexual-minority parents may be particularly likely to value school diversity (e.g., with regard to family structure and race). Second, it suggests that sexual-minority parents may perceive mistreatment at their children’s schools, although such experiences may be mitigated by sending one’s child to a private school (Kosciw & Diaz, 2008; Lindsay et al., 2006), and, indirectly, by the presence of financial and educational resources, as well as living in a “progressive” or gay-friendly area (Casper & Schultz, 1999).

1.5. The current study

This exploratory study examines a sample of 210 parents in 105 couples – 35 lesbian two-parent families, 30 gay two-parent families, and 35 heterosexual two-parent families – all of whom had adopted a child three years earlier. Both partners in each couple were surveyed. At the time of assessment, children in the sample were between 3.0 and 5.5 years (M age = 3.47) and enrolled in a preschool. Based on the literature, we proposed the following exploratory research questions.

1.6. School selection considerations

Research Question 1

Does the frequency by which parents consider school selection factors such as cost, location, educational philosophy, religion/language, racial diversity, and the presence of adoptive families vary according to parent sexual orientation, child race, family income, and parent education level?

We hypothesize that sexual-minority parents will be more likely to consider racial diversity and the presence of other adoptive families; parents of children of color will be more likely to consider racial diversity; and higher levels of financial/educational resources will be related to greater consideration of “quality” factors such as educational philosophy and religion/language offerings, greater consideration of diversity factors such as racial diversity and the presence of other adoptive families, and lesser consideration of “practical” factors such as cost and location.

Research Question 2

Among sexual-minority parents, does the frequency by which parents consider the presence of other lesbian/gay-parent families and school gay-friendliness vary by parent gender, child race, family income, or parent education level?

These are exploratory questions, and we do not have hypotheses about the role of gender or race. We expect higher levels of financial/educational resources to be related to greater consideration of lesbian/gay-parent families and greater consideration of school gay-friendliness.

1.7. Perceived experiences of mistreatment

Research Question 3

Do parents’ perceptions of mistreatment related to adoptive status vary by parent sexual orientation, child race, family income, parent education level, or school type?

We do not have hypotheses about the role of parent sexual orientation. We hypothesize that parents of children of color will perceive more mistreatment; parents with fewer financial/educational resources will perceive more mistreatment; and parents whose children attend public preschools will perceive more mistreatment.

Research Question 4

Among sexual-minority parents, do parents’ perceptions of mistreatment related to sexual orientation vary by parent gender, child race, the perceived gay-friendliness of one’s community, family income, parent education, or school type?

We have no hypotheses about parent gender. We hypothesize that parents of children of color will perceive more mistreatment; parents who live in less gay-friendly areas will perceive more mistreatment; parents with fewer resources will perceive more mistreatment; and parents whose children attend public preschools will perceive more mistreatment.

2. Method

2.1. Description of the sample

Data were taken from a longitudinal study of the transition to adoptive parenthood. All 105 couples had adopted their first child three years earlier, and in all cases it was a single child. Respondents’ data were included in the current study if their adopted child was in preschool.

Descriptive data for the full sample, and by family type, appear in Table 1. ANOVA revealed that the average family incomes for lesbian-, gay-, and heterosexual-parent families differed significantly, F(2,103) = 5.29, p = .006, such that gay male couples (M = $193,572, Mdn = $150,000, SD = $132,641) had a significantly higher annual combined income than lesbian couples (M = $121,268, Mdn = $105,000, SD = $64,795), p < .001, and heterosexual couples (M = $138,377, Mdn = $120,000, SD = $85,483), p = .012. The sample as a whole is more affluent compared to national estimates for same-sex and heterosexual adoptive families, which indicate that the average household incomes for same-sex couples and heterosexual married couples with adopted children are $102,474 and $81,900, respectively (Gates et al., 2007). The sample as a whole was well-educated, M = 4.49 (SD = 1.01), where 4 = bachelor’s degree and 5 = master’s degree. Hierarchical linear modeling (HLM, in which parents were nested within couples) revealed no differences in education level by family type.

Table 1.

Demographics for the full sample and by family type.

	Family type
	Full sample (M, SD, or %)	Lesbian (M, SD, or %)	Gay (M, SD, or %)	Heterosexual (M, SD, or %)
Family income	$148,827 ($99,753)	$121,268 ($64,795)	$193,572 ($132,641)	$138,377 ($85,483)
Education	4.49 (1.01)	4.46 (1.04)	4.50 (1.01)	4.48 (.99)
Work hours	34.87 (17.70)	31.99 (16.72)	37.84 (15.81)	33.01 (19.32)
Hours in school (Child)	26.12 (12.39)	25.07 (12.48)	31.16 (11.81)	23.82 (12.05)
Child age	3.47 (.99)	3.37 (.98)	3.39 (.96)	3.56 (1.04)
Parent race (White)	91%	91%	95%	88%
Child race (White)	35%	25%	48%	33%
Child gender (Boys)	52%	49%	58%	53%

The adoptive parents in the sample were mostly White (91%), whereas the children in the sample were disproportionately of color (i.e., non-White, including biracial children): namely, 65% of couples adopted children of color. Fifty-two percent of couples adopted boys, and 48% of couples adopted girls. Chi square analyses indicated that the distributions of parent race, child race, and child gender did not significantly differ by family type.

The average age of the children in the sample was 3.47 years (SD = .99); ANOVA showed that child age did not differ by family type. ANOVA revealed that the number of hours that children spent in preschool differed by family type, F(2,103) = 3.48, p = .034, such that the children of gay male couples were in school for significantly more hours per week (M = 31.16, SD = 11.81) than the children of lesbian couples (M = 25.07, SD = 12.48), p = .047, and heterosexual couples (M = 23.82, SD = 12.05), p = .012. Somewhat paralleling these data, MLM revealed that the number of hours parents worked per week differed by family type, F(2,103) = 3.22, p = .044, such that gay male parents worked significantly more hours than lesbian parents (M = 37.84, SE = 1.81, versus M = 31.99, SE = 1.64), p = .034, but not heterosexual parents (M = 33.01, SE = 1.60).

The types of preschool environments that children were enrolled in varied. One quarter of the sample reported that their children attended public preschools (e.g., YMCA-based programs), and three quarters of the sample reported that their children attended private preschools. Within the latter group, 33% were described as private day care-based programs, 20% were Montessori schools, 9% were religiously oriented or affiliated preschools (e.g., Catholic, Christian, Lutheran), and the remainder were given a wide range of descriptors (e.g., Waldorf; French Immersion; country day school; university-based).

2.2. Recruitment and procedures

Inclusion criteria for the larger study from which this sample was drawn were: (a) couples must be adopting their first child; and (b) both partners must be becoming parents for the first time. Participants were originally recruited during the pre-adoptive period (i.e., while couples were waiting for a child placement). Adoption agencies throughout the US were asked to provide study information to clients who had not yet adopted, typically in the form of a brochure which invited them to participate in a study of the transition to adoptive parenthood. We explicitly invited both same-sex and heterosexual couples to participate, since a major goal of the study was to understand how same-sex couples, specifically, experienced the transition to adoptive parenthood. Toward this end, U.S. census data were utilized to identify states with a high percentage of same-sex couples (Gates & Ost, 2004) and effort was made to contact agencies in those states. We recruited both heterosexual and same-sex couples through these agencies, in an effort to match couples roughly on geographic status and financial resources. Over 30 agencies provided information to their clients, and interested couples were asked to contact the principal investigator for details. Because some same-sex couples may not be “out” to agencies about their sexual orientation, several national gay organizations also assisted in disseminating study information.

Three years after they had been placed with a child, parents in the original study were contacted and asked to complete an in-depth questionnaire packet that focused on their experiences with regard to their children’s preschools. Questionnaires included closed- and open-ended items that addressed parents’ school-related values and experiences. Lesbian and gay parents – that is, parents who had identified themselves as being in a same-sex relationship at the first assessment point – were mailed questionnaires with several additional questionnaires and items aimed at identifying unique aspects of their experience as sexual-minority parents. The data for the current study are drawn from this three-year post-adoptive placement assessment point.

2.3. Measures

2.3.1. Outcomes

2.3.1.1. School selection factors

Parents were presented with the question, “What factors did you consider in choosing a school?” and asked to circle all that apply: 1. Cost; 2. Educational philosophy; 3. Religion or language emphasized at school; 4. Racial diversity; 5. Presence of other adoptive families; 6. Presence of other lesbian/gay-parent families; 7. Gay-friendliness of school; 8. Other (please list). Items 6 and 7 were asked of sexual minorities only. In the “other” category, the only factor that was listed frequently enough to treat as an outcome was “Location.” Thus, we explore how family type and child race relate to parents’ consideration of eight selection factors.

2.3.1.2. Mistreatment due to adoptive status

School mistreatment due to adoptive status was assessed using an eight-item measure developed for this study. Prior research and relevant popular press literature informed the development of this measure (Child Welfare Information Gateway, 2012; Kosciw & Diaz, 2008). Further, it was reviewed by several adoption scholars who provided feedback; several items were revised in response to their feedback. The measure assesses exclusion and mistreatment by teachers, school personnel, and other parents, related to the child’s adoptive status. Parents responded to these six items using a 1–5 scale (1 = not at all true, 5 = very true): 1. I have felt that my parenting skills were questioned because I am an adoptive parent; 2. I have felt mistreated by school staff because I am an adoptive parent; 3. I have felt that staff members/school personnel treat my child differently because he/she is adopted; 4. My child’s teacher uses language that acknowledges adoptive families (reverse scored); 5. My child’s teacher sensitively handles assignments that could be hurtful to adoptive families (e.g., family trees) (reverse scored); 6. My child’s school uses forms that allow families to identify themselves in the way that they choose (reverse scored). In addition, parents responded to the following two items using a different 1–5 scale (1 = not at all excluded, 5 = very excluded): 1. To what degree do you feel excluded from your child’s school on the basis of your status as an adoptive family?; 2. To what degree do you feel excluded by the parents of your children’s peers on the basis of your status as an adoptive family? An exploratory factor analysis (EFA) using principal components analysis (PCA) showed acceptable factor loadings from .41 to .64. A composite score was created by taking the average across the individual items. Alpha = .71 for this scale (.70 for lesbians, .68 for gay men, and .73 for heterosexual parents).

2.3.1.3. Mistreatment due to parental sexual orientation

School mistreatment due to sexual orientation was assessed using an eight-item measure developed for this study. Prior work informed the development of this measure (Casper & Schultz, 1999; Kosciw & Diaz, 2008). The measure was reviewed by several scholars who studied sexual-minority parent families, and several items were revised in response to their input. This measure assesses exclusion and mistreatment by teachers, school personnel, and other parents, related to parents’ sexual orientation, and was completed by lesbian and gay parents only. Parents responded to the following six items using a 1–5 response scale (1 = not at all true, 5 = very true): 1. I have felt that my parenting skills were questioned because I am a lesbian/gay parent; 2. I have felt mistreated by school staff because I am a lesbian/gay parent; 3. I have felt that staff members/school personnel treat my child differently because his/her parents are lesbian/gay; 4. My child’s teacher uses language that acknowledges lesbian/gay parent families (reverse scored); 5. My child’s teacher sensitively handles assignments that could be hurtful to lesbian/gay-parent families (e.g., Mother’s Day/Father’s Day) (reverse scored); 6. My child’s school uses forms that allow families to identify themselves in the way that they choose (reverse scored). Parents also responded to the following two items using 1–5 scale (1 = not at all excluded, 5 = very excluded): 1. To what degree do you feel excluded from your child’s school on the basis of your sexual orientation?; 2. To what degree do you feel excluded by the parents of your children’s peers on the basis of your sexual orientation? An EFA using PCA for extraction showed factor loadings from .36 to .63. The one item with a loading lower than .40 was item 3, referring to treatment of the child. As the factor loading was close to .40, and we wanted the two mistreatment scales to contain parallel items, all items were retained for the analyses. A composite score was created by taking the average across the individual items. Alpha = .73 for the scale (.75 for lesbians, .70 for gay men).

2.3.2. Predictors

2.3.2.1. Sexual orientation

We examined differences by parent sexual orientation by creating a dummy variable where 1 = heterosexual parent, and 0 = sexual-minority parent. In follow-up analyses, we included two dummy variables (1 = gay male, 0 = not gay/everyone else), lesbian (1 = lesbian, 0 = not lesbian/everyone else), which enabled us to detect differences across all groups. By including the lesbian and gay male dummy variables (but not heterosexual), we could test whether lesbian couples differed from heterosexual couples and whether gay couples differed from heterosexual couples. By changing the default group to include the gay male and heterosexual (but not lesbian) dummies, we could test for differences between gay couples and lesbian couples.

2.3.2.2. Child race

Child race was dummy coded such that 1 = of color and 0 = White, where “of color” includes biracial and multiracial children. We also considered other codings of child race, namely: Black child (1) versus non-Black child (0); internationally adopted from non-Anglo country (i.e., child with cultural and racial differences) (1) versus others (0); and transracial (1) versus inracial (0) adoption, where “transracial” refers to adoptions where the parent and child are of different races, and “inracial” refers to adoptions where the parent and child are of the same race. We examined the effect of these other codings of child race in follow-up analyses given the possibility that the salience and meaning of race may vary depending upon how it is coded or operationalized.

2.3.2.3. Perceived community homophobia

We asked parents, “How gay-friendly is your immediate community?” and provided these response options: 1 = extremely gay-friendly, 2 = somewhat gay-friendly, 3 = neutral, 4 = not very gay-friendly, and 5 = not at all gay-friendly. Higher scores indicate greater perceived homo-phobia/less gay-friendliness. We included this as a predictor in our analyses predicting perceived mistreatment due to sexual orientation given evidence that lesbian parents who perceive their communities as progressive report more positive experiences with their children’s schools (Casper & Schultz, 1999). Of note is that we also asked parents about their perceptions of neighborhood gay-friendliness specifically (see Goldberg & Smith, 2011, for an examination of the relationship between neighborhood gay-friendliness and parents’ mental health); thus, respondents were explicitly prompted to consider the gay-friendliness of the broader community in which they lived.

2.3.3. Controls

2.3.3.1. Family income

We used parents’ combined annual income as a control given its association with parents’ school consideration criteria, whereby, for example, parents with less income are more likely to consider cost, and less likely to consider school philosophy/quality, in their selection process (Early & Burchinal, 2001; Leslie, Ettenson, & Cumsille, 2002); although some studies have not found associations between income and child care/preschool considerations or values (Cryer & Burchinal, 1997).

2.3.3.2. Education

Parents’ education ranged from 1 to 6 where 1 = less than high school education, 2 = high school diploma, 3 = associate’s degree or some college, 4 = bachelor’s degree, 5 = master’s degree, and 6 = Ph.D./M.D./JD. We included education as a control given its association with some school selection criteria. For example, parents with less education have been found to place more emphasis on cost (Leslie et al., 2002).

2.3.3.3. School type

Parents were asked to indicate whether their child attended a public or private preschool. School type was dummy coded such that 1 = private school and 0 = public school. We included school type as a control given that lesbian parents have been found to expect less mistreatment in private schools than public schools (Lindsay et al., 2006).

2.3.3.4. Number of hours in school

Number of hours in school was included as a control in follow-up analyses examining school selection criteria and perceived mistreatment.

2.3.3.5. Child age

Child age, in months, was included as a control in follow-up analyses examining school selection criteria and perceived school mistreatment.

2.3.3.6. Parent race

Parent race (1 = of color and 0 = White) was included as a control in follow-up analyses examining school selection criteria and perceived school mistreatment.

2.4. Analytic strategy: hierarchical linear modeling

2.4.1. Hierarchical linear modeling

As both parents in each couple reported on their own school selection criteria and experiences, hierarchical linear modeling (HLM) was used to analyze the data. HLM permits examination of dyadic data (such as partners nested in couples) and provides accurate standard errors for testing the regression coefficients relating predictors to outcome scores (Kenny, Kashy, & Cook, 2006; Smith, Sayer, & Goldberg, 2013). The variance in the outcome is partitioned into the variance that occurs within couples (how partners differ from each other) and the variance that occurs between couples (how couples differ from each other). Predictors, both those that vary by couples (e.g., family income), and by partner (e.g., education), can then be added to explain this variance.

The hierarchical linear models tested were two-level random intercept models such that both parents’ reports (Level 1) were nested within the couple (Level 2). A single intercept was used as there was no characteristic meaningful to the analyses (such as parent gender) available to distinguish between the two parents in a couple (Smith et al., 2013). In Level 1 of the unconditional model, the intercept, β_0j^, represents average outcome score for each couple, and r_ij represents the deviation of each member of the couple from the couple average. This intercept is treated as randomly varying; that is, it is allowed to take on different values for each couple. The intercepts that are estimated for each couple are treated as an outcome variable at Level 2. The intercept in the Level-2 equation, γ₀₀, provides an estimate of the average outcome score across couples and u_0j represents the deviation of each couple from the overall average across all couples.

Level 1 (within couples):

$$Y_{ij}={\beta }_{0j}+r_{ij}$$

Level 2 (between couples):

$${\beta }_{0j}={\gamma }_{00}+u_{0j}$$

where Y_ij represents the outcome score of partner i in dyad j, where i = 1, 2 for the two members of the dyad. In addition to the above “fixed effect” estimates (e.g., the γ₀₀’s), estimates of the variance of the “random effects” both within and between couples are provided (e.g., the variance of the r_ij’s and the u_0j’s), as well as the covariance between partners. Predictors can then be added to the model, with those that vary within couples (e.g., education level) added at Level 1 and those that vary between couples (e.g., family income) added at Level 2. HLM was also used to examine mean differences by family type on the descriptive variables for which there was more than one report per family (e.g., parents’ work hours).

HLM of dyadic data does not, however, always derive accurate parameter estimates, when applied to dichotomous outcomes (or other outcomes requiring a link function; Raudenbush, 2008; Smith et al., 2013). We used hierarchical general linear modeling (HGLM) with a La Place transformation to improve parameter estimation (Raudenbush, Yang, & Yosef, 2000) to address Questions 1 and 2 regarding whether or not parents considered each of the eight selection factors, for which all outcomes were dichotomous (i.e., whether or not parents considered the factor). As a check, we refit all models examining dichotomous outcomes using logistic regression. Continuous variables (i.e., income) were grand mean-centered. Dichotomous variables (sexual-minority versus heterosexual couple; White child versus child of color; public versus private school) were dummy coded (0, 1). To examine for collinearity, each predictor was entered alone as well as in combination with each other variable.

2.4.2. Follow-up analysis

In our analyses of school selection considerations, we first substituted several alternate codes for child race (i.e., Black versus non-Black; internationally adopted from non-Anglo country versus not; transracial adoption versus inracial), out of an awareness that complexities of race, ethnicity, and culture may be difficult to capture. Second, we conducted follow-up analyses controlling for the number of hours children were in school, child age, and parent race. Third, to ascertain whether there were differences in findings for lesbian versus gay male parents, all models were fit using separate dummy variables (0, 1) for lesbian and gay men, and then refit, changing the default category (i.e., gay male, heterosexual), so that differences between all groups could be examined. In our analyses of perceived mistreatment, we conducted follow-up analyses controlling for the number of hours children were in school, child age, parent race, and type of adoption (inracial/transracial). We then examined whether there were differences in findings for lesbian versus gay male parents.

3. Results

3.1. School selection factors: descriptives

Table 2 presents the percentage of parents who endorsed considering each factor in selecting a school for their child, for the full sample and broken down by family type, child race, and school type (private versus public school: 75% of the sample had their child enrolled in a private school). As Table 2 reveals, educational philosophy was the most frequently endorsed consideration (82%), for the full sample and for each group. Cost was the second most frequently cited consideration, for the full sample (60%) and for heterosexual parents (62%). Among sexual-minority parents, however, the gay-friendliness of the school was the second most common consideration, with 66% of lesbians and 69% of gay men indicating gay-friendliness as a concern in their selection of a school. Cost followed close behind, with 62% of lesbians and 55% of gay men indicating that cost factored into their selection of a school. Racial diversity was the fourth most commonly endorsed consideration, for the full sample (40%), and for each group – although there was significant variability by group, with 57% of lesbians, 38% of gay men, and 28% of heterosexuals indicating that they considered racial diversity. Location was not offered as a specific option, however, it was the fifth most frequently endorsed consideration (26%) according to responses to the “Other” category. Fewer than 25% of sexual-minority participants (16% of lesbians, 24% of gay men) considered the presence of other lesbian/gay-parent families, and just 11% of the full sample considered the presence of other adoptive families in their search for a school. Finally, just 10% of the sample considered the religion or language offerings of the school.

Table 2.

School selection considerations by family type, child race, and school type.

	Family type				Child race		School type
	Full sample (%)	Lesbian (%)	Gay (%)	Hetero (%)	White (%)	Of color (%)	Private (%)	Public (%)
Cost	60%	62%	55%	62%	58%	61%	59%	63%
Location	26%	24%	25%	28%	25%	27%	27%	24%
Educational philosophy	82%	81%	87%	79%	78%	83%	90%	61%
Religion/language	10%	5%	15%	10%	13%	8%	11%	7%
Racial diversity	40%	57%	38%	28%	26%	48%	43%	32%
Adoptive family presence	11%	14%	13%	6%	10%	11%	12%	7%
LG family presence	n/a	16%	24%	n/a	20%	20%	23%	11%
Gay-friendliness	n/a	66%	69%	n/a	71%	66%	74%	52%

Table 2 also shows the breakdown by child race. As Table 2 reveals, there is a noticeable difference in the consideration of racial diversity: 26% of parents of White children compared to 48% of parents of children of color considered racial diversity. Table 2 shows the breakdown by school type. All considerations, with the exception of cost, were more likely to be named by parents of children who attended private schools, while cost was cited more often by parents of children attending public schools.

In response to the request to list other considerations in choosing a school, in addition to mentioning location (26%), small percentages of parents identified the reputation of the school (4.5%), special needs accommodations (2.5%), and the teacher-child ratio (1.5%) as shaping their school selection. Other less-often cited considerations were cleanliness and flexibility.

The intraclass correlation (ICC) for mistreatment due to adoptive status (i.e., the correlation between partners’ reports of mistreatment) was .42. The ICC for mistreatment due to sexual orientation was .68, and for community gay-friendliness, it was .62. Thus, partners within couples tended to have more similar perceptions of sexuality-related stigma than adoption-related stigma. For each of the school selection considerations (all of which were dichotomous variables), we determined intracouple agreement by calculating the percentage of couples in which both partners considered, or did not consider, a given factor. The percentage of couples with convergent (as opposed to discrepant) reports for each factor is as follows: cost (67%); location (58%); educational philosophy (88%); religion/language (88%); racial diversity (81%); adoptive family presence (86%); lesbian/gay-parent family presence (73%); school gay-friendliness (60%). Thus, the majority of partners within couples had convergent reports for each type of selection criteria.

3.2. School selection factors by sexual orientation and child race

Hierarchical general linear modeling (HGLM) for dichotomous outcomes was used to examine whether parent sexual orientation (sexual-minority parent versus heterosexual parent) and child race predicted each type of school selection factor (Table 3). Two control variables, parents’ income and education level, were entered given that financial and educational resources may influence their selection criteria (Goyette, 2008; Peyton et al., 2001). Income was related to cost as a selection factor, such that parents with less income were more likely to consider cost, B = −.005, SE = .002, OR = .99, t(165) = −2.45, p = .016. Education was related to location as a selection factor, such that parents with less education were more likely to consider location, B = −.43, SE = .21, OR = .65, t(165) = −2.06, p = .041.

Table 3.

Predicting consideration of school selection factors (full sample).

Predictors	Cost β (SE)	Odds ratio	Locat β (SE)	Odds ratio	Educ Phil β (SE)	Odds ratio	Relig/Lang β (SE)	Odds ratio	Racial divers β (SE)	Odds ratio	Adopt Fam β (SE)	Odds ratio
Constant	−.60 (.81)	.54	.55 (.94)	1.74	1.53 (1.25)	4.63	−2.55 (1.45)+	.08	−2.42 (1.02)*	.09	−8.16 (2.04)***	.0003
Child race (of color)	.32 (.36)	.138	.38 (.45)	1.47	.65 (.52)	1.91	−.53 (.62)	.58	.97 (.45)*	2.65	.08 (.62)	1.08
Sexual or (hetero)	.03 (.34)	1.03	.04 (.41)	1.03	.03 (.51)	1.03	−.07 (.60)	.93	−1.37 (.43)**	.25	−1.32 (.62)*	.27
Family income	−.005 (.002)*	.99	.003 (.002)	1.00	−.002 (.002)	.99	−.001 (.003)	.99	−.002 (.002)	.98	−.002 (.002)	.99
Education	.18 (.18)	1.19	−.43 (.21)*	.65	−.04 (.27)	.96	.12 (.31)	1.12	.43 (.22)*	1.52	1.32 (.39)**	3.75

^*p < .05.

^**p < .01.

^***p < .001.

⁺p < .10.

In predicting racial diversity as a selection factor (Table 3), we found a significant effect of child race, such that parents of children of color were more likely to consider racial diversity than parents of White children, B = .97, SE = .45, OR = 2.65, t(165) = 2.12, p = .035. Parent sexual orientation was also significant, such that heterosexual parents were less likely to consider racial diversity than sexual-minority parents, B = −1.37, SE = .43, OR = .25, t(165) = −3.20, p = .002. Parent education was also significant, such that more educated parents were more likely to consider racial diversity, B = .43, SE = .22, OR = 1.52, t(165) = 2.00, p = .047.

Regarding the presence of other adoptive families (Table 3), we found that parent sexual orientation was associated with the presence of adoptive families as a selection factor, such that heterosexual parents were less likely than sexual-minority parents to consider the presence of adoptive families at the school, B = −1.32, SE = .62, OR = .27, t(165) = −2.11, p = .036. The effect of education was also significant, such that more educated parents were more likely to consider whether there were other adoptive families at the school than less educated parents, B = 1.32, SE = .39, OR = 3.75, t(160) = 3.37, p = .001. There were no significant findings for parent sexual orientation, child race, income, or education in predicting educational philosophy or religion/language as selection factors.

3.3. School selection factors by parent gender and child race (sexual-minority parents only)

Turning to the sexual minority-specific selection factors (Table 4), we found that education level was associated with the presence of other lesbian/gay-parent families as a selection factor, such that sexual-minority parents with higher levels of education were more likely to consider this than less educated sexual-minority parents, B = .82, SE = .37, OR = 2.28, t(85) = 2.18, p = .031. Likewise, we found that education level was also related to the gay-friendliness of the school as a selection factor, at the level of a trend, such that sexual-minority parents with higher levels of education were somewhat more likely to consider the gay-friendliness of the school than less educated sexual-minority parents, B = .45, SE = .24, OR = 1.57, t(85) = 1.85, p = .067.

Table 4.

Predicting consideration of presence of lesbian/gay-parent families and school gay-friendliness (sexual-minority parents only).

Predictors	Presence of LG- parent families β (SE)	Odds ratio	Gay-friendliness β (SE)	Odds ratio
Constant	5.82 (1.99)	.003	−1.39	.25
Female	−.57 (.66)	.57	.30 (.55)	1.34
Child race (of color)	1.10 (.75)	3.02	−.11 (.51)	.90
Family income	−.002 (.003)	1.00	−.004 (.003)	1.00
Education	.82 (.37)*	2.28	.45 (.24)+	1.57

^*p < .05.

⁺p < .10.

3.3.1. Exploratory follow-up analyses

3.3.1.1. Additional controls

We added several additional controls in follow-up analyses predicting school selection criteria: namely, hours in school per week, child age, and parent race. None of the findings changed with the addition of these controls (or when the original controls were removed), and none of the additional variables were significant, with one exception. Parents whose children were in preschool more hours per week were more likely to consider the religion/language of the school, B = .06, SE = .03, OR = 1.06, t(162) = 2.10, p = .037. Further, none of the findings changed with two out of the three alternate codings for child race. Both Black versus non-Black and transracial versus inracial significantly predicted consideration of racial diversity. Specifically, parents who adopted Black children were more likely to consider racial diversity, B = 1.05, SE = .48, OR = 2.86, t(165) = 2.17, p = .031, and parents who adopted transracially were more likely to consider racial diversity, B = 1.03, SE = .44, OR = 2.79, t(165) = 2.32, p = .021. Non-Anglo versus Anglo did not predict racial diversity as a selection factor, however, indicating that adopting a child with both racial and cultural differences did not necessarily lead parents to consider racial diversity.

3.3.1.2. Differences between lesbian versus gay parents

To determine whether school selection considerations differed for lesbian parents and gay parents, we substituted lesbian and gay dummy codes (lesbian: 1, 0, where 1 = lesbian and 0 = not a lesbian; gay: 1, 0, where 1 = gay and 0 = not a gay man) for sexual orientation. This enabled us to probe for differences across all three family groupings (lesbian, gay, heterosexual), by altering the default category. Models showed no differences between lesbians and gay men in predicting any outcome.

In addition, we examined whether the findings for same-sex couples held up when lesbian and gay male parents were examined separately (lesbian and gay dummy variables were substituted for heterosexual versus sexual minority). These analyses showed fairly consistent findings for both lesbians and gay men. The significance for gay men fell to the level of a trend in predicting racial diversity as a school selection consideration, B = .88, SE = .46, OR = 2.39, t(165) = 1.88 p = .062, while the effect for lesbians remain significant, B = 1.60, SE = .47, OR = 4.97, t(165) = 3.40, p = .001. In addition, both lesbian, B = 1.22, SE = .66, OR = 3.39, t(165) = 1.83, p = .069, and gay, B = 1.53, SE = .78, OR = 4.63, t(165) = 1.95, p = .052, emerged as marginally significant in predicting the presence of other adoptive families as a selection consideration.

3.4. Perceptions of mistreatment due to adoptive status

Of interest was whether parent sexual orientation or child race predicted perceived school mistreatment due to adoptive status, controlling for income, education, and school type (Table 5). HLM analyses revealed that parent sexual orientation was marginally related to perceived mistreatment due to adoptive status, F(1, 83) = 3.39, p = .069; B = .15, SE = .08, t(83) = −1.84, p = .069, such that heterosexual parents perceived somewhat higher levels of mistreatment than sexual-minority parents. Of note is that parents reported relatively low levels of mistreatment due to adoptive status overall, with an overall M of 1.69 on a scale of 1–5 (Mdn = 1.63, SD = .50; range 1.00–3.71), and M = 1.81 (Mdn = 1.75, SD = .52, range 1.00–3.71) for heterosexual parents, M = 1.65 (Mdn = 1.62, SD = .46, range 1.00–2.88) for lesbian parents, and M = 1.52 (Mdn = 1.50, SD = .42, range 1.00–3.00) for gay male parents. While several individual items were skewed, the overall distribution for the scale was only mildly skewed (skewness = .66, SE of skewness = .21).

Table 5.

Predicting perceived mistreatment related to adoption (full sample) and sexual orientation (sexual-minority parents only).

	Due to adoption (estimate, SE)	Due to sexual orientation (estimate, SE)
Intercept	1.72 (.20)***	1.21 (.35)**
Same-sex	−.15 (.08)+
Male		15.16
White	.06 (.08)	−.20 (.15)
Public	.02 (.09)	.07 (.16)
Education	.007 (.04)	.05 (.06)
Family income	.001 (.004)*	−.003 (.007)
Perceived community homophobia		.18 (.08)*

Note: Sexual orientation was dummy coded (0, 1), with heterosexual as the default group. Child race was dummy coded (0, 1), with of color as the default group. Follow-up analyses utilized gender (0, 1), with female as the default group, when comparing just gay men and lesbians. Family income was mean centered. MLM was not used for logistic regression because MLM provides biased estimates when using link functions (such as a log link) in modeling dyadic data (Raudenbush, 2008).

^*p < .05.

^**p < .01.

^***p < .001.

⁺p < .10.

3.5. Perceptions of mistreatment due to sexual orientation

When perceived mistreatment due to sexual orientation was treated as the outcome, gender (gay men versus lesbian women), child race, and perceived community gay-friendliness were included as predictors, with income, education, and school type included as controls (Table 5). Community gay-friendliness predicted perceived mistreatment related to sexual orientation, such that parents who perceived their communities as more unfriendly (more homophobic) reported more mistreatment, F(1, 86) = 5.07, p = .027; B = .18, SE = .08, t(86) = 2.25, p = .027. Sexual-minority parents reported relatively low levels of perceived mistreatment due to their sexual orientation, on average: M = 1.66 (Mdn = 1.62, SD = .48, range 1.00–3.13) for lesbian parents, and M = 1.75 (Mdn = 1.50, SD = .81, range 1.00–5.00) for gay parents. While several individual items were highly skewed, the overall distribution for the scale was moderately skewed (skewness = 1.03, SE of skewness = .20).

3.5.1. Exploratory follow-up analyses

3.5.1.1. Additional controls

Several additional controls were included in predicting the two school mistreatment outcomes (hours in school; child age; parent race: of color versus White; type of adoption: transracial versus inracial/same-race adoption). None were significant, either when the original controls were retained or when they were removed.

3.5.1.2. Differences between lesbian versus gay parents

To determine whether perceived mistreatment due to adoptive status differed for lesbian parents and gay parents, we substituted lesbian and gay dummy codes (lesbian: 1, 0, where 1 = lesbian and 0 = not a lesbian; gay: 1, 0, where 1 = gay and 0 = not a gay man) for sexual orientation, which, again, enabled us to probe for differences across all three family groupings. Perhaps due to insufficient power, once lesbian and gay were split into separate groups, neither dummy code was significant.

4. Discussion

The current exploratory study is the first that we know of to examine school selection considerations and perceived mistreatment among lesbian, gay, and heterosexual adoptive parents of preschool-aged children. The findings hold implications for early childhood educators.

In considering our descriptive data regarding the frequency by which parents considered each of the various school criteria, we found that educational philosophy was the most frequently named consideration for the full sample and for each group. This is consistent with prior research showing that educational philosophy often weighs heavily in parents’ school selection process –especially among middle-class parents, such as the current sample (Peyton et al., 2001). Cost was the second most frequently endorsed concern, for the full sample and for heterosexual parents; however, among lesbian and gay parents, the gay-friendliness of the school was the second most common consideration, with cost following close behind. This finding is interesting, especially contrasting it with the findings of the GLSEN survey, which found that less than one-fifth of parents emphasized the school’s reputation for being welcoming of LGBT-parent families in selecting a school (Kosciw & Diaz, 2008). This difference may reflect the fact that whereas most of the parents in the GLSEN sample had elementary school-age children, the majority of whom were in public schools, the parents in the current study were selecting preschools, most of which were private, and thus felt that they could weigh school gay-friendliness more heavily in their school selection. This finding carries implications for early childhood educators, in that it suggests that sexual-minority parents often consider the gay-friendliness of preschools in making decisions about where to send their children.

Indeed, the relatively low level of perceived mistreatment due to sexual orientation in our sample is perhaps in part due to the fact that the gay-friendliness of the school was an important school selection consideration for sexual-minority participants. In turn, their children typically ended up at schools that were at least relatively gay-friendly. This finding is consistent with prior qualitative research showing that sexual-minority parents often seek out progressive, gay-friendly schools in an effort to decrease the likelihood their children will be exposed to stigma based on their family structure (Casper & Schultz, 1999; Mercier & Harold, 2003). As the parents in this sample were fairly affluent, future work should examine patterns and predictors of perceived school mistreatment due to sexual orientation in samples that, for financial or geographic reasons, ultimately have less access to progressive and gay-friendly schools.

Racial diversity was the fourth most often endorsed consideration, for the full sample, and for each group – although lesbians were the most likely to consider racial diversity and heterosexuals were the least likely to consider it. This is consistent with the finding, in our multivariate analysis, that sexual orientation significantly predicted the degree to which parents considered racial diversity in selecting a school for their child. Location –which we did not explicitly ask about, but some participants identified as a salient variable influencing their decision – was the fifth most commonly endorsed consideration. Less than a quarter of sexual-minority parents considered the presence of other lesbian/gay-parent families in selecting a school, and about one-tenth of the overall sample considered the presence of other adoptive families. That these factors were rarely recalled as significant determinants of parents’ choice of a school may reflect the overall paucity of these families in their communities, and, thus, their understanding that it was unrealistic to prioritize these in their selection process. Religion/language offerings were also rarely considered. This is somewhat consistent with the GLSEN survey, which found that only 10% of the parents in their sample emphasized the language programs offered in selecting a school; also, the children in the GLSEN sample were less likely than national percentages to be attending a religiously-affiliated school (Kosciw & Diaz, 2008).

We also examined the factors that predicted the degree to which parents considered each of these school criteria. Our finding that parents with less income were more likely to consider cost in selecting a school was unsurprising, and is consistent with prior work showing that practicality-based school considerations are more common among individuals with fewer resources (Leslie et al., 2002; Smrekar & Goldring, 1999). Affluent parents may feel less constrained by cost in choosing a preschool, and/or they may be more willing to spend money on a preschool, especially to gain valued aspects of schools – such as the presence of diverse families (Smrekar & Goldring, 1999). Similarly, less educated parents were more likely to consider location in choosing a preschool, again echoing previous findings showing that fewer resources are associated with greater consideration of practical factors in school selection (Smrekar & Goldring, 1999).

We found that more educated parents were significantly more likely to consider both the presence of racial diversity and other adoptive families in choosing a preschool for their children – and, among sexual-minority parents, the presence of other lesbian/gay-parent families and the gay-friendliness of the school. This likely reflects the role of education in shaping ideas about the value and role of schooling, whereby middle-class parents are more likely to prioritize the degree to which children’s educational environments match or reflect their own family’s values and attributes (Goyette, 2008; Wells & Crain, 1997). Our finding that, among sexual minorities, parents with more education were more likely to consider the presence of other lesbian/gay-parent families provides quantitative support for prior qualitative work suggesting that social class plays a role in shaping lesbian mothers’ values and priorities with regard to schooling (Casper & Schultz, 1999; Nixon, 2011).

Sexual-minority parents were more likely than heterosexual parents to consider the presence of adoptive families in choosing a school. These parents were likely aware of the multiple ways that their child would be different from their peers at school, and were thus motivated to find a school that seemed to reflect, normalize, and perhaps value at least this aspect of their family’s diversity. They may also simply place a higher value on diversity in general, as suggested by the fact that they were also more likely to value racial diversity, regardless of their children’s race. This notion is consistent with previous findings that sexual minorities often value schools’ inclusion of diversity because they believe that it bodes well for the school’s attitude toward lesbian/gay-parent families (Gartrell et al., 2000; Kosciw & Diaz, 2008). That heterosexual parents were less likely to consider the presence of adoptive families is interesting in light of the finding that they were also more likely to perceive mistreatment due to adoptive status than sexual-minority parents. Prior work has found that heterosexual adoptive parents possess greater internalized stigma surrounding adoption (Goldberg et al., 2011) and are more concerned about not appearing visibly different from the heterosexual nuclear family model (Goldberg, 2009), compared to lesbian/gay adoptive parents. In turn, heterosexual adoptive parents may not be as likely to emphasize their adoptive status and thus seek out the presence of adoptive families.

Consistent with prior work (Kosciw & Diaz, 2008; Vonk et al., 2010), adoptive parents of children of color were more likely to consider the school’s racial diversity than parents of White children. Indeed, this finding is consistent with Vonk et al.’s finding that over half of their sample of heterosexual parents of adopted children of color chose teachers and child care providers in part out of awareness of racial socialization issues. Future work is needed that (a) examines how parents of adopted, racial minority children make decisions about their children’s schooling, and (b) identifies the parent-, child-, and social-contextual factors that distinguish those parents of children of color who consider racial diversity in the selection process from those who do not.

Next we turn to the findings related to parents’ experiences of school mistreatment. Sexual-minority parents were somewhat less likely to perceive school mistreatment due to their adoptive status than heterosexual parents. Sexual-minority parents may simply be more attuned to other aspects and areas of potential mistreatment – namely, mistreatment related to their sexual orientation. For lesbian and gay parents, adoption is not an alternative route to family building; it is one of the typical or expected routes to family building (Goldberg, 2010). Thus, they may expect and perceive less mistreatment related to their adoptive status, because they do not view this as the most remarkable aspect of their families. Heterosexual adoptive parents, on the other hand, may be more sensitive to how school administrators are responding to their adoptive status because they more clearly deviate from some (heterosexual, nuclear biologically-related family) norm. Further, prior research has documented higher levels of internalized stigma surrounding adoption in heterosexual adoptive parents as compared to same-sex adoptive parents (Goldberg et al., 2011). This higher level of stigma may predispose heterosexual parents to greater sensitivity regarding how their families are being treated by outsiders. Educators should thus be aware that heterosexual adoptive families may be especially sensitive to perceived exclusion or stigma related to their adoptive family status. In turn, they should be sensitive to how language and curricular materials may alienate all adoptive parents (e.g., by failing to acknowledge that not all mothers gave birth; by failing to acknowledge that not all children have been with their parents from birth).

Regarding school mistreatment due to sexual orientation, sexual-minority parents who perceived their communities as less gay-friendly were more likely to perceive school bias due to sexual orientation. This provides some indication of how aspects of the broader social context (e.g., community attitudes) may shape or “trickle down” into more localized school attitudes and practices (Bronfenbrenner, 1986). Prior qualitative research has also pointed to the significance of geographic context for parents’ experiences of their children’s schools, with lesbian and gay parents who lived in more “progressive” areas describing more positive experiences with their children’s schools (Casper & Schultz, 1999). Parents who perceive their communities as gay-friendly may tend to find that their children’s schools reflect and embody the communities’ values; likewise, parents who feel that their communities are homophobic may face similar issues in the microcosm of their children’s schools. Educators who teach in geographic areas that are relatively conservative and homophobic should be sensitive to the need to offset the stigma that lesbian/gay adoptive families may face in their broader communities.

In sum, we found that limited resources were related to practical concerns (cost, location) in choosing a school, such that parents with more financial and educational constraints were more likely to consider these factors. On the other hand, more educated parents tended to emphasize all aspects of diversity – adoptive families, racial diversity, and the presence of other lesbian/gay parents, among sexual minorities – illustrating how education may shape values and concerns about schooling. Sexual-minority parents were also more likely to consider racial diversity. Finally, we found that heterosexual parents were more likely to perceive mistreatment due to their adoptive status than sexual-minority parents, and sexual minority parents living in less gay-friendly communities were more likely to perceive mistreatment due to sexual orientation than sexual minority parents living in more gay-friendly communities.

4.1. Limitations

The current study has several key strengths, when comparing it to the prior literature, including its examination of sexual-minority parents of young children, its inclusion of gay-father families, and its quantitative design. But there are a number of limitations to the study. First, we did not assess parents’ decision-making prospectively (i.e., before they selected a preschool) as some research has done (Galotti & Tinkelenberg, 2009). Rather, we asked parents who had children currently enrolled in preschools about their selection considerations. Thus, it is possible that parents’ choice of a preschool ultimately influenced what factors they recalled as most important to them. Reporting retrospectively on their consideration of their school selection process, they may have been motivated to identify a coherent and logical narrative whereby they chose a school that fit with their values and priorities. In essence, their recollections may have been influenced by their actual experiences with their children’s schools. Second, we only measured perceived community gay-friendliness; we did not include an objective measure of community political attitudes or practices, for example. Further, our measurement of community gay-friendliness was based on a single item. Although this was not ideal, it is notable that other studies have also used one-item measures to assess contextual aspects of gay-friendliness (e.g., neighborhood gay-friendliness; Goldberg & Smith, 2011). A third limitation is that we did not provide definitions of “private” and “public” schools to participants; thus, it is possible that some parents may have misunderstood and misrepresented the type of school that their children attended. Fourth, we did not ask parents directly about location in our initial survey; they volunteered this as an additional factor that they considered, and thus we included it in analyses. Thus, the number of parents who considered location is likely an undercount. Fifth, our measures of mistreatment were created for the study, and lack established psychometric properties; thus, future research utilizing these measures is needed. Sixth, the sample was well-educated, financially well-off, and mostly White. Greater diversity in parent race, for example, might have allowed us to discern interesting differences for White parents versus parents of color, and to probe differences among parents of different racial and ethnic groups.

4.2. Conclusions and implications for practice

Early childhood educators and administrators should be aware that sexual-minority adoptive parents are likely to be attentive to issues of family, racial, and sexual diversity in evaluating and selecting preschools for their children. Parents of preschoolers may be particularly sensitive to these issues, insomuch as they have more contact with and control over their children’s school environments than parents of school-age children (Beveridge, 2005). Educators who wish to create a welcoming atmosphere for sexual-minority parents, then, should consider how to demonstrate best their commitment to diversity in their schools and classrooms. This may require additional professional training, insomuch as childhood educators and preschool teachers often receive minimal preparation for working with sexual-minority parent families, and may experience uncertainty about how to enact a more inclusive and affirming approach to these families (Kintner-Duffy, Vardell, Lower, & Cassidy, 2012). Early childhood educators may also benefit from training that encourages awareness and knowledge of adoptive-parent families – particularly heterosexual adoptive-parent families, who may be especially sensitive to adoption-related stigma and exclusion. On a more general level, it is critical that early childhood educators recognize the growing diversity of contemporary families in the US, and consider issues of adoption, family structure, and race in thinking about curriculum development and in ensuring classroom environments that are welcoming of all families.

Insomuch as preschool is a key stage during which children are beginning to develop an understanding of basic concepts such as race, ethnicity, gender, and families (Trolley, Magerkorth, & Fromme, 1999), and family structures are becoming increasingly diverse (Gates et al., 2007; Goldberg, 2010), it is critical that early childhood educators are educated about the benefits of creating a more inclusive curriculum and overall approach to teaching about families and diversity. Preschool is the stage during which children are often beginning to comprehend, and are thus receptive to, lessons about diversity and difference (Trolley et al.). In the absence of such lessons, even very young children may develop heterosexist and racist sentiments (Robinson & Ferfolja, 2002), creating an environment in which children with sexual-minority parents – as well as children of color and adopted children – may feel stigmatized, excluded, or ignored.