Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 1.
Published in final edited form as: J Fam Psychol. 2023 Jan 12;37(2):195–202. doi: 10.1037/fam0001056

Measurement Equivalence of Family Acceptance/Rejection among Sexual and Gender Minority Youth by Disclosure Status

Amanda M Pollitt 1, Jessica N Fish 2, Ryan J Watson 3
PMCID: PMC9928906  NIHMSID: NIHMS1856854  PMID: 36634006

Abstract

Family acceptance is a crucial component of healthy development during adolescence, especially for sexual and gender minority youth (SGMY) who often fear rejection from family members. Studies focused on SGMY family environments often utilize broad measures that fail to capture SGMY-specific aspects of family acceptance and rejection. Less research has considered how the measurement of family acceptance and rejection might differ depending on whether SGMY have disclosed their sexual and/or gender identities to their parents. We used data from a national non-probability sample of 9,127 SGMY in the U.S. who had either disclosed (n = 6,683) or not disclosed (n = 2,444) their sexual and/or gender identities to parents to test the factor structure of an 8-item measure of family acceptance and rejection and differences by disclosure status. A two-factor, negatively correlated model reflecting constructs of family acceptance and family rejection was equivalent across disclosure groups. Youth who had disclosed their identity reported greater acceptance and less rejection and showed a stronger negative association between the two constructs than non-disclosed youth. Family acceptance, but not rejection, had higher variability among disclosed youth than non-disclosed youth. Results suggest that the family environments of SGMY are simultaneously characterized by accepting and rejecting behaviors. Though families of disclosed youth appear to be more accepting and less rejecting, the experiences of these youth are complex. Findings suggest that research on SGMY family environment must consider both supportive and undermining behaviors and that the measures assessed here operate similarly for youth based on disclosure.

Keywords: Sexual and gender minority youth, identity disclosure, adolescence, acceptance and rejection, family environments

Family support is essential for the healthy development of young people (Chu et al., 2010). However, sexual and gender minority youth (SGMY) experience deficits in family support (Fish et al., 2020; Fish & Russell, 2018), particularly regarding their sexual and gender identities (Allen et al., 2022; D’Augelli & Grossman, 2001; Fish et al., 2020; Grossman et al., 2021; Savin-Williams & Ream, 2003). Research on understanding the family relationships of SGMY is critical, considering that family behaviors related to SGMY’s sexual identity are strongly associated with mental health, substance use, and well-being (Ryan et al., 2010; Snapp et al., 2015).

Scholarship on SGMY’s family environment has largely conceptualized family and parental behaviors that reflect either “acceptance” or “rejection” of youth’s sexual identity. These studies – typically focused on parents – have shown that more positive reactions to disclosure of sexual/gender identities and supportive relationships with family members are associated with better health outcomes, and more negative reactions/relationships are associated with worse health. Studies that have examined SGM-specific family behaviors find that this type of social support is more beneficial to mental health and self-esteem than general support (Doty et al., 2010; Ryan et al., 2010). Caitlin Ryan of the Family Acceptance Project created an extensive SGM-specific measure of family acceptance and rejection, which consisted of 106 items based on family behaviors related to SGMY identities (Ryan, 2010). Studies from this project found that family members’ accepting behaviors were associated with lower risk for depression, suicidality, and substance use (Ryan et al., 2010), while rejecting behaviors predicted greater risk for these outcomes (Ryan et al., 2009).

Previous work has measured family acceptance and rejection along a single continuum; high levels of family acceptance are often interpreted and operationalized as low levels of rejection, and vice versa (e.g., D’Augelli & Grossman, 2001; Savin-Williams & Ream, 2003). As the field has grown, researchers have investigated family acceptance and rejection as distinct constructs that reflect family processes related to SGMY’s identity within the family: SGMY experience both accepting and rejecting behaviors from family members that may contradict one another and send mixed messages to youth (see Allen et al., 2022). Thus, acceptance is not the absence of rejection, nor rejection the absence of acceptance. For example, in one study, sexual minority adults described how acceptance from their families came with negative undertones or microaggressions (Mena & Vaccaro, 2013). Some participants minimized their family members’ rejecting behaviors, sometimes interpreting them as supportive, by contrasting their behavior to the possibility of more extreme reactions such as disownment (p. 12). Another study found that reactions from gay and bisexual men’s fathers were often contradictory in that they accepted their sons but rejected SGM (sexual and gender minority) people in general (Jadwin-Cakmak et al., 2015). These studies illustrate that SGMY’s family relationships are nuanced, such that family accepting and rejecting behaviors are not mutually exclusive, unidimensional constructs. Approaches that consider modeling these behaviors together may provide a better understanding of family ambiguity or ambivalence. Thus, examining the dimensionality of quantitative measures of family acceptance and rejection could improve our understanding of SGMY’s family relationships and the degree to which these processes are interrelated.

The family environments of SGMY can vary based on many different contextual factors, including their religion, geographical location, and intersecting identities. In the current study, we consider the role of an experience that is specific to SGMY: disclosure of a sexual or gender identity. Moreover, many of these studies have only included SGMY who have disclosed their identity to their parents, where family interactions around SGMY are often more overt (e.g., D’Augelli et al., 1998) given the shared knowledge of youth’s sexual identity. Researchers often exclude youth who have not disclosed from their analyses to avoid introducing bias or conflating their experiences with disclosed youth. This exclusion may be appropriate when studying specific SGMY experiences in which youth must have disclosed to family members, such as embarking on a gender transition or introducing family members to a same-sex partner. However, some research on family interactions has found differences between SGMY who have and have not disclosed their sexual and/or gender identity (D’Augelli et al., 1998), suggesting a selection effect that excludes youth who have yet to disclose their identity and, therefore, may have unique family experiences and related outcomes. Furthermore, SGMY who have yet to disclose their sexual and/or gender identity likely monitor their parents closely for clues about their parents’ attitudes toward SGM people to decide whether to disclose. For example, some SGMY report that they expect their family members will be rejecting their sexual orientation based on previous homophobic comments (Jadwin-Cakmak et al., 2015). It is likely that the way that SGMY respond to measures designed to capture family behaviors around youth’s disclosed or perceived sexual orientation or gender identity differ by disclosure status. SGMY who have disclosed would likely evaluate whether their family members are accepting or rejecting based on explicit reactions to their disclosure; in comparison, SGMY who have not disclosed must rely on implicit cues about SGM people from their families. There are also implications for validity when items in measures of family acceptance and rejection imply that parents or caregivers are aware of their child’s sexual or gender identity. Thus, depending on disclosure status, family experiences may differ for SGMY and influence how SGMY qualitatively and quantitatively respond to family environment measures. This could lead to biased estimates of associations with health outcomes.

In the current study, we sought to address current limitations in the conceptualization of family environment by examining the dimensionality of an SGMY-specific measure of family acceptance and rejection of youth’s sexual identity. Moreover, we test the degree to which youth may respond to family environment measures as a function of their disclosure to parents by examining differences in measurement by disclosure status in a large sample of SGMY.

Methods

Data and Measures

Data come from a subset of a large online national sample (N = 17,112) of 13-17-year-old SGMY living in the U.S. Data were collected in 2017. The Institutional Review Board at the University of Connecticut approved the original study procedures. All youth provided electronic assent, and a waiver of parental consent was obtained due to the sensitive nature of study questions and to avoid youth needing to “out” themselves to parents to take the survey. The data were collected in partnership with the largest LGBTQ+ organization in the United States, the Human Rights Campaign (HRC). Partnering with the HRC allowed researchers to sample diverse youth who utilized community-serving organizations (e.g., local drop-in youth centers) and followed/interacted with the HRC or their partners (e.g., CenterLink, Trevor Project) on social media. With the help of the HRC, researchers leveraged influencers on Twitter (e.g., Jazz Jennings) to reach diverse youth audiences by tweeting their support of the study. To ensure diversity in the sample, the researchers utilized paid advertisements on Facebook targeted toward youth with minoritized social positions. Before data were collected, a series of steps were taken to eliminate the potential for bots and mischievous responders to participate in the survey, including a multistep consent and sorting process. This process included a response tree protocol that diverted ineligible participants based on age and residence. After data were collected, a rigorous post-hoc data cleaning process was undertaken to ensure the quality of data (see Robinson-Cimpian, 2014), which excluded participants with multiple extreme responses, including impossible (e.g., weight of 8 pounds) and/or implausible (e.g., a gender identity of “Donald Trump”) responses.

Materials and analysis code for this study are available by emailing the corresponding author. The study was not preregistered. Youth were included (n = 13,909) if they were currently living with a parent and identified as gay, lesbian, bisexual, queer, or pansexual (n = 13,668) or heterosexual if they also identified as transgender (n = 241). We excluded youth who did not complete the survey beyond the demographics section (n = 1,018) and were missing on measures capturing disclosure (n = 2,820) and on all indicators of family acceptance and rejection (n = 944). This resulted in a final sample of 9,127. Demographic information by disclosure status is shown in Table 1. The sample was majority White, with a plurality of cisgender female and bisexual youth.

Table 1.

Disclosure Status and Missingness on All Family Rejection/Acceptance Items by Demographic Characteristics

Missing on All
Family
Rejection/Acceptance
Items		 Non-Disclosed Disclosed Total
n % n % n % X2 P n %
Race/ethnicity
 White 558 8.56 1488 24.95 4475 75.05 29.18 < 0.001 5963 65.38
 Black 65 13.27 134 31.53 291 68.47 5.14 0.02 425 4.66
 Native American 2 5.26 7 19.44 29 80.56 0.99 0.32 36 0.39
 Asian American 42 10.29 187 51.09 179 48.91 114.99 < 0.001 366 4.01
 Hispanic/Latino 122 11.32 280 29.29 676 70.71 3.44 0.06 956 10.48
 Biracial/Multiracial 133 9.81 305 24.94 918 75.06 2.43 0.12 1223 13.41
 Different race 21 12.21 41 27.15 110 72.85 0.01 0.92 151 1.66
Gender Identity
 Cisgender male 240 10.35 567 27.27 1512 72.73 0.34 0.56 2079 22.78
 Cisgender female 336 7.69 1395 34.57 2640 65.43 224.12 < 0.001 4035 44.21
 Transgender male 98 11.53 80 10.64 672 89.36 108.87 < 0.001 752 8.24
 Transgender female 6 5.5 20 19.42 83 80.58 2.88 0.09 103 1.13
 Nonbinary 264 10.90 382 17.70 1776 82.23 118.74 < 0.001 2158 23.64
Sexual Identity
 Gay/lesbian 335 8.2 786 20.97 2962 79.03 109.36 < 0.001 3748 41.06
 Bisexual 418 11.08 1191 35.52 2162 64.48 206.62 < 0.001 3353 36.74
 Heterosexual 15 8.98 30 19.74 122 80.26 3.91 0.05 152 1.67
 Queer 37 7.68 88 19.78 357 80.22 11.70 0.001 445 4.88
 Pansexual 139 8.86 349 24.42 1080 75.58 4.79 0.03 1429 15.66
Open in a new tab

We categorized SGMY as non-disclosed (n = 2444) or disclosed (n = 6683) based on their responses to the Outness Inventory (Mohr & Fassinger, 2000), which asked, “For each of the following groups, how many people currently do you think know of your sexual orientation?” and “For each of the following groups, how many people currently do you think know of your gender identity?” on a five-point scale from “none” to “all”. Those who reported greater than “none” to the specific item about parents were categorized as disclosed to parents.

Family rejection and acceptance measures were modified from items from the Family Acceptance Project (Abreu et al., 2022; Miller et al., 2020). Participants were asked to report how often their parents or caregivers behaved in rejecting (n items = 4; α = 0.88) or accepting (n items = 4; α = 0.82) ways on a four-point scale from “never” to “often.” Participants could also respond with “doesn’t apply to me”; these responses were coded as missing. The wording of these items and descriptive information by SGMY who were categorized as non-disclosed or disclosed are shown in Table 2.

Table 2.

Observed Means and Standard Deviations of Family Rejection and Family Acceptance Items for the Non-Disclosed and Disclosed Groups

Non-Disclosed Disclosed
How often do your parents or caregivers… M SD M SD t(df) p
Family rejection
 Taunt or mock you because you are an LGBTQ person 1.68 1.00 1.66 0.92 0.53(6829) 0.30
 Say negative comments about you being an LGBTQ person 1.75 1.07 1.75 0.99 0.19(6911) 0.43
 Say bad things about LGBTQ people in general 2.66 1.03 2.22 1.08 17.40(8900) < 0.001
 Make you feel like you are bad because you are an LGBTQ person 2.33 1.19 1.84 1.06 15.08(7541) < 0.001
Family acceptance
 Say that they like you as you are in regards to being an LGBTQ person 1.65 1.01 2.44 1.14 18.86(6730) < 0.001
 Say they were proud of you for being an LGBTQ person 1.31 0.77 1.97 1.14 16.34(6847) < 0.001
 Get involved in the larger LGBTQ community 1.26 0.65 1.64 0.97 14.07(7594) < 0.001
 Tell you that you are a role model as an LGBTQ person 1.14 0.51 1.40 0.84 8.91(6762) < 0.001
Open in a new tab

Note. One-tailed t-tests shown (H0: non-disclosed = disclosed), in which Ha: non-disclosed > disclosed for family rejection and Ha: non-disclosed < disclosed for family acceptance.

Analysis Plan

We conducted measurement invariance testing procedures in Mplus 8 (Muthén &Muthén, 1998-2017). We evaluated whether initial models had acceptable model fit by following standard cut-off values (i.e., CFI > 0.90; TLI > 0.90; RMSEA < 0.10) (Hu & Bentler, 1999). We first tested a two-factor model in which separate latent factors of family acceptance and rejection consisting of four items each were simultaneously modeled and allowed to covary. Next, we compared this two-factor model to a one-factor model where all eight items were included as factor indicators of a rejection latent factor, with acceptance items reverse coded. We used a fixed factor method of setting the scale of the latent factors in which we constrained the latent variances of each construct to 1 and allowed all indicator factor loadings to be freely estimated. Our final model was based on whether the one-factor model resulted in a significantly worse fit to the data when compared to the two-factor model.

After selecting a final model, we examined standardized expected parameter change (SPEC) and model modification indices (MI) to determine whether model fit would be improved by correlating the residual variances of indicators (Byrne et al., 1989; Whittaker, 2012). We correlated the residuals of items one by one if they had (1) had a large SPEC, (2) had a corresponding MI greater than 3.85, and (3) if they were theoretically plausible until model fit no longer significantly statistically improved, as indicated by a change in chi-square and CFI. Research shows that freeing parameters based on SPEC, MI, and theory when modifying models results in less misspecification than based on MI alone (Whittaker, 2012).

We examined configural (unconstrained; overall patterns of factor loadings freely estimated across groups), metric (factor loadings constrained to equality), and scalar (intercepts constrained to equality) invariance between youth who were disclosed or non-disclosed. We determined whether models passed measurement invariance based on model CFI (ΔCFI < .01), as recommended by Cheung & Rensvold (2002). Change in CFI is a more appropriate indicator for evaluating measurement invariance because, unlike other goodness-of-fit indicators (e.g., chi-square, RMSEA), CFI is not affected by model complexity or sample size (Cheung & Rensvold, 2002). Steps described above (i.e., examining SPEC and MI for change in CFI) were followed for unconstraining model factor loadings (metric invariance) or intercepts (scalar invariance) to improve model fit if an invariance test failed. In subsequent metric models, as long as the model passes the metric invariance test, constraining the factor loadings to equality allows for the estimation of the latent variances of the factors in the non-disclosed group compared to the disclosed group. We conducted Wald tests to determine whether the latent variance of each factor is equivalent in each group. Similarly, once the intercepts are constrained to equality in the scalar models, the latent means in the disclosed group models can be constrained to zero for interpretation: This specification sets the disclosed group as the reference group and the latent means of the non-disclosed group are interpreted as differences in latent means between each group.

Results

Missing Data

Participants who responded to at least one item on the family rejection or acceptance scales had their data included in the models. Their missing responses were handled using full information maximum likelihood (FIML). Adjusting the models using FIML for missing data from youth who selected “does not apply to me” for all items is not possible because only the information from these items is used for analysis. We conducted logistic regression analyses to determine whether some groups were more likely to be missing (full analyses are available upon request). Of the 944 youth who were missing on all items, disclosed youth had 0.47 lower odds (p < 0.001) to be missing on all items (n = 532, 7.37% of all disclosed youth) than non-disclosed youth (n = 412, 14.43% of all non-disclosed youth). The number and percentage of youth missing on all items by demographic characteristics are shown in Table 1. Compared to White youth, Black (OR = 1.63, p < 0.001) and Latino (OR = 1.36, p = 0.003) youth were more likely to be missing on all items. Bisexual youth were more likely to be missing than gay/lesbian youth (OR = 1.39, p < 0.001), and cisgender females were less likely to be missing than cisgender males (OR = 0.72, p < 0.001).

Confirmatory Factor Analyses

The model fit of the two latent factor model was acceptable (see Table 3). The one factor model fit the data poorly and was significantly worse than the two-factor model, Δχ2 = 6244.34, Δdf = 1, p < 0.001, ΔCFI = −0.259, ΔTLI = −0.358. Moreover, the unconstrained covariance between the rejection and acceptance factors was statistically different from zero (φ = −0.44, se = 0.01, p < 0.001), demonstrating that the factors were neither redundant nor orthogonal. Thus, we proceeded with the two-factor model for further analyses.

Table 3.

Model Fit Statistics of the Full Sample Confirmatory Factor Analyses

Model tested χ2 df p Δ χ2 Δ df p RMSEA RMSEA
90% CI		 CFI ΔCFI TLI ΔTLI
Single latent factor 8771.22 20 < 0.001 -- -- -- 0.219 (0.215, 0.223) 0.687 -- 0.562 --
Correlated two factor 1526.88 19 < 0.001 -- -- -- 0.093 (0.089, 0.097) 0.946 -- 0.920 --
Like with proud 1157.34 18 < 0.001 369.54 1 < 0.001 0.083 (0.079, 0.087) 0.959 −0.013 0.937 −0.017
Mock with negative comments 523.87 17 < 0.001 633.47 1 < 0.001 0.057 (0.053, 0.061) 0.982 −0.023 0.970 −0.033
Role model with involve 376.77 16 < 0.001 147.10 1 < 0.001 0.050 (0.045, 0.054) 0.987 0.005 0.977 0.007
Open in a new tab

Note. RMSEA=root mean square error of approximation; CFI=comparative fit index; TLI=Tucker-Lewis index.

The largest modification index suggested correlating the residuals of the proud and like items (SPEC = 0.25; MI = 359.47), which we considered appropriate given the similar wording of the two items. This residual correlation between items improved model fit (see Table 3). The next highest SPEC (0.18) was for freeing the residual correlation between the mock and negative comments items (MI = 730.61); both items similarly ask about negative statements made by parents that are directly made to youth. Next, the modification indices suggested correlating the residual variances of the role model and involve items (SPEC = 0.14; MI = 151.08). However, we could not determine a theoretical reason for correlating these items, and model fit did not significantly improve, so we did not include this correlation. The remaining SPECs were too low or theoretically implausible.

Measurement Invariance by Disclosure Status

Table 5 shows the factor loadings, intercepts, and residual variances for each family acceptance and rejection indicator for the non-disclosed and disclosed group models. Model fit statistics for each invariance test are shown in Table 6. The fit statistics showed that the model passed configural and metric invariance tests, indicating an equivalent factor structure (configural) and factor loadings (metric) across groups. We can directly compare the variances and covariances of the latent rejection and acceptance constructs with metric equivalency. The latent variance of family rejection was similar between the disclosed (ψ1,1 = 1.00; reference group) and non-disclosed (ψ1,1 = 1.01) group models (z = 0.01, p = .92); however, the latent variance of family acceptance for the non-disclosed group (ψ2,2 = 0.45) was roughly half that of the disclosed (ψ2,2 = 1.00) models (z = 96.07, p < .001), suggesting that there was lower variability in family acceptance among non-disclosed SGMY compared to disclosed SGMY. Further, the covariance between family acceptance and rejection was twice as strong (z = 54.07, p < 0.001) in the disclosed model (ψ1,2 = −0.47) compared to the non-disclosed model (ψ1,2 = −0.23). Thus, a one-unit increase in family rejection was associated with a −0.47 unit decrease in family acceptance for disclosed SGMY but only a −0.23 unit decrease in family acceptance for non-disclosed SGMY.

Table 5.

Factor Loadings, Intercepts, and Residual Variances of Family Rejection and Family Acceptance Items for the Correlated Two Factor Model by Disclosure Status

Non-disclosed (n =2444) Disclosed (n = 6683)
λ τ θ λ τ θ
Family rejection Mock 0.66 1.77 0.53 Mock 0.68 1.68 0.39
Negative comments 0.75 1.84 0.54 Negative comments 0.83 1.76 0.29
Bad things 0.82 2.66 0.40 Bad things 0.85 2.21 0.45
Feel bad 1.00 2.27 0.38 Feel bad 0.91 1.84 0.29
Family acceptance
Like 0.50 1.64 0.75 Like 0.62 2.42 0.93
Proud 0.61 1.33 0.21 Proud 0.91 1.96 0.48
Involved 0.49 1.27 0.19 Involved 0.73 1.63 0.41
Role model 0.41 1.15 0.10 Role model 0.62 1.40 0.32
Open in a new tab

Note. λ = item loading; τ =item intercept; θ = item residual variance.

Table 6.

Model Fit Statistics of Measurement Invariance Tests for the Single Latent Factor and Two Latent Factors Models by Disclosure Status

Model χ2 df p Δχ2 Δp RMSEA 95% CI CFI ΔCFI TLI Pass?
Configural 535.04 34 < 0.001 -- -- 0.057 (0.053, 0.061) 0.982 -- 0.970 ✓
Metric 566.03 40 < 0.001 −30.99 < 0.001 0.054 (0.050, 0.058) 0.981 0.001 0.973 ✓
Scalar 905.85 46 < 0.001 −339.83 < 0.001 0.056 (0.064, 0.068) 0.969 0.012 0.962 ✕
Freed like item 796.79 45 < 0.001 −230.76 < 0.001 0.061 (0.057, 0.064) 0.973 0.008 0.966 ✓
Open in a new tab

Note. RMSEA=root mean square error of approximation; CFI=comparative fit index; TLI=Tucker-Lewis index.

The model did not pass the test of scalar invariance. The largest MI suggested freeing the intercept of the like item (MI = 100.07, SEPC = −0.27 for the non-disclosed group, SPEC = 0.06 for the disclosed group). The model passed the scalar invariance test with this item freely estimated for both groups. Compared to the model for non-disclosed SGMY, in which the latent means were constrained to zero, the non-disclosed to parents group model had significantly higher family rejection (α1 = 0.46, se = 0.03, p < .001) and lower family acceptance (α2 = −0.54, se = 0.03, p < .001).

Discussion

To better understand the family environment of SGMY, we examined the dimensional properties of a measure of family acceptance and rejection behaviors and measurement differences between youth who had and had not disclosed their sexual orientation or gender identity to their parents. Major strengths of the study include the use of a large sample of SGMY, the use of an SGM-specific measure of family acceptance and rejection behaviors, and the ability to compare the measurement structure between disclosed and non-disclosed SGMY. Below, we highlight key takeaways from the results of the study.

Consistent with recent theorizing and commentary (Catalpa & McGuire, 2018; Fish, 2020), our results suggest that acceptance and rejection from family members comprise two distinct constructs in the family environment rather than a single unidimensional construct. These findings suggest that researchers should be measuring positive and negative familial behaviors and interactions as distinct constructs and operationalizing them as such in studies. Studies that reverse code items based on valence and calculate sum scores with accepting and rejecting behaviors combined may miss essential nuances in family dynamics that could uniquely shape SGMY development, health, and family relationships. Moreover, the covariance between family acceptance and rejection showed that these two constructs were correlated but not strongly negatively correlated, suggesting that many SGMY simultaneously experience both accepting and rejecting behaviors from parents and caregivers. Given recent research highlighting the deleterious impacts of ambiguous and inconsistent family responses to children’s sexual orientation and gender identity (Allen et al., 2022; Catalpa & McGuire, 2018; Tomlinson, 2021), the simultaneous modeling and exploration of these constructs in SGMY family environment could lead to important implications for future research and family intervention.

Research shows that family acceptance and rejection are strongly associated with SGMY well-being, including mental health (Ryan et al., 2009; Ryan et al., 2010; Shilo & Savaya, 2011) and suicidality (Hatchel et al., 2021). High-quality relationships with parents during adolescence protect health even when these relationships are strained (Umberson & Thomeer, 2020). Future research should consider to what degree parental rejection may nullify the positive impacts of acceptance, even when both co-occur within families (Tomlinson, 2021); these findings could have important clinical implications for supporting families as they navigate the coming out process and as families learn to support their SGMY. We also note that this correlation might suggest that many SGMY experience apathy or ambiguity from parents. Parental ambiguity toward SGMY’s identities can be common (Mena & Vaccaro, 2013), particularly among transgender youth (Allen et al., 2022; Catalpa & McGuire, 2018); previous work has shown that family ambiguity has negative impacts on health (Khaleque & Rohner, 2002). For example, study results in Allen et al. (2022) suggested that, among transgender adults, family environments characterized by ambiguity–that is, high levels of accepting and rejecting behaviors–were most harmful to mental and physical health, even more so than environments characterized by high levels of rejection and low levels of acceptance. Taken together, previous work suggests that strong, clear displays of acceptance are the most beneficial for the mental health of SGMY.

Measurement between groups was largely equivalent by disclosure status. We found this particularly interesting because the wording of the items imply that parents have some knowledge of youth’s sexual orientation or gender identity. Coming out as SGM is a complex process in which SGMY experience varying degrees of awareness, knowledge, and disclosure related to their sexual and/or gender identities (Bishop et al., 2020; Caba et al., 2022). Youth who have not directly told their parents about their identities may be carefully paying attention to their parents’ attitudes toward LGBTQ people to gauge how they might react to their disclosure (Jadwin-Cakmak et al., 2015), and thus respond to these items from that perspective. It is common for researchers who study SGMY’s family environment to exclude – intentionally or unintentionally – non-disclosed youth, which assumes that measures of familial experiences might not be as accurate in capturing the experiences of youth whose parents are unaware of their sexual or gender identity. Our study was also likely affected by the unintentional exclusion of non-disclosed youth, considering that non-disclosed youth were more likely to be missing on family acceptance and rejection because they selected “does not apply” to items that implied parents’ knowledge of their identities. However, the results of the current study suggest that these measures have utility for understanding family dynamics even when youth have not disclosed their sexual orientation and/or gender identity to their parents. Thus, we encourage researchers to consider how their study designs might be more inclusive of youth who have yet to disclose their sexual orientation and/or gender identity to parents and the degree to which their measures address these distinctions. Generally, the inclusion of youth who have yet to disclose their identity will provide understanding about the family environments of these youth and how family dynamics may influence when, how, and why youth disclose (or not) their identity with parents and others in their social network.

We observed marginal mean level differences in parental accepting and rejecting behaviors between disclosure groups: SGMY who had disclosed to parents reported higher acceptance and lower rejection than SGMY who had not disclosed. Moreover, the inverse association between family rejection and acceptance was stronger among disclosed compared to non-disclosed youth. Past research on parental reactions to disclosure has shown that SGMY report both higher parental acceptance (e.g., D’Augelli et al., 1998) and rejection (D'Augelli et al., 2010) compared to those who have not disclosed. Recently, longitudinal studies have begun to consolidate these inconsistent findings to show that, although initial disclosure is linked to rejection and poor outcomes, SGMY report greater support over time (Samarova et al., 2014). That is, SGMY may have more opportunities to cultivate parental support once disclosed. However, it is interesting that the two rejection items that differed between groups at the observed level seemed to reflect parents’ broader attitudes about SGM people (your parents say bad things about SGM people in general, your parents make you feel bad about being an SGM person); SGMY did not differ on items related to negative comments that were directly about their identities. We hesitate to speculate too strongly about these findings, but they could suggest that SGMY may struggle to call out personal rejection from family members; it may also be that SGMY are more likely to disclose in an environment in which family members appear less rejecting of SGM people in general than of SGMY’s own identities. Regardless of the reasons for this finding, rejection seems to be common in families, even in the presence of accepting behaviors and for youth who have disclosed their identity; further research is needed to better understand family dynamics around disclosure, particularly in the years following disclosure (see (Huebner et al., 2019). Research in this area will provide clinicians greater perspective on interventions that may support SGMY and families during this time.

Interestingly, variability in family rejection was similar between the two groups, but there was much more variability in family acceptance among disclosed youth than non-disclosed youth. Though it is difficult to fully understand these findings without including covariates that might account for some of this variability (e.g., time since disclosure, sociopolitical context), we highlight possible explanations that could be explored in future research. Research shows that parents who respond to disclosure in neutral or ambiguous ways are sometimes seen by SGMY as accepting of their identities because their parents have not responded explicitly or exclusively rejecting ways (Mena & Vaccaro, 2013). Some of this variability in acceptance among the disclosed group may not be a result of disclosure, but a precursor to disclosure. If parents present some affirmative language and behaviors towards SGM communities, youth may be more likely to disclose their identity. Research suggests that SGMY disclose their sexual or gender identity to parents for various reasons, such as a desire to be authentic to their true selves, even if their parents have not expressed high levels of acceptance (Grafsky, 2018). We also acknowledge that disclosure, particularly as measured here, may not necessarily reflect an active, agentic disclosure by SGMY, considering that some youth may have experienced accidental or forced disclosure (being “outed” by others). These suppositions further illustrate the need for more research on family dynamics that span the pre- and post-disclosure to parents.

Limitations and Future Directions

Our study is not without limitations. First, it is difficult to accurately measure youth's disclosure (Caba et al., 2022); thus, categorizing SGMY based on disclosure status will be naturally imprecise. We dichotomized SGMY as being disclosed to parents if they said that at least “a few” of their parents knew about their identity. Therefore, we might have identified measurement differences if we had more accurate disclosure measures. Second, these data are cross-sectional; we could not assess predictive validity and whether there was longitudinal measurement invariance in family acceptance and rejection. Longitudinal data and analyses are necessary to understand better the direction of associations between disclosure and family behaviors specific to sexual orientation and gender identity and how these associations change over time. In the absence of longitudinal data, disclosure timing can illuminate whether family relationships improve after youth disclose, whether youth disclose in the context of better family relationships, or perhaps a more reciprocal process. Third, we did not ask SGMY for the timing of when they disclosed their identities to their parents. When and how youth disclose could influence the degree to which family members engage in specific accepting and/or rejecting behaviors and, subsequently, how measures of family dynamics around sexual orientation and gender identity operate for SGMY who have and have not disclosed their identity. Last, we tested a few items that tapped into broad behaviors from family around youth sexual orientation and/or gender identity. More nuanced measures of family dynamics and environment might provide additional perspectives on measurement and the degree to which the family environment and measures differ for youth who have and have not disclosed their identity to their parents. Further examination of the dimensionality of other family acceptance and rejection measures would provide additional evidence for or against the multidimensionality of the measures found in the current study.

We also acknowledge that our critiques about how the field has approached the conceptualization of outness (including disclosure) and family environments also apply to the current study, particularly regarding intersectionality. Many studies examine a dominant group (i.e., White, gay, male) understanding of outness as explicitly disclosing a sexual or gender minority identity to others, which often does not reflect the realities of SGM youth of color and SGMY with other marginalized identities. For example, some SGM youth of color say that their parents are implicitly aware of their SGM identity and that a lack of acknowledgment about it means that their parents are accepting (Pollitt et al., 2018). Though these experiences are not just isolated to SGM people of color, binary categories based on whether youth have disclosed to others cannot capture differences in family acceptance and rejection related to such nuanced experiences of outness. Policy and clinical practices developed to improve the family environments of SGMY must be built on an accurate, methodologically rigorous, and inclusive body of research that adequately considers multiple aspects of youth’s identities.

Conclusion

Despite being key predictors of adolescent wellbeing, there remains limited understanding surrounding youth’s family environment and behaviors that characterize acceptance or rejection of SGMY’s sexual orientation and/or gender identity. The current study found that family acceptance and rejection comprise two separate, measurable constructs among SGMY. We noted few measurement differences between disclosed and non-disclosed SGMY, suggesting that studies on SGMY’s family environment should include and explore the experiences of SGMY who may or may not be out to parents and other family members. Continued measurement refinement and studies designed to explore the complex family dynamics of SGMY will provide valuable insight into factors influencing SGMY development and the development of interventions designed to support SGMY and their families.

Table 4.

Factor Loadings, Intercepts, and Residual Variances of Family Rejection and Family Acceptance Items for the Final Two-Factor Model

λ τ θ
Family Rejection
Mock 0.68 1.73 0.42
Negative comments 0.82 1.82 0.34
Bad things 0.86 2.33 0.45
Feel bad 0.95 1.95 0.30
Family Acceptance
Like 0.73 2.29 0.80
Proud 1.01 1.85 0.22
Involved 0.67 1.55 0.50
Role model 0.52 1.35 0.38
Open in a new tab

Note. λ = item loading; τ =item intercept; θ = item residual variance. The final two-factor model consisted of correlated residuals between the like and proud items and the mock and negative comments items.

Acknowledgments

Research reported in this publication was supported by the National Institute on Minority Health and Health Disparities of the National Institutes of Health under Award Number U54MD012388. This work was supported by grant P2CHD042849, Population Research Center, and by grant T32HD007081, Training Program in Population Studies, both awarded to the Population Research Center at The University of Texas at Austin by the Eunice Kennedy Shriver National Institute of Child Health and Human Development. The research was also supported in part by the National Institute on Alcohol Abuse and Alcoholism grant F32AA025814 (awarded to Pollitt) and the National Institutes on Drug Abuse grant K01DA047918 (awarded to Watson). Fish acknowledges support from the Eunice Kennedy Shriver National Center for Child Health and Human Development grant P2CHD041041, Maryland Population Research Center and the University of Maryland Prevention Research Center cooperative agreement #U48DP006382 from the Centers for Disease Control and Prevention (CDC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the CDC.

An earlier version of the results reported in this manuscript were presented at a conference. Study data and materials are available upon request. This study was not preregistered.

References

  1. Abreu RL, Lefevor GT, Gonzalez KA, Barrita AM, & Watson RJ (2022). Bullying, depression, and parental acceptance in a sample of Latinx sexual and gender minority youth. Journal of LGBT Youth, 1-18. 10.1080/19361653.2022.2071791 [DOI] [Google Scholar]
  2. Allen SH, Fish JN, McGuire JK, & Leslie LA (2022). Beyond the (family) binary: Family environment heterogeneity among transgender adults. Journal of Marriage and Family. 10.1111/jomf.12835 [DOI] [Google Scholar]
  3. Bishop MD, Fish JN, Hammack PL, & Russell ST (2020). Sexual identity development milestones in three generations of sexual minority people: A national probability sample. Dev Psychol, 56(11), 2177–2193. 10.1037/dev0001105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Byrne BM, Shavelson RJ, & Muthén B (1989). Testing for the equivalence of factor covariance and mean structures: The issue of partial measurement invariance. Psychological Bulletin, 105(3), 456–466. 10.1037/0033-2909.105.3.456 [DOI] [Google Scholar]
  5. Caba AE, Mallory AB, Simon KA, Rathus T, & Watson RJ (2022). Complex outness patterns among sexual minority youth: A latent class analysis. Journal of Youth and Adolescence, 51(4), 746–765. 10.1007/s10964-022-01580-x [DOI] [PubMed] [Google Scholar]
  6. Catalpa JM, & McGuire JK (2018). Family boundary ambiguity among transgender youth. Family Relations, 67(1), 88–103. 10.1111/fare.12304 [DOI] [Google Scholar]
  7. Cheung GW, & Rensvold RB (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9(2), 233–255. 10.1207/S15328007SEM0902_5 [DOI] [Google Scholar]
  8. Chu PS, Saucier DA, & Hafner E (2010). Meta-analysis of the relationships between social support and well-being in children and adolescents. Journal of Social and Clinical Psychology, 29(6), 624–645. 10.1521/jscp.2010.29.6.624 [DOI] [Google Scholar]
  9. D'Augelli AR (2002). Mental health problems among lesbian, gay, and bisexual youths ages 14 to 21. Clinical Child Psychology and Psychiatry, 7(3), 433–456. 10.1177/1359104502007003039 [DOI] [Google Scholar]
  10. D'Augelli AR, & Grossman AH (2001). Disclosure of sexual orientation, victimization, and mental health among lesbian, gay, and bisexual older adults. Journal of Interpersonal Violence, 16(10), 1008–1027. 10.1177/088626001016010003 [DOI] [Google Scholar]
  11. D'Augelli AR, Grossman AH, Starks MT, & Sinclair KO (2010). Factors associated with parents’ knowledge of gay, lesbian, and bisexual youths’ sexual orientation. Journal of GLBTFamily Studies, 6(2), 178–198. 10.1080/15504281003705410 [DOI] [Google Scholar]
  12. D'Augelli AR, Hershberger SL, & Pilkington NW (1998). Lesbian, gay, and bisexual youth and their families: Disclosure of sexual orientation and its consequences. American Journal of Orthopsychiatry, 68(3), 361–371. 10.1037/h0080345 [DOI] [PubMed] [Google Scholar]
  13. Doty ND, Willoughby BLB, Lindahl KM, & Malik NM (2010). Sexuality related social support among lesbian, gay, and bisexual youth. Journal of Youth and Adolescence, 39(10), 1134–1147. 10.1007/s10964-010-9566-x [DOI] [PubMed] [Google Scholar]
  14. Fish JN (2020). Sexual minority youth are at a disadvantage: what now? The Lancet Child & Adolescent Health, 4(1), 3–4. 10.1016/s2352-4642(19)30383-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fish JN, Russell BS, Watson RJ, & Russell ST (2020). Parent-child relationships and sexual minority youth: Implications for adult alcohol abuse. Journal of Youth and Adolescence, 49(10), 2034–2046. 10.1007/s10964-020-01299-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fish JN, & Russell ST (2018). Queering methodologies to understand queer families. Family Relations, 67(1), 12–25. 10.1111/fare.12297 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gamarel KE, Watson RJ, Mouzoon R, Wheldon CW, Fish JN, & Fleischer NL (2020). Family rejection and cigarette smoking among sexual and gender minority adolescents in the USA. International Journal of Behavioral Medicine, 27(2), 179–187. 10.1007/s12529-019-09846-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Grafsky EL (2018). Deciding to come out to parents: Toward a model of sexual orientation disclosure decisions. Family Process, 57(3), 783–799. 10.1111/famp.12313 [DOI] [PubMed] [Google Scholar]
  19. Grossman AH, Park JY, Frank JA, & Russell ST (2021). Parental responses to transgender and gender nonconforming youth: Associations with parent support, parental abuse, and youths' psychological adjustment. Journal of Homosexuality, 68(8), 1260–1277. 10.1080/00918369.2019.1696103 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hatchel T, Polanin JR, & Espelage DL (2021). Suicidal thoughts and behaviors among LGBTQ youth: Meta-analyses and a systematic review. Archives of Suicide Research, 25(1), 1–37. 10.1080/13811118.2019.1663329 [DOI] [PubMed] [Google Scholar]
  21. Hu LT, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  22. Huebner DM, Roche KM, & Rith KA (2019). Effects of family demographics and the passage of time on parents' difficulty with their lesbian, gay, or bisexual youth's sexual orientation. Archives of Sexual Behavior, 48(5), 1581–1588. 10.1007/s10508-019-1430-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jadwin-Cakmak LA, Pingel ES, Harper GW, & Bauermeister JA (2015). Coming out to dad: Young gay and bisexual men’s experiences disclosing same-sex attraction to their fathers. American Journal of Men's Health, 9(4), 274–288. 10.1177/1557988314539993 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Khaleque A, & Rohner RP (2002). Perceived parental acceptance-rejection and psychological adjustment: A meta-analysis of cross-cultural and intracultural studies. Journal of Marriage and Family, 64(1), 54–64. 10.1111/j.1741-3737.2002.00054.x [DOI] [Google Scholar]
  25. McConnell EA, Birkett M, & Mustanski B (2016). Families matter: Social support and mental health trajectories among lesbian, gay, bisexual, and transgender youth. Journal of Adolescent Health, 59(6), 674–680. 10.1016/j.jadohealth.2016.07.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mena JA, & Vaccaro A (2013). Tell me you love me no matter what: Relationships and self-esteem among GLBQ young adults. Journal of GLBT Family Studies, 9(1), 3–23. 10.1080/1550428X.2013.746052 [DOI] [Google Scholar]
  27. Miller KK, Watson RJ, & Eisenberg ME (2020). The intersection of family acceptance and religion on the mental health of LGBTQ youth. In Halkitis PN (Ed.), Annals of LGBTQ Public and Population Health (Vol. 1, pp. 27–42). Springer. 10.1891/LGBTQ.2019-0005 [DOI] [Google Scholar]
  28. Mohr J, & Fassinger R (2000). Measuring dimensions of lesbian and gay male experience. Measurement and Evaluation in Counseling and Development, 33(2), 66–90. 10.1080/07481756.2000.12068999 [DOI] [Google Scholar]
  29. Muthén LK, & Muthén BO (1998-2017). Mplus user's guide (8th ed.). Muthén & Muthén [Google Scholar]
  30. Pollitt AM, Muraco JA, Grossman AH, & Russell ST (2017). Disclosure stress, social support, and depressive symptoms among cisgender bisexual youth: Bisexual disclosure stress and social support. Journal of Marriage and Family, 79(5), 1278–1294. 10.1111/jomf.12418 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Robinson-Cimpian JP (2014). Inaccurate estimation of disparities due to mischievous responders: Several suggestions to assess conclusions. Educational Researcher, 43, 171–185. 10.3102/0013189X1453429 [DOI] [Google Scholar]
  32. Rosario M, Schrimshaw EW, & Hunter J (2009). Disclosure of sexual orientation and subsequent substance use and abuse among lesbian, gay, and bisexual youths: Critical role of disclosure reactions. Psychology of Addictive Behaviors, 23(1), 175–184. 10.1037/a0014284 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Ryan C, Huebner D, Diaz RM, & Sanchez J (2009). Family rejection as a predictor of negative health outcomes in White and Latino lesbian, gay, and bisexual young adults. PEDIATRICS, 123(1), 346–352. 10.1542/peds.2007-3524 [DOI] [PubMed] [Google Scholar]
  34. Ryan C, Russell ST, Huebner D, Diaz R, & Sanchez J (2010). Family acceptance in adolescence and the health of LGBT young adults. Journal of Child and Adolescent Psychiatric Nursing, 23(4), 205–213. 10.1111/j.1744-6171.2010.00246.x [DOI] [PubMed] [Google Scholar]
  35. Samarova V, Shilo G, & Diamond GM (2014). Changes in youths' perceived parental acceptance of their sexual minority status over time. Journal of Research on Adolescence, 24(4), 681–688. 10.1111/jora.12071 [DOI] [Google Scholar]
  36. Savin-Williams RC, & Ream GL (2003). Sex variations in the disclosure to parents of same-sex attractions. Journal of Family Psychology, 17(3), 429–438. 10.1037/0893-3200.17.3.429 [DOI] [PubMed] [Google Scholar]
  37. Shilo G, & Savaya R (2011). Effects of family and friend support on LGB youths' mental health and sexual orientation milestones. Family Relations, 60(3), 318–330. 10.1111/j.1741-3729.2011.00648.x [DOI] [Google Scholar]
  38. Snapp SD, Watson RJ, Russell ST, Diaz RM, & Ryan C (2015). Social support networks for LGBT young adults: Low cost strategies for positive adjustment. Family Relations, 64(3), 420–430. 10.1111/fare.12124 [DOI] [Google Scholar]
  39. Tomlinson K (2021). Family support, rejection, and connectedness: Family context and SGM youth mental health and substance use [master's thesis, University of Maryland College Park; ]. http://hdl.handle.net/1903/3 [Google Scholar]
  40. Umberson D, & Thomeer MB (2020). Family matters: Research on family ties and health, 2010-2020. Journal of Marriage and Family, 82(1), 404–419. 10.1111/jomf.12640 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Watson RJ, Grossman AH, & Russell ST (2019). Sources of social support and mental health among LGB youth. Youth & Society, 51(1), 30–48. 10.1177/0044118X16660110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Whittaker TA (2012). Using the modification index and standardized expected parameter change for model modification. The Journal of Experimental Education, 80(1), 26–44. 10.1080/00220973.2010.531299 [DOI] [Google Scholar]

RESOURCES