Abstract
In LGBTQIA+ adults, 36% are parents, but there are a limited number of parenting scales that have undergone psychometric evaluation in LGBTQIA+ parents. This study assessed the Multidimensional Assessment of Parenting Scale (MAPS) in LGBTQIA+ parents, addressing a gap in parenting scales considering both positive and negative aspects for this group. Our study includes two racially and ethnically diverse national samples of parents (n = 2,299, mean age = 37.05 years, 32.84% LGBTQIA+). We explored measurement invariance between LGBTQIA+ and non-LGBTQIA+ parents. The analysis revealed a modest fit for configural and metric invariance, indicating the MAPS structure and item loadings were consistent across groups, although scalar invariance was only partially supported. Subscales showed good internal consistency for both parent groups. Importantly, following false discovery rate corrections, interactions between LGBTQIA+ status and specific MAPS subscales (supportiveness, physical control) were significantly, but differentially, related to various child psychopathology dimensions, suggesting parenting impacts may differ for LGBTQIA+ families. These findings underscore the importance of including LGBTQIA+ perspectives in parenting research to develop tailored interventions that address their unique strengths and challenges. Acknowledging and supporting the diverse experiences of LGBTQIA+ parents can enhance targeted resources and interventions, contributing to better outcomes for all families.
Keywords: psychometrics, LGBTQIA+, parenting
Understanding the influence of parenting on youth development and psychosocial adjustment is crucial (Rose et al., 2018) to inform effective interventions and support strategies for families (Chen et al., 2019). Given the significant role of parenting interventions in mitigating youth mental health symptoms (Cohen et al., 2012; Herschell & McNeil, 2005), the need for psychometrically robust, evidence-based assessments of parenting is essential. Psychometrically robust assessments are not only vital for research but also to guide interventions aimed at enhancing youth development. In the Lesbian, Gay, Bisexual, Transgender, Queer, Intersex, Asexual+ (LGBTQIA+) adult community, 36% are parents, and 37% have had a child at some point in their lives (Brainer et al., 2020; Gates, 2013). As parenthood introduces new challenges for families (Moore & Brainer, 2013), it is important to include LGBTQIA+ parents in parenting research to ensure that underrepresented families are adequately represented and supported (Rodriguez et al., 2024; Rodriguez, La Barrie, et al., 2023). However, there are a limited number of parenting scales that have undergone psychometric scrutiny in LGBTQIA+ parents, which compromises our understanding of parenting in LGBTQIA+ parents (Rodriguez et al., 2024). Such research is necessary to know how to best support LGBTQIA+ parents, and in turn, their children.
Irrespective of sexual orientation or gender identity, parenting brings forth a spectrum of challenges and experiences that may differentially shape parenting (Moore & Brainer, 2013). It is essential to recognize and address these unique aspects in LGBTQIA+ parents to promote positive parenting practices and minimize negative ones. Thus, the inclusion of LGBTQIA+ parents in research is not just a matter of representation but a necessity for developing a comprehensive understanding of parenting across diverse family structures. Psychometric evaluations of parenting scales are particularly important given that various conceptualizations of parenting exist, and parenting itself is dynamic, varying across settings and developmental stages (Hurley et al., 2014; Locke & Prinz, 2002). An additional challenge in the field of parenting research is the limited psychometric data available for parenting measures (Hurley et al., 2014; Locke & Prinz, 2002). Even when psychometric data exist, reviews have indicated generally low quality in these measures and classified the psychometric evidence as “dismal” (Hurley et al., 2014). These psychometric limitations are worse in underrepresented groups, including LGBTQIA+ parents (Rodriguez et al., 2024; Rodriguez, La Barrie, et al., 2023).
Combined, such findings underscore the urgent need for valid and reliable assessments in the field as well as more careful attention to how we can use robust psychometric evaluations to prevent mispresenting individuals who are vulnerable to experiencing marginalization (Rodriguez, Cadet, et al., 2023). Through the lens of minority stress theory, it is known that there are several adverse effects on mental and emotional well-being caused by living in a stigmatizing society (Meyer, 1995, 2003). This persistent stress arises from being part of a marginalized group and is more intense than everyday life stressors (Meyer, 2003). Minority stress theory further suggests that individuals from minority groups face unique stressors due to their marginalized identities, as well as the societal systems that foster marginalization for these communities (Meyer, 2003). Furthermore, minority stress theory focuses on the various societal pressures minoritized individuals endure, shedding light on the potential causes of their adaptive challenges. Minority stress can undermine a person’s sense of safety and invulnerability (Meyer, 2003). As Meyer (2007) pointed out, this stress is specific, ongoing, and rooted in societal factors. Through the lens of minority stress theory, therefore, especially within the LGBTQIA+ community, parenting may take on unique dimensions, making psychometric evaluations of parenting scales particularly necessary for this community.
Historically, LGBTQIA+ families have often been portrayed in a negative light, with heterosexual and cisgender parents being considered the ideal (Goldberg, 2023). In fact, previous studies have shown that for LGBTQIA+ parents, becoming parents intensifies their minority stress due to increased social visibility and societal pressure (Amodeo et al., 2018). This approach has generally implied that children might fare worse with LGBTQIA+parents than those with non-LGBTQIA+parents. This is a belief that likely itself stemmed from the intense scrutiny LGBTQIA+ parents face due to societal and political biases expecting negative outcomes compared to cisgender heterosexual families. In addition to this scrutiny, LGBTQIA+ families face stigma and discrimination in various societal and political realms. However, research shows that children of LGBTQIA+ parents generally exhibit similar psychosocial adjustment levels as those with cisgender and heterosexual parents (Goldberg et al., 2014). This predominant hetero-cis-normative perspective in research has limited our understanding of how to effectively support LGBTQIA+ families and youth (Goldberg, 2023). Nevertheless, experiencing rejection and the constant need to justify their parental role may also impact the quality of parenting for LGBTQIA+ parents. While much of the comparative research is hetero-cis-normative, simply ignoring or overlooking LGBTQIA+ status in parenting studies is not the answer. Rather, parenting research must begin to define and operationalize the relevant constructs more precisely to better support LGBTQIA+ families.
In light of the absence of robust psychometric evaluations of parenting as well as the need to increase the representation of LGBTQIA+ parents in parenting research, this article focuses on measurement invariance, reliability, and validity of the Multidimensional Assessment of Parenting Scale (MAPS) when applied to LGBTQIA+ parents (Parent & Forehand, 2017). Whereas measurement invariance refers to whether the scale functions equivalently for participants in one group relative to another, differential validity refers to whether, and to what degree, an assessment is predictive for different groups (Linn, 1978). That is, how do different MAPS subscales, comprised of negative and positive domains of parenting, relate to youth mental health symptoms (e.g., emotional symptoms, conduct problems, hyperactivity/inattention, relationship problems, prosocial behavior) among parents of children aged 3–17 years? A lack of measurement invariance can lead to inaccuracies in measuring specific constructs. Establishing equivalent measures is a prerequisite to meaningful comparisons of observed scores across different populations. A lack of measurement invariance can result in observed score differences that do not truly reflect differences in the underlying attributes but rather the biases of the instrument itself. If differential validity exists for the MAPS for LGBTQIA+ parents and non-LGBTQIA+, it could inform parenting interventions and how they might be tailored to meet the specific needs of each parent and their child. Furthermore, observed differences could be indicative of the differential relative utility of the individual MAPS subscales in predicting child psychopathology. Conversely, the absence of differential validity would be similarly informative, as it would indicate that parenting subscales are similarly predictive of psychopathology across groups.
By exploring the psychometric properties of the MAPS in the context of LGBTQIA+ families, we aim to contribute to the calls for psychometrically robust parenting scales, ensuring that the tools used are inclusive, reliable, and valid across diverse family structures. Such efforts are crucial in advancing our understanding of parenting in LGBTQIA+ parents and in developing tailored interventions that effectively support them and their children. Enhancing the robustness of our approach, our study used data from two national U.S. samples, both oversampling for racial and ethnic minorities, to examine the psychometric properties of the MAPS.
Method
Ethical Approvals
Our study was comprised of two national samples of parents. The first study, which was mostly comprised of heterosexual and cisgender parents from racially and ethnically diverse backgrounds, received ethical approval from the University of Georgia Institutional Review Board. The second, comprised of LGBTQIA+ parents, received approval from the University of Illinois Urbana Champaign.
Participant Eligibility
Eligible parents for the survey were at least 18 years of age, parenting at least one child aged 3–17 years, and living with their target child at least 50% of the time, consistent with the MAPS development and validation sample. For the purposes of the study, if the parent reported having more than one child, the parent was asked to enter the name, age, and sex of each child. If the parent had more than one child, they were instructed to choose the one with the most recent birthday. Only one parent per family was allowed to complete the surveys to ensure data independence and integrity. To ensure this, Qualtrics only allows one member per household could participate. For the second study, parents had an additional eligibility criterion of identifying as a member of the LGBTQIA+ community. Parents were not aware of the eligibility criteria; instead, parents were screened out if they did not meet criteria based on a preliminary survey.
Recruitment
The survey was distributed by Qualtrics XM, a data collection and management company. Available participants in Qualtrics panels are nationally representative samples of adults who volunteer to participate in surveys online in exchange for monetary compensation. Qualtrics contacted their pre-screened participant panels to assess eligibility and emailed the survey to participants to complete via a unique link, then disbursed compensation as a participation incentive. Recruitment ran from November to December 2022 for the first sample, and from November to December 2023 for the second sample. For quality purposes, only eligible participants who correctly and appropriately answered attention and quality checks embedded throughout the survey were included in the final dataset. Examples of these validity checks included instructions to select a specific response (e.g., “Select ‘Strongly Agree’ to indicate you are paying attention”) embedded as part of questionnaires. Furthermore, the surveys, for both samples, were piloted in n = 150 participants initially, and an average was taken of their survey completion time. Participants whose survey completion time was faster than 1.5 standard deviations from the pilot average were excluded. For the first sample, n = 147 participants were excluded due to standard deviations faster than 1.5 standard deviations and not passing three out of three validity checks. From the second sample, n = 70 were excluded using similar criteria.
Measures
Sociodemographic Characteristics.
Participants were asked to complete a sociodemographic questionnaire about their age, gender, race, ethnicity, sexual and gender identity, educational attainment, relationship status and length, employment status, source of income, personal and household income, and current living arrangements (i.e., time spent with child). Parenting-related questions include the number of children and gender and age of the children.
Multidimensional Assessment of Parenting Scale.
As previously noted, the 34-item MAPS (Parent & Forehand, 2017) is composed of seven factors of parenting practices within two broadband domains (i.e., positive and negative parenting). The original MAPS has a 16-item broadband positive parenting subscale composed of items related to proactive parenting, positive reinforcement, warmth, and supportive parenting practices. The original MAPS also includes an 18-item negative parenting subscale addressing hostility, lax control, and physical control. Items are rated on a 5-point Likert-type scale ranging from 1 (never) to 5 (always), with greater scores reflecting higher positive or negative parenting, depending on the specified domain or subscale.
Child Symptoms of Psychopathology.
Parents completed the Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997), a 25-item behavioral screening questionnaire that assesses conduct problems, hyperactivity-inattention, emotional symptoms, peer problems, and prosocial behavior. The SDQ uses a 3-point Likert-type scale ranging from 1 (not true) to 3 (certainly true) to describe how much an item describes the target child’s behavior. For parents of target children ages 2–4 years, reliability was acceptable in this sample (v = .79). For parents of target children ages 5–11 and 12–17 years, reliability was acceptable in both groups (v = .83).
Analysis
Descriptive statistics were used to summarize the sample characteristics, including demographic information and scores on the MAPS subscales. Preliminary checks for normality, missing data, and outliers will be conducted to ensure data integrity and suitability for further analyses. No participants were removed due to missing data, although n = 35 participants were removed due to multivariate outliers (measured by Mahalanobis distance probabilities calculated using a chi-square cumulative probability distribution) and univariate outliers (measured as Z scores greater than 3.5).
First, we tested measurement invariance of the MAPS using multi-group confirmatory factor analyses (CFAs) to test the factor structure across different groups (heterosexual and cisgender parents vs. LGBTQIA+ parents). Multi-group CFA involves comparing the factor structure across these groups to assess whether the scale functions similarly. We then investigated measurement invariance across groups by sequentially constraining parameters (configural, metric, scalar invariance) and examining changes in fit indices. Model fit was evaluated using χ2 tests, comparative fit index (CFI), Tucker–Lewis Index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR). A CFI value ⩾ .95 indicates an “excellent” fit of the model to the observed data (Brown, 2014; Kline, 2015; van Zyl & Ten Klooster, 2022). Similarly, an RMSEA value ⩽ .05 indicates a close fit of the model to the data (Brown, 2014; Kline, 2015). If the lowest level of invariance (i.e., configural) was not met, we will test the factor structures of the groups separately.
Next, multiple regression analyses were conducted to assess the predictive validity of each MAPS subscale for youth mental health (SDQ subscales). Interaction terms (e.g., MAPS subscale * LGBTQIA+ status) were included in the regression models to test for differential predictive validity across LGBTQIA+ status. In designing our analytic plan, we balanced comprehensiveness and parsimony by including several interactions per outcome (e.g., emotional symptoms, conduct problems) within a single model. This approach limited the number of models, simplifying the presentation of results while providing a detailed assessment of the MAPS’ psychometric properties. We focused on SGM status and excluded other variables to avoid introducing extraneous factors, thereby enhancing the robustness of our analysis. This strategy ensures each predictor or interaction term is interpreted within the context of related constructs, offering comprehensive understanding of the MAPS’ performance between LGBTQIA+ and non-LGBTQIA+ groups. Standardized effect sizes were reported to interpret the magnitude of observed relations. Given significant interactions, we conducted simple slope analyses to examine within-group relations between the independent and dependent variables. All analyses were conducted using R, using packages “psych” and “lavaan.”
Results
Sociodemographic Characteristics
Combining two racially and ethnically diverse national samples, participants were n = 2,299 parents. Demographic characteristics of the sample, as well as comparisons by LGBTQIA+ status, are provided in Table 1.
Table 1.
Sociodemographic Characteristics of Parents and Their Children (N = 2,089).
| Variable | N (%)/M (SD) | Non-LGBTQIA+ (n = 1,532) | LGBTQIA+ (n = 556) | F/χ2, p |
|---|---|---|---|---|
|
| ||||
| Parents | ||||
| Age | 37.38 (9.14) | 38.26 (9.12) | 34.96 (8.78) | 54.81, <.001 |
| Sex | ||||
| Male | 985 (47.2%) | 816 (53.2%) | 169 (30.4%) | |
| Female | 1104 (52.8%) | 717 (46.8%) | 387 (69.6%) | 85.37, <.001 |
| Ethnoracial identity | ||||
| White | 459 (22.0%) | 400 (26.1%) | 59 (10.6%) | |
| Asian | 437 (20.9%) | 346 (22.6%) | 91 (16.4%) | |
| Black | 457 (21.9%) | 353 (23.0%) | 104 (18.7%) | |
| Hispanic | 546 (26.1%) | 385 (25.1%) | 161 (29.0%) | |
| Native | 107 (5.1%) | 30 (2.0%) | 77 (13.8%) | |
| Other | 83 (4.0%) | 19 (1.2%) | 64 (11.5%) | 278.794, <.001 |
| Relationship status | ||||
| Never married | 383 (18.3%) | 259 (16.9%) | 124 (22.3%) | |
| Separated or divorced | 176 (8.4%) | 122 (8.0%) | 54 (9.7%) | |
| Married | 1142 (54.7%) | 922 (60.1%) | 220 (39.6%) | |
| Widowed | 23 (1.1%) | 14 (0.9%) | 9 (1.6%) | |
| In a committed relationship | 350 (16.8%) | 210 (13.7%) | 140 (25.2%) | |
| Prefer not to answer | 15 (0.7%) | 6 (0.4%) | 9 (1.6%) | 82.10, <.001 |
| Education | ||||
| Post-graduate degree | 297 (14.2%) | 242 (15.8%) | 55 (9.9%) | |
| Bachelor’s degree | 547 (26.2%) | 424 (27.7%) | 123 (22.1%) | |
| Associate’s degree | 265 (12.7%) | 191 (12.5%) | 74 (13.3%) | |
| Partial college | 404 (19.3%) | 275 (17.9%) | 129 (23.2%) | |
| High school or GED | 515 (24.7%) | 367 (23.9%) | 148 (26.6%) | |
| Partial high school | 61 (2.9%) | 34 (2.2%) | 27 (4.9%) | 31.73, <.001 |
| Employment | ||||
| Not employed outside the home | 425 (20.3%) | 304 (19.8%) | 121 (21.8%) | |
| Work part-time | 337 (16.1%) | 216 (14.1%) | 121 (21.8%) | |
| Work full-time | 1290 (61.8%) | 993 (64.8%) | 297 (53.4%) | |
| Prefer not to answer | 37 (1.8%) | 20 (1.3%) | 17 (3.1%) | 31.24, <.001 |
| Household yearly income | ||||
| Less than $10,000 | 179 (8.6%) | 109 (7.1%) | 70 (12.6%) | |
| $10,000–$30,000 | 283 (13.5%) | 200 (13.0%) | 83 (14.9%) | |
| $30,000–$50,000 | 356 (17.0%) | 241 (15.7%) | 115 (20.7%) | |
| $50,000–$70,000 | 364 (17.4%) | 242 (15.8%) | 122 (21.9%) | |
| $70,000–$90,000 | 256 (12.3%) | 202 (13.2%) | 54 (9.7%) | |
| $90,000–$100,000 | 196 (9.4%) | 153 (10.0%) | 43 (7.7%) | |
| Over $100,000/year | 409 (19.6%) | 355 (23.2%) | 54 (9.7%) | |
| Prefer not to answer | 46 (2.2%) | 31 (2.0%) | 15 (2.7%) | 74.85, <.001 |
| Target child | ||||
| Age | 9.84 (4.38) | 9.79 (4.36) | 9.98 (4.41) | 0.77, .379 |
Note. Age is presented as mean (standard deviation). Percentages may not total 100% due to rounding. Household income categories are in USD.
Measurement Invariance
Measurement invariance testing between LGBTQIA+ and non-LGBTQIA+ groups was conducted using a series of CFA models. The tests aimed to determine whether the MAPS scale measured the same constructs of the MAPS in a similar way across LGBTQIA+and non-LGBTQIA+parents and are summarized in Table 2.
Table 2.
Tests of Measurement Invariance by LGBTQIA+ Status.
| Models | df | AIC | BIC | χ2 | Δχ2 | ΔRMSEA | ΔCFI | Δdf | p (Δχ2) |
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| Configural model | 1078 | 182,872 | 184,294 | 4,017.3 | |||||
| Metric vs. Configural | 1106 | 182,855 | 184,119 | 4,055.9 | 38.627 | 0.02 | 0.00 | 28 | .087 |
| Metric vs. Scalar | 1134 | 182,894 | 184,000 | 4,150.8 | 94.919 | 0.05 | 0.00 | 28 | <.001 |
Note. Δχ2 represents the chi-square difference test comparing nested models. ΔRMSEA and ΔCFI represent changes in the root mean square error of approximation and comparative fit index, respectively. Configural invariance tests whether the basic factor structure is consistent across groups. Metric invariance tests whether factor loadings are equivalent across groups. Scalar invariance tests whether item intercepts are equivalent across groups.
Configural Invariance.
The first model, assessing configural invariance, provided a baseline model for comparison. The configural model had a modest fit (χ2 = 4,043.6, df = 1,078, RMSEA = 0.051, SRMR = .050, CFI = 0.92, TLI = 0.91) to the data. The configural invariance model, which does not impose any equality constraints across groups, suggested that the basic factor structure of the MAPS scale was consistent across the LGBTQIA+ and non-LGBTQIA+ groups.
Metric Invariance.
Subsequently, metric invariance was assessed by constraining the factor loadings to be equal across groups. The metric model had a similarly good fit (χ2 = 4,082.2, df = 1,106, RMSEA = 0.051, SRMR = .052, CFI = 0.92, TLI = 0.92). The difference in model fit from the configural model (Δχ2 = 38.61, Δdf = 28, p = .087, ΔRMSEA = .02) indicated no significant worsening of model fit. These results combined suggest that the way variables load onto MAPS factors is similar across groups, supporting metric invariance.
Scalar Invariance.
The scalar invariance model, which constrained the intercepts to be equal across groups, was then tested. The scalar model had a similarly good fit (χ2 = 4,176.96, df = 1134, RMSEA = 0.051, SRMR = .052, CFI = 0.92, TLI = 0.92). However, comparing this model to the metric invariance model revealed a significant worsening of fit (Δχ2 = 94.725, Δdf = 28, p < .001, ΔRMSEA = 0.05). The significant difference indicates that scalar invariance does not hold, suggesting differences in item intercepts between the LGBTQIA+ and non-LGBTQIA+ groups.
Partial Scalar Invariance.
Given that scalar invariance was not achieved, we proceeded to test for partial scalar invariance. To address this, modification indices suggested the relaxation of equality constraints on the intercepts for Item 25 (“I spank my child when I am angry”). Notably, allowing the intercept for Item 25 to vary across groups significantly improved the model fit, indicating partial scalar invariance (Δχ2 = 94.919, df = 28, p < .001). The model fitting the partial scalar invariance criteria (with the intercept of Item 25 free to vary) demonstrated acceptable fit indices: χ2 = 4,055.92, df = 1,106, CFI = .92, TLI = .917, RMSEA = .051, 90% CI: [.049, .052], SRMR = .051. The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) for the partial scalar model were 182,854.592 and 184,118.947, respectively, indicating a preferred model over the strict scalar model (AIC = 182,893.511, BIC = 183,999.821) when considering the balance of model fit and complexity. Thus, except for Item 25, which is part of the Physical Control subscale, scalar invariance was supported across the remaining subscales.
Reliability
Overall, the MAPS subscales demonstrated robust reliability across most subscales for both LGBTQIA+ and non-LGBTQIA+ parents, as noted in Table 3. The Omega coefficients indicate a high level of internal consistency. The most notable difference was in the Supportiveness subscale, for which the alpha coefficient varied slightly between groups, with a higher Omega = 0.82 for LGBTQIA+ parents compared to 0.77 for non-LGBTQIA+ parents.
Table 3.
Omega Reliability Coefficients in LGBTQIA+ Versus Non-LGBTQIA+ Parents.
| MAPS subscales | LGBTQIA+ | Non-LGBTQIA+ |
|---|---|---|
|
| ||
| MAPS hostility | 0.88 | 0.87 |
| MAPS lax parenting | 0.85 | 0.83 |
| MAPS physical control | 0.90 | 0.90 |
| MAPS proactive parenting | 0.83 | 0.82 |
| MAPS reinforcement | 0.85 | 0.83 |
| MAPS supportiveness | 0.82 | 0.77 |
| MAPS warmth | 0.85 | 0.83 |
Note. Omega reliability coefficients provide an estimate of the internal consistency of the subscales. Higher values indicate greater internal consistency.
Validity
To examine differential predictive validity, a series of multiple regression analyses were conducted to assess the predictive validity of each MAPS subscale for each youth mental health subscale (SDQ subscales) on observed scores. First, interaction terms (e.g., MAPS subscale * LGBTQIA+ status) were included in the regression models to test for differential predictive validity across LGBTQIA+ status. Our regression analyses explored the relationships between several psychosocial constructs—hostility, lax parenting, physical control, reinforcement, warmth, and supportiveness (measured by MAPS subscales)—and outcomes measured by the SDQ, considering the moderating role of parental LGBTQIA+ status (1 = LGBTQIA+; 0 = not LGBTQIA+). The study’s findings detail the associations between parenting constructs on emotional symptoms, conduct problems, hyperactivity, prosocial behavior, peer problems, and overall child psychopathology, with a focus on interaction effects between each MAPS subscale and LGBTQIA+ status. Although not preregistered, before interpreting our results, we applied a false discovery rate (FDR) correction, as presented in Table 4. Estimated marginal means (using the emmeans package in R), presented in Table 5, examined the difference in estimated SDQ subscale scores between low and high levels of Supportiveness and Physical Control (1.5 standard deviations below and above mean).
Table 4.
Interactions Between LGBTQIA+ Status and MAPS Subscales Predicting SDQ Subscales.
| Outcome: Emotional symptoms | |||||
|---|---|---|---|---|---|
| B | SE | t | p | FDR p | |
|
| |||||
| Intercept | 6.298 | 0.407 | 15.484 | <.001 | <.001 |
| LGBTQIA+ status | 0.417 | 0.650 | 0.641 | .522 | .659 |
| Hostility | 0.085 | 0.014 | 6.121 | <.001 | <.001 |
| Lax parenting | 0.090 | 0.012 | 7.440 | <.001 | <.001 |
| Physical control | 0.138 | 0.020 | 6.854 | <.001 | <.001 |
| Pp | −0.015 | 0.017 | −0.896 | .370 | .501 |
| Reinforcement | −0.040 | 0.027 | −1.454 | .146 | .226 |
| Supportiveness | −0.051 | 0.039 | −1.292 | .196 | .290 |
| Warmth | −0.065 | 0.024 | −2.679 | .007 | .022 |
| LGBTQIA+ status × hostility | 0.012 | 0.024 | 0.514 | .607 | .720 |
| LGBTQIA+ status × lax parenting | 0.008 | 0.021 | 0.388 | .698 | .779 |
| LGBTQIA+ status × physical control | −0.077 | 0.035 | −2.210 | .027 | .065 |
| LGBTQIA+ status × proactive parenting | 0.059 | 0.029 | 2.062 | .039 | .082 |
| LGBTQIA+ status × reinforcement | −0.083 | 0.050 | −1.650 | .099 | .170 |
| LGBTQIA+ status × supportiveness | 0.188 | 0.069 | 2.735 | .006 | .019 |
| LGBTQIA+ status × warmth | −0.095 | 0.043 | −2.212 | .027 | .065 |
|
Outcome: Conduct problems | |||||
| B | SE | t | P | FDR p | |
|
| |||||
| (Intercept) | 4.936 | 0.309 | 15.985 | <.001 | <.001 |
| LGBTQIA+ status | 0.088 | 0.494 | 0.178 | .859 | .916 |
| Hostility | 0.059 | 0.011 | 5.580 | <.001 | <.001 |
| Lax parenting | 0.068 | 0.009 | 7.455 | <.001 | <.001 |
| Physical control | 0.168 | 0.015 | 10.975 | <.001 | <.001 |
| Proactive parenting | −0.015 | 0.013 | −1.198 | .231 | .322 |
| Reinforcement | −0.059 | 0.021 | −2.854 | .004 | .014 |
| Supportiveness | −0.044 | 0.030 | −1.456 | .146 | .226 |
| Warmth | −0.039 | 0.019 | −2.104 | .036 | .076 |
| LGBTQIA+ status × hostility | 0.008 | 0.018 | 0.455 | .649 | .746 |
| LGBTQIA+ status × lax parenting | −0.001 | 0.016 | −0.067 | .947 | .967 |
| LGBTQIA+ status × physical control | −0.053 | 0.027 | −1.995 | .046 | .091 |
| LGBTQIA+ status × proactive parenting | 0.027 | 0.022 | 1.233 | .218 | .317 |
| LGBTQIA+ status × reinforcement | −0.005 | 0.038 | −0.122 | .903 | .942 |
| LGBTQIA+ status × supportiveness | 0.035 | 0.052 | 0.665 | .506 | .656 |
| LGBTQIA+ status × warmth | −0.015 | 0.033 | −0.455 | .649 | .746 |
|
Outcome: hyperactivity | |||||
| B | SE | t | P | FDR p | |
|
| |||||
| Intercept | 4.419 | 0.357 | 12.387 | <.001 | <.001 |
| LGBTQIA+ status | −0.326 | 0.570 | −0.572 | .567 | 0.689 |
| Hostility | 0.059 | 0.012 | 4.843 | <.001 | <.001 |
| Lax parenting | 0.024 | 0.011 | 2.236 | .025 | 0.065 |
| Physical control | 0.095 | 0.018 | 5.374 | <.001 | <.001 |
| Proactive parenting | 0.009 | 0.014 | 0.627 | .531 | 0.662 |
| Reinforcement | −0.011 | 0.024 | −0.450 | .653 | .746 |
| Supportiveness | −0.085 | 0.035 | −2.471 | .014 | .038 |
| Warmth | −0.003 | 0.021 | −0.132 | .895 | .942 |
| LGBTQIA+ status × hostility | 0.041 | 0.021 | 1.962 | .050 | .096 |
| LGBTQIA+ status × lax parenting | <.001 | 0.018 | 0.024 | .981 | .981 |
| LGBTQIA+ status × physical control | −0.123 | 0.031 | −3.988 | <.001 | <.001 |
| LGBTQIA+ status × proactive parenting | 0.048 | 0.025 | 1.894 | .058 | .106 |
| LGBTQIA+ status × reinforcement | −0.086 | 0.044 | −1.937 | .053 | .100 |
| LGBTQIA+ status × supportiveness | 0.191 | 0.060 | 3.174 | .002 | .006 |
| LGBTQIA+ status × warmth | −0.036 | 0.038 | −0.965 | .335 | .459 |
|
Outcome: Prosocial behavior | |||||
| B | SE | t | p | FDR p | |
|
| |||||
| Intercept | 6.615 | 0.383 | 17.292 | <.001 | <.001 |
| LGBTQIA+ status | 1.356 | 0.612 | 2.217 | .027 | .065 |
| Hostility | −0.033 | 0.013 | −2.552 | .011 | .031 |
| Lax parenting | −0.024 | 0.011 | −2.127 | .034 | .075 |
| Physical control | 0.013 | 0.019 | 0.668 | .504 | .656 |
| Proactive parenting | 0.076 | 0.016 | 4.893 | <.001 | <.001 |
| Reinforcement | 0.083 | 0.026 | 3.237 | .001 | .005 |
| Supportiveness | 0.201 | 0.037 | 5.420 | <.001 | <.001 |
| Warmth | 0.047 | 0.023 | 2.049 | .041 | .083 |
| LGBTQIA+ status × hostility | −0.007 | 0.022 | −0.314 | .754 | .832 |
| LGBTQIA+ status × lax parenting | 0.017 | 0.020 | 0.881 | .379 | .505 |
| LGBTQIA+ status × physical control | −0.002 | 0.033 | −0.076 | .940 | .967 |
| LGBTQIA+ status × proactive parenting | −0.015 | 0.027 | −0.568 | .570 | .689 |
| LGBTQIA+ status × reinforcement | −0.065 | 0.047 | −1.363 | .173 | .260 |
| LGBTQIA+ status × supportiveness | −0.078 | 0.065 | −1.213 | .225 | .322 |
| LGBTQIA+ status × warmth | 0.063 | 0.040 | 1.546 | .122 | .200 |
|
Outcome: Peer problems | |||||
| B | SE | t | p | FDR p | |
|
| |||||
| Intercept | 4.703 | 0.270 | 17.443 | <.001 | <.001 |
| LGBTQIA+ status | 0.519 | 0.431 | 1.204 | .229 | .322 |
| Hostility | −0.001 | 0.009 | −0.054 | .957 | .967 |
| Lax parenting | 0.075 | 0.008 | 9.370 | <.001 | <.001 |
| Physical control | 0.109 | 0.013 | 8.151 | <.001 | <.001 |
| Proactive parenting | −0.003 | 0.011 | −0.239 | .811 | .875 |
| Reinforcement | −0.050 | 0.018 | −2.773 | .006 | .018 |
| Supportiveness | −0.046 | 0.026 | −1.747 | .081 | .141 |
| Warmth | −0.023 | 0.016 | −1.416 | .157 | .239 |
| LGBTQIA+ status × hostility | 0.034 | 0.016 | 2.149 | .032 | .074 |
| LGBTQIA+ status × lax parenting | −0.021 | 0.014 | −1.510 | .131 | .210 |
| LGBTQIA+ status × physical control | −0.068 | 0.023 | −2.940 | .003 | .011 |
| LGBTQIA+ status × proactive parenting | 0.012 | 0.019 | 0.652 | .514 | .658 |
| LGBTQIA+ status × reinforcement | −0.052 | 0.033 | −1.544 | .123 | .200 |
| LGBTQIA+ status × supportiveness | 0.144 | 0.045 | 3.157 | .002 | .006 |
| LGBTQIA+ status × warmth | −0.061 | 0.029 | −2.131 | .033 | .075 |
|
Outcome: Overall child psychopathology | |||||
| B | SE | t | p | FDR p | |
|
| |||||
| Intercept | 20.356 | 1.025 | 19.869 | <.001 | <.001 |
| LGBTQIA+ status | 0.697 | 1.638 | 0.426 | .670 | .757 |
| Hostility | 0.203 | 0.035 | 5.784 | <.001 | <.001 |
| Lax parenting | 0.257 | 0.030 | 8.445 | <.001 | <.001 |
| Physical control | 0.511 | 0.051 | 10.045 | <.001 | <.001 |
| Proactive parenting | −0.023 | 0.042 | −0.562 | .574 | .689 |
| Reinforcement | −0.160 | 0.069 | −2.324 | .020 | .055 |
| Supportiveness | −0.225 | 0.099 | −2.272 | .023 | .062 |
| Warmth | −0.130 | 0.061 | −2.116 | .034 | .075 |
| LGBTQIA+ status × hostility | 0.095 | 0.060 | 1.590 | .112 | .189 |
| LGBTQIA+ status × lax parenting | −0.013 | 0.052 | −0.255 | .799 | .871 |
| LGBTQIA+ status × physical control | −0.321 | 0.088 | −3.641 | <.001 | .001 |
| LGBTQIA+ status × proactive parenting | 0.146 | 0.072 | 2.021 | .043 | .087 |
| LGBTQIA+ status × reinforcement | −0.225 | 0.127 | −1.773 | .076 | .136 |
| LGBTQIA+ status × supportiveness | 0.557 | 0.173 | 3.222 | .001 | .005 |
| LGBTQIA+ status × warmth | −0.207 | 0.108 | −1.912 | .056 | .103 |
Note. MAPS = Multidimensional Assessment of Parenting Scale; SDQ = Strengths and Difficulties Questionnaire; FDR = false discovery rate. Bold values indicate FDR p < .05.
Table 5.
Estimated Marginal Means for the Interaction Effects Between LGBTQIA+ Status and Supportiveness/Physical Control at ± 1.5 Standard Deviations.
| Outcome measure | Predictor | Group | EMM | SE | 95% CI | Estimate difference | p |
|---|---|---|---|---|---|---|---|
|
| |||||||
| Emotional symptoms | Supportiveness | Non-LGBTQIA+ | 5.23 | 0.09 | [5.05, 5.40] | −1.17 | <.001 |
| Emotional symptoms | Supportiveness | LGBTQIA+ | 6.39 | 0.15 | [6.09, 6.69] | — | – |
| Hyperactivity | Supportiveness | Non-LGBTQIA+ | 5.23 | 0.09 | [5.05, 5.40] | −1.17 | <.001 |
| Hyperactivity | Supportiveness | LGBTQIA+ | 6.39 | 0.15 | [6.09, 6.69] | — | — |
| Peer problems | Physical control | Non-LGBTQIA+ | 4.51 | 0.07 | [4.38, 4.64] | −0.559 | <.001 |
| Peer problems | Physical control | LGBTQIA+ | 5.07 | 0.11 | [4.84, 5.29] | — | — |
| Peer problems | Supportiveness | Non-LGBTQIA+ | 4.51 | 0.07 | [4.38, 4.64] | −0.558 | <.001 |
| Peer problems | Supportiveness | LGBTQIA+ | 5.06 | 0.11 | [4.84, 5.29] | — | — |
| Hyperactivity | Physical control | Non-LGBTQIA+ | 5.22 | 0.09 | [5.05, 5.40] | −1.17 | <.001 |
| Hyperactivity | Physical control | LGBTQIA+ | 6.39 | 0.15 | [6.10, 6.69] | — | — |
| Child psychopathology | Physical control | Non-LGBTQIA+ | 21.9 | 0.25 | [21.4, 22.4] | −3.61 | <.001 |
| Child psychopathology | Physical control | LGBTQIA+ | 25.5 | 0.43 | [24.6, 26.3] | — | — |
| Child psychopathology | Supportiveness | Non-LGBTQIA+ | 21.9 | 0.25 | [21.4, 22.4] | −3.6 | <.001 |
| Child psychopathology | Supportiveness | LGBTQIA+ | 25.5 | 0.43 | [24.6, 26.3] | — | — |
Note. EMM = estimated marginal mean; SE = standard error; 95% CI = confidence interval. Provides a range within which the true mean is expected to lie with 95% confidence. Estimate: Difference in EMMs between non-LGBTQIA+ and LGBTQIA+ groups for the interactions.
Overall, no significant associations emerged predicting conducting problems or prosocial behavior; however, interactions emerged between LGBTQIA+ status and supportiveness and physical control predicting emotional symptoms, hyperactivity, peer problems, and overall child psychopathology. Significant interactions are discussed below and plotted in Figures 1 and 2.
Figure 1.
Interaction Between Supportiveness and LGBTQIA+ Status Predicting Child Psychopathology.
Note. This figure demonstrates the interaction effect between parental supportiveness and LGBTQIA+ status on child psychopathology. For children with LGBTQIA+ parents, higher levels of supportiveness are associated with lower levels of child psychopathology, as indicated by the steeper negative slope.
Figure 2.
Interaction Between Physical Control and LGBTQIA+ Status Predicting Child Psychopathology.
Note. This figure illustrates the interaction effect between parental physical control and LGBTQIA+ status on child psychopathology. For children with LGBTQIA+ parents, higher levels of physical control are significantly associated with higher levels of child psychopathology, as indicated by the more pronounced positive slope. This suggests that physical control has a more substantial adverse impact on psychopathology symptoms in children with LGBTQIA+ parents than those with non-LGBTQIA+ parents.
A significant interaction between Supportiveness and LGBTQIA+ status suggested the association between Emotional Symptoms and Supportiveness is moderated by LGBTQIA+ status (B = 0.188, SE = 0.069, FDR p = .019). Specifically, the association between increased supportiveness and decreased emotional symptoms is stronger for children with LGBTQIA+ parents than for children with non-LGBTQIA+ parents. This is depicted in Figure 1 (top right), where the slope for the LGBTQIA+ group is steeper, indicating a more robust negative relationship for the LGBTQIA+ group.
In terms of hyperactivity symptoms, the interaction between LGBTQIA+ status and Physical Control was found to be significant (B = −0.123, SE = 0.031, t = 23.99, FDR p < .001). This suggests that the negative association between physical control and hyperactivity symptoms is more pronounced in children with LGBTQIA+parents than in those with non-LGBTQIA+parents, indicating a stronger effect for the LGBTQIA+group. In addition, a significant interaction was found between LGBTQIA+status and Supportiveness on hyperactivity (B = 0.19, SE = 0.06, t = 3.17, FDR p = .006). Increased supportiveness is associated with decreased hyperactivity symptoms more strongly for children with LGBTQIA+ parents than for those with non-LGBTQIA+parents. These interactions are illustrated in Figure 1 (bottom left for Physical Control and bottom right for Supportiveness), where the slopes for the LGBTQIA+ group are steeper, showing a stronger association.
For Peer Problems, two significant interactions emerged. The first interaction was between LGBTQIA+ status and Physical Control, which showed a significant negative association (B = −0.07, SE = 0.02, t = −2.94, FDR p = .011). This finding suggests that the negative effect of physical control on peer relationships is worse in children with LGBTQIA+ parents. Furthermore, a significant interaction was found between LGBTQIA+ parental status and Supportiveness (B = 0.14, SE = 0.05, t = 3.16, FDR p = .006). This interaction indicates that supportiveness is more negatively associated with peer problems for children with LGBTQIA+ parents than their non-LGBTQIA+ counterparts. Figure 2 (left panel for Physical Control and middle panel for Supportiveness) shows these interactions, with steeper slopes for the LGBTQIA+ group, indicating stronger associations.
In terms of overall child psychopathology, two significant interactions emerged. The interaction between LGBTQIA+ status and Physical Control was found to be significantly negative (B = −0.32, SE = 0.09, t = −3.64, FDR p = .001). This suggests that for children with LGBTQIA+ parents, the negative effect of physical control on psychopathology symptoms is higher. Furthermore, our results showed a significant positive interaction between LGBTQIA+ status and Supportiveness on overall child psychopathology (B = 0.56, SE = 0.17, t = 3.22, FDR p = .005). This interaction indicates that supportiveness from parents has a more significant mitigating effect on psychopathology in children with LGBTQIA+ parents than in those with non-LGBTQIA+ parents. These interactions are illustrated in Figure 2 (middle panel for Physical Control and right panel for Supportiveness), with steeper slopes for the LGBTQIA+ group, indicating a stronger relationship.
Although not pre-registered, we conducted identical interaction analyses on factor scores based on the partial scalar invariance model, which is presented in Supplemental Table 1. Findings were identical for Conduct Problems and Prosocial Behavior, which both showed no significant interactions with LGBTQIA+ status. For all other outcomes, differences emerged in terms of which interactions with LGBTQIA+ status were statistically significant after FDR adjustment. However, Supportiveness showed significant interactions with LGBTQIA+ status across both observed scores and adjusted factor scores.
Discussion
The focus of parenting research has traditionally been on identifying the best ways to support children, leading to ongoing questions about the definition of positive and negative parenting. The MAPS, with multiple dimensions of positive and negative parenting, has facilitated the exploration of this question. However, key questions of these lines of research remain, including what is universally applicable to parenting across diverse family structures, and how to measure parenting with-out pathologizing parenting in groups experiencing marginalization (Rodriguez et al., 2024; Rodriguez, La Barrie, et al., 2023). Given the limited number of psychometric studies in LGBTQIA+ parents, our study presents an examination of the psychometric properties of the MAPS in LGBTQIA+ parents (Rodriguez et al., 2024). We examine the psychometric properties of MAPS, specifically its measurement invariance, reliability, and differential predictive validity across LGBTQIA+ and non-LGBTQIA+ parents.
We tested measurement invariance between LGBTQIA+ and non-LGBTQIA+ parents to examine whether MAPS consistently measured parenting across LGBTQIA+ and non-LGBTQIA+ parents. The configural invariance model demonstrated that the factor structure of MAPS remained consistent across both groups. The metric invariance model’s good fit reinforced this finding, suggesting a similar pattern of factor loadings across LGBTQIA+ and non-LGBTQIA+ parents. This finding suggests that MAPS factors resonate similarly across LGBTQIA+ and non-LGBTQIA+ parents. However, scalar invariance was not supported, suggesting differences in item intercepts between the LGBTQIA+ and non-LGBTQIA+ groups, specifically for Item 25. In fact, further tests suggested that allowing the intercept of Item 25 to vary was sufficient to achieve partial scalar invariance. These findings suggest that while most items function equivalently across groups, the intercept of Item 25 (“I spank my child when I am angry”) differs, reflecting potential group-specific interpretations or responses to this item. While this decision was made based on modification indices, previous psychometric studies of the MAPS have pointed to invariance issues related to items assessing physical control (Rodriguez, Cadet, et al., 2023). The successful establishment of partial scalar invariance is promising, as studies note scalar invariance is often unattainable (Muthén & Asparouhov, 2018), particularly considering the vast sociodemographic differences between the samples. These findings are also more promising than those related to findings previously reported by race and ethnicity, for example, with the MAPS and other parenting scales, like the Alabama Parenting Questionnaire (Rodriguez, La Barrie, et al., 2023; Shaffer et al., 2022).
Overall, these findings on measurement invariance indicate that while the MAPS generally functions similarly across LGBTQIA+ and non-LGBTQIA+ parents, certain items, such as those related to Physical Control, may be interpreted differently by these groups. Specifically, the intercept difference for Item 25 suggests that LGBTQIA+ parents might interpret or respond to this item differently compared to non-LGBTQIA+ parents. This does not necessarily mean the entire scale is unreliable for one group over another, but it highlights the need for cautious interpretation of comparisons made by Physical Control. For example, when conducting comparisons, it may be prudent to test for testing measurement invariance before conclusions can be made, as psychometric properties are sample-dependent. For researchers and clinicians, it is crucial to consider these variations to avoid potential biases in assessments and to ensure accurate representations of parenting across diverse populations. In such cases, it is advisable to use the MAPS alongside other measures or to incorporate qualitative assessments to ensure a comprehensive understanding of parenting behaviors in LGBTQIA+ families.
Furthermore, our analysis revealed strong internal consistency across most MAPS subscales for both groups. The most notable (albeit small) difference was observed in the “Supportiveness” subscale, on which LGBTQIA+ parents exhibited a higher alpha coefficient than their non-LGBTQIA+ counterparts. As such, the reliability of the MAPS was supported in both groups. Our investigation into differential predictive validity using multiple regression analyses provided another layer about the psychometric properties of the MAPS across groups. Seven significant interaction terms emerged after FDR corrections, notably in terms of physical control and supportiveness, in relation to the Emotional Symptoms, Hyperactivity, Peer Problems, and Overall Child Psychopathology scores of the SDQ. A particularly compelling finding was the interaction between supportiveness and LGBTQIA+ status, indicating a moderating effect of LGBTQIA+ status on the association between supportiveness and emotional symptoms, which was consistent with both sum and factor scores. Moreover, the interactions involving physical control and LGBTQIA+ status suggested a stronger negative association with hyperactivity and peer problems among children of LGBTQIA+ parents. These findings imply that parenting may have different implications for the mental health of children in LGBTQIA+ families. We must also acknowledge that while our findings contribute valuable insights into LGBTQIA+ parenting, they should be interpreted with caution, considering the potential influence of uncontrolled confounding variables and the unique LGBTQIA+ parenting context, which remains understudied. We advocate for further empirical exploration and replication of these findings, emphasizing the need for longitudinal studies to better understand parenting over time and across diverse LGBTQIA+ family structures.
Given the differences in how physical control and supportiveness interact with child outcomes between LGBTQIA+ and non-LGBTQIA+ parents, it is important to recognize that the MAPS may need to be adapted or interpreted differently for these groups. Specifically, the partial scalar invariance found in this study highlights the need for careful application of the MAPS, particularly Physical Control, in LGBTQIA+ parents. Researchers and clinicians should consider additional qualitative data or complementary measures to fully capture the parenting dynamics in LGBTQIA+ families. These adaptations could enhance the tool’s utility and ensure that it performs equitably across groups.
Future Research
Despite the promising findings of the strong psychometric properties of the MAPS in LGBTQIA+ parents as well as how parenting research can point to the need to support LGBTQIA+ parents, we must note that further studies are needed separating sexual (e.g., gay, lesbian, bisexual) and gender (e.g., transgender, intersex, asexual) minorities, as LGBTQIA+ parents are not a monolith. However, our study was underpowered to assess such potential subgroup differences and merely provides a foundation and call for future psychometric research to adequately support this population and for consideration of LGBTQIA+ parents in parenting research. Therefore, our results should be interpreted as a preliminary step to optimizing the measurement of parenting more carefully in LGBTQIA+ parents as well as constructs that commonly related to parenting behavior. For example, previous studies have suggested using qualitative interviews, such as cognitive interviews, to more adequately examine potential sources of bias and interpretation, as well as complement quantitative methods (Rodriguez, La Barrie, et al., 2023).
Limitations
The findings of this study should be considered in light of certain limitations. First, although we used data from two large racially and ethnically diverse national samples, there was an imbalance in the subgroup sizes in terms of the number of LGBTQIA+ and non-LGBTQIA+ parents. In addition, also due to sample size limitations, we did not incorporate other demographic variables that might influence the measurement invariance of parent-reported measures like the MAPS. However, previous studies have examined the measurement invariance of the MAPS by racial and ethnic identity, parent and child gender, and developmental stages, which all have generally found negligible effects on its psychometrics (Loiselle et al., 2021; Parent & Forehand, 2017; Rodriguez, Cadet, et al., 2023). Future research, particularly those with large sample sizes might delve into the intersectionality of various aspects of parental identity, as well as explore the potential impacts of societal stressors and minority stress over time. In addition, our study was cross-sectional and single-informant in nature. As such, multi-informant (e.g., child, co-parent) and longitudinal investigations may provide additional insights, including enabling tests of longitudinal invariance to the assessment of timing effects. Future studies might also consider examining the psychometric properties of the MAPS by child characteristics other than those that have already been examined, including the child’s sex and age (Loiselle et al., 2021; Parent & Forehand, 2017).
Conclusion
Our study highlights the nuanced complexities of parenting across diverse groups, especially within the LGBTQIA+ community. The differential interactions observed in parenting and their potential effects on child psychopathology may suggest that a tailored approach to parenting interventions and supports may be needed. As we move forward, recognizing and integrating these diverse parenting experiences will be pivotal in developing inclusive and effective parenting support frameworks, ultimately contributing to healthier family dynamics and child outcomes. While the MAPS demonstrates robust psychometric properties across LGBTQIA+and non-LGBTQIA+parents, the observed differences in item intercepts and interaction effects underscore the importance of careful application and interpretation. Researchers and clinicians should be aware of these nuances to avoid potential misinterpretations and to provide more accurate assessments of parenting.
Supplementary Material
Supplemental Material
Supplemental material for this article is available online.
Transparency Statement.
Our study was preregistered on the Open Science Framework. All methods follow our registered report except where noted. Data and R analysis code are also available on the Open Science Framework https://osf.io/q7kr5/?view_only=a9cbdb96cb0c45bf9d2e4b77eefbb7e8. A view only link is provided for review.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: One of the samples included in this study was partially collected with support from a grant from the National Institute of Mental Health (R36MH127838). Violeta J. Rodriguez’s work was supported by a DP5 Early Independence Grant from the National Institutes of Health Office of the Director DP5OD03650801.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References
- Amodeo AL, Esposito C, Bochicchio V, Valerio P, Vitelli R, Bacchini D, & Scandurra C (2018). Parenting desire and minority stress in lesbians and gay men: A mediation framework. International Journal of Environmental Research and Public Health, 15(10), 2318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brainer A, Moore MR, & Banerjee P (2020). Race and ethnicity in the lives of LGBTQ parents and their children: Perspectives from and beyond North America. In Goldberg AE & Allen KR (Eds.), LGBTQ-parent families: Innovations in research and implications for practice (pp. 85–103). Springer. [Google Scholar]
- Brown TA (2014). Confirmatory factor analysis for applied research. The Guilford Press. [Google Scholar]
- Chen Y, Haines J, Charlton BM, & VanderWeele TJ (2019). Positive parenting improves multiple aspects of health and well-being in young adulthood. Nature Human Behaviour, 3(7), 684–691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cohen JA, Mannarino AP, & Deblinger E (2012). Trauma-focused CBT for children and adolescents: Treatment applications. The Guilford Press. [Google Scholar]
- Gates GJ (2013). LGBT parenting in the United States. Williams Institute UCLA. [Google Scholar]
- Goldberg AE (2023). LGBTQ-parent families: Diversity, intersectionality, and social context. Current Opinion in Psychology, 49, 101517. 10.1016/j.copsyc.2022.101517 [DOI] [PubMed] [Google Scholar]
- Goldberg AE, Gartrell NK, & Gates GJ (2014). Research report on LGB-parent families. [DOI] [PubMed] [Google Scholar]
- Goodman R (1997). The strengths and difficulties questionnaire: A research note. Journal of child Psychology and Psychiatry, 38(5), 581–586. [DOI] [PubMed] [Google Scholar]
- Herschell AD, & McNeil CB (2005). Theoretical and empirical underpinnings of parent-child interaction therapy with child physical abuse populations. Education and Treatment of Children, 28(2), 142–162. [Google Scholar]
- Hurley KD, Huscroft-D’Angelo J, Trout A, Griffith A, & Epstein M (2014). Assessing parenting skills and attitudes: A review of the psychometrics of parenting measures. Journal of Child and Family Studies, 23(5), 812–823. [Google Scholar]
- Kline RB (2015). Principles and practice of structural equation modeling. The Guilford Press. [Google Scholar]
- Linn RL (1978). Single-group validity, differential validity, and differential prediction. Journal of Applied Psychology, 63(4), 507–512. [Google Scholar]
- Locke LM, & Prinz RJ (2002). Measurement of parental discipline and nurturance. Clinical Psychology Review, 22(6), 895–929. [DOI] [PubMed] [Google Scholar]
- Loiselle R, Parent J, Georgeson AR, Thissen D, Jones DJ, & Forehand R (2021). Validation of the Multidimensional Assessment of Parenting: An application of item response theory. Psychological Assessment, 33(9), 803–815. [DOI] [PubMed] [Google Scholar]
- Meyer IH (1995). Minority stress and mental health in gay men. Journal of Health and Social Behavior, 36(1), 38–56. [PubMed] [Google Scholar]
- Meyer IH (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129(5), 674–697. 10.1037/0033-2909.129.5.674 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meyer IH (2007). Prejudice and discrimination as social stressors. In Meyer IH & Northridge ME (Eds.), The health of sexual minorities: Public health perspectives on lesbian, gay, bisexual and transgender populations (pp. 242–267). Springer. [Google Scholar]
- Moore MR, & Brainer A (2013). Race and ethnicity in the lives of sexual minority parents and their children. In Goldberg AE & Allen KR (Eds.), LGBT-parent families: Innovations in research and implications for practice (pp. 133–148). Springer. 10.1007/978-1-4614-4556-2_9 [DOI] [Google Scholar]
- Muthén B, & Asparouhov T (2018). Recent methods for the study of measurement invariance with many groups: Alignment and random effects. Sociological Methods & Research, 47(4), 637–664. [Google Scholar]
- Parent J, & Forehand R (2017). The Multidimensional Assessment of Parenting Scale (MAPS): Development and psychometric properties. Journal of Child and Family Studies, 26(8), 2136–2151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rodriguez VJ, Cadet GD, Sisitsky M, Cooley C, Acosta J, Coles E, Charity-Parker B, Walters A, Shaffer A, & Parent J (2023). Assessing parenting in racially and ethnically diverse families: A lack of measurement equivalence. Journal of Family Psychology, 37(6), 753–762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rodriguez VJ, LaBarrie DL, Adams S, & Liu Q (2024). Parenting measures and their psychometrics in LGBTQIA+ families: A systematic review. Child Psychiatry & Human Development, 1–12. [DOI] [PubMed] [Google Scholar]
- Rodriguez VJ, La Barrie DL, Zegarac MC, & Shaffer A (2023). A Systematic Review of Parenting Scales measurement invariance/ equivalence of by race and ethnicity: Recommendations for inclusive parenting research. Assessment, 30(1), 22–36. 10.1177/10731911211038630 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rose J, Roman N, Mwaba K, & Ismail K (2018). The relationship between parenting and internalizing behaviours of children: A systematic review. Early Child Development and Care, 188(10), 1468–1486. [Google Scholar]
- Shaffer A, Rodriguez VJ, Kolko DJ, Pilkonis PA, & Lindhiem O (2022). Measurement invariance in the Alabama Parenting Questionnaire: Implications for culturally inclusive parenting research. Journal of Family Issues, 43(8), 2040–2062. [Google Scholar]
- Van Zyl LE, & Ten Klooster PM (2022). Exploratory structural equation modeling: Practical guidelines and tutorial with a convenient online tool for Mplus. Frontiers in Psychiatry, 12, 795672. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.