Multigroup Invariance Tests Research Articles

Examining the Factor Structure and Validity of the Depression Anxiety Stress Scale-21.

The prevalence of mental health disorders calls for valid and reliable instruments that are easy to administer and assess for clinicians and researchers. The Depression Anxiety Stress Scale-21 (DASS-21) is a commonly used instrument to assess psychological distress; however, model fit and internal reliability issues have been reported. Our objective was to assess the factorial and structural validity of the DASS-21. A confirmatory factor analysis (CFA) was conducted on the full sample (n = 1036) to assess the proposed three-factor DASS-21 using a priori cut-off values. Because model fit indices were not met, an exploratory factor analysis (EFA) was conducted to identify a parsimonious model. The resulting three-factor structure (i.e., DASS-9) was then assessed using CFA and multigroup invariance testing procedures. The proposed three-factor DASS-21 did not meet model fit criteria. The DASS-9 did meet recommended model fit criteria and was invariant between sex, injury status, mental health diagnosis, and activity level groups. Statistically different group means were found between mental health diagnosis and activity level groups, while no differences between sex or injury status groups were found. The current study provides support to use a condensed DASS-21 instrument, such as the DASS-9. Future research is necessary to establish the DASS-9 prior to its adoption in research and clinical practice. Additionally, there is a need to identify and review all condensed versions of the DASS-21, so individuals know which instrument can be used for clinical or research purposes.

Need-support facilitates well-being across cultural, economic, and political contexts: A self-determination theory perspective

BackgroundSelf-determination theory (SDT) posits that teachers who engage in need-supportive teaching through satisfying students’ basic needs for autonomy, competence, and relatedness facilitate optimal well-being. However, there are debates about the purported applicability and relevance of need-supportive teaching across cultural, economic, and political contexts. AimsThis study examined whether need-supportive teaching was associated with students’ subjective, eudaimonic, and cognitive well-being. These relationships were tested across different macro-contexts, including cultural, economic, and political systems. SampleWe drew on data from 535,512 students across 70 countries. These students came from diverse cultural groups (Western Europe, Eastern-Central Europe, Eastern Europe, Latin America, English-speaking, Confucian, Southeast Asia, and Africa and the Middle East), economic systems (high, upper-middle, and lower-middle-income), and political climates (full democracies, flawed democracies, hybrid, and authoritarian regimes). MethodsConfirmatory factor analyses, structural equation modelling, and multi-group invariance tests were conducted. ResultsBy and large, need-supportive teaching was associated with better well-being across cultural, economic, and political contexts. However, the magnitude of associations was somewhat different across macro-contexts. Minor deviations from the general pattern were also found in a few cultural groups. ConclusionsAcross the globe, students who perceived their teachers to engage in need-supportive teaching were also more likely to experience better well-being. The results supported the universalist perspective, which recognizes the existence of broad universal patterns alongside contextual differences.

Truth on ProQOL and Burnout Assessment—Development and Intercultural Validation of the Brief ProQOL-12 from the Professional Quality of Life, Version 5

Objective The prevalence of burnout is becoming a public health issue across the world. This study developed and validated Brief ProQOL-12, formed from the existing, but yet unvalidated 30-item ProQOL-5. Methods Study 1—Eight intercultural samples of helping professionals from four continents (n = 4,129). Validation included step-wise appraisal and multi-group invariance testing. Study 2—Ethnically and occupationally-balanced sample (n = 453). Rasch modeling, factor analysis, analysis of measure correlates, and scale refinements. Results ProQOL-5 (30-item) did not fit among any continential or national samples, including North America. Study 1—Provisional ProQOL-12 showed good internal structure and measurement invariance. Study 2—Brief ProQOL-12 had an excellent fit in the ethnically and occupationally diverse sample. Included ProQOL 5 items were: 9, 10, 12-14, 18, 19, 21, 24-26, and 30. The reliability and validity of the Brief ProQOL-12 were significantly improved over the 30-item measure. Rasch modeling and factor analysis indicated that the measure was reliable, valid interculturally and occupationally. Conclusions The ProQOL-5 (30-item) reliability and validity concerns were resolved in this 12-item measure which was refined to include a more realistic time context of seven days and time definite Likert ranges. These enhancements increased validity, as evidenced by the improvement in all model fit indicators in the excellent range in Study 2. Transcending culture and ethnicity with proven psychometric robustness, the Brief ProQOL-12 serves as a valuable resource for evaluating burnout risk and well-being across heterogeneous ethnic, cultural, and occupational landscapes.

Assessing the Structural Validity of the Knee Injury and Osteoarthritis Outcome Score Scale.

The Knee Injury and Osteoarthritis Outcome Score (KOOS) scale is used to assess patient perspectives on knee health. However, the structural validity of the KOOS has not been sufficiently tested; therefore, our objective was to assess the KOOS in a large, multi-site database of patient responses who were receiving care for knee pathology. A cross-sectional study was conducted using the Surgical Outcome System (SOS) database. A confirmatory factor analysis (CFA) was conducted to assess the proposed five-factor KOOS using a priori cut-off values. Because model fit indices were not met, a subsequent exploratory factor analysis (EFA) was conducted to identify a parsimonious model. The resulting four-factor structure (i.e., KOOS SF-12) was then assessed using CFA and subjected to multigroup invariance testing. The original KOOS model did not meet rigorous CFA fit recommendations. The KOOS SF-12 did meet model fit recommendations and passed all invariance testing between intervention procedure, sex, and age groups. The KOOS failed to meet model fit recommendations. The KOOS SF-12 met model fit recommendations, maintained a multi-factorial structure, and was invariant across all tested groups. The KOOS did not demonstrate sound structural validity. A refined KOOS SF-12 model that met recommended model fit indices and invariance testing criteria was identified. Our findings provide initial support for a multidimensional KOOS structure (i.e., KOOS SF-12) that is a more psychometrically sound instrument for measuring patient-reported knee health.

Psychometric analysis of the Hip Disability and Osteoarthritis Outcome Score Joint Replacement (HOOS-JR)

ObjectiveCenters for Medicare and Medicaid Services (CMS) has requested hospitals collect and report patient-reported outcomes (PROs) beginning in 2024 including the Hip Disability and Osteoarthritis Outcome Score Joint Replacement (HOOS-JR). However, scale structural validity of the HOOS-JR has minimally been assessed. The purpose of this study was to assess internal consistency, structural validity, and multi-group invariance properties of the HOOS-JR in a large sample of patients who underwent a total hip arthroplasty (THA). MethodsA cross-sectional study using the Surgical Outcomes System was retrospectively queried for patients who underwent a THA. Internal consistency was assessed using Cronbach's alpha and McDonald's Omega. A confirmatory factor analysis (CFA) was performed on the HOOS-JR using a priori cut-off values. Multi-group invariance testing was also performed on the sample of patients across sex and age groups. ResultsInternal consistency was acceptable for 6-item (alpha ​= ​0.88; omega ​= ​0.88) and 5-item (alpha ​= ​0.86; omega ​= ​0.86) HOOS-JR. The one-factor, 6-item CFA did not meet the recommended fit indices. The one-factor, 5-item CFA had acceptable fit for the sample data. Invariance testing criteria were met between the age groups; however, scalar invariance was not met for sex. ConclusionThe 6-item HOOS-JR did not meet contemporary model fit indices indicating that scale refinement is warranted. The 5-item met most goodness-of-fit indices and invariance criteria. However, further scale refinement may be warranted as localized fit issues were identified.

The impact of brand trust on consumers' behavior toward agricultural products' regional public brand.

The importance of agricultural products' regional public brands, owned by relevant organizations and jointly used by several agricultural production and operation entities, is increasing in contemporary marketing research. Based on a survey of 544 consumers, this study investigates the influence of brand trust, attitude, subjective norms, and perceived behavioral control on their purchase intention and behavior toward agricultural products' regional public brand. Through SEM, we find that brand trust toward agricultural products' regional public brand positively influences consumer attitude, purchase intention, and purchase behavior. In addition, attitude and purchase intention mediate the relationship between brand trust and purchase behavior. While attitude and perceived behavioral control positively affect purchase intention, no similar effect is found for subjective norms. Moreover, multigroup invariance tests demonstrate that consumer behavior can be influenced by factors such as gender, age, monthly income, marital status, previous visits to the region, and purchase purpose. We therefore recommend that to bolster competitiveness, regional public brand stakeholders maintain agricultural product quality, ensure reliable purchasing and transportation channels, and enhance brand trust.

Validating the Negative Context of Reception Scale for Mainland Puerto Ricans

This study examined the extent to which the Negative Context of Reception Scale generates valid and reliable scores with Puerto Rican Hurricane Maria survivors. A sample of 319 adult Hurricane Maria survivors completed measures of the negative context of reception, discrimination, language stress, depressive and anxiety symptoms, optimism, and life satisfaction. Confirmatory factor analyses (CFAs) indicated that the previously validated one-factor solution fit the data adequately, and multigroup invariance tests indicated that this solution fit largely equivalently across gender, age groups, year of arrival in the United States, and self-reported English proficiency. Negative context of reception was correlated positively with discrimination and language stress, positively with symptoms of depression and anxiety, and negatively with optimism and life satisfaction. It may be important to adjust item order to reduce intercorrelations among some item responses. This study offers an instrument that counselors can use with Hurricane Maria survivor clients.

From racial awakening to collective action: Asian Americans' pathways to activism and benevolent support during COVID-19.

In response to increased anti-Asian discrimination and violence during the COVID-19 pandemic, this study examined pathways from discrimination experiences to own-group collective action in a diverse sample of 689 Asian Americans. Informed by theories of ethnoracial identity, critical consciousness, and collective action and utilizing structural equation modeling, we examined the associations among discrimination, psychological distress, critical awareness and motivation (CAM) to resist racism, and two types of own-group collective action: political activism and benevolent support. Multigroup invariance tests also examined whether these associations differed by ethnic subgroup, immigrant generation, and age. Results supported our integrated model in which distress mediated the relationship between discrimination and CAM, and CAM mediated the relationship between discrimination and collective action. The structural pathways from discrimination to own-group collective action generally did not differ by ethnic subgroup and immigrant generation, although path coefficients for the effect of discrimination on distress did vary by age (p < .01). Further, discrimination did not appear to have the same catalyzing effect on CAM for South and Southeast Asians compared to East Asians. While anti-Asian discrimination was associated with both distress and engagement in collective action during the COVID-19 pandemic, group differences in mediational processes highlight the importance of disaggregating analyses to explore both similarities and differences in Asian Americans' responses to discrimination. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Confirmatory Factor Analysis of the Hip Disability and Osteoarthritis Outcome Score (HOOS) and Associated Sub-scales.

Hip Disability and Osteoarthritis Outcome Score (HOOS), HOOS-Joint Replacement (JR), HOOS Physical Function (PS), and HOOS-12 item scale have been suggested as reliable and valid instruments for assessing hip disability. However, factorial validity, invariance across subgroups, and repeated measures of the scale across different populations have not been well supported in the literature. The primary study objectives were to: (1) assess model fit and psychometric properties of the original 40-item HOOS scale, (2) assess model fit of the HOOS-JR, (3) assess model fit of the HOOS-PS, and (4) assess model fit of the HOOS-12. A secondary objective was to perform multigroup invariance testing across physical activity level and hip pathology subgroups for models that met recommended fit indices. Cross-Sectional Study. Individual confirmatory factor analyses (CFAs) were conducted for the HOOS, HOOS-JR, HOOS-PS, and HOOS-12. Additionally, multigroup invariance testing (i.e., activity level, injury type) was conducted on the HOOS-JR and HOOS-PS. Model fit indices did not meet contemporary recommendations for the HOOS and the HOOS-12. Model fit indices for the HOOS-JR and the HOOS-PS met some, but not all, contemporary recommendations. Invariance criteria was met for the HOOS-JR and HOOS-PS. The scale structure of the HOOS and HOOS-12 were not supported; however, preliminary evidence to support the scale structure of the HOOS-JR and HOOS-PS was found. Clinicians and researchers who utilize the scales should do so with caution due to their limitations and untested properties until further research establishes the full psychometric properties of these scales and recommendations for their continued use.

Multi-group invariance testing of the knee injury osteoarthritis outcome score for joint replacement scale

ObjectiveThe Knee Osteoarthritis Outcome Score for Joint Replacement (KOOS-JR) scale is commonly used to assess patient progress. Scale structural validity has not been completely assessed. The purpose of this study was to assess the internal consistency, structural validity, and multi-group invariance properties of the KOOS-JR in a large sample of patients receiving knee arthroplasty or non-operative care. MethodsA cross-sectional study using the Surgical Outcome System (SOS) database. Patients receiving care for degenerative knee conditions were included in the study. Internal consistency was assessed using Cronbach's alpha and McDonald's Omega. A confirmatory factor analysis was conducted to confirm scale structure of the KOOS-JR using a priori cut-off values (Comparative Fit Index [CFI], Tucker-Lewis Index [TLI], Incremental Fit Index [IFI] ​≥ ​0.95, Root Mean Square Error of Approximation [RMSEA] ​≤ ​0.06 preferred and ≤0.08 acceptable). Multigroup invariance testing was conducted across sex, age, and intervention groups. ResultsInternal consistency was acceptable (alpha ​= ​0.83; omega ​= ​0.83). The unidimensional structure of the KOOS-JR exceeded most contemporary model fit recommendations (CFI ​= ​0.976, TLI ​= ​0.964, IFI ​= ​0.976, RMSEA ​= ​0.067). The KOOS-JR was invariant across groups, allowing for comparison of variances and means between sex, age, and intervention groups. ConclusionThe KOOS-JR met or exceeded most of the recommendations for model fit. The scale can be used to assess differences between males and females, middle and older aged adults, and between baseline measures of patients who received total knee arthroplasty or non-operative care.

Sex differences in conduct and emotional outcomes for young people with hyperactive/inattentive traits and social communication difficulties between 9 and 16 years of age: a growth curve analysis.

The purpose of this paper is to identify the trajectory of conduct and emotional problems for young people within the general population at four time points (between 9 years 7 months and 16 years 6 months), investigate their relationship with hyperactive/inattentive traits and explore the moderating effect of autistic social traits (ASTs). Data from 9305 individuals involved in The Avon Longitudinal Study of Parents and Children (ALSPAC) study were included. Conduct and emotional problems and hyperactive/inattentive traits were measured by the Strengths and Difficulties Questionnaire. ASTs were assessed using the Social Communication Disorder Checklist. Individual trajectories for conduct and emotional problems were identified via growth curve modelling. Hyperactive/inattentive traits were included within the growth curve model as a time-varying covariate to determine their effect on these outcomes. Finally, participants were split into two groups (below and above clinical threshold ASTs Groups) and multi-group invariance testing was conducted on the data to identify the moderating effect of ASTs on the relationship between hyperactive/inattentive traits and outcomes (i.e. conduct and emotional problems). Hyperactive/inattentive traits were associated with higher rates of conduct and emotional problems for both boys and girls. The presence of ASTs moderated these relationships for boys, but not for girls, by increasing the risk of boys with hyperactive/inattentive traits developing greater conduct and emotional problems. These findings underscore the importance of identifying hyperactive/inattentive traits and ASTs in young people and addressing the increased risk of conduct and emotional problems. Research and clinical implications are explored.

Measurement invariance of the Perceived Stress Scale-10 across gender in Korean university students

This study examined the factorial structure and gender invariance of the 10-item Perceived Stress Scale (PSS-10) among 582 Korean university students. Data were collected from 582 university students enrolled in a four-year university in the central region of Korea. Confirmatory factor analysis was used to test four distinct factor models identified in previous studies. Analyses of multigroup invariance were also conducted to determine whether the PSS-10 items’ factorial structure exhibited gender invariance. Results indicated that the PSS-10 was the most adequately represented by a bifactor structure, composed of one general factor reflecting overlap across all 10 items and two separate factors of perceived stress and perceived coping. Multigroup invariance tests indicated that PSS-10 items were interpreted similarly between men and women and that responses could be meaningfully compared across gender. The study confirms that the PSS-10 is a theoretically sound instrument that is feasible for both research and university settings.

A multi-group assessment of safety culture among engineering students in the United Arab Emirates

PurposeThe purpose of this paper is to identify the factor structure of safety culture construct among engineering students at university context and to examine the measurement invariance of this instrument across different socio-demographic groups in a sample of engineering students in the United Arab Emirates (UAE).Design/methodology/approachAn exploratory online questionnaire was completed by 770 undergraduate and postgraduate engineering students across the UAE. Data were analyzed using a diversified multi-group and a robust and sophisticated cross-validation testing strategy. Confirmatory factor analysis (CFA) was used to test factor structures identified in previous studies. Multi-group invariance testing was conducted to determine the extent to which factor structure is comparable across groups (i.e. gender, educational and experiential background).FindingsThree-factor model was preferred for its parsimony. The results showed that the level of safety awareness and attitude is relatively satisfactory, whereas safety behaviour is inadequate. No significant difference was showed in multi-group invariance between demographic groups.Research limitations/implicationsThis research is a cross-sectional study and limited to the views of engineering students (informal group). The study would benefit from both informal and formal groups in assessing safety culture at university for a robust empirical evidence. The research highlights relevant implications for policy and program development, by pointing to the need to promote safety culture and mitigate safety-related accidents among engineering students.Originality/valueThis paper offers insight into benefit of understanding the level of safety culture among engineering students and extend knowledge of informal group involvement in safety-related accidents at university level.

Development and validation of a volunteer nostalgia scale: assessing measurement invariance across cultures and types of volunteering

ABSTRACT Volunteers’ nostalgia, which is a multifaceted concept, has not been fully explored, while the concept of nostalgia is receiving increased attention in many fields for its role and effect on individuals’ behaviour. In this study, the authors applied multigroup invariance tests to develop a volunteer nostalgia scale. This study consists of three phases. In the first phase, 180 participants with previous nostalgic volunteer experience were recruited in Singapore for the pilot study. In the second phase, data were collected from the United States (n = 201) and Singapore (n = 203) for the assessment of psychometric properties and cross-cultural validation of the scale. In the third phase, invariance tests were conducted to compare tourism event volunteers with volunteers for non-tourism sectors (n = 420). Results showed consistency between the data, indicating the robustness of the scale, which can be a useful tool in future research involving nostalgia.

Loss of financial well-being in the COVID-19 pandemic: Does job stability make a difference?

This article aims to assess the loss of financial well-being in the COVID-19 pandemic. The developed theoretical model identifies the impacts of the perception of financial risk and financial anxiety on financial well-being. It also seeks, through a comparative analysis, to assess whether public servants, due to their status of job stability in Brazil, are less likely to have the effects of the pandemic than private employees. A survey was carried out on 1222 Brazilians with structural equation modeling and multi-group invariance tests. The results indicate that lower financial well-being is influenced by the level of financial anxiety and financial risk. Public servants perceive fewer losses in financial well-being, anxiety and risks than other professions. In the pandemic context, where the risks of unemployment and loss of income are increased, job stability works like an insurance, allowing public servants greater financial security and then minor losses of financial well-being. Evidence indicates that in countries where a large percentage of workers have temporary or informal jobs, the challenge of reducing the financial impacts of the pandemic will be great. Interventions to alleviating anxiety and public policies of income transfer and reduction of unemployment are instruments to reduce the loss of financial well-being.

Measuring psychological pain: psychometric analysis of the Orbach and Mikulincer Mental Pain Scale

BackgroundSuicide is a public health concern, with an estimated 1 million individuals dying each year worldwide. Individual psychological pain is believed to be a contributing motivating factor. Therefore, establishing a psychometrically sound tool to adequately measure psychological pain is important. The Orbach and Mikulincer Mental Pain Scale (OMMP) has been proposed; however, previous psychometric analysis on the OMMP has not yielded a consistent scale structure, and the internal consistency of the subscales has not met recommended values. Therefore, the primary purpose of this study was to assess the psychometric properties of the OMMP in a diverse sample.MethodsA confirmatory factor analysis (CFA) on the 9-factor, 44-item OMMP was conducted on the full sample (n = 1151). Because model fit indices were not met, an exploratory factor analysis (EFA) was conducted on a random subset of the data (n = 576) to identify a more parsimonious structure. The EFA structure was then tested in a covariance model in the remaining subset of participants (n = 575). Multigroup invariance testing was subsequently performed to examine psychometric properties of the refined scale.ResultsThe CFA of the original 9-factor, 44-item OMMP did not meet recommended model fit recommendations. The EFA analysis results revealed a 3-factor, 9-item scale (i.e., OMMP-9). The covariance model of the OMMP-9 indicated further refinement was necessary. Multigroup invariance testing conducted on the final 3-factor, 8-item scale (i.e., OMMP-8) across mental health diagnoses, sex, injury status, age, activity level, and athlete classification met all criteria for invariance.ConclusionsThe 9-factor, 44-item OMMP does not meet recommended measurement criteria and should not be recommended for use in research and clinical practice in its current form. The refined OMMP-8 may be a more viable option to use; however, more research should be completed prior to adoption.

The Moderating Roles of Sensation Seeking and Worry among Nature-Based Adventure Tourists.

The adventure tourism subsector continues to be popular today. Both industry and academia define adventure tourism’s scope from either the physical (e.g., outdoor activity and physical activity) or psychological aspects (e.g., thrill seeking and challenges). Recent studies have pointed out that adventure tourism can be interpreted differently across cultures and markets. Still, risk has always been universally accepted as an essential characteristic of adventure tourism. Thus, most empirical research has studied the role of risk as one of the motivations. However, attempts to investigate related elements that are either a response to or a coping mechanism for the presence of risk are scarce. This present study adopted one of the most prominent frameworks in explaining behavioral intentions, the theory of planned behavior, and included involvement and knowledge variables to extend it. Furthermore, the sensation-seeking and worry constructs were tested for their moderating impact on intentions to participate in adventure tours. The results of structural equation modeling and multigroup invariance tests revealed that subjective norms were not a significant predictor of intentions, while both sensation seeking and worry significantly moderated the relationships between the study variables.

Invariance Testing of the Disablement in the Physically Active Scale.

The increased emphasis on implementing evidence-based practice has reinforced the need to more accurately assess patient improvement. Psychometrically sound, patient-reported outcome measures are essential for evaluating patient care. A patient-reported outcome instrument that may be useful for clinicians is the Disablement in the Physically Active Scale (DPAS). Before adopting this scale, however, researchers must evaluate its psychometric properties, particularly across subpopulations. To evaluate the psychometric properties of the DPAS in a large sample using confirmatory factor analysis procedures and assess structural invariance of the scale across sex, age, injury status, and athletic status groups. Observational study. Twenty-two clinical sites. Of 1445 physically active individuals recruited from multiple athletic training clinical sites, data from 1276 were included in the analysis. Respondents were either healthy or experiencing an acute, subacute, or persistent musculoskeletal injury. A confirmatory factor analysis was performed on the full sample, and multigroup invariance testing was conducted to assess differences across sex, age, injury status, and athletic status. Given the poor model fit, alternate model generation was used to identify a more parsimonious factor structure. The DPAS did not meet contemporary fit index recommendations or the criteria to demonstrate structural invariance. We identified an 8-item model that met the model fit recommendations using alternate model generation. The 16-item DPAS did not meet the model fit recommendations and may not be the most parsimonious or reliable measure for assessing disablement and quality of life. Use of the 16-item DPAS across subpopulations of interest is not recommended. More examination involving a true cross-validation sample should be completed on the 8-item DPAS before this scale is adopted in research and practice.

Applicability of belief measures for advertising to sponsorship in sport

PurposeWhile advertising and sponsorship are conceptually different, many studies have used the same measures for both constructs. The assumption is that respondents perceive both domains similarly. The purpose of this study was to test the invariance of the belief measures between the advertising and sponsorship measurement models across different consumer segments and to provide empirical justification for the practice.Design/methodology/approachTwo independent samples were recruited from two different consumer segments: university student consumers (n = 290) and general consumers (n = 324). This study conducted multigroup invariance tests using LISREL 8.80. The measurement and structural invariances were concerned with factor loadings (λ), factor variance and covariance (f) and error variance (θ) metrics.FindingsThe factor patterns of the belief model were generally invariant between the advertising and sponsorship models in both consumer groups. However, the respondents interpreted three items between advertising and sponsorship in different ways: one annoyance/irritation item in the generic consumer group and two falsity/no sense items in the student consumer groups.Originality/valueWhile the invariance test reveals three problematic items, the majority of items seem to be invariant, concluding that the advertising belief scale could be applicable to the sponsorship context.

It's like herding cats: Atheist minority stress, group involvement, and psychological outcomes.

Using minority stress theory with a sample of 522 atheist people from the United States, the present study examined the associations of discrimination, proximal minority stressors (stigma consciousness, internalized antiatheism, outness as atheist), and atheist group involvement with psychological distress and self-esteem. Atheist group involvement was associated positively with outness and self-esteem, and negatively with discrimination. Structural equation modeling indicated that discrimination and stigma consciousness yielded significant positive direct relations with distress, whereas outness yielded a significant negative direct relation with distress. Relatedly, discrimination yielded a significant negative direct relation with self-esteem and outness yielded a significant positive direct relation with self-esteem. There was a significant positive unique indirect relation of antiatheist discrimination with distress via the mediating role of stigma consciousness, but no other proximal variables. Multigroup invariance testing of this model did not yield evidence that the pattern of relations of the minority stressors with mental health outcomes differed significantly between participants who were and who were not involved in an atheist group. Implications of these findings for research, practice, and advocacy are discussed. (PsycINFO Database Record (c) 2019 APA, all rights reserved).