Measuring Perceived Discrimination and Its Consequences for Latino Health

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study makes a robust contribution by addressing measurement of perceived discrimination among Latino subgroups using NLAAS data. Below are refined suggestions to enhance clarity, rigor, and presentation:  

1. **Clarify Depression Measure Specification**: Explicitly state that the 10 categorical indicators for the depression latent factor (e.g., reduced appetite, guilt) correspond to a validated scale (e.g., adapted PHQ-9) to ensure methodological transparency.  

2. **Justify Chronic Conditions Dichotomization**: Acknowledge in the methods that dichotomizing chronic conditions (vs. using a count variable) may obscure severity/quantity but was necessitated by sample size constraints for subgroup analyses. Briefly discuss trade-offs (e.g., loss of granularity) to contextualize the decision.  

3. **Enhance Table 5 Transition**: Introduce Table 5 with a brief rationale (e.g., "To address collinearity between subscales and isolate individual effects, we tested each dimension separately") to ensure logical flow from prior multivariate models.  

4. **Deepen Discussion of Cuban-Depression Null Finding**: Speculate on potential explanations for the lack of discrimination-depression association among Cubans, such as their lower reported discrimination levels, stronger ethnic identity (Table 1), or historical/migration context (e.g., enclave protection), citing relevant literature on Latino health disparities.  

5. **Correct Typographical and Terminological Issues**:  
   - Standardize terminology (e.g., "Blatant" instead of "Bidantt" in tables and text).  
   - Fix typos (e.g., "strong-est" to "strongest" in Table 3 footnotes) to maintain professionalism.  

6. **Format Consistency with Journal Guidelines**:  
   - Convert tables to **three-line formats** (horizontal lines for headers, separators, and footers) to align with academic standards.  
   - Review and unify **reference formatting** (e.g., author names, publication dates, journal abbreviations) to meet the target journal’s style requirements, ensuring consistency in citations like doi formatting and capitalization.  

These revisions will strengthen the manuscript’s methodological clarity, interpretive depth, and adherence to scholarly publishing norms.

Author Response

Reviewer 1:

Open Review

**This study makes a robust contribution by addressing measurement of perceived discrimination among Latino subgroups using NLAAS data. Below are refined suggestions to enhance clarity, rigor, and presentation:  

Thank you for the insights and recommendations. The paper is much stronger as we were able to address your recommendations.

1. **Clarify Depression Measure Specification**: Explicitly state that the 10 categorical indicators for the depression latent factor (e.g., reduced appetite, guilt) correspond to a validated scale (e.g., adapted PHQ-9) to ensure methodological transparency.  

We identify the source of the depression scale and provide citations to its validation. Specifically, we wrote:

Depression is measured with ten questions from Radolph’s (1977) Center for Epidemiological Studies Depression Scale (CES-D), which is one of the most widely used and validated screening scales of depression (Blodgett et al. 2021).  We also cite Radloff and their work on the original scale: The CES-D Scale. A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement. 1977; 1:385-401.We explicitly state the 10 categorical indicators and discuss how to analyze these with IRT.

 

2. **Justify Chronic Conditions Dichotomization**: Acknowledge in the methods that dichotomizing chronic conditions (vs. using a count variable) may obscure severity/quantity but was necessitated by sample size constraints for subgroup analyses. Briefly discuss trade-offs (e.g., loss of granularity) to contextualize the decision.  

Thank you for this suggestion. We indicated and acknowledged this in the method’s section. We added the following:

We recognize that dichotomization reduces granularity and may obscure variation in severity or number of conditions. However, this approach was necessary to ensure adequate cell sizes for subgroup analyses across Latino ethnic groups. Given the relatively small sample sizes in certain subgroups, a count variable would have produced sparse cells. Our use of dichotomization thus represents a trade-off: it facilitates more reliable comparisons across groups but at the cost of some precision. Future studies with larger samples should examine discrimination’s effects on both the presence and number of chronic conditions to capture the full range of health impacts.

In addition, we provided several citations on the benefits and pitfalls of this coding scheme.

3. **Enhance Table 5 Transition**: Introduce Table 5 with a brief rationale (e.g., "To address collinearity between subscales and isolate individual effects, we tested each dimension separately") to ensure logical flow from prior multivariate models.  

This is a great suggestion, we include your recommended text and added: In Table 5, we test the effects of each dimension of discrimination separately. When two independent variables are correlated, they can be entered in the equation separately to address collinearity between subscales and isolate individual scale effects.

4. **Deepen Discussion of Cuban-Depression Null Finding**: Speculate on potential explanations for the lack of discrimination-depression association among Cubans, such as their lower reported discrimination levels, stronger ethnic identity (Table 1), or historical/migration context (e.g., enclave protection), citing relevant literature on Latino health disparities.  

We reference Table 1 when discussing the Cuban null findings and added our own insights based on the history of this group.

At the end of the of the article, in the discussion, we also  added the following text to inform this Null finding: The absence of a significant relationship between the second-order discrimination scale and depression may stem from several factors. As shown in Table 1, Cubans reported the lowest levels of discrimination relative to other Latino groups, the highest levels of ethnic identity, and were more likely to be first-generation and/or foreign-born. Each of these has been documented as a protective factor. This null finding is consistent with the literature on the Hispanic Paradox, which shows that more recent and first-generation immigrants often have better health outcomes than later generations despite lower socioeconomic status (Fuentes-Mayorga and Burgos 2017). It also aligns with research demonstrating that a strong ethnic identity can buffer the effects of stressors among Latinos (Burgos 2025; Sassa et al. 2023). In addition, Cubans benefit, relative to other Latino groups, from other protective resources, including higher levels of education, greater household income, lower poverty rates, and strong social support within ethnic communities, as discussed above.

 

 

5. **Correct Typographical and Terminological Issues**:  
    - Standardize terminology (e.g., "Blatant" instead of "Bidantt" in tables and text).  
    - Fix typos (e.g., "strong-est" to "strongest" in Table 3 footnotes) to maintain professionalism.  

We carefully fixed the spelling and grammar throughout the document and proofread the manuscript several times.  It should be clear of these errors.

 

6. **Format Consistency with Journal Guidelines**:  
    - Convert tables to **three-line formats** (horizontal lines for headers, separators, and footers) to align with academic standards.  
    - Review and unify **reference formatting** (e.g., author names, publication dates, journal abbreviations) to meet the target journal’s style requirements, ensuring consistency in citations like doi formatting and capitalization.  

These revisions will strengthen the manuscript’s methodological clarity, interpretive depth, and adherence to scholarly publishing norms.

Thank you for the suggestion. We entered each reference into a bibliograph program. We also provide hyperlinked DOI’s and checked that the DOI’s are correct and the URL’s point to the correct citation. We added the lines to the tables. The citation style is consistent throughout in the Journal’s expected format.

Reviewer 2 Report

Comments and Suggestions for Authors

The substance of the article is very important and timely.

The corrections/updates/amendments now made by the authors have made the article appealing.

Point 2 (Materials and Methods - 490) is well defined, as is the statistical analysis carried out, as well as the results obtained.

Comments on the Quality of English Language

Not being from a country whose mother tongue is English, I believe the article could be revised in terms of its language.

Author Response

REVIEWER TWO:

Comments and Suggestions for Authors

**The substance of the article is very important and timely.

**The corrections/updates/amendments now made by the authors have made the article appealing.

**Point 2 (Materials and Methods - 490) is well defined, as is the statistical analysis carried out, as well as the results obtained.

 

**Comments on the Quality of English Language

**Not being from a country whose mother tongue is English, I believe the article could be revised in terms of its language.

We are please that you view the article as very important and timely. We reworked the introduction and conclusion to speak to the larger issues that are affecting this population.

Importantly, we fix all grammatical and spelling issues and worked on the flow of the transitions.

Reviewer 3 Report

Comments and Suggestions for Authors

The study is scientifically robust: it explores a representative and disaggregated national sample (the use of NLAAS data, a nationally representative sample with oversampling of specific groups, is a crucial strength), allowing for the examination of heterogeneity within the selected pan-ethnic population. The methodological approach is sophisticated and appropriate to the study’s objectives. The study reveals particular care in validating the scale structure in subgroups, showing how different ways of operationalising discrimination affect other dimensions of health. 
The introduction to the study should include a brief note characterising the situation of these communities, referring to updated bibliography, with particular attention to the complex social geography that the sample implies. The inclusion of this contextualisation strategy could increase the impact of the study on the scientific community, contributing, in particular, to reinforcing the "comprehensive" dimension of the data analysed. This contextualisation could strengthen the possibilities for sociological discussion of the results of the analysis.

Author Response

REVIEWER THREE

 

Top of Form

**The study is scientifically robust: it explores a representative and disaggregated national sample (the use of NLAAS data, a nationally representative sample with oversampling of specific groups, is a crucial strength), allowing for the examination of heterogeneity within the selected pan-ethnic population. The methodological approach is sophisticated and appropriate to the study’s objectives. The study reveals particular care in validating the scale structure in subgroups, showing how different ways of operationalizing discrimination affect other dimensions of health. 

Thank you. Great care, time, and much work went into this study. It’s encouraging to hear this comment and acknowledgment.

**The introduction to the study should include a brief note characterizing the situation of these communities, referring to updated bibliography, with particular attention to the complex social geography that the sample implies. The inclusion of this contextualization strategy could increase the impact of the study on the scientific community, contributing, in particular, to reinforcing the "comprehensive" dimension of the data analyzed. This contextualization could strengthen the possibilities for sociological discussion of the results of the analysis.

 

This insight really made the paper much stronger, and we are very appreciative. We really homed in on the complex social geography of the ethnic groups and included a brief history to highlight that the ethnic groups have different experiences. In the introduction, we slightly reframe the paper so that it speaks to the developing political situation of Latinos. In the background section, we added a few paragraphs on how these groups are different. Importantly, we theorize the importance of distinguishing subtle from blatant discrimination and provide several theoretical propositions of why it is important to distinguish these types of discrimination. In the conclusion, we bring the study full circle  with a thoughtful discussion of how this study is informed by what is happening politically in the U.S. and what the findings means for the understanding of racial and ethnic inequalities in genera. We fully contextualize the sociological discussion and we hope that you find it much more engaging and pertinent.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

1.**Address Data Timeliness and Contextual Relevance** The study relies on 2002–2003 NLAAS data, which is over two decades old. Significant societal shifts (e.g., heightened anti-immigrant policies, the COVID-19 pandemic, and evolving discourse on structural racism) have likely altered Latino populations’ experiences of discrimination and health outcomes. It would strengthen the paper to explicitly discuss how this temporal gap might limit the generalizability of findings, and to suggest future research using more recent nationally representative data (e.g., updated HCHS/SOL waves or post-2020 surveys) to validate the EDS factor structure and discrimination-health relationships in contemporary contexts.   2. **Deepen Subgroup Analyses of Intersectional Factors** While the study disaggregates Latino groups (Mexicans, Puerto Ricans, Cubans, etc.), it could further explore how intersectional variables (e.g., skin tone, immigrant generation, or language proficiency) moderate the link between discrimination (subtle vs. blatant) and health within each subgroup. For example, the paper notes that darker skin tone correlates with more discrimination, but it does not unpack whether this dynamic differs between Mexican Americans (with higher Indigenous ancestry) and Puerto Ricans (with more racial diversity). Adding such analyses would clarify why blatant discrimination impacts Puerto Ricans’ chronic health more strongly than Mexicans’, enhancing the study’s theoretical and practical insights.   3. **Validate EDS Measurement Invariance Across Language Groups** The NLAAS includes both English and Spanish interviews, and the paper acknowledges language proficiency affects discrimination reports. However, it does not test whether the EDS (used in English and Spanish) demonstrates measurement invariance across language groups—critical for ensuring that “subtle/blatant discrimination” is interpreted consistently by respondents of different language backgrounds. Supplementing the CFA with tests of scalar invariance for English vs. Spanish EDS versions would confirm that factor loadings and item thresholds are equivalent, strengthening the reliability of cross-ethnic comparisons and reducing potential bias from language-related measurement differences.

Author Response

See Attached File

Author Response File: Author Response.pdf