Unmeasured Confounding and Racial or Ethnic Disparities in Disability Identification - January 2021 Article

Oct 15, 2026

Uploaded by dabgoudw

Students identified as Black or Hispanic are often reported as less likely than White students to be identified as having disabilities, though replication has led many to treat this pattern as erroneous. The study applies the E-value method to estimate how likely unmeasured confounding could account for observed associations between race/ethnicity and disability identification. Evidence from three population-based studies indicates unmeasured confounding is unlikely, including scenarios consistent with federal overidentification requirements.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Unmeasured Confounding and Racial or Ethnic Disparities in Disability Identification

The dispute over under- vs. over-identification findings

Importance of assessing unmeasured confounding

Policy implications and risk-ratio monitoring

Briefs Educational Evaluation and Policy Analysis June 2021, Vol. 43, No. 2, pp. 351–361 DOI: 10.3102/0162373721991575 Article reuse guidelines: [sagepub.com/journals-permissions](sagepub.com/journals-permissions)[ ](sagepub.com/journals-permissions)© 2021 AERA. [https://journals.sagepub.com/home/epa](https://journals.sagepub.com/home/epa) Unmeasured Confounding and Racial or Ethnic Disparities in Disability Identification Paul L. Morgan The Pennsylvania State University Students who are Black or Hispanic are now reported to be less likely to be identified as having disabilities than similarly situated students who are White. Although repeatedly replicated, this finding is often characterized as in error. I use a new statistical technique, the E-value, to quantify the likelihood that unmeasured confounding explains observed associations between race or ethnicity and disability identification. Results based on calculations across three population-based studies using extensive statistical controls suggest that unmeasured confounding is an unlikely explanation for the observed associations. Unmeasured confounding that would result in levels of overidentification consistent with federal law and regulation is especially unlikely. Keywords: at-risk students, disparities, educational policy, special education, effect size, observational research, regression analyses, secondary data analysis, statistics StudentS who are Black or Hispanic have been reported to be more likely to be identified as having disabilities while attending U.S. schools including for specific conditions (e.g., Grindalet al., 2019; U.S. Department of Education, 2016a) . Reports of overidentification have resulted in suggestions that the U.S. special education system may be racially biased (Blanchett, 2006; Codrington & Fairchild, 2012; Grindalet al., 2019) . Federal law and regulation require monitoring for significant disproportionality in disability identification based on race or ethnicity (the Individuals with Disabilities Education Act [IDEA], 2004, §300.646) based on “clear evidence that overrepresentation on the basis of race and ethnicity continues to exist at both the national and local levels” (U.S. Department of Education Equity in IDEA Rule, 2016, p. 10977) . Yet other empirical work is now reporting that students who are Black or Hispanic are less likely to be identified as having disabilities than similarly situated students who are White (e.g., Morgan et al., 2015), suggestive of differential treatment on the basis of race and ethnicity (U.S. Department of Education, 2016a) . Although repeatedly replicated, findings of under-identification have been dismissed as “in error”(Skiba et al., 2016, p. 221; for a reply, see Morgan & Farkas, 2016) and resulting from “simplistic investigations that overreach both their data set and their own analyses”(Skiba et al., 2016, p. 224) . Dismissals of findings of under-identification have been justified based on methodologically flawed data collection procedures, sampling, and statistical analysis (Grindal et al., 2019; Skiba et al., 2016; Whitford & Carrero, 2019) . Replication studies continue to find evidence of under-identification including in analyses of other nationally representative samples (Morgan et al., 2017) as well as both student- and district-level data sets that indicate overidentification in unadjusted analyses, helping to address criticisms about data collection and sampling (Farkas et al., 2020; Morgan et al., 2017) . The replication work has also statistically controlled for a wide range of student- , family- , and school-level confounds including family socioeconomic status (SES) and student achievement (Morgan et al., 2017) . Student achievement is considered an especially strong confound of disability identification disparities (Donovan & Cross, 2002) . The Importance of Assessing for Unmeasured Confounding in Studies Reporting Disability Under-Identification To d

FAQs

What issue does the article address in disability identification?

It addresses whether observed differences in disability identification by race or ethnicity (Black/Hispanic vs. White) could be explained by unmeasured confounding rather than true disparities.

How does the article evaluate the likelihood of unmeasured confounding?

It uses the E-value approach and computes results across three population-based studies with extensive statistical controls.

Why is unmeasured confounding especially important for policy monitoring?

Federal law and new regulations require schools and states to monitor racial and ethnic disproportionality, using risk-ratio thresholds, so misattribution could affect compliance decisions.

Oct 15, 2026

Uploaded by dabgoudw