UNIVERSITY PARK, Pa. -- U.S. school districts may be flagged as over-identifying students of color as having disabilities when other factors, such as achievement gaps, may explain these disparities, according to new Penn State research published in Exceptional Children.
Federal legislation and regulations require U.S. school districts to monitor whether students of color are overrepresented in special education. School districts reporting overrepresentation that exceeds pre-specified risk ratio thresholds are required to take corrective action including reallocating up to 15 percent of their federal funding for special education. Earlier work led to the belief that U.S. school districts were systemically over-identifying students as having disabilities based on their race or ethnicity.
According to Paul Morgan, Harry and Marion Eberly Fellow, professor of education and demography, and director of Penn State's Center for Educational Disparities, new work is now consistently finding that students of color are less likely, on average, to be identified as having disabilities than similarly situated students who are white.
"We sought to extend this knowledge base by expanding our analysis to include black-to-white and Hispanic-to-white student achievement gaps, as well as other variables measured at the district level for several thousand districts nationwide, as alternative explanatory factors for the reported risk ratios," said Morgan.
The research team used data sets from the U.S. Department of Education, the Civil Rights Data Collection survey on student enrollment, services, and outcomes, and the new Stanford Education Data archive that includes school district-level mathematics and reading achievement scores.
They analyzed about 2,000 districts with black students and white students, and about 2,500 districts with Hispanic students and white students. "In many U.S. school districts, the lack of racial diversity in student enrollment makes it difficult to conduct this type of analysis," said Morgan.
After merging the data sets, the researchers found that the larger a district's black or Hispanic-to-white achievement gap, the larger a district's black or Hispanic-to-white risk ratio. Statistical control for district-level achievement gaps indicated that, on average, black and Hispanic students were less likely to be receiving special education than white students.
"This provides further evidence in the debate regarding whether non-white students are overrepresented in special education due to systemic bias," Morgan said. "While inappropriate over-identification may sometimes be occurring, our analyses of district-level data suggests that the greater representation of black or Hispanic students in special education is largely explained by black or Hispanic students being more likely to experiencing academic difficulties. U.S. school districts may be attempting to provide students experiencing academic difficulties with greater academic supports, including through special education. "
The researchers were able to demonstrate a method that can be used to identify U.S. school districts where significant disproportionality is occurring that is not explained by achievement gaps. This method could be used by state and federal authorities to better monitor for U.S. districts for significant disproportionality based on race or ethnicity.
"Doing so would provide a more methodologically and substantively justifiable method for identifying school districts where significant disproportionality based on race or ethnicity may be occurring, and where further civil rights investigations may be warranted," Morgan said.
Other researchers on the project were George Farkas, distinguished professor of education, University of California, Irvine; and Marianne Hillemeier, professor emerita of health policy and administration; Cynthia Mitchell, researcher analyst and programmer; and Adrienne Woods, postdoctoral research scholar in educational policy, all at Penn State.
Funding was provided by Penn State's Center for Educational Disparities Research, part of the Social Science Research Institute; an infrastructure grant to Penn State's Population Research Institute; and the Eunice Kennedy Shriver National Institute of Child Health and Human Development.