Understanding Differential Growth During School Years and Summers for Students in Special Education

Abstract

Under the Individuals with Disabilities Education Act, schools are required to provide a free appropriate public education to students with disabilities and show that the students are making academic progress. This study compares within- and across-years academic growth from kindergarten to 4^th grade for students who were ever in special education (ever-SPED) to students who were never in special education (never-SPED). We use math and reading assessment scores from NWEA’s Growth Research Database. Our dataset contains one cohort of about 4,200 students, assessed up to three times per year for five years. Although ever-SPED students started kindergarten with lower math and reading test scores and grew less in both subjects than never-SPED students during the kindergarten school year, ever-SPED students grew more than never-SPED students during the 1^st and 2^nd grade school years in math and 1^st, 3^rd, and 4^th grade school years in reading. However, ever-SPED students lost more learning during every summer than never-SPED students. This led the test score disparities between the two to grow from under 0.5 standard deviations in kindergarten to 1.0 standard deviation in 4^th grade. These findings suggest that summer learning opportunities are crucial for improving educational outcomes for students with disabilities.

Acknowledgments

First authorship is shared between the authors, who contributed equally to this work. The authors are grateful for helpful feedback from Meg Guerreiro, Nate Jensen, Andrew McEachin, and Erik Ruzek and for support with data analysis from Art Katsapis.

Disclosure statement

The authors have no conflict of interest to declare.

Statement of human subjects research

Data used in this study were deidentified and requirements for approval by the independent ethnics committee waived.

Notes

1 The 13 categories are: (a) autism; (b) deaf-blindness; (c) deafness; (d) emotional disturbance; (e) hearing impairments; (f) intellectual disability; (g) multiple disabilities; (h) orthopedic; (i) other health impairments; (j) specific learning disability; (k) speech or language impairment; (l) traumatic brain injury; (m) visual impairment including blindness (34 CFR 300.8(a)(1)). Because some disabilities are difficult to diagnose, states may choose to designate a student as “developmentally delayed”; however, there is typically a timeline for this particular designation, and students may then transition out of SPED or into one of the 13 aforementioned categories.

2 If a student’s disability does not impact their academics, the student will not qualify for SPED. However, they could still receive services and protections under Section 504 Rehabilitation Act of 1973.

3 We determine whether a district submitted complete SPED program data by comparing the total number of students in SPED services reported for each district in the GRD to the CCD (National Center for Education Statistics, 2017). For about 8% of districts in the GRD, the numbers of students in SPED reported to GRD and CCD were within 20% of each other; for about 4% of districts, the two numbers were within 10% of each other.

4 We examine the quality of the SPED indicator in our data in two ways. First, we compare the total number of students in SPED services reported for each district in the GRD to the CCD. We retain districts for which the reported number of students in SPED from the two data sources were within 10% of each other. Second, we examine the data files, which contain binary indicators for service receipt as well as text fields for identification results or program participation. In this qualitative check, we verify that the text fields provided descriptions that were relevant to SPED services. For instance, many of the SPED text field values included disability categories or identification notes.

5 Average test duration, standard error of measurement, and percentage of rapidly-guessed items (items to which students responded in a very short amount of time, insufficient to read/comprehend the item) were similar for ever-SPED and never-SPED students.

6 For comparison, estimated Black-White achievement disparities in math and reading are 0.54 SD and 0.41 SD, respectively (Kuhfeld et al., 2021).

7 For In a typical year, about 10,000 schools across the nation administer NWEA assessments to kindergarten students. This means that the sample with complete SPED data is only about 1% of schools with MAP assessment data.