Recently, an article published in the journal Pediatrics argued that inclusion in the general education classroom was not an evidence-based indicator of improved postsecondary outcomes for autistic students. The Autistic Self Advocacy Network is profoundly alarmed by the serious scientific limitations in the design and analysis present in the study in question and the ways in which these methodological flaws have not been acknowledged in media coverage describing the publication.
Many of the challenges faced by the study are acknowledged by the authors themselves. These include “fairly extensive missing data,” a “crude” measure of classroom inclusivity, and “high imprecision” of their adjustment for measured potential confounds. The authors recognize that the analysis fails to include a direct measure of students’ academic performance, and ignores “school and district characteristics” like “funding and eligibility requirements for special education,” “processes determining the availability of supportive services,” or other “school policies and strategies.” Despite these problems, the study assumed that the level of school inclusion was random among students with the same set of measured characteristics, which the authors acknowledge is “impossible” to test, even though many families make active decisions about level of inclusive placement based on factors like services.
The measures used to ascertain two of the three outcomes the study measured – not dropping out of high school and attending college – reflect significant biases that challenge the validity of the study’s conclusions. The authors group together high school graduation, any kind of completion certificate, and still being in high school in participants’ early 20’s under the outcome of “not dropping out of high school.” The measure of “attending college” grouped together not only colleges of any length or type but also taking a postsecondary class to earn a high school degree. In short, the study makes no distinction between a high school graduate who is attending a 4-year college in order to obtain an undergraduate degree and a student taking a community college course on track for a certificate of completion rather than a high school diploma. These are significantly different paths for autistic students – treating them as identical as a way of diminishing the role inclusion can play in enhancing higher education access and high school graduation is scientifically questionable. This calls into serious question the authors’ claim that inclusive educational opportunities do not impact post-secondary results.
In addition to these core flaws, the study has problems with its “rather small heterogenous” sample. With the small number of participants for the type of analyses performed, the statistics may not have enough power to be significant. The results did find that inclusion was associated with better outcomes for all three measures. However, the study determined this association to no longer be significant after controlling for potentially confounding variables. Thus the authors’ decisions as to what to and not to include as a confounding variable is extremely relevant to the conclusions reflected. Yet the authors neglected to implement the key step of making sure their potential confounders did not independently relate to their level of inclusion, even though some of them were extremely predictive of level of inclusion.
Finally, the authors’ decision to track inclusion by splitting students into three categories – those who spent 0% of their time in inclusive settings, those who spent 1% to 74%, and those who spent 75% or above – was curious and likely had a negative impact on the quality of the study’s results. Department of Education data collected on State and Local Education Authority progress in including students with Individualized Education Plans in the general education classroom tracks the percentage of students who spend 80% or more of their time in general education; those who spend between 40-79%, those who spend less than 40%; and those in separate schools, residential facilities and other non-public school settings. This allows for a far more meaningful representation of the continuum of access to the general education classroom than offered by the Pediatrics study.
ASAN believes this study is so riddled with critical flaws that it should not have been published. A wide scope of research on both students with disabilities in general and autistic students in particular supports the idea that students have better outcomes from inclusive settings. The question of whether or not to include students with disabilities has been asked and answered many times over. The time has come to change the conversation from “if” to “how”. We look forward to research studying how to improve inclusion, not trying to prove that separate can be equal (let alone superior).
ASAN would like to thank Steven Kapp for his assistance in the analysis and drafting that went into this statement.
1 Soukup, J. H., Wehmeyer, M. L., Bashinski, S. M., & Bovaird, J. A. (2007). Classroom variables and access to the general curriculum for students with disabilities. Exceptional Children, 74(1), 101-120.
2 Ryndak, D. L., Morrison, A. P., & Sommerstein, L. (1999). Literacy before and after inclusion in general education settings: A case study. Research and Practice for Persons with Severe Disabilities, 24(1), 5-22.
3 Freeman, S. F., & Alkin, M. C. (2000). Academic and social attainments of children with mental retardation in general education and special education settings. Remedial and Special Education, 21(1), 3-26.
4 Cole, C. M., Waldron, N., & Majd, M. (2004). Academic progress of students across inclusive and traditional settings. Mental Retardation, 42(2), 136-144.
5 Bailey, D. B., McWilliam, R. A., Buysse, V., & Wesley, P. W. (1998). Inclusion in the context of competing values in early childhood education. Early Childhood Research Quarterly, 13(1), 27-47.
6 Kurth, J. A. & Matergeorge, A. M. (2010). Academic and cognitive profiles of students with autism: Implications for classroom practice and placement. International Journal of Special Education, 25 (2), 8-14.
7 Kurth, J., & Mastergeorge, A.M. (2012). Impact of setting and instructional context for adolescents with autism. Journal of Special Education, 46(1), 36-48.