PROOF POINTS: New research review questions the evidence for special education inclusion

Analysis unable to disentangle which students benefit from being taught alongside general education peers

For the past 25 years, U.S. policy has urged schools to keep students with disabilities in the same classrooms as their general education peers unless severe disabilities prevent it. It seems a humane policy not to wall off those with disabilities and keep them apart from society. Who would argue against it?

Schools have embraced inclusion. According to the most recent data from the 2020-21 school year, two-thirds of the 7 million students with disabilities who receive special education services spent 80 percent or more of their time in traditional classrooms. Separation is less common today; only one out of every eight students with disabilities was taught separately in a special-needs-only environment most of the time.

But a recent international analysis of all the available research on special education inclusion found inconsistent results. Some children thrived, while others did very badly in regular classrooms. Overall, students didn’t benefit academically, psychologically, or socially from the practice. Math and reading scores, along with psychosocial measures, were no higher for children with disabilities who learned in general education classrooms, on average, compared to children who learned in separate special education classrooms.

“I was surprised,” said Nina Dalgaard, lead author of the included study for the Campbell Collaboration, a nonprofit organization that reviews research evidence for public policy purposes. “Despite a rather large evidence base, it doesn’t appear that inclusion automatically has positive effects. On the contrary, for some children, it appears that being taught in a segregated setting is beneficial.”

Many disability advocates balked at the findings, published in December 2022 on social media. An influential lobbying organization, the National Center for Learning Disabilities, said it continues to believe that inclusion is beneficial for students and that this study will “not change” how the disability community advocates for students.

“Students with disabilities have a right to learn alongside their peers, and studies have shown that this is beneficial not only for students with disabilities but also for other students in the classroom,” said Lindsay Kubatzky, the organization’s director of policy and advocacy. “Every student is different, and ‘inclusion’ for one student may look different from others. For some, it could be a classroom separate from their peers, but that is rarely the case.”

The Campbell Collaboration study is a meta-analysis, which means it is supposed to sweep up all the best research on a topic and use statistics to tell us where the preponderance of the evidence lies. Dalgaard, a senior researcher at VIVE—The Danish Centre for Social Science Research, initially found over 2,000 studies on special education inclusion. But she threw out 99 percent of them, many of which were quite favorable to inclusion. Most were qualitative studies that described students’ experiences in an inclusion classroom but didn’t rigorously track academic progress. Among those that did monitor math or reading, many noted how much students improved in an inclusive setting but didn’t compare those gains with how students might have otherwise fared in a separate special-needs-only setting.

Fewer than 100 studies had comparison groups, but still, most of those didn’t make the cut because the students in inclusive settings were vastly different from those in separate settings. Special education is a particularly difficult area to study because researchers cannot randomly assign students with disabilities to different treatments. Schools tend to keep children with milder disabilities in a regular classroom and teach only those with the most severe disabilities separately. In comparing how both groups fare, it should be no surprise that students with milder disabilities outperform those with more severe disabilities. But that’s not good evidence that inclusion is better. “It’s a serious, confounding bias,” Dalgaard said.

In the end, Dalgaard was left with only 15 studies where the severity of the disability was somehow noted so that she could compare apples to apples. These 15 studies covered more than 7,000 students across nine countries, ages six through 16. Four of the studies were conducted in the United States, with the others in Europe.

The disabilities in the studies ranged widely, from the most common ones, such as dyslexia, ADHD, speech impairments, and autism, to rarer ones, such as Down syndrome and cerebral palsy. Some students had mild versions; others had more severe forms. I asked Dalgaard if she found clues in the results as to which disabilities were more conducive to inclusion. I was curious if children with severe dyslexia, for example, might benefit from separate instruction with specially trained reading teachers for the first couple of years after diagnosis.

Dalgaard said there wasn’t enough statistical evidence to untangle when inclusion is most beneficial. But she did notice in the underlying studies that students with autism seem to be better off in a separate setting. For example, their psychosocial scores were higher. But more studies would be needed to confirm this.

She also noticed that how a school goes about including students with disabilities matters. In schools that used a co-teaching model, one regular teacher and one trained in special education, students fared better in inclusion classrooms. Again, more research is needed to confirm this statistically. And, even if co-teaching proves to be effective over multiple studies, not every school can afford to hire two teachers for every classroom. It’s particularly cost-prohibitive in middle and high school as teachers specialize in subjects.

Instead, Dalgaard noted that inclusion is often a cost-cutting practice because schools save money when they no longer run separate classrooms or schools for children with disabilities. “In some cases, children with disabilities no longer had access to the same resources. It’s not supposed to happen this way, but it does in some places,” said Dalgaard. “That is probably why the results of the meta-analysis show that some children actually learn more in segregated settings.”

I was surprised to learn from Dalgaard that no sound meta-analysis has found “clear” benefits for special education inclusion. Indeed, previous meta-analyses have found precisely the same inconsistent or very small positive results. This latest Campbell Collaboration study was commissioned to see if newer research, published from 2000 to September 2021, would move the dial. It did not.

As a nation, we spend an estimated $90 billion a year in federal, state, and local taxpayer funds on educating children with disabilities. We ought to know more about how to help them learn best.