Perkins School for the Blind Transition Center

Evidence-Based Practices for Educating Children with ASD

The use of interventions derived from principles of applied behavior analysis (ABA) has a decades-long history in providing effective methods of ameliorating some of the symptoms of autism spectrum disorders. As a notable example, functional communication training (FCT)(Durand, 1990) has been shown to be an effective way to decrease behaviors that interfere with learning, including behaviors that can be significantly or severely harmful. Briefly, FCT assumes that all challenging behaviors have communicative intent, and FCT decreases those challenging behaviors by providing an individual with alternate, more appropriate ways to communicate that intent. So a child who throws himself on the floor when asked to do a difficult task, may be provided with a way to communicate “I need help.”

Based on the example provided, one may wonder whether to consider functional communication training in a clinical or educational intervention. The elimination or decreasing of challenging behaviors has traditionally been under the purview of the medical, psychiatric or psychological fields and thus may be seen as a clinical intervention. Helping an individual learn to state “I need help” has traditionally been under the purview of speech pathology, and can also be considered a clinical intervention. Clinical fields like medicine and psychology can demonstrate, through a history of research, that providing this intervention can result in certain improved outcomes. These interventions may be defined as evidence-based and it is the expectation of insurance companies that an intervention prove its ability to reach certain outcomes before it will be paid for by them.

However, schools are required by federal law (IDEA) to complete a functional behavioral assessment for an individual on any behavior that interferes with learning. The determination that a challenging behavior functions to help a child escape from difficult tasks, and then the development of a program to help that child communicate “I need help” may in some circumstances, be completed by licensed clinical staff, such as a psychologist and a speech pathologist. For this intervention to be truly effective, it must be implemented as well by teachers, aides, other classroom support staff and parents so that the new behavior of requesting a break is demonstrated across all settings. And in all likelihood, it is probably the teacher, aide, support staff and parents who complete the FBA and develop FCT as an intervention in most schools anyway. That FCT is now thankfully provided by educational teams, has resulted in more students with autism included in less restrictive settings and tremendous advancements for individual students to have better behavioral and social outcomes.

However, it also may have contributed to the current debates over whether insurance companies will cover ABA as an intervention with autism. Any good intervention plan using ABA as a model will include increasing skills. It is usually a school’s responsibility to teach new skills. Why would a medical insurance company take on the task of education? To the best of my knowledge, only six states have a law in place requiring private insurance companies to cover ABA as a service for individuals with autism (Arizona, Florida, Indiana, Minnesota, South Carolina, and Texas). Over 20 states have no such laws, and no bills on the horizon to address this issue.

It is important for autism intervention that educational and related service providers become more attuned to the need for empirical evaluation of teaching strategies. As a field, education is rife with “fad” instructional practices (Jacobson, Foxx, & Mulick, 2005); special education less so than general education, as the development of strategies for students who struggle to learn in traditional ways has required research analysis to some extent. For autism in particular, the type of strategies that are needed to ameliorate symptoms can be so complex and varied and can need contributions from medical, psychiatric, behavioral, speech, occupational and physical therapies, that true analysis of what is effective for each individual should be a required component of the child’s educational plan.

It would greatly benefit autism services if all providers were required to learn how to conduct and consume research in single-case design. When a medical intervention is tested, it is demonstrated effective if the average amount of change in all individuals combined pre-to post-treatment is greater for the sample who received the intervention than those who didn’t. This may be acceptable in general education, where the intention is to treat as many students as possible with the same instruction, and get the largest number of those students graduating. The drawbacks of large group research designs (in general and for autism in particular), is that you cannot predict how any one individual will respond to treatment when you are looking at averages.

To make this a concrete example, let’s say we have five individuals who score a “20” on some test pre-treatment. The average pre-test score would then be (20+20+20+20+20)/5 * 100 = 20. Let’s then say that post-treatment, that four of the individuals increase their score to a 30, and one individual decreases their score to a 5, making the post-treatment average (30+30+30+30+5)/5 * 100= 25. There are more complex statistical analyses for this, but for sake of argument, since an average score of 25 post-treatment is higher than an average score of 20 pre-treatment, a group research design would determine this was an effective intervention. How would you feel though, if you were the individual whose score decreased to five? This is not an intervention that you would want, as you got worse with it than without. Group designs to empirically-validate a treatment do not give you information about how any one individual will respond.

Single-case design research, however, does provide a methodology for evaluating instructional practices, behavior interventions, or other medical or environmental modifications that may impact on the education and outcome of a student with autism. These research designs allow the evaluation of the effectiveness of an intervention for one specific individual, and some common ones are described briefly below (Barlow, Hayes & Nelson, 1984)):

  • Reversal design: This research design involves taking data while systematically providing and withdrawing an intervention, to see if the rate of behavior changes as a function of the intervention.
  • Alternating treatments: This research design allows you to compare two (or more) interventions provided in alternating sessions within the same time period.
  • Multiple baseline: This research design allows you to evaluate effects of an intervention across individuals, across settings, or across different behaviors.

With decades of single-case design research behind it, components of ABA (like FCT mentioned above) have emerged as the only empirically-validated interventions for individuals with autism. As such, insurance companies are increasingly by required to take ABA services seriously as interventions to address communication, social, and behavioral deficits in autism. Schools that support students with autism, whether public or self-contained, should be required to demonstrate that services provided are resulting in improved outcomes. These students already have individualized education plans that list services to be provided and goals to be addressed. Measures of the empirical basis of a service and accountability for outcomes should be standard practice. ABA provides a methodology to do this, even for interventions that are not based on those same principles.

As an example, students with autism routinely receive occupational therapy as a service on their individual education plans in New York. These services can include “sensory integration” activities, which is par-for-the-course in occupational therapy, but does not have research support as an effective intervention for autism (NYSDOH, 1999). I have conducted single-case research designs with individuals in the school for whom sensory activities are recommended by an occupational therapist to improve reduce impulsivity and distraction. I have been able to demonstrate for a few individuals, that these sensory activities can be provided contingently as effective reinforcers to increase attention to task.

Again, it would benefit autism services if there was a greater emphasis placed on the use of evidence-based practices in education. First, we could truly determine those strategies that will help an individual reach their goals and improve their lives in the community. Second, it would add some accountability to education that is not based on inappropriate standardized tests or teacher-created portfolios of student work. Third, it would assist in legitimizing services that are provided in schools as necessary for some students with autism and would allow a mechanism for families to obtain those services when the school can or does not provide it.

Have a Comment?