EBP and Autism

In terms of evidence-based practice for students with autism, there are a variety of definitions as well. According to a guide published by the North Carolina Department of Public Instruction, a method that is based on scientifically-based research means that there exists some reliable evidence that the method is effective for some students with autism (KATC, 2008, as cited in North Carolina Department of Public Instruction, 2011).

The guide goes on to say that, in order for a method to be considered an evidence-based practice for students with autism, it must have certain amounts of published research (in scientific journals) that are either randomized or quasi-experimental research designs or single-subject research designs.

So, according to this particular guide, to be considered an evidence-based practice (EBP), a method must have one of the following amounts of research:

Type of Research	Number of Studies Needed
Randomized or quasi-experimental	2
Single-subject	5 (by at least 3 different researchers)
Combination	At least 1 randomized or quasi-experimental study and at least 3 single-subject studies (by at least 3 different researchers)

Reichow and colleagues also propose some different, specific criteria for determining whether a method meets EBP standards in their book:

Evidence-Based Practices and Treatments for Children with Autism

FYI Randomized or quasi-experimental studies can be thought of as a traditional group study, where one group receives a treatment and another does not. The two groups are then compared. One example: two groups of students with autism participate in a study. One group receives a social skills training class, and the other does not. The levels of social skills for the groups are then compared. A single-subject study involves one or a small number of participants, in which each person’s skill levels or performance are compared before and after the intervention takes place. So, each person’s scores during the intervention are compared to his or her scores before the intervention. One example: Jake, a 6-year-old student with autism has difficulty during some work tasks in school. He sometimes throws his pencil and workbook on the floor. Jake’s teacher takes some data on how often this happens before she makes any changes (this is called the baseline phase). Then, she starts giving Jake a small reward for each workbook page completed. She takes data during this time (the intervention phase) to see if the behavior of throwing objects on the floor decreases. The amounts of “object throwing” behavior are compared before and during/after the intervention. Please note that in actual single-subject research, there are generally more research phases included in order to demonstrate experimental control, but the above example gives a basic overview of the process.

References:

Evidence-Based Practices in Educating Children with Autism