We, as behavior analysts, are on the verge of either dramatically improving behavioral treatment for children with autism or hopelessly tainting our research with ambiguous terms and petty distinctions.

In my experience, discussion of behavioral treatment for children with autism can be broken into three different time periods. From the 1960's through the 1980's, research was steadily compiled demonstrating the effectiveness of behavioral treatment for many children with autism, as well as the limitations of this treatment. During the 1990's, different approaches were branded either as "tried and true" methods of behavioral treatment (e.g., Lovaas, traditional ABA) or "cutting-edge" methods (e.g., Verbal Behavior, PRT). With the start of the new millennium, research has started to move beyond specific terms and more systematically focus on different aspects of behavioral treatment and the likelihood they will best help a particular child.

Unfortunately, the terminology used to describe procedures, methods, approaches, etc. in behavioral treatment has not always been defined precisely enough. If we are to truly research what works best for a particular child, it is important that we don't confuse the issue by using ambiguous terminology.

This is the first in a series of discussions on refining the behavioral terminology we use in research. My discussion on behavioral terminology may not be flawless, but I have enough background to get the conversation going and am more than willing to learn from other people's insights. So, here we go...

Question 1:
Which of the following are basic, precise procedures that are used in behavioral treatment for children with autism: Verbal Behavior, Pivotal Response Teaching, Discrete Trial Teaching (note the use of capital letters) Natural Environment Teaching, incidental teaching, fluency-based instruction, time-delay?

My Answer:
None of the above are basic, precise procedures that are used in behavioral treatment. Some of these are ambiguous terms – they aren't precise. Others are specific variations of other procedures – they aren't basic. We'll return to this question in a later discussion.

Question 2:
So what are some basic, precise procedures to increase behaviors that are used in behavioral treatment for children with autism?

My Answer:
There are two basic, precise procedures often used in behavioral treatment. The best names for these procedures are discrete trial teaching and mand training.

Discrete trial teaching is a potentially five-part unit of instruction consisting of:

1. the discriminative stimulus (i.e., what the instructor says or does)
2. a prompt (i.e., any help the instructor gives to the child)
3. a response (i.e., what the child does)
4. a consequence (i.e., whether or not the response is reinforced)
5. an inter-trial pause (i.e., the few seconds before the next discriminative stimulus is presented).

Here is an example of discrete trial teaching.

1. Instructor asks, "What's your name?"
2. Instructor prompts, "My name is Vi..."
3. Child responds, "My name is Vince."
4. Instructor shouts "hurrah!" and spins child around.
5. Instructor pauses before asking the next question.

One example of a discrete trial procedure is included in the following Journal of Applied Behavior Analysis research article: "Teaching Spontaneous Responses to Young Children with Autism" by Emily A. Jones, Kathleen M. Feeley, and Jennifer Takacs.
http://seab.envmed.rochester.edu/jaba/articles/2007/jaba-40-03-0565.pdf

Mand training is a potentially four-part unit of instruction consisting of:

1. establishing operations (i.e., environment is created in which objects become valuable)
2. a prompt (i.e., any help the instructor gives to the child)
3. a behavior (i.e., what the child does)
4. a consequence (i.e., whether or not the behavior is reinforced)

Here is an example of mand training.

1. The instructor places the child's favorite juice on a high shelf where he can see it, but can't reach it. Child eventually walks over and looks up at the juice.
2. Instructor prompts, "I want..."
3. Child says, "I want juice"
4. Instructor gives the juice to the child.

One example of mand training is included in the following Journal of Applied Behavior Analysis research article: "Manipulating Establishing Operations to Verify and Establish Stimulus Control During Mand Training" by Anibal Gutierrez and colleagues.
http://seab.envmed.rochester.edu/jaba/articles/2007/jaba-40-04-0645.pdf

Question 3:
Are there other basic, precise procedures that could be used in behavioral treatment for children with autism?

My Answer:
Yes—One example would be video modeling. Another would be differential reinforcement of other behaviors. However, the above two procedures (discrete trial teaching and mand training) encapsulate the vast majority of teaching that occurs in behavioral treatment, particularly for young children with autism.

Your thoughts and comments?