What is ABA?

What is applied behavior analysis?

What is ABA? There are probably a hundred people who have written answers, but please bear with me.

"Applied" means practice, rather than research or philosophy. "Behavior analysis" may be read as "learning theory," that is, understanding what leads to (or doesn't lead to) new skills. (This is a simplification: ABA is just as much about maintaining and using skills as about learning.) It may seem odd to use the word "behavior" when talking about learning to talk, play, and live as a complex social animal, but to a behaviorist all these can be taught, so long as there are intact brain functions to learn and practice the skills. (That is the essence of the recovery hypothesis--that for many children, the excesses and deficits of autism result largely from a learning 'blockage,' which can be overcome by intensive teaching.)

Typically developing children learn without our intervention--that is, the 'typical' environment they are born into provides the right conditions to learn language, play, and social skills. (After a few years, however, this breaks down, and we no longer learn everything 'naturally'--it takes a very structured environment, for example, for most of us to learn to read, write, and do arithmetic.)

Children with autism learn much, much less from the environment. They are often capable of learning, but it takes a very structured environment, one where conditions are optimized for acquiring the same skills that typical children learn 'naturally.' ABA is all about the rules for setting up the environment to enable our kids to learn.

Behavior analysis dates back at least to Skinner, who performed animal experiments showing that food rewards (immediate positive consequences to a target behavior) lead to behavior changes. This is accepted by everyone who wants to train their dog to 'go' outside, but we are not so inclined to want to believe the same of ourselves. Part of the problem is that people do respond to a broad range of reinforcements (rewards), but it is really true that an edible treat is among the most reliable, especially at first. (The skills that we more often think lead to learning--motivation, self-discipline, curiosity--are marvelous, and really do set us apart from other animals--but those are truly sophisticated 'behaviors' that fully develop only after more basic language and social skills are in place.)

Conversely, any new behavior that an animal (or you or I) may try, but is never rewarded, is likely to die out after a while (how often will you dial that busy number?). And, as common sense would have it, a behavior that results in something unpleasant (an aversive) is even less likely to be repeated. These are the basics of behavioral learning theory. ABA uses these principles to set up an environment in which our kids learn as much as they can as quickly as possible. It is a science, not a 'philosophy.' (Even the "as quickly as possible" part is based on science, since there is some--not conclusive--evidence that the developmentally disordered brain "learns how to learn" best if the basic skills are taught in early childhood.)

Behavioral learning is not the only type of learning. Most learning in schools is from an explanation or from a model, what people call 'natural' learning. Typically developing children learn from their environment (other people) at an astounding rate, completely unassisted. The whole point of ABA is to teach the prerequisites to make it possible for a child to learn 'naturally.' If our kids could learn without assistance in the first place they wouldn't have autism!

The most common and distinguishing type of intervention based on applied behavior analysis is discrete trial teaching. It is what people most often think of when you say "ABA" or "Lovaas method." This is partly because there are so many hundreds of hours of DT teaching, and partly because it looks so odd. But it is what it is because that's what works--every aspect has been refined (and is still being refined) to result in maximum learning efficiency.

(Briefly: the student is given a stimulus--a question, a set of blocks and a pattern, a request to go ask Mom for a glass of water--along with the correct response, or a strong 'hint' at what the response should be. He is rewarded (an M&M, a piggy-back ride, a happy "good job!") for repeating the right answer; anything else is ignored or corrected very neutrally. As his response becomes more reliable, the 'clues' are withdrawn until he can respond independently. This is usually done one-on-one at a table (thus the term table-top work), with detailed planning of the requests, timing, wording, and the therapist's reaction to the student's responses.)

It is a mistake, however, to think of an ABA program as just DT teaching. Lovaas (among others) notes very clearly that a behavioral program is a comprehensive intervention, carried out in every setting, every available moment. The skills that are taught so efficiently in discrete trial drills must be practiced and generalized in 'natural' settings. A child who does not know the difference between 'ask' and 'tell' may slowly get a higher and higher percentage of right answers during table-top drills until he is considered to have 'mastered' that skill; but he will not go on to use 'ask' and 'tell' appropriately without additional support in natural situations; it takes time to go from 'mastery' to 'ownership.' It takes trained and supportive people--parents, teachers, relatives, even peers--to help reinforce a wide range of appropriate behaviors in a variety of settings, until the level of reinforcement fades to a typical level (such as the smile you get when you greet someone).

Here is a child's interaction with a teacher or other adult who is being as helpful as possible without appropriate training:

Teacher: Hi, Alex, are you excited about Christmas?
A: [no response]
Teacher: What are you going to do on Christmas?
A: I don't know.
Teacher: Are you going to get presents?
A: Yes.
Teacher: What else are you going to do?
A: [no response]
Teacher: Do you have a tree?
A: Yes.
Teacher: Who's going to bring presents on Christmas?
A: I don't know.
Teacher: Is it Santa Claus?
A: Yes.
Teacher: [smile] Thanks, Alex!

This is the child's half of the conversation:

"I don't know, Yes, Yes, I don't know, Yes."

Any learning going on? (By the way, I've watched people have conversations like this, then comment "He's talking so much more!")

Here's how a trained person might make this an opportunity for practicing conversation skills:

Teacher: Hi, Alex, are you excited about Christmas?
A: [no response]
Teacher: Are you excited about Christmas? Say, Yeah, I want to open my...
A: Yeah, I want to open my presents!
Teacher: [Smile] Me too! What presents did you ask for?
A: I asked for presents.
Teacher: What presents did you ask for? Say, For Christmas, I asked for...
A: I asked for a bike. For Christmas.
Teacher: Cool! [Small tickle] Are you excited about Christmas?
A: Yeah, I want a bike.
Teacher: [Bigger tickle] A bike! That's great! I've got my tree all decorated with ornaments. I put lots of ornaments on MY tree. [Point to A's tree.]
A: I put heart ornaments on my tree.
Teacher: Alex, that's so great! [Great big tickle]
A: Ahhhhh! Cut it out!