Components of a discrete trial

Each skill is taught in very small and very brief units called "trials". Each trial consists of an instruction, a prompt, an opportunity/response, and feedback. The instruction is given in very clear language that the child can understand, and, as the child becomes able to handle more complex language, it moves toward a more natural language base. They may begin with a direct stimulus "look at me" then move to a more natural stimulus such as just saying the child’s name to get his attention.

Prompts are not always needed, but they will use "anything that will facilitate the desired response". These can be visual prompts such as a gesture or taking the child’s hand and moving it through the desired motions (a physical prompt), but the aim is to reach a point where the child can do the desired task with no prompt at all. The response should be evident within about 5 seconds. If there is no response, or an incorrect response, the trial is considered to be over and a new one begins (although it may be the same task – in fact, many people would not be able to tell where one "trial" ends and the next begins. It is broken up this way so that data can be obtained on the progress of the child).

Feedback on each trial is given immediately. The more information that is provided to a child here, the faster the learning can take place. It is important, he says, not to praise every response; some will be terrific, some barely adequate, some not quite accurate and some simply unacceptable (such as throwing things). Use the feedback, Dr. McEachin says, to indicate where the response falls. ABA has a reputation for using strong aversives because that was the way the program was originally designed years ago, however, that has changed and they no longer use very strong negative feedback. He points out though, that "some people think that merely saying ‘no’ to a child constitutes an aversive. What I would say about that is, hearing the word ‘no’ is a very common, everyday event. It is part of the world, part of life, and if you’re a person who cannot handle simply hearing the word ‘no’, you’re going to have a really, really rough go of it. I think it’s important for us to teach children to be able to handle the word ‘no’ simply at an informational level." That doesn’t, he stresses, imply that screaming at children is OK, but firmness can be conveyed without disrupting the situation. There are times, he says, to be very supportive even though they have given the wrong answer. They worked hard, they are sitting appropriately, they are careful about their response, but it is incorrect. Clear information is conveyed in the words and the vocal inflection "Oh, no. Good try." If they are not attending, it’s OK to say "No. You need to look."

The only way to tell if learning is actually taking place is if the child gives some kind of response "that he can only make if he, in fact, understands." Evaluation of the effectiveness of trials is a critical part of the therapy, he says. Data tells whether a student is making progress or not, and if not, "I conclude there’s something wrong with what I’m doing, rather than ‘this student doesn’t learn very well’." He stated that behavior experts need data to demonstrate the cause-and-effect relationship between the intervention and the outcome, "we want to see measurable, observable changes."

The hallmark of the Applied Behavior Analysis program is the intensity with which it is delivered. It takes many, many hours a day, and is very directive and instructionally based. Dr. McEachin says that if a family is not comfortable with a directive-type program, they should find another program that would better suit their needs. He says, "our goal, our responsibility to children is to direct them because they are not able to direct themselves. And, in fact, I’ll bet that there are many times a day when you are quite directive of your children." Brushing their teeth is a good example. "They would rather not do that. But I make a decision as a parent that I know, from a health standpoint, this is important and I’m going to insist that it be done."

A lot of people fault the ABA program because they believe that generalization does not occur, and if a child can’t take information from one situation and use it in another then it’s not really learning; it’s just mimicking. But Dr. McEachin says that generalization is part of the ABA package "and if you (the therapist) understand that, and if you include that in part of the work that you do, you will get generalization." Strong parental involvement in the ABA program is stressed as a very large component of its success.