Tuesday, September 19, 2006

Clinic Notes: Discrete Trial Training

I have posted below the general procedure for discrete trial training from my web site. This is a very useful procedure for teaching children with autism or other neuropsychological disorders. In future posts I will have specific programs. If you are having problems teaching a child let me know and I will help.

10. Applied Behavioral Analysis (ABA) Program for Teaching
Children Shapes, Colors, Numbers, and Letters

Some children with neuropsychological disorders fail to learn different shapes, colors, numbers, or letters because of compliance problems, while other children with neuropsychological disorders do not learn because of sensory and/or neurological processing problems associated with certain disorders. If a reward is given to the child for a correct response, performance often dramatically improves in noncompliant children. On the other hand, the rate of improvement in children with actual sensory/neurological problems is much, much slower. In the beginning the best thing to do when we are trying to teach a child to learn the differences between stimuli, such as two colors or two shapes, is to run intensive drills using bite size rewards for all children, which we then fade out. The rate by which the child improves tells us if we are dealing with a noncompliant child or a true sensory processing problem. See Case History Number 1 in Little Bubba’s Not Ready For Nashville Yet for an example. (Available at http://wwwABA4autism.com).
Discrete Trial Training
Teaching discriminations between stimuli is easier if only two stimuli are used in the beginning. For example, take two objects, which are different colors. Start with the easiest discriminations first. The difference between blue and red should be easier to learn than the difference between red and orange.
Behavioral Assessment
Use objects which are identical but have different colors, such as pieces of cloth or pieces of construction paper. (It is better not to use different colored toys which the child would want to play with.) Ask the child to point to or hand you one color and record the number of correct responses and incorrect responses on the attached form. Do this for several days without comment or reward in order to obtain a baseline. (If you run blocks of ten trials, then it is easy to calculate the percentage of correct responses.) Only wait 3-5 seconds for a response and 3-5 seconds between trials. Be sure to switch the different colored objects from side to side so the child will not learn a position habit. (In other words, a child can learn right or left side rather than the color and still be correct.)
Behavioral Intervention
After you have a baseline, reward the child immediately for each time he/she correctly picks the right color with a bite size treat. (A Tupperware bowl full of bite size goodies is what we use as tangible reinforcers for most kids in our clinic. The child only gets to pick one bite size piece.) If the child gets it wrong say, “no,” and ask the child again. After the child receives a “no” response twice in a row, prompt the child with the correct answer (give no reward). Point to the correct color; move it closer--anything that will make the discrimination between the two colors easier for the child. Then fade the prompt on later trials by gradually moving the correct colored object back so it is in line with the colored object. Continue to wait 3-5 seconds for a response and 3-5 seconds between trials. And be sure to switch the objects from side to side so the child will not learn a position habit.
Once the child is picking the correct color, ask the child to point to or hand you the other color. When the child can discriminate between the two colors, move to another color. Again, make it easy for the child. Red and green would be easier than red and yellow. Children with neuropsychological disorders may not generalize from one set of colors to another so be prepared to spend a long time in discrete trial training. If you get no improvement after a large number of trials, try black and white discrimination because the child may be color blind.
Maintenance and Generalization
Deliver social reinforcement in the form of praise, hugs, pats on the back, etc. as often as you can once the child learns the difference between stimuli. Edible reinforcers, such as candy or other treats, may be faded out once a significant change from baseline is achieved. Only give the tangible reinforcer every other time, then every third time, etc. until it is no longer required to maintain appropriate behavior. Of course, social reinforcement should be given as often as possible.
If you are not getting anywhere after several weeks, contact a psychologist versed in ABA in your area, if at all possible. (Go to http://www.aabt.org/ to locate ABA therapists.) One possibility for the ABA program not working is satiation. In other words, your child is getting tired of whatever reinforcer you are using.
(See http://www.polyxo.com/discretetrial/ for a good overview of discrete trial training.)