by Smith, T.
Children with Autism Spectrum Disorders do not readily learn in the manner of their typically developing peers. Therefore it is necessary to implement specialized instruction in a manner that suits the learner with autism. Applied Behavior Analysis is one method of instruction that has received a lot of attention as being successful with learners with autism. The specific component of discrete trial training (DTT) is the most extensively studied. DTT is an important component of ABA but is not the only component and should be combined with other ABA procedures. DTT is a small unit of instruction typically lasting between 5 and 20 seconds implemented by a trained person in a one-to-one distraction free environment. DTT consists of five parts which are the cue, prompt, response, consequence, and intertrial interval. The amount of time a learner spends doing DTT is dependent on the treatment program that has been designed for their individual needs. Three aspects of DTT have been studied as potential factors contributing to the learners increased motivation to learn and actual learning. Language and communication skills are often targets of DTT instruction and have a sound research base to support their use. Additional skills that DTT may be used for are expanding children’s skills, such as vocabularies and managing disruptive behavior. Despite the value of DTT in teaching children with autism, DTT is not without significant limitations which include cue dependency, lack of generalizability of skills, and the labor intensive requirements to effectively implement DTT. These limitations suggest that DTT be used in conjunction with other methods such as incidental teaching, peer models, videotapes. Overall DTT is the only instructional method that is demonstrated by empirical evidence to be effective for teaching new behaviors and discriminations to children with ASD. It is important to consider the limitations of DTT and combine its use with other instructional methods that lend themselves to better generalization and less prompt dependency.