The effects of a training package (modeling, role-playing, and feedback) were evaluated using a multiple baseline across skill areas. Two trainers taught two teachers to use basic intervention skills that included components of both discrete trial teaching (DTT) and the Natural Language Paradigm (NLP). Training occurred in the context of one task. Generalization was assessed with two untrained tasks. Teachers' responses in the target task increased following training, as did appropriate responding from the child. Some generalization of the teaching skills occurred. Teachers were instructed to generalize acquired skills as a possible method to promote generalization. The results of these findings and implications for training of ABA providers are discussed.

Few studies have investigated methods for establishing neutral stimuli as conditioned reinforcers in human subjects. Conditioned reinforcers, however, can alleviate some of the problems encountered in applied behavior analytic (ABA) therapy for children with autism, such as satiation and suitability of reinforcers for specific environments. A series of reversals evaluated the effects of a conditioning procedure involving pairing a neutral stimulus, the remote control stimulus (RCT), with an identified reinforcer. Phase 1 demonstrated that the RCT was neutral. In Phase 2, alternating pairing and testing conditions were run. During testing the effects of pairing were evaluated by the effectiveness of the RCT in maintaining a response in the absence of a previously available reinforcer (extinction test) and in increasing a new response over a baseline level (learning test). Results from the extinction test suggest that under some pairing conditions the RCT can acquire properties of a reinforcer.