Association for Behavior Analysis International

The goal was to assess preference between massed and alternating matching-to-sample trials in children with autism. Two participants were presented with a matching-to-sample task on a computer. Sessions consisted of six trials of an unknown set of word-picture stimuli relations and six trials of known picture-picture identity relations. Prior to the presentation of the 12 trials, participants were asked to choose between three different colored stimuli randomly located on the computer screen. The participants choice of stimuli determined the condition for the session either mass (6 known/6 unknown stimuli relations, or vice-versa), alternating (alternation of known and unknown stimuli relations), or extinction (no reinforcement delivered). Once a condition preference to a condition was established a new set of colors were presented and preference was reassessed. Throughout a session, no prompts were delivered during the presentation of known stimuli relations, and a spoken word was used as prompt for the unknown stimuli relations. Prescribed prompts were faded systematically to assess acquisition of skill. All correct responses produced a token and once three tokens were earned a trade-in period occurred (unless the extinction condition was chosen). A preference for alternating trials developed during every assessment within and across participants.

An accurate tact repertoire often requires control by multiple features of nonverbal stimuli over the verbal response. Inappropriate stimulus control can develop when learners are exposed to compound stimuli. The current study evaluated the effects of a differential observing response (DOR) on typically developing childrens (3-5 years old) acquisition of tacts of overlapping compound stimuli. When initial instruction without a DOR did not produce acquisition, an instruction format was introduced which required participants to engage in a DOR in the form of an identity match to sample response prior to tacting the stimuli. If the DOR intervention did not produce mastery, further manipulations were evaluated, including a verbal DOR and the establishment of vocal precurrent responses. Implications of the effectiveness of such interventions on tact acquisition are discussed.

Some children with autism spectrum disorders (ASD) display persistent errors during discrete-trial training. Fisher and colleagues showed that combining a second-order reinforcement schedule with response cost produced improved accuracy in completion of intraverbal and receptive tasks among children with ASD (Fisher, Pawich, Dickes, Paden, & Toussaint, 2014). The purpose of the current study was to extend the findings of Fisher et al. (2014) by evaluating the efficiency and effectiveness of their second-order reinforcement and response-cost procedure with a common error-correction procedure during acquisition of receptive-identification tasks. Two children diagnosed with an ASD participated. During baseline, we implemented a VR2 reinforcement schedule for cooperative behavior but not for correct responses. During treatment, in one condition we reinforced correct responses on a second-order FR3 [FR1] schedule and removed all accumulated reinforcers contingent on an error. In the other condition, we reinforced correct responses on an FR1 schedule, and following an error, we repeated the trial until the participant emitted an independent correct response. Results showed that the error-correction procedure proved to be more efficient with both participants and more effective with one participant. We discuss these results relative to the discriminative effects of reinforcement and punishment during acquisition of new tasks. We will continue to replicate these procedures with additional participants.