|Efficacy and Efficiency in Skill Acquisition: Novel Approaches to Measurement and Procedural Refinement
|Sunday, May 24, 2020
|11:00 AM–12:50 PM
|Walter E. Washington Convention Center, Level 2, Room 202A
|Area: AUT; Domain: Applied Research
|Chair: Daniel E Conine (Georgia State University)
|Discussant: Dorothea C. Lerman (University of Houston-Clear Lake)
|CE Instructor: Daniel E Conine, Ph.D.
A wide variety of principles and teaching strategies exist for establishing new skills in the repertoires of children with autism. Each of the studies in this symposium extends previous research in this area with a focus on improving the overall efficacy and efficiency of intervention through the use of novel measurement strategies or through the refinement of specific procedural variables. Topics investigated include the use of continuous (all trials) versus discontinuous (probe) data collection, evaluating prompt dependence when teaching behavior chains, the role of task interspersal in error correction procedures, and the use of a screening condition to predict treatment efficacy when teaching response to name. Findings from these studies have implications for clinical practice and future research in the area of skill acquisition for children with autism.
|Instruction Level: Intermediate
|Keyword(s): error correction, probe data, skill acquisition, task analysis
Clinicians, researchers, students
|Learning Objectives: At the conclusion of this presentation, attendees will be able to describe: 1) the potential impacts of evaluating mastery criteria using continuous or discontinuous data collection when teaching skills, 2) ways to assess for prompt dependence when teaching behavior chains, 3) the impact of task interspersal during error correction, and 4) procedures to teaching response to name to children with autism.
|A Comparison of Continuous and Discontinuous Data Collection in Discrete Trial Teaching
|CRYSTAL M. SLANZI (University of Florida), Timothy R. Vollmer (University of Florida), Daniel E Conine (Georgia State University), James E. Carr (Behavior Analyst Certification Board)
|Abstract: Previous studies comparing continuous and discontinuous measurement have reported different outcomes regarding the effects of each type on sessions to acquisition and maintenance of skills. Specifically, some studies have found that skills mastered using a mastery criterion based on discontinuous data were acquired in fewer sessions than those based on continuous data, but were less likely to be maintained, whereas other studies have found no difference. This may be due to procedural differences such as the percent correct required for mastery or variations in prompts or target selection. Mastery criteria in previous studies were based on a single target rather than a set of targets. The purpose of the current study is to extend previous research in this area by comparing the effects of continuous and discontinuous measurement by when mastery criteria are based on correct responding across all three targets as a set rather than with individual teaching targets. In the preliminary results, there has been a considerable difference in rate of acquisition in 1 out of 3 participants and almost no difference in 2 of the 3 participants. Levels of maintenance have been variable across all 3 participants.
|An Evaluation of Prompting Procedures on Prompt Dependence and Task Mastery
|EMMA GRAUERHOLZ-FISHER (University of Florida), Jonathan K Fernand (Aurora University), Brandon C. Perez (University of Florida), Haleh Amanieh (West Virginia University), Timothy R. Vollmer (University of Florida)
|Abstract: Horner and Keilitz (1975) demonstrated that individuals with intellectual disabilities could learn a very complex self-care task with targeted training and established the least-to-most prompt sequence as an effective procedure for teaching chained responses. However, Horner and Keilitz considered a step to have been completed with “no help,” and thus differentially reinforced, independent responses and responses that occurred after the general prompt. Because true independent responding was not differentially reinforced, prompt dependence could have emerged at the general prompt. The purpose of this study is to evaluate whether there is a temporal difference between when a subject masters a daily living skill at the general prompt level versus at the independent level and whether prompt dependence at the general prompt can emerge under the reinforcement contingencies used by Horner and Keilitz. Initial results from six children with autism show that a majority of tasks were mastered at the general prompt before the independent level. Of the tasks that were mastered in the general prompt first, the average number of sessions between mastery at the general prompt and mastery at the independent level was 13.50. One subject was found to be prompt dependent at the general prompt for one task.
|The Efficacy and Efficiency of Error-Correction: An Examination of Dependent Measures During Instruction
|JESSI REIDY (Marquette University), Tiffany Kodak (Marquette University), Mary Halbur (Marquette University), Lauren Debertin (Marquette University), Alyssa P. Scott (Marquette University ), Courtney Lyn Meyerhofer (Marquette University), Xi'an Maya Williams (Marquette University), Marisa E. McKee (Marquette University)
|Abstract: Various error correction procedures have proven to be effective across individuals when correcting learner error. However, there are still inconsistent results concerning which error correction procedures lead to more efficacious and efficient acquisition (McGhan and Lerman, 2013). This may be due to the possibility of multiple sources of stimulus control present during error correction trials (Carroll, Joachim, St. Peter, & Robinson, 2015). Literature has shown that overt repeated responses between trials may inhibit learning, as the learner may be rehearsing the target response (Kodak, Campbell, Bergmann, LeBlanc, & Kurtz-Nelson, 2016). To prevent this rehearsal, tasks could be interspersed between trials; however, there is minimal empirical evidence on the efficacy of interspersed tasks between error-correction trials. Therefore, the goal of the current study was to compare the efficacy and efficiency of three commonly-used error correction procedures as compared to a control. Results indicated that the interspersal of either nonverbal, or verbal, tasks between error correction trials led to more efficacious and efficient acquisition for participants across both sets. However, the overt repeated responses and responses given without attending indicated mixed results. Future research could use different verbal operants and further investigate the relationship between overt repeated responses and independent correct responses.
Assessment and Treatment of Response to Name in Children With Autism
|DANIEL E CONINE (Georgia State University), Timothy R. Vollmer (University of Florida), Molly A Barlow (University of Florida), Cynthia Dela Rosa (Florida Autism Center), Abigail Petronelli (University of Florida; Florida Autism Center), Emma Grauerholz-Fisher (University of Florida)
Response to name (RTN) is an early developmental milestone, deficits in which are associated with autism spectrum disorder (ASD). The current study extends previous research on this target behavior by evaluating an abbreviated assessment and treatment model for RTN with thirteen children with ASD. In phase one, a naturalistic social baseline was conducted with all participants. In phase two, a series of treatment conditions involving the use of tangible reinforcement was evaluated with all children for whom RTN did not meet mastery criteria in phase one. In phase three, treatment components were removed, schedules of tangible reinforcement were thinned, and generalization to other people was assessed. Results indicate that tangible reinforcement procedures can produce rapid increases in discriminated response to name, sometimes without the addition of response prompts. The total number of trials to mastery was reduced in the current study relative to previous research. This study also investigates whether baseline patterns of RTN from phase one could be used to predict the treatment that was ultimately necessary in phase two. Results provide preliminary evidence to suggest that using this sort of baseline as a screening to predict treatment effects could further reduce trials to mastery.