|Analyses of Equivalence-Based Instruction Using Three Different Training Structures
|Monday, May 28, 2018
|8:00 AM–8:50 AM
|Manchester Grand Hyatt, Seaport Ballroom H
|Area: AUT; Domain: Translational
|Chair: Colleen Yorlets (RCS Behavioral & Educational Consulting; Simmons College)
|CE Instructor: Christina M. King, Ph.D.
Equivalence-based instruction has utilized a one to many, many to one, and linear series training structure, with varying degrees of effectiveness. These three studies demonstrate the efficacy of equivalence-based instruction using three different training structures. In Experiment One, a participant diagnosed with autism will be taught to sort a variety of physically dissimilar items, followed by visual-visual match-to-sample training. It is expected that posttests will demonstrate the formation of generalized equivalence classes and generalization of money skills to the natural setting. Experiment Two will assess for the emergence of selection and topography-based verbal and non-verbal behavior in two children diagnosed with autism. Visual-visual conditional discrimination training and tact training will be conducted through an equivalence-based format. It is hypothesized that nine additional relations will be demonstrated following the training of three relations for each stimulus class. Experiment Three demonstrated that learners emitted substantially more errors and formed fewer equivalence classes with a trial-and-error protocol compared to an errorless learning protocol. It is anticipated that these results will be replicated when the reinforcement density is kept constant across both training conditions. These three experiments will expand upon the existing equivalence research through the use of different training structures.
|Instruction Level: Intermediate
|Keyword(s): errorless learning, stimulus equivalence, verbal behavior
This presentation is appropriate for behavior analysts of an intermediate and advanced skill level.
Emergent Coin Relations and Stimulus Generalization Following Conditional Discrimination Training
|MEGAN BREAULT (RCS Learning Center; Simmons College), Christina M. King (RCS Learning Center; Simmons College), Colleen Yorlets (RCS Behavioral & Educational Consulting; Simmons College), Russell W. Maguire (Simmons College)
Equivalence based instruction has been demonstrated to be an efficient strategy for teaching a variety of individuals functional money skills; however, the generalization of the emerged relations in the natural environment has yet to be assessed. Several equivalence-based studies, conducted in laboratory settings, have utilized a variety of pictures of the stimulus class members during conditional discrimination training to form generalized equivalence classes. The purpose of the current study is to demonstrate the emergence of a minimal generalized equivalence class in an applied setting. In the current study a 14 year-old boy, diagnosed with autism, will be taught to sort a variety of physically different items (C) that can be purchased at a school store based on price. Followed by training the participant to match coins (B) to their corresponding written values (A) and items that can be purchased in a school store (C) to their corresponding assigned coin values (B). After acquisition of trained relations, all tests for a minimal generalized equivalence class and generalization probes of purchasing a variety of items in a school store will be conducted. These data will be discussed in terms of maximizing student learning and programming for stimulus generalization during conditional discrimination training.
Categorization and the Emergence of Selection and Topography-Based Verbal and Non-Verbal Behavior
|CHRISTINA M. KING (RCS Learning Center; Simmons College), Colleen Yorlets (RCS Behavioral & Educational Consulting; Simmons College), Megan Breault (RCS Learning Center; Simmons College), Lauren Donovan (RCS Learning Center), Jessica Byrne (RCS Learning Center), Russell W. Maguire (Simmons College)
Teaching children with autism to select members of a class by category name (e.g. selecting drum in the presence of the spoken word instrument), tact the class of a stimulus (e.g. saying furniture when shown a chair or bed), and match members within a class to one another (e.g. fork to knife; guitar to piano) are three skills that are often addressed in language acquisition programming. The applied literature, however, lacks evidence of participants demonstrating this type of class formation, as well as efficient teaching procedures to produce. The purpose of this study is to assess the efficiency and efficacy of training one arbitrary visual-visual conditional discrimination (D-B) and two tacts (B-Name and C-Name) and then testing for the emergence of nine additional untrained relations: tacting by class name (D-Name), selecting members of the class in the presence of the auditory stimulus (A-B, A-C, & A-D) and arbitrarily matching class members to one another (C-B, B-D, C-D, D-C). The participants included two children diagnosed with autism spectrum disorders. It is expected that the results of this study will demonstrate the emergence of these nine untrained relations across three stimulus classes, with only three directly trained relations.
|Comparison of an Errorless Learning to a Trial-and-Error Protocol on Equivalence Class Formation
|Russell W. Maguire (Simmons College), Megan Breault (RCS Learning Center; Simmons College), COLLEEN YORLETS (RCS Behavioral & Educational Consulting; Simmons College), Christina M. King (RCS Learning Center; Simmons College)
|Abstract: Errors emitted during instruction pose a number of potential risks, particularly for learners with developmental disabilities. While it is preferable to utilize errorless protocols to minimize the occurrence of errors, practitioners often rely on traditional trial and error protocols. Experiment One compared the effects of errorless versus trial-and-error protocols to form equivalence classes via conditional discrimination training. Participants 1 and 2 emitted errors, on average, during 73% of trials in the trial and error training condition. They emitted errors for an average of 5% of trials within the errorless learning condition. Participant 1 formed 4 of 9 equivalence classes in the errorless condition and 3 of 9 classes in the trial and error condition. Participant 2 formed 9 of 9 equivalence classes in the errorless condition and 2 of 9 classes in the trial and error condition. Participant 3 completed only the errorless learning condition and formed 9 of 9 equivalence classes. The effects of density of reinforcement on equivalence class formation will be further evaluated within Experiment Two. Errorless and trial-and-error protocols will be compared for Participants 1 and 2, while holding the density of reinforcement constant across both protocols. This change in protocol from Experiment One will allow for evaluation of the effects of errors on skill acquisition while eliminating reinforcement density as a variable between errorless and trial and error protocols.