Association for Behavior Analysis International

The three talks in this symposium describe basic research studies in equivalence-based instruction (EBI) in which different instructional variables were investigated. The first study examined the effects of lecture, interdependent group contingency EBI, and individualized EBI to teach equivalence classes of abstract stimuli to elementary-school-age children. Results showed that the interdependent group contingency EBI yielded better results in less time than individualized EBI and lecture. The second study compared effects of two types of compound configurations for incorrect (negative) comparisons during equivalence class formation using abstract compound stimuli with adults. Negative comparison compounds contained either stimuli from only one class or from both classes. Results showed that class formation was high and equally likely for both groups. The third study compared video-based observational learning (OL) to traditional match-to-sample (MTS) to establish equivalence classes of abstract stimuli with adults. OL training used video models depicting progressive increases in correct responding (OL-Mixed) or 100% correct responding (OL-Expert). Results showed that OL-Mixed training resulted in the highest yield of equivalence classes when compared to traditional MTS and the OL-Expert training. Collectively, these studies further our understanding of procedures that can be used to facilitate equivalence class formation.

Intermediate Attendees should have a working knowledge of what equivalence classes are.

In a previous study, Ostrosky et al. (2022) compared the effects of equivalence-based instruction (EBI) with an interdependent group contingency, individualized EBI (no group contingency), and a PowerPoint lecture to teach equivalence classes with adults. All three effectively established classes, but written multiple-choice tests performances for the two EBI groups were significantly higher than for the lecture group, and the group-contingency EBI was significantly more effective for a fill-in written test than was lecture. The present study extended Ostrosky et al. by evaluating the effectiveness of EBI using three group-based instructional models. The results compared lecture-based instruction with Interdependent Group Contingency and Individualized EBI with 10- and 11-year-old children in an elementary school. The study used a multiple baseline design across three groups and included pre-assessment, EBI training, testing, lecture instruction (control), and post-assessment. Training used a linear class structure with a simple-to-complex training/testing format using match-to-sample tasks on Quizizz.com. Findings confirmed EBI's superiority over lectures in improving test performance, with Interdependent EBI yielding better results in less time than Individualized EBI or lecture instruction.

In a recent study, Colasurdo et al. (2022) taught equivalence classes in which the samples and comparisons were compound stimuli (i.e., pairs of abstract symbols). One potential problem when presenting compound stimuli during the establishment of equivalence classes, however, is that one element may overshadow the other, thereby interfering with class formation. To address this, the current study replicated and extended Colasurdo et al. by using two different types of compound configurations for incorrect (negative) comparisons during MTS. One group of participants was exposed to negative comparison compounds containing stimuli from only one class (referred to as Type 1 trials) while the other group was exposed to negative comparison compounds containing stimuli from both classes (Type 2 trials) along with Type 1 trials. Class formation was assessed with a test block containing single-element-to-single-element, single-element-to-compound, compound-to-single element, and compound-to-compound derived relations trials. A card sorting test with single elements was also conducted. Results showed that class formation was high and equally likely for both groups; therefore, overshadowing did not occur for either group. Thus, Type 2 trials may be unnecessary during class formation with compound stimuli, and their omission would likely increase efficiency of instruction.

The current study compared the efficacy and efficiency of video-based observational learning (OL) to traditional match-to-sample (MTS) training to establish three 3-member equivalence classes of abstract visual stimuli with adult participants. Video models used during OL depicted a first-person perspective of a computer screen with modeled on-screen movements of a mouse cursor making comparison selections. Participants were randomly assigned to three different groups. Participants in the OL-Mixed group received training in which the video models depicted progressive increases in correct responding of modeled baseline relations across five blocks of trials. Participants in the OL-Expert group received identical training except that the models always depicted 100% correct responding across the five modeled trial blocks. Participants in the MTS group actively responded while completing five blocks of traditional MTS training. Posttest results showed that OL-Mixed training resulted in the highest mean percentage of class-consistent responding and yield of equivalence classes when compared to traditional MTS and the OL-Expert group, despite minimal differences in total duration during training. Across the groups, nearly all who successfully demonstrated equivalence class formation during the computerized posttest also showed generalization of class-consistent during the card sorting posttest. Implications of these results are discussed.