Association for Behavior Analysis International

A common characteristic of acquired brain injury (ABI) are deficits of social cognition. These deficits can affect the lives of people with brain injury in negative ways, such as damaging social circles or limiting the person's ability to receive therapy due to aggression. Previous research has demonstrated the stimulus equivalence approaches are effective in training the detection of facial emotions, yet this literature base is narrow. Given that the literature does show promise in the area, the current study evaluated the feasibility of the PEAK-Equivalence (PEAK-E) Relational Training System to improve social cognition deficits. PEAK-E training programs that targeted a variety of social cognition deficits were presented to three participants. The programs focused on recognizing and interpreting facial expressions, relating feelings to contexts, and understanding the emotions of others. A multiple baseline design, across programs and participants, was used with a pre and post-test measure of social cognition performance; The Awareness of Social Inference Test short version (TASIT-S). Results, future research and clinical implications are discussed with regards to social cognition, mastery issues of derived relations during PEAK-E training, and the effectiveness PEAK-E training to generalize to performance on the TASIT-S.

A bidirectional naming repertoire consists of responding both as speaker and listener and is demonstrated when one is only taught a speaker repertoire and can then derive listener behavior without direct training and vice versa. Bidirectional naming is foundational to complex human language but many children with autism and other developmental delays may not develop this repertoire without proper instruction. A substantial amount of research by Douglas Greer’s research group has found that multiple exemplar instruction (MEI) is effective in establishing bidirectional naming in young children but very little research has attempted to replicate this finding outside of that particular research group. The purpose of this study is to evaluate MEI for establishing bidirectional naming in children with autism in a community-based autism clinic. We are using a multiple baseline design across participants to evaluate the effects of MEI on the emergence of bidirectional naming in young children with autism. The procedure is currently underway with the first participant and acquisition data are positive thus far. Results will be discussed in terms of real-life replication of university-based research in community-based settings.

Several behavioral analytic researchers report that one of the main features of a good instructional system is “it must be effective in helping students learn more rapidly than they would on their own” (Barrett et al., 1991, p.80). Since the 1980’s, there has been an elaborate amount of research on stimulus equivalence protocols (SEP) in teaching reading, spelling (De Souza & Rehfeldt, 2013), grammar (Connell, 2004), and complex academic skills with individuals with and without disabilities (LeBlanc, Miguel, Cummings, Goldsmith, & Carr; 2003; Lovett, Rehfeldt, & Dunning, 2011). A study conducted by O’Neill et al. (2015) compared an online SEP to an assigned reading study method to teach Skinner’s (1957) verbal operants to college students. O’Neill et al. found the online SEP group outperformed the assigned reading group by 10 percentage points (one full letter-grade difference). The current study aims to systematically replicate the O’Neill et al. (2015) study by examining the effectiveness of this SEP for teaching undergraduate college students to identify and understand the verbal operants. Current pilot data indicates the SEP was effective in deriving multiple untrained selection-based and topography-based relations across multiple verbal operants.

Reading and writing skills may be conceived as a network of equivalence relations between stimuli (e.g., printed words, dictated words, and pictures) and stimuli and responses (e.g., naming, writing, etc.). This conceptualization has been foundational to the development of evaluation tools and teaching programs. The Reading and Writing Network Assessment (Avaliação da Rede de Leitura e Escrita [ARLE]) is an online 15-task instrument that evaluates matching-to-sample, naming, reading, and writing skills. The present study sought to describe the empirical network of relations, measured by the ARLE, that characterizes performance in beginning readers. The records of 2388 students, 6 to 12 years old, were assessed through an online platform. All of the tests used the R programming language. Generalized additive models (GAMs) were used to assess potential relations between skills using the MGCV package. Generalized cross validation was used to calculate variable importance in multiple regression models using the Caret package. After calculating the importance of each variable, a single model was constructed using all variables with the GAM. The least important variables were progressively removed until all remaining variables were statistically significant, evaluated by p values. The graphs were constructed using the ggplot2 package. The data analysis showed that all of the skills were significantly related to reading and writing measures. Matching printed words to dictated words and naming consonants were strongly related to reading. Matching printed words to pictures was the skill that was most related to spelling. An index that was created to compare the network of relations for students with different repertoires indicated increasing integration between skills as the repertoire evolved, which was predicted by the stimulus equivalence paradigm. The integration index may be a useful tool for the prediction and control of effects of teaching procedures that seek to establish reading and writing in non-readers.