Training Observers in Class-Wide Behavior Data Collection Using Behavior Skills Training

Training individuals to conduct classroom observations typically requires practice in an analogue setting using one or a combination of strategies, including behavior vignettes, role-play, or video clips of target behaviors (Cooper, Heron, & Heward, 2007; Hartmann & Wood, 1990; Bass, 1987). However, researchers have yet to determine if these strategies are effective and to develop a standard approach to training observers. This investigation was a part of a larger study that looked at the effectiveness of tiered teacher training to facilitate implementation of an interdependent group contingency in the classroom. For this study, I utilized a concurrent multiple-baseline design to help determine the effectiveness of behavior skills training to teach graduate students to take classroom behavior and teacher fidelity data. Students were trained in an analogue setting using videos of elementary classrooms. Interobserver agreement (IOA) data were collected in both the training and the generalization setting (i.e., live classroom). Effectiveness of BST was measured via visual analysis, percentage of all non-overlapping data to determine an effect size, and a within-case comparison model to determine significance. Results indicate the BST framework was an effective approach to quickly training classroom observers in an analogue setting and training gains maintained into the generalization setting