TY - JOUR TI - Evaluating teacher implementation of discontinuous data collection in the classroom DO - https://doi.org/doi:10.7282/T3P55RKN PY - 2017 AB - Discontinuous data collection procedures such as momentary time sampling (MTS) and partial interval recording (PIR) provide ABA practitioners with an alternative to tedious and oftentimes unfeasible continuous data collection. Discontinuous data is especially practical for classroom teachers who must collect behavioral data while also implementing instructional protocols. However, the existing literature on MTS and PIR come from simulated or controlled laboratory studies rather than applied settings. Furthermore, most studies focus on methodological error and do not consider human error in discontinuous data collection. The present study compared four discontinuous data collection procedures: 10-s MTS for 10 min, 30-s MTS for 30min, 10-s PIR for 10 min, and 30-s PIR for 30 min in a classroom setting using three teacher-student dyads. This study aimed to identify the procedure that had the least methodological and human error when used by teachers who were collecting duration data on stereotypy. Methodological error was measured by comparing teacher-collected estimates to duration data coded from video. Human error was quantified by calculating teachers’ treatment integrity (TI) of an instructional protocol and their interobserver agreement (IOA) for each discontinuous data collection method. In addition, this study compared the social validity of these procedures by examining teacher perceptions and preference. With regards to methodological error, results indicated that 10-s PIR, and especially 30-s PIR, significantly overestimated the occurrence of stereotypy, while both 10-s and 30-s MTS yielded very accurate estimates. All three teachers, however, erroneously perceived PIR to be more accurate than MTS. Results for human error were less conclusive, but indicated that these teachers could multitask while maintaining high TI and IOA. Lastly, findings from the present study suggest that the factors that affect preference are complex and vary across individuals. KW - Clinical Psychology LA - eng ER -