Machine Learning to Measure Vocal Stereotypy: An Extension

Repeated measurement of behavior is a process central to behavior analysis, but its implementation occasionally requires hiring observers dedicated exclusively to data collection, which may increase the cost of providing services and conducting research. One potential solution to reduce resources necessary to conduct behavioral observations involves machine learning. Using data previously published by Dufour et al. (2020), we developed and tested novel models to automatically measure vocal stereotypy in eight children with autism. In addition to accuracy, we examined session-by-session correlation between values measured by machine learning and those recorded by a human observer. Nearly all our models produced correlations similar to those between continuous and discontinuous methods of measurements (i.e., .90 or more) and resulted in better metrics than those reported by Dufour et al. (2020). Although practitioners and researchers should continue examining their accuracy in measuring vocal stereotypy, the adoption of the proposed models may prove useful.