Comparison of direct observational methods for measuring stereotypic behavior in children with autism spectrum disorders.

We compared partial-interval recording (PIR) and momentary time sampling (MTS) estimates against continuous measures of the actual durations of stereotypic behavior in young children with autism or pervasive developmental disorder-not otherwise specified. Twenty-two videotaped samples of stereotypy were scored using a low-tech duration recording method, and relative durations (i.e., proportions of observation periods consumed by stereotypy) were calculated. Then 10, 20, and 30s MTS and 10s PIR estimates of relative durations were derived from the raw duration data. Across all samples, PIR was found to grossly overestimate the relative duration of stereotypy. Momentary time sampling both over- and under-estimated the relative duration of stereotypy, but with much smaller errors than PIR (Experiment 1). These results were replicated across 27 samples of low, moderate and high levels of stereotypy (Experiment 2).