Accuracy of Long Momentary Time-Sampling Intervals: Implications for Classroom Data Collection

Abstract

The accuracy of various momentary time-sampling (MTS) interval lengths in estimating continuous measures of behavior and the implications for classroom data collection procedures were examined. Continuous interval recording records were simulated to provide standards against which to assess MTS interval lengths of 30 seconds and 5, 10, and 20 minutes. Standards were constructed to simulate behavior that occurred during 20%, 40%, 60%, and 80% of the total intervals observed. Each MTS percent occurrence estimate was compared to the percent occurrence represented by the standard. The data revealed that short MTS interval lengths were more accurate and that level of percent occurrence did not differentially affect MTS estimates. The ability of MTS to reflect increasing and decreasing behavior trends and the interaction of MTS interval lengths with various percent occurrence levels also were examined.