|The Assessment and Application of Momentary Time Sampling and Partial Interval Recording in Classroom Settings
|Tuesday, May 27, 2008
|12:00 PM–1:20 PM
|Area: AUT; Domain: Applied Research
|Chair: Kate E. Fiske Massey (Rutgers University)
|Discussant: Rebecca P. F. MacDonald (The New England Center for Children)
|CE Instructor: Lara M. Delmolino Gatley, Ph.D.
Direct observational data recording methods are utilized widely in ABA instructional settings, and interval sampling methods such as partial interval recording (PIR) and momentary time sampling (MTS) are commonly used when behavior occurs at high frequencies and/or have ambiguity in their onset and offset. Gardenier, MacDonald, and Green (2004) compared the use of PIR and MTS in data collection for stereotypy exhibited by children with autism in an assessment setting, and found that MTS often produced a more accurate measure of the true occurrence of stereotypy than did PIR. To replicate and extend this finding, we conducted a series of studies to further assess the accuracy of MTS and PIR measurement for collecting data on stereotypy, and also to evaluate the application of these data collection methods to the classroom setting. Understanding that a careful balance of accuracy and applicability is desirable for the selection of a data collection method, these studies, taken together, inform the selection and implementation of data collection methods in the classroom setting.
|Comparing the Use of Momentary Time Sampling and Partial Interval Recording in Stereotypy Data Collection.
|LARA M. DELMOLINO GATLEY (Rutgers University), Kate E. Fiske Massey (Rutgers University), Melissa Dackis (Rutgers University)
|Abstract: In their research, Gardenier, MacDonald, and Green (2004) concluded that momentary time sampling (MTS) was more accurate than partial interval recording (PIR) when collecting data on the stereotypy exhibited by children with autism. The authors reported PIR grossly overestimated the occurrence of behavior in all students. To replicate this finding and begin to evaluate its utility in a classroom setting, we compared the use of MTS and PIR for collecting data on the stereotypy exhibited by 10 children with autism in the classroom setting.
Three 10-minute classroom observations were collected for students aged 3 to 20 years. Duration data was collected for stereotypy during each videotaped sample, and we collected PIR data and MTS using 10- to 300-second intervals using these same videotaped samples. Results indicated that, for each student, PIR data collection methods significantly overestimated the occurrence of stereotypy, even at the shortest interval. MTS resulted in data consistent with that collected using duration coding, and was often accurate using intervals as long as 60 seconds.
|Calibration of Stereotypy Data Collection Methods Based on Frequency of Behavior and Episode Length.
|SUZANNAH J. FERRAIOLI (Rutgers University), Kate E. Fiske Massey (Rutgers University), Lara M. Delmolino Gatley (Rutgers University)
|Abstract: Collecting data on measurable outcomes is a basic tenet of the development of behavioral interventions, and when data collection methods are selected without sufficient consideration of the frequency of a student’s behavior, even the most effective interventions may be compromised. Momentary time sampling (MTS) and partial interval recording (PIR) are methods frequently used to collect data on stereotypy in classroom settings, and may require calibration based on the frequency of a student’s stereotypy and the average duration of episode to increase accurate evaluation of intervention outcome.
Three 10-minute classroom observations were collected for students aged 3 to 20 years. Duration data was collected for stereotypy during each videotaped sample, and we collected PIR data and MTS using 10- to 300-second intervals using these same videotaped samples. Students will be grouped into high- and low-frequency stereotypy groups and long- and short-episode groups based on the occurrence of behavior recorded during observation. Examining the MTS and PIR intervals best fit to each student will indicate which data collection methods and interval lengths might be most appropriate for students based on the total frequency and episode length of exhibited stereotypy. Implications for calibrating data collection methods in the classroom environment will be discussed.
|Teacher Perceptions of and Accuracy in Data Collection Using Momentary Time Sampling and Partial Interval Recording.
|KATE E. FISKE MASSEY (Rutgers University), Lara M. Delmolino Gatley (Rutgers University)
|Abstract: Though research indicates that momentary time sampling (MTS) is more accurate than partial interval data (PIR) in collecting data on stereotypy, evaluating teacher perceptions of the utility of these data collection methods may be helpful in ensuring accurate implementation of these methods in the classroom setting. Additionally, the accuracy with which teachers collect MTS data as compared to PIR data has not yet been determined, and could be integral to implementing these methods in the classroom environment.
We collected two 10-minute classroom observations of 10 students with autism aged 3 to 20 years. For each student, a classroom teacher collected data on stereotypy using MTS and PIR at an interval length calibrated to the students’ frequency of behavior. Teachers were then asked to rate the ease and accuracy with which they were able to collect data using each method. Results indicate that, on average, teachers rated MTS as easier to use but less accurate than PIR. In contrast, comparison to duration data indicated that MTS was more accurate than PIR, though teacher error was high using each method. Implications for the use of these methods in classroom settings will be addressed.