Evaluation of behavior sampling procedures for event recording

Document Type

Thesis

College

College of Arts and Sciences

Department

Psychology

Degree

M.S. Applied Behavior Analysis

Date Completed

2016

First Committee Member

Graff, Richard B.

Second Committee Member

Bourret, Jason C.

Third Committee Member

Hanley, Gregory P.

Abstract

"One of the defining characteristics of applied behavior analysis is direct observation and repeated measurement of behavior. Obtaining a complete record of behavior may not be possible in some settings, and although behavior analysts use behavior sampling procedures that are thought to be representative of behavior that is measured continuously, there has been little research suggesting how behavior sampling procedures could be used with event recording. In this study, post-hoc data were collected from eight individuals with autism spectrum disorders. The number of target responses per week for nine different measurement periods were compared to continuous data collection. Pearson-product correlation coefficients were calculated between continuous data recording and each measurement period, and the optimal measurement period (i.e., the smallest measurement period that had a correlation coefficient that was significant at the 0.001 level) was identified for each individual. Twenty-eight Board Certified Behavior Analysts® then evaluated data from the optimal measurement periods using visual inspection. The results of the study suggested that continuous and observational samples of behavior were highly correlated, and visual inspection of graphs suggested that using data from shorter measurement periods did not result in any significant loss of information that would be used for clinical decision-making."

This document is available upon request to Western New England University faculty, students, and staff. Please contact D'Amour Library at dref@wne.edu for access.

Share

COinS