Clinical decision making : comparison of data presentation methods

Document Type

Thesis

College

College of Arts and Sciences

Department

Psychology

Degree

M.S. Applied Behavior Analysis

Date Completed

2017

First Committee Member

Gould, Daniel

Second Committee Member

Karsina, Allen

Third Committee Member

Thompson, Rachel

Abstract

"Visual inspection of data displays is commonly used by behavior analysts. The data presentation method can affect decision making. In this study, cumulative record and latency graphs were evaluated by 13 behavior analysts. In Experiment 1, decision making was assessed through the number of sessions viewed when data were presented point-by-point as cumulative record versus latency graphs for a Test and a Control condition constructed from the same data sets. Experiment 2 assessed whether participants made the same data interpretation when viewing the complete data set as when viewing the partial data sets in Experiment 1. Results showed that in Experiment 1, participants viewed the lowest number of data points for the low variability latency graph; average number of data points viewed for the 3 other graphs were within +/- 1 point of each other. In Experiment 2, participants’ determination of the effect of the independent variable varied by graph type, with opposite conclusions reached for the cumulative record and low variability latency graphs. Comparison of conclusions reached across experiments showed that participants reached the same conclusion when viewing full data sets in Experiment 2 as they had when viewing partial data sets in Experiment 1. Differences in conclusions occasioned by the different graph types suggest that behavior analysts should use multiple methods of visual inspection of data."

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

Share

COinS