Quantification of behavioral data with effect sizes and statistical significance tests

This article describes the use of statistical significance tests and distance‐based effect sizes with behavioral data from single case experimental designs (SCEDs). Such data often are interpreted only with visual analysis. However, a growing movement in the field is to quantify results to improve d...

Full description

Saved in:
Bibliographic Details
Published in:Journal of applied behavior analysis 2022-10, Vol.55 (4), p.1068-1082
Main Authors: Costello, Mack S., Bagley, Raymond F., Fernández Bustamante, Laura, Deochand, Neil
Format: Article
Language:English
Subjects:
Applied behavior analysis
behavior analysis
Design of experiments
effect size
single‐case experimental design
single‐case research
Statistical significance
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This article describes the use of statistical significance tests and distance‐based effect sizes with behavioral data from single case experimental designs (SCEDs). Such data often are interpreted only with visual analysis. However, a growing movement in the field is to quantify results to improve decision‐making and communication across studies and sciences. The goal of the present study was to assess the agreement between visual analysis and various statistical tests. We recruited visual analysts to judge 160 pairwise data sets from published articles and compared these analyses to significance tests and effect sizes. One‐tailed significance testing of Tau z and the percentage of pairwise differences in the predicted direction (PWD) generally agreed with each other, and complemented the effect sizes of Ratio of Distances (RD) and g. Visual analysis was somewhat unreliable and should be combined with statistical complements to maximize decision accuracy.
ISSN:0021-8855
1938-3703
DOI:10.1002/jaba.938