An Examination of Linear and Ratio Graphs (Richard M. Kubina Jr.)

Line graphs and visual analysis have served as the engine of evaluation and decision making for applied behavior analysis. The advantages of line graphs include providing treatment data visually and summarizing a person’s performance across a given time interval. Line graphs also communicate the sequence of treatments, the time spent in treatment phases, and the magnitude of behavioral change. Yet visual analysis has a history of low interrater reliability and inconsistent judgments. Other criticisms include the lack of universal decision rule and the lack of any meaningful statistics. The solution to all of the previously listed problems may lie in a standard ratio graph. The following experiment examines the extent to which behavior analysts could accurately detect a trend and reliability make a decision based on three conditions: a linear graph with a trend, a linear graph with a quantified slope, and a ratio graph with a celeration value. The results and implications of the study suggest a healthy path forward for visual analysis and the analysis, evaluation, and communication of data via ratio graphs.
hide abstract
Richard M. Kubina Jr.
Presented at the 2021 Michigan Autism Conference
show abstract
×
This is the open-access version of this video and your completion of it is not recorded. It also does not contain embedded questions or interactions like the CEU and coursepack versions. If you purchased the CEU or coursepack version, you should access it from the myLearning dashboard.
Shopping cart
Your cart is empty.