|
Dinosaurs of Behavior Analysis: Interval Time-Sampling, Interobserver Agreement, and Linear Graphs |
Monday, May 28, 2018 |
10:00 AM–10:50 AM |
Manchester Grand Hyatt, Harbor Ballroom C |
Area: TBA/PCH; Domain: Applied Research |
CE Instructor: Douglas E. Kostewicz, Ph.D. |
Chair: Jessica E. Frieder (Western Michigan University) |
DOUGLAS E. KOSTEWICZ (University of Pittsburgh) |
RICHARD M. KUBINA (Penn State) |
DONNY NEWSOME (Fit Learning) |
Abstract: Counting, measurement quality, and displaying behavioral data often occurs via interval time-sampling, interobserver agreement, and linear graphing. Each practice has a long history within behavior analysis. However, each has evoked debate as to its usefulness. Interval time-sampling, for example, epitomizes a discontinuous time-based method of measurement. Adopting such a system of data collection leads to over or underestimating the occurrence of behavior. The most common metric to determine measurement quality calculates the agreement between two independent observers for the same phenomena. In fact, the measure establishes believability rather than quality between observations. Once measured and assessed, behavior data appear on linear graphs. Yet linear graphs visually display an absolute amount of change and do not offer important behavior change statistics.The panel will address questions about each of the previously mentioned measurement and data display practices. A discussion surrounding the weakness of each practice and potential solutions will occur. Audience members will have opportunities to interact with the panel. |
Instruction Level: Basic |
Target Audience: Practicing/Certified Behavior Analysts, Instructors of Behavior Analytic Content, Behavioral Researchers |
Learning Objectives: Learning Objective 1: Attendees will learn from panel presentation and discussion the weaknesses inherent in collecting data via interval time-sampling methods, using interobserver agreement, and displaying behavioral data on linear graphs. Learning Objective 2: Attendees will learn from panel presentation and discussion the potential pitfalls of employing the aforementioned practices. Learning Objective 3: Attendees will learn from panel presentation and discussion specific ways to more effectively measure behavioral data, establish measurement quality, and display behavioral data. |
Keyword(s): Interobserver Agreement, Interval Time-Sampling, Linear Graphs |
|
|