|
| Effect Sizes for Single Case Experimental Designs: Demonstrating an Undesirable Baseline Trend |
| Tuesday, November 14, 2017 |
| 11:30 AM–1:30 PM |
| Studio F, Niveau 2 |
| Area: EAB/EDC; Domain: Theory |
| CE Instructor: Kimberly Vannest, Ph.D. |
| KIMBERLY VANNEST (Texas A&M University) |
| Description: Overlap in Visual Analysis is well established and understood by the field, less widely accepted but gaining momentum is the addition of effect sizes to the interpretation of data. Effects size calculations can enhance analysis particularly when undesirable baseline trend or the variability characteristic of applied settings is present. Non-overlap effect sizes such as Tau U provide strong statistical power, control trend in a moderate way, data are directly interpretable, respect design features, and the calculations are easy, even by hand with short data series. This session will train behavior analysts to use effect sizes to enhance interpretation and allow behavior analyst remain in control of data analysis and interpretation. |
| Learning Objectives: At the conclusion of the workshop, participants will be able to: (1) successfully identify difference between desirable or undesireable trend and lack of experimental control; (2) use two methods to calculated and adjust for trend in experimental data; (3) discuss implications of trended data in individual studies and meta-analytic research. |
| Activities: This two hour work shop will include brief lecture, demonstration, practice opportunities and discussion. |
| Audience: Graduate students, Scholars, Researchers, Clinicians in need of defensibility in practice. |
| Content Area: Methodology |
| Instruction Level: Intermediate |
| Keyword(s): Date Interpretation, Effect Sizes, Visual Analysis |