On the effect size and power analysis in single-case experimental research

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K17267
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Educational psychology
• Research Institution: Joetsu University of Education
• Principal Investigator: Okumura Taichi 上越教育大学, 大学院学校教育研究科, 准教授 (90547035)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: ABデザイン

Outline of Final Research Achievements

Based on the standardized effect size by Hedges et al. (2012), this study investigated a standardized effect size which evaluates the "pure" treatment effect by excluding the a natural linear trend over both the baseline and treatment phases if it exists.
Under some reasonable constraints, this study shows that a treatment effect corresponds to the interaction between time and treatment, which can be standardized by the square root of the sum of inter- and intra-individual variances.
Through this study, a new perspective on the statistical power analysis was also proposed; that is, the larger population effect size a researcher assumes, the wider confidence interval the researcher allows. This perspective is significant when we shed a light on the power analysis by comparing it with the precision analysis.

Free Research Field

心理統計学