Calculating effect sizes for meta-analysis: the case of the single case

Area of Science:

• Behavioral research methods
• Quantitative psychology
• Educational research

Background:

• Interrupted time-series (ITS) designs are crucial for single-case research.
• Existing methods for analyzing ITS data have limitations in accurately capturing treatment effects.
• Controlling for baseline trends is essential for valid effect estimation.

Purpose of the Study:

• To review and improve methods for deriving effect measures in single-case interrupted time-series designs.
• To address limitations of prior statistical models used in single-case research.
• To provide a practical, step-by-step algorithm for enhanced data analysis.

Main Methods:

• Review of existing regression-based approaches for interrupted time-series analysis.
• Modification of the Center, Skiba, and Casey (1985-86) regression model.
• Development of a computational algorithm focusing trend estimation on baseline data.
• Application of the refined method across various single-case design implementations.

Main Results:

• The proposed modifications allow simultaneous accounting for treatment effects on level and slope.
• The method effectively controls for the influence of pre-existing trends in the data.
• A clear, step-by-step algorithm facilitates the implementation of the improved analytical technique.
• The approach is adaptable to diverse single-case research designs.

Conclusions:

• The enhanced regression approach provides more accurate effect size measures for single-case interrupted time-series designs.
• Accurate trend control using baseline data improves the validity of treatment effect estimations.
• The provided algorithm simplifies the application of advanced statistical methods in single-case research.
• This methodology offers a robust tool for researchers evaluating interventions in individual cases.