Power of a randomization test in a single case multiple baseline AB design.

Area of Science:

• Behavioral research methods
• Single-case experimental designs
• Statistical analysis

Background:

• Multiple baseline designs are common in single-case research.
• Visual inspection is the traditional analysis method.
• Statistical hypothesis testing can complement visual analysis.

Purpose of the Study:

• To investigate the statistical power of randomization tests in multiple baseline designs.
• To identify factors influencing the power of randomization tests.

Main Methods:

• A crossed factor simulation study was conducted.
• The study simulated various parameters of multiple baseline designs.
• The power of randomization tests was evaluated under different conditions.

Main Results:

• Autocorrelation, number of participants, effect size, intervention start overlap, baseline-to-intervention ratio, effect emergence, and measurement count significantly impact power.
• Interactions between participant number and effect size, and measurement count and intervention start moments, also strongly influence power.

Conclusions:

• Randomization tests are a viable statistical tool for multiple baseline designs.
• Understanding design characteristics is crucial for maximizing statistical power.
• An online tool was developed to aid in power calculations for these designs.