About this applet
This applet allows scientists to calculate the sample size or power for a Micro-Randomized Trial (MRT) with a binary proximal outcome
How can a behavioral scientist use this applet?
Behavioral intervention scientists can use this applet to calculate the sample size for a Micro-Randomized Trial (MRT) with a binary proximal outcome given a desired power. Or they can use this to calculate the power given a sample size. For those who prefer to use R, the package MRTSampleSizeBinary implements the same sample size calculator as this Shiny applet.
What method does this applet implement?
This applet implements a sample size formula based on a test of the null hypothesis that there is no proximal effect (that is, relative risk is a constant 1 over time). The method incorporates a few small sample adjustments.
Related References
Eric R. Cohn, Tianchen Qian, Susan A. Murphy (2023). Sample Size Considerations for Micro-Randomized Trials with Binary Proximal Outcomes. Statistics in Medicine. https://doi.org/10.1002/sim.9748
The sample size calculation methodology is described in this paper.
Tianchen Qian, Eric R. Cohn, Susan A. Murphy (2021). Statistical designs for developing personalized mobile treatment interventions. In Digital Therapeutics: Scientific, Statistical, Clinical, and Regulatory Development Aspects, O. Sverdlov and J. van Dam, Eds. Chapman & Hall/CRC, 2021. https://doi.org/10.1201/9781003017288
This book chapter reviews the method and illustrates the use of the sample size calculator.