Biostatistics, Epidemiology, and Research Design (BERD) Program – Robin Baudier, M.S.P.H. M.F.A.
 
If you’re struggling to find a power and sample size (PSS) calculation appropriate for your proposed analysis or data structure, Monte Carlo power simulations are an endlessly adaptable solution. In this seminar, we will cover how to simulate datasets that resemble your study’s data distributions, correlation structures, and type including repeated measures, correlation with covariates, and time to event. Then, we will cover how to use those simulated datasets with the analytic approach of your choice to find the sample and effect sizes needed to reach your power requirements.

Recommended background: This seminar builds on material from the introductory seminar in this series, PSS 101, and will recap those concepts only briefly (power, sample size, effect size, level, Type I & II errors). The recording and slides from the PSS 101 seminar are available on the OCTRI Research Forum webpage. Experience with R is recommended, as this seminar will use R code to simulate data, generate statistical models, create loops, and generate PSS estimates using Monte Carlo simulation models. Previous knowledge of Monte Carlo simulation methods is not required.
 
Questions? Contact Amy Laird: laird@ohsu.edu