How to use Simulation Methods to Conduct Power Analysis

The frequency with which applications of multivariate procedures, such as structural equation modeling or multilevel models, in the research literature and federal grant proposals has increased over the last decade. Furthermore, applied researchers often collect data for which model assumptions are not tenable, have missing values, or otherwise require challenging decisions. Determining the target sample size when planning studies that will use advanced statistical modeling can be difficult because existing power analysis programs simply cannot accommodate these complexities. In this workshop, we demonstrate how to use simulation methods in R to conduct power analyses for advanced but commonly used procedures.