Using Simulation to Understand Mixed Models in the Wild

Real-world data present a number of challenges for data analysis. Simulation provides a way to examine the impact of the quirks of real-world data on your analysis. In this tutorial, I will introduce simulation with mixed models as a tool for planning your analysis (e.g., power analysis) and as a way to re-consider your plan after contact with real-world data (e.g., what impact does imbalance in my sample have on my inferences?). Simulation and real-world data analysis should not be rivals, but rather partners in inference.