Strong inferences require ample data collected from well-planned studies. Data collected from such studies contain information to determine which plausible statistical model is most appropriate and provides good power to detect real effects.  The generalized linear mixed models framework provides a wealth of models (whether the data are measured on a continuous scale or are discrete counts), and are suitable for a wide variety of experimental designs, including those with random effects. We will explain concepts underlying generalized linear mixed models and discuss approaches for powering experiments so that inferences can be made with confidence. Examples using SAS and R will be provided.