Week 8

Gregory Hancock — Professor and Distinguished Scholar-Teacher, University of Maryland

Power Analysis Has Lost Its Way: New Methods To Bring It Back Home

In a time when the alarms of research replicability are sounding louder than ever, the process of a priori power analysis has devolved from its intended origins as an honest, planful insurance policy for our scientific endeavors to a collection of near-delusional yard sale hagglings whose goal is slinking away as cheaply as possible. The current presentation discusses on-going work, with Yi Feng of UCLA, to derive new methods that attempt (1) to restore appropriate conservatism and robustness to the sample size planning process, and (2) to greatly simplify that process. Derivations and suggestions for practice will be presented using the framework of structural equation modeling.