Overview
In this lecture we look at transformations of variables for normality.
Readings
Readings: Christensen Chapter 15 and Hoff Chapter 9
Wakefield 6
In this lecture, we revist model assumptions, in particular normality and examine ways to transform models so that the normality assumption applies. We also consider nonlinear regression (models that are nonlinear in the parameters) for cases where the model is intrinsically linear (can be converted to a model that is linear under a reparameterization with additive error and intrinsically nonlinear where it is not possible to transform it, in which case nonlinear least squares is needed to obtain MLEs.