BL16 Applications

This lecture is in two parts.

Part 1: Summarizes the critical calculations required for inference for a sample of IID Normal Data. This is both for conventional inference and for Bayesian Inference. It is shown that conventional inference is a special case of the Bayesian inference when we use and un-informative prior.

Part 2: Explain how this theory is applied to real data sets. The KEY is to search for data structures which can reasonably be modeled as IID Normals. Most economic time series DO NOT fit this assumption directly. How do we recognize match and mis-match and how we can use transformations of the data to achieve a better match?