Measurement error: Models, identification and inference

This class will review approaches aimed at eliminating biases due to the presence of measurement error in covariates. Special emphasis will be given to consistent methods that circumvent distributional assumptions and the need for validation data. Both classical and nonclassical measurement error models will be considered. We will discuss a recent approach to latent variable problems, including measurement error. The class will conclude with examples of the use of measurement error techniques in related fields, such as panel data, auctions and factor models.