Bielby (1986) argues that conventional practices for normalizing latent variable models can lead the applied researcher astray. He presents two examples to show how this may occur. In the first example, he considers the regression of income on education. In the second, he considers a panel model...