This article introduces a method of improving annual regression estimates of income and expense by using historical information to extract latent effects. Historical information is incorporated in a panel data structure, which gives access to the class of latent variable models that estimate and test the presence of group effects. The method suggests that inclusion of latent effect estimates improves income and expense predictions for filers and nonfilers.
Multiple regression analysis
Quintos, C. (2015). Estimating latent effects in commercial property models. Journal of Property Tax Assessment & Administration, 12(2), 37-66. Retrieved from https://researchexchange.iaao.org/jptaa/vol12/iss2/3