Using MLS data for hedonic price modeling: An experiential learning activity

Abstract

Multiple Listing Service (MLS) data are notoriously cumbersome for empirical analysis due to inaccuracies and incompleteness. Applied real estate market analysis using MLS data requires a diligent effort to identify and address data limitations. This paper describes a learning activity that provides students with an opportunity to work with a large, real-world MLS dataset to answer research questions about house price determinants using hedonic price modeling with OLS regression. The activity is designed for use with Microsoft Excel due to its ready availability. Available instructor resources include a dataset with 20,126 residential listings from a well-defined market covering a three-year time period with 85 data fields, directions for students, and a model answer.