Analysis of the Revere, Quincy and Stamford Structure Data Bases for Predicting Building Material Distribution

摘要

Data bases on buildings in Revere and Quincy, Massachusetts, and Stamford, Connecticut, were studied to determine if a measure of building material distribution could be calculated for a city using land use, census tract and the Corps' data on buildings. Statistical measures of chi-square asymmetric lambda, uncertainty coefficient, F ordinate, as well as the R sub 2 and eta sub 2 statistics were calculated for the three data bases. The Corps definition of building type was found to be the best predictor of the building surface area. However, all indicators (including building type) explained only low percentages of the variability in the dependent variable (building surface area). Our results indicated that other variables are required to explain the variability of building surface area adequately.