Time Stability of Zonal Trip Production Models

Abstract

Using trip production information data gathered in the 1958 Pittsburgh Area Transportation Study and the 1967 Southwestern Pennsylvania Regional Planning Commission Transportation Study, zonal trip production models were constructed using socio-economic data compatible over both time periods. Both multiple regression and cross classification techniques were employed. Models were calibrated on the 1958 data, applied to 1967 observed productions and tested statistically for observable failure in predictive ability. Both types of models were found to be unstable over time. While only marginal differences were observed in the total trip production equation model, highly significant variations were observed in the case of the work, shopping and school purpose equations. Where auto ownership was used as an equation variable it was found to provide an underestimate of future trips. Variables found to overestimate trips included number of residents and residential acreage.