A single data bank, comprising over 60000 trips from sample household surveys in the reading area in 1962 and 1971, was used to compare two alternative types of trip generation model, one based on household trip rates and the other on person trip rates for each household. Their performance was found to be similar, each accounting for over 50 per cent of the variability in household trip rates, but the person trip rate model has been shown to be simpler to use and statistically more acceptable. The most important variables for modelling home-based trips were household size and car ownership in both types of model. Work trips required only household employment in a household rate model and car ownership in an employed person rate model. Household location and the year of study has a small but discernible effect on trip rates due to some reduction in the inner and middle area rates between the two years. Although too small to incorporate into the models, this effect illustrates the value of the statistical procedures for detecting potentially important changes in travel behaviour so that they can be monitored as further data become available. (Author/publisher)
Samenvatting