Topic: 8.1 Transport Planning: General

For external validity, SP experiments need to present choices that are realistic for the respondents. The use of portable computers in interviewing has many advantages.

This article examines methods of predicting numbers of trips made by car and by public transport, and shows how closely the model is influenced by the size and composition of the household and by its income.

This paper models the endogenous frequency of shopping in a linear town, where the costs of travel and parking are affected by the degree of congestion. A simple model of rival shopping centres is also used to illustrate trip diversion and trip generation, when the parking provision of one centre is improved.

Urban transport is modelled so that prices and investment levels maximise a social welfare function. First-order optimality equations are derived.

The empirical model fuses the shopping destination choices made by individuals with shopping expenditure decisions. The joint estimation of discrete-continuous choices, in contrast to sequential estimation, has wide applicability to many problems.

This paper constructs a model of shopping behaviour in which the frequency of shopping is endogenous.

The paper aims to link chaos theory to spatial interaction analysis. Lags are incorporated. It is shown that unstable systems behaviour may emerge for particular lag values. The theoretical analysis is illustrated by means of various simulation experiments.

The Negative Binomial distribution, a generalisation of the Poisson distribution, is found to be useful in constructing a statistical model of trip frequency. It indicates that expenditure is a more important variable than the number of individuals in the household.

Choice of functional form for the demand (forecasting) model is very important for empirical research. Of the five models examined, the translog demand system with theoretical constraints imposed upon the parameters performs the best. But further investigation is needed.

Income and car ownership affect the demand for all three modes (car, public transport and bicycle), and these are all interrelated. Public transport demand is lowest at the middle of the three income categories. Further research is suggested.