This paper attempts to serve two purposes with respect to trip generation analysis. Research findings concerning data variations and aggregative effects are outlined. Also, the implications of the continuing phase of transportation planning for trip generation methodology are discussed. Trip generation estimating procedures have generally relied heavily on statistical methodology. Such procedures require a sound knowledge of the structural relationships contained in the basic data set. Several suggestions are offered for the operational transportation planner concerning trip generation methods. By utilizing standard statistical techniques, it is possible to identify the sampling variations associated with zonal aggregate data and to illustrate the realistic level of accuracy to be expected from zonal estimating procedures. Other simple statistical procedures make explicit the effects of aggregating trip- making relationships which are, in actuality, most meaningful at the individual house-hold level. It is concluded that the typical 'fine tuning' of a multiple regression trip generation analysis with aggregate data may be of marginal value. The compelling logic of a disaggregate analysis of trip generation relationships is indicated. Trip generation estimating relationships are derived from cross- sectional data and are subject to change with time. These relationships must be evaluated for stability periodically. Little has been written concerning the trip generation phase of the continuing urban transportation planning process. Several of the existing and proposed procedures are particularly valuable as a basis for an efficient monitoring and reevaluation program. Three trip generation analysis procedures are suggested for use in the continuing planning phase. Rather than implementing an extensive resurvey, well- designed small sample cross-sections and on-site surveys appear to hold promise for a periodic reevaluation of basic trip generation estimating procedures. /author/.
Abstract