Analysis of Household Survey Sample Size in Trip Modelling Process

Abstract

Large cities are developing sophisticated trip models that consist of demand and supply part. Since supply model is quite easy to obtain and develop, demand model needs more input data. In particular Comprehensive Travel Surveys (CTS) are conducted to feed the demand model. The core part of the CTS are household trip questionnaires which are used, among others, to identify residents’ trip patterns. These surveys are very expensive and often determine CTS frequency. Many cities, also large ones, cannot afford to conduct these kinds of survey as often as it should be done.In this paper the authors analyzed possible ways to reduce CTS cost. They verified whether survey sample size reduction would not cause rapid error increase. Based on CTS for Krakow metropolitan area authors compared results between full sample (6000 households inquired) and reduced samples (2400, 1000 and 750 households inquired). The differences between results obtained from different samples are perceptible but not significant. Therefore, the authors proved that household survey sample size might be reduced even to 1000 inquired households with acceptable loss of results quality.