CONSIDERATION OF NONRESPONSE EFFECTS IN LARGE-SCALE MOBILITY SURVEYS

This paper continues the line of investigation of nonresponse problems previously presented. After a brief review of the context of the problem, namely the nonresponse effects on measured behavior in spite of demographic weighting, and the results of the previous research on this topic, the paper documents a broadening of the insights gained into the effects of nonresponse. These insights were applied to a large-scale nonresponse analysis of approximately 100000 trips. The analysis included the nonresponse effects for the number of trips, trip purpose, travel mode, and seasons. Also, nonresponse effects are compared for written and interview surveys. Experience with the characteristics and impacts of nonresponse for intercity travel is presented. The insights gained could be used to clear up and correct past and present survey efforts and also to ensure that future data-collection efforts are conducted at lower costs, since corrections can also be made for smaller rates of return. (Author)