The paper is concerned with the problem of non-sampling errors in written mail-back household travel surveys arising from refusals of information to specific questions and incorrect entries by the respondents. The paper describes empirical results of a research project for which the entries of a written household survey were re-examined by interviews. By this experiment the frequency of incorrect and incomplete entries about personal and trip characteristics could be estimated. Emphasis is put on the problem of non-reported trips. based on statistical tests of relationships between non-reporting on the one hand and travel and personal characteristics of the respondents on the other, correction procedures for re-estimating average numbers of trips can be developed. (Author/publisher) for the covering abstract of the conference see IRRD 286978.
Abstract