SAMPLE BIAS IN TELEPHONE INTERVIEW TRAVEL SURVEYS

The methodological aspects of travel surveys have been subjected to much recent investigation, with a major emphasis being placed on the biases inherent in a variety of survey instruments. This paper examines the bias in one particular type of survey, the travel survey conducted by telephone interview. More specifically it examines the bias created by the omission of non-telephone households from the sample, but does not address the bias in responses received over the phone. The study is conducted by examining the characteristics of phone-owning households in the 1978/79 Melbourne home interview survey. In this way, a sample is created which approximates that which would have been obtained had the interview been conducted by telephone. The characteristics of phone-owning and non-phone-owning households are compared with respect to socio-economic characteristics, and areas of potential bias in the survey results are identified. The study then examines differences in travel patterns and suggests ways in which such biases may be corected. (Author/TRRL)