Limitations of telephone based selection and interview procedures.

The use of telephone based sampling and interview techniques has become more popular in epidemiological research in recent years. Previous studies of these designs have shown good response rates and reliability of information which is comparable to that obtained from the use of mail and personal interview. This paper reports on the results of a pilot study, designed to examine the possible limitations of the use of telephone interviews, in population based studies in Australia. The findings show that as many as 31 per cent of people listed on the Federal electoral rolls in the Hunter Valley region of New South Wales may be inaccessible to interview by telephone, and that these people differ from those who are accessible on a number of important socioeconomic and health indicators. The potential biases introduced by telephone surveys should be considered and explored by those wishing to use them.