Intensity of follow-up. Effects on estimates in a population telephone survey with an extension of Kish's (1965) approach.

Information from a population telephone survey of attitudes of residents of the Province of Ontario, Canada, toward legislation to restrict smoking was used to examine the effects of intensity of follow-up on population estimates. It was found that the bias introduced to the attitude estimates by less intensive follow-up was smaller than the bias introduced to estimates of sociodemographic characteristics and smoking status. An algebraic determination of the components contributing to the bias in the attitude estimates was used to investigate how the relation between smoking status and certain attitudes influenced the relation between attitude and intensity of follow-up. This required an extension of the approach described by Kish (Survey Sampling. New York: John Wiley and Sons, 1965.)