Sampling in Social Gerontology: a Method of Locating Specialized Populations

Abstract

This paper describes a two-stage sampling design for obtaining probability samples of the elderly and other specialized populations. The first stage enumerates a sample of elderly individuals residing in a probability, sample of households; the second stage involves the mailing of a questionnaire to the eligible respondents located in the first stage. The results of this method for a study of noninstitutionalized elderly in Washington State are reported. These results indicate that (1) elderly individuals will respond in the two-stage sampling design does not appear to increase sample bias above that expected in normal mailed-questionnaire studies, and (3) the method is extremely inexpensive. Use of the two-stage design by social gerontologists is recommended.