National Estimates of Nonlisted Telephone Households and Their Characteristics

Telephone directories are frequently used in market research as a frame for sample surveys of households. This is especially true for telephone surveys, but such directories are also often used to select samples for mail surveys and studies using a personal interview. Unless special procedures are employed, the use of telephone directories excludes households that do not have telephones as well as telephone households not listed in current directories. A telephone household may be nonlisted because the household asks not to be listed or because the household has moved too recently to be included in the directory. One way of including nonlisted telephone households that has received attention is random-digit dialing 1, 2, 3. To assess the adequacy of telephone directories as a frame for sampling, or in estimating the bias from the use of such a frame, or in deciding whether to use random-digit dialing, it is important to have some idea of the characteristics of nonlisted telephone households. A previous article 1 provided national estimates on the incidence of nonlisted telephone households and a few of their demographic characteristics based on a March 1970 study. This article updates and expands those results based on subsequent annual surveys through March 1974.