Skip to main content icon/video/no-internet

Directory Sampling

Directory sampling is one of the earliest versions of telephone sampling. Telephone directories consist of listings of telephone numbers. The residential numbers are generally placed in a section of the directory separate from business numbers. Each telephone listing is generally accompanied by a name and an address, although the address is not always present. Households may choose not to have their telephone number published in the directory. These are referred to as unpublished numbers, most of which also are unlisted numbers.

In the original application of directory sampling, a set of telephone directories covering the geopolitical area of interest to the survey were assembled. After the sample size of telephone numbers was determined, a random selection procedure was used to draw the required number of residential directory-listed telephone numbers for each directory. The actual selection method ranged from using systematic random sampling of listed telephone numbers to first selecting a sample of pages from the directory and then sampling one or more telephone numbers from the selected pages.

Directory samples provide samples only of telephone numbers that are directory listed. Directory samples will yield biased samples of a population, because all unlisted households are given a zero probability of selection, and unlisted households generally differ from listed households on key characteristics. For example, persons with unlisted numbers are more likely to be minorities, recent movers, and single female adults. In some geographic areas, a substantial percentage of households may have unlisted telephone numbers, for example, larger central city areas and Western states.

Today, directory-listed sampling is rarely used alone, having been replaced by list-assisted random-digit dial sampling. But in other ways, directory sampling has made a comeback. Telephone directories are now entered into national databases of listed residential telephone numbers that are updated on an ongoing basis. A fairly common random-digit dialing sample design involves forming two strata. The first stratum consists of directory-listed residential telephone numbers. The second stratum consists of telephone numbers in the list-assisted sampling frame that are not residential directory-listed telephone numbers. Thus two mutually exclusive strata are formed, and a sample of telephone numbers is drawn from each stratum.

The presence of an address for most residential directory listed telephone numbers in national databases makes it possible to assign geographic codes to the addresses. Typical geographic codes include county, zip code, census tract, block group, and census block. This makes it possible to sample directory-listed telephone numbers from small geographic areas, for example, from a reverse directory. The presence of a name with each listed number also enables the matching of the names to lists of ethnic surnames. This makes it possible to sample directory-listed households with specific surnames.

Michael P.Battaglia

Further Readings

Lavrakas, P. J. (1993). Telephone survey methods: Sampling, selection, and supervision (
2nd ed.
). Newbury Park, CA: Sage.
  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading