Sampling by Ethnic Surnames: The Case of American Jews

It is often difficult and costly to locate members of numerically small minorities using standard probability sampling. Consequently, nonprobability sampling techniques of various sorts are commonly used. This paper analyzes the differences between samples chosen by two such techniques—the use of ethnic surnames and the use of organization lists—tand compares them with probability samples. Using data from the National Jewish Population Study, we find that Jews with one of 35 so-called distinctive Jewish names do not differ substantially from the general population of Jews in demographic characteristics or indicators of Jewish identification, and that this technique produces a sample which is more similar to the general population of Jews than does the organization list sample technique.