Coverage Bias in Telephone Surveys in Israel

Potential coverage bias in telephone surveys among the general population in Israel was studied in a random sample of 6,405 adults drawn from the population register. Registered telephone status (landline and mobile, mobile-only, no telephone) of each sampled individual was matched with his or her sociodemographic characteristics obtained from diverse databases. This facilitated estimation of the coverage properties of the landline telephone directory sampling list in Israel and identification of subpopulations potentially undercovered in telephone surveys. Nonveteran immigrants, individuals in the 30—44 age group, and those living in low socioeconomic areas proved systematically underrepresented in a telephone survey based on the landline telephone directory sampling list. Bias due to the exclusion of mobile-only individuals from the sampling list was also tested, and implications for sampling list selection are discussed.