A methodological study of the relative validity of personal and telephone surveys for studying political and communications behavior. Data collected in telephone and personal surveys in seven western Pennsylvania cities are used to illustrate the impact of the joint effects of sampling telephone households, added-digit dialing, within-household selection procedures, refusal rates, and other factors that might affect the sampling power of telephone surveys. Results indicate that available techniques of random-digit and added-digit dialing provide representative samples of telephone households and that respondents will provide detailed personal information on a variety of topics over the telephone as well as they will in personal interviews. It is concluded that telephone surveys provide representative samples of the general population and can obtain reliable answers on personal as well as factual subjects. Moreover, telephone surveys have greater potential for quality control and flexibility than personal interviews, and at lower cost. 65 pp. Bibliog.