CALLBACKS Attempts to reach a respondent who was not available earlier to be interviewed or to check on inconsistent or missing information on the questionnaire. Callback techniques vary according to the interviewing mode being used--repeat telephone calls for telephone polls; return home visits for face-to-face interviews; and follow up mailing for mail surveys ( Alreck and Settle, 1985:404). Callbacks are time-consuming, can produce reluctant, even hostile respondents, and are often futile. Nevertheless, they are essential. If they are not done, the sample becomes unrepresentative, slanted toward people who are at home more often, or are easier to interview ( Backstrom and Hursh- Cesar , 1981:115; Martin, 1983:701).
Virtually every poll requires some callbacks, since up to one-half of all initial attempts to interview are unsuccessful for one reason or another. But callbacks are a technique in which more is not better. In fact, callbacks are not efficient after the second or third call. About 95 percent of respondents who will ever be reached have already been reached by the third call. After that, the completion rate improves only marginally with more call backs ( M. L. Young, 1987:83). See alsoNOT AT HOMES; RESPONDENT.
CALL-IN POLL A type of straw poll conducted by radio and television outlets. Call-in polls register public reaction to recent events by inviting listeners to call a designated number to record their opinion ( Weisberg, Krosnick and Bowen, 1989:33). Cable News Network (CNN) runs a feature called the news night poll that illustrates the style of call-in polls. During the 1987 Iran- contra hearings, CNN viewers were invited to call a 900 number and record their answer to this question: "Did Oliver North tell the truth when he said President Reagan didn't know about the arms for hostages deal?" CNN added