call back to talk to a respondent at greater length about his or her views. This respondent could even end up being quoted by name (and with a photograph) in a subsequent newspaper story. The interviewer will usually have given each respondent a pledge of confidentiality at the beginning of the interview, so respondents will have to be asked to waive this guarantee before giving their names or addresses.
Respondents should understand that the polling organization often knows something about their household before the interview begins. A sample for a mail survey is based on a list of names and addresses. In face-to-face polling at a person's home, the interviewer will always know the address of the home, but not necessarily the household's or respondent's name, unless it is asked and given. In telephone polls that use random-digit dialing, the name of the household is never known unless it is asked and given, or unless the telephone number is a listed number and the pollster later uses a telephone number database to cross-check the numbers of households in which interviews were completed against those listed in the database. Polls that sample individuals from voter registration lists, whether conducted by mail, in person, or via telephone, will know the sampled respondent's name and past voting record in advance, and the questionnaire will either be addressed to that person or the interviewer will ask to speak with that person (and no one else).
Nevertheless, you do want to be concerned if a private "poll" is really a fund-raising gimmick or a way to construct a mailing list for a subsequent sales pitch. And you do not have to supply any personal information that might be used for such purposes.
If you are contacted and asked to participate in an election poll, it is only prudent to think at least briefly about whether the pollster is likely to know your name and, if so, whether you should have any reasonable concern about this. We can neither automatically encourage everyone to release their names to pollsters nor discourage everyone from doing so. We do encourage readers to use common sense in deciding whether or not they should give out their names if they are asked in an election poll. Here, a good rule of thumb is: if you are uncertain, ask the interviewer to clarify why your name is being requested. If you are mailed a questionnaire and it asks for identifying information, you may want to contact the polling organization to ask about this. If you do give your name, make certain you understand whether or not you also are waiving your right to have your responses kept confidential.
DILLMAN, DONALD A. 1978. Mail and Telephone Surveys: The Total Design Method. New York: Wiley.
The most widely respected book written to date about the "practice" or application of survey research methods. This is very much a "how to do a survey" book via mail