Designated Respondent

Designated respondents are the individuals chosen specifically to be interviewed for a survey. Surveys often are conducted in two stages: first, selecting a sample of household units and, second, selecting persons within the households with whom to speak. Survey researchers' and interviewers' jobs would be easier if they could question the persons first answering the phone or first coming to the door or simply any adult resident in the unit who was willing to talk. This usually is an acceptable idea only if the researchers simply need to know the basic characteristics of the household; however, much of the time researchers need to gather data from one specifically chosen person in the household—that is, translate the sample of units into a sample of individuals. In contrast, if the respondent is merely the most likely person to answer the phone or to be home, his or her characteristics may be overrepresented in the sample, meaning that the sample will be biased. These more willing or available individuals tend to be older and/or female. Such biases mean that survey researchers are likely to get an inaccurate picture of their samples and can come to some incorrect conclusions. Information quality depends on who is providing it.

Researchers try to avoid such bias by using a within-household selection procedure likely to produce a more representative sample at the person level. These tend to be more expensive than interviewing any available person in the household, but they are also more precise. It takes more time to find the [Page 190]“right person” and to gain an interview when that person is available. As a result, refusal rates can, and often do, increase. The informant (person who answers the door or phone) may be put off by some of the questions interviewers have to ask in order to pick the designated respondent—for example, a complete list of household residents—and may refuse to proceed further. If informants are cooperative but are not the designated respondent, a handoff must occur, and interviewers may have to keep calling back if the designated respondent is not immediately available. Survey researchers have to make trade-offs when they choose a respondent selection method. Different kinds of respondent selection methods have been devised to identify the correct person for interviewing and obtain his or her cooperation, and each has advantages and disadvantages with respect to costs and precision.

Respondent designation techniques have consequences for errors of nonresponse, such as not finding the correct person, inability of the person selected to participate because she or he does not qualify (e.g. because of language barriers, ill health, illiteracy), or that person's unwillingness to be interviewed. Ways to compensate for these problems exist, such as callbacks, interviewing a secondary person in the household who also meets appropriate criteria (e.g. speaks English, is able-bodied, literate), or weighting responses by appropriate criteria. Among principal concerns are within-unit coverage errors; for instance, when the wrong types of respondents consistently are interviewed or when the selected respondents consistently do not meet the survey requirements and another qualified person is available but not interviewed. Survey researchers need to think out solutions to these issues in advance.

...