ABSTRACT
The chief aim of this chapter is to point out the need for directing effort toward all of them when planning a survey, and the futility of concentrating on only one or two of them. Another aim is to point out the need for theories of bias and variability in response. In the planning of a survey, effort should be directed toward the reduction of all of the errors that it is possible to reduce, but the effort should be apportioned with a view to producing the greatest possible usefulness with the funds available. It is a mistake, however, to take refuge in the assumption that errors in response are going to cancel each other, and thus to excuse poorly designed questionnaires and inexpert interviewing. Errors and differences in interpretation sometimes arise from misunderstanding the questionnaire or failure to take into consideration the form of the questions as written on it or as actually used in the interview.
