pilot study

Any small-scale test of a research instrument (such as a questionnaire, experiment, or interview-schedule), run in advance of the main fieldwork, and used to test the utility of the research design. Pilot studies therefore very in size and nature, Interview-schedules for a large-scale survey may be piloted on a substantial sub-sample of the relevant population-perhaps as many as 500 respondents. In-depth interviewing techniques, on the other hand, may be adequately piloted on only a few acquaintances and friends. Quantitative sociologists and social survey research agencies will insist on full briefing and debriefing of interviewers who conducted the pilot, in order to identify and rectify logistical and analytical difficulties in the research design, including (for example) ambiguous questions, incorrect routeing or filtering through the questionnaire , replies which cannot be coded , and items about which there appears to be no variance (see variation, statistical ) in the responses. If the difficulties are substantial a second pilot may also be required in order to confirm that the amendments to the original design have proved effective.