Definition
Questionnaire design is the methodology of constructing survey instruments that elicit accurate, comparable, and analysable responses from respondents. It is the bridge between an abstract research question — what do you want to know? — and the concrete data that gets collected. A well-designed questionnaire reduces measurement error, controls for known biases, and produces answers that can be aggregated across respondents without comparing apples to oranges.
Bad questionnaire design is a silent saboteur of empirical work. The data may look clean — neat rows, no missing values — yet measure something other than what the analyst intended. By the time errors surface, the field work is done and unrepeatable.
Why it matters
How it works
Sound questionnaire design follows a sequence. First, define the constructs being measured — a vague target like "satisfaction" must be operationalised into specific indicators that can be asked about. Second, choose between closed questions (multiple-choice, rating scales, yes/no) and open questions (free text); closed yields comparable data faster, open captures unanticipated answers. Third, draft the wording with care: questions should be neutral (no leading framing), specific (one issue per question, no double-barrelled compounds), and use vocabulary the target respondent will understand. Avoid negations — do you disagree with becomes a parsing puzzle that depresses data quality.
Fourth, control for the major systematic biases. Social-desirability bias drives respondents toward answers that flatter them — anonymity and indirect phrasing reduce it. Acquiescence bias — the tendency to agree — is countered by balancing positively- and negatively-keyed items. Order effects mean that earlier questions can prime later answers, so sensitive questions usually go late and demographic items at the end. Fifth, design the response scale deliberately: Likert scales of 5 or 7 points handle most attitude measurement, with verbal anchors at the extremes and ideally the midpoint. Sixth, pilot test the instrument with a small group from the target population, watch for confusion, time the completion, and revise. Finally, pair the questionnaire with a sampling strategy — even a perfect instrument administered to a biased sample produces biased results. Questionnaire design and sampling design are two halves of the same measurement system.