Questionnaires should include:
- Good questions - must be clear, so that the respondent understands what is being asked (so no ambiguity, with numerous meanings, or double negatives, or double barrelled questions) and not biased so the respondent is lead to a particular answer (such as social desirabiity bias, which is where someone chooses an answer to make them more attractive/nice/generous rather than being truthful).
- Also good questions could be written so the answers are easy to analyse. Open questions are harder to analyse than closed questions because they result in qualitative data being produced, but respondents may be forced to select answers that don't represent their true thoughts/behaviours in closed questions
- Irrelevant filler questions to distract respondents from the main purpose of the study and reduce demand characteristics.
- A sequence to the questions- starting with the easy ones so questions that might make someone feel anxious or defensive are left until after the respondent is relaxed.
- Often use stratified sampling.
- Questions can be tested on a small group of people in a pilot study so they can later be refined in response to any difficulties encountered.
Important things to consider with interviews: