Questionnaire Design

Length

Number of questions

Ability to extract sensitive/embarassing information

Cost

Method of distribution

Can it show interviewer bias?

Force a choice or allow neutral?

Start with easy questions. Then move on to asking more difficult questions and asking for personal info.

Hide the sensitive questions in a group of questions.

Provide response categories instead of specific figures.

No loaded questions ever. Never lead the respondent into answering in a way you want. Leading questions give the respondent clues as to what the answer should be.

Is the question necessary?

Can you get the same information from the respondent with less questions?

is the respondent informed?

Can the respondent remember?

Can the respondent articulate?

How much effort does the respondent need to put in?

Is the purpose of the questionnaire clear?

Do the respondents know the context?

Question types to consider: Unstructured, Structured, Multiple Choice, Dichotomous, Scale

Define the issue, who what when where why how

Be unambiguous. Use simple, ordinary words.

Avoid implicit alternatives and assumptions

Avoid generalisations and estimates

Avoid implicit assumptions in the questions. Make clear all the assumptions.

Pilot testing a questionnaire is a good idea.

Also good to have a template or grid of the questionnaire.

Should look professional.