A poorly considered and designed questionnaire can yield extremely unreliable results. Elaine McColl examines the steps a researcher has to take to ensure that only the most valid instrument is created.