COMMON MISTEAKS MISTAKES IN USING STATISTICS: Spotting and Avoiding Them

Suggestions for Wording Questions

Summary

Elaboration

Suggestions

• Use familiar wording. Avoid technical terminology or words that are distinctive to a particular subculture (unless the questionnaire is aimed at that particular subculture).
• Avoid indefinite words such as "regularly," "recently" unless the text of the questionnaire specifies what they mean (e.g., "at least once a week" or "within the past week")
•  Avoid other ambiguous words (e.g, "recreational facilities")
• Avoid long questions. (This needs to be balanced with avoiding ambiguity)
• Try to avoid questions requiring the respondent to recall information.
• Avoid leading questions (e.g., "Do you agree with most people that ..."). The leading part of the question might appear in a preceding sentence --e.g., "Most people believe that ...  Do you agree?"
• Avoid "loaded" words in questions -- e.g., "inconsiderate people who ...",   "good citizens"
• Avoid asking precise questions about things respondents can't reasonably be expected to know precisely. (e.g., "How many servings of vegetables do you eat in a typical week?") A question with several ranges as options is better.
• Avoid sensitive questions if possible. Answers are likely to be inaccurate. If sensitive questions are necessary, phrase them in ways that  reduce embarrassment or shame. For example, "We all get angry sometimes. How often have you gotten angry in the past week?"
• Avoid questions that ask two things at once. (e.g., "Do you think you should exercise more often and eat fewer desserts?")
• Avoid questions with two negatives. (e.g., "Do you oppose a ban on smoking?") They may be hard to unravel and therefore subject to misinterpretation.
• Avoid hypothetical questions (e.g., "If you were president, what would you do to ..."). It is hard to say just what information the answer to such questions is really telling you.