Whether you are soliciting input through interviews, surveys, discussions, or mindreading, the quality of your questions determines the quality of your data.
- Identify what you want to know, and what you want to do with that information.
- Use the Ideation Card to generate lots of possible questions.
- Find the most promising questions by looking at how well each aligns with your goals.
- Test your questions with a variety of people. Were your respondents confused or uncomfortable? Do their answers produce the information you want?
- Revise your questions and test them with new people.
Closed questions ask people to rank or rate options, choose from predetermined options, or enter a numeric answer. Use them to collect facts, support an existing conclusion or prioritize set of known options. They generate quantitative data that often translates well into graphs or charts. They’re quick to answer and simple to analyze.
Open-ended questions ask people to explain their experiences, offer an opinion or contribute ideas. Use them to seek new information and insights, to understand cause and effect, and to explore uncharted territory. Open-ended questions require more commitment from respondents and more analysis from the researcher, but they generate compelling stories and rich insights.
- Keep questions brief, but provide any background people need to know in order to answer.
- Ask respondents about their own thoughts, beliefs and experiences. Don’t invite or expect respondents to speak for other people.
- Don’t give the reader any reason to believe you prefer one answer over another, or prime their responses by providing examples they might parrot back to you.
- Avoid words with multiple meanings, and subjective response options such as “regularly” or “a lot.”
- Keep your language gender-neutral, objective, and inclusive. Would your questions be equally friendly and relevant regardless of the respondent’s age, race, culture, faith, income, ability, marital status, sexual preference, language, or literacy? Particularly when you’re exploring a sensitive topic or a community outside your lived experience, ask experts and community members to review your wording.
- Unless you need to pre-screen participants, save the demographic questions for last. In numeric questions, ask people to select a range. People would rather give an age range (e.g., 45-55 years) than disclose their exact age, and a household income range than their take-home pay. Let people opt out of questions.
- Always finish with something open- ended like “Is there anything you’d like to add?”