The tips below are based on a simple premise: An effective survey is merely a conversation with respondents who are willing and able to share accurate information. Questions that are approachable, interesting, easy-to-answer, and that invite thoughtful responses can be satisfying. On the other hand, a confusing, bureaucratic form that demands responses with no clear direction is frustrating and will likely be ignored or skipped (or worse, the responses will be inaccurate).
Writing Good Survey Questions: Tips & Considerations
Have Empathy for Your Respondents
- Consider the burden and cognitive effort you are asking of respondents. Reducing burden increases the likelihood of completion and the accuracy of responses.
- Respondents must be able to know or recall what you are asking about. More recent time frames and less common or unusual events are easier to recall.
- Ask about firsthand experiences—what they have done, their current situations, perceptions, etc.
Clarity and Comprehensiveness
- Questions must be clearly written in the simplest possible language and grammar so that all respondents understand the meaning. NSSE often uses a sixth-grade reading level as a guide.
- When the meanings of words are not necessarily known or shared by all respondents, provide a definition.
Response Options
- Response options must include all reasonably possible answers for participants and must be mutually exclusive unless using “select all that apply.”
- Include “not applicable,” “don’t know,” and “I prefer not to respond” options, if necessary, to avoid having respondents guess or become annoyed.
- Response options must agree with the question stem (e.g., do not ask a “yes/no” question and provide “how often” response options).
Equity and Inclusiveness
- Be sure your question wording is inclusive of all groups of respondents in your population. Be cautious about cultural assumptions, gendered language, terms that may not be universally understood, etc.
- Sometimes for inclusivity you need to allow for multiple responses (e.g., select all that apply).
Things to Watch Out For
- Double-Barrels: Ask about only one subject.
- Social Desirability: Be sure questions do not cause a desire to respond in socially desirable ways.
- Trigger Words: Avoid highly charged, controversial, or strong terms.
- Ranking: Avoid asking to rank order items. Ranking is burdensome and you get better data quality from asking them to rate each item individually.
Always Specify Context
- Specify an unambiguous, appropriate time frame and location for clearer recall (during the current school year, while attending this institution, etc.).
- Phrase the context at the beginning of the question, not at the end where it might be overlooked.
Question Order
- Earlier questions affect responses to later questions.
- Group questions that cover related topics together. It eases cognitive effort.
- Place sensitive or potentially objectionable questions near the end of the questionnaire.
- Ask questions about events in the order the events occurred.
Additional Tips
- For questions that use vague quantifiers in the response set (very much, very often, etc.) use of “about” (About how often have you…?) relieves the respondent from the burden of precise recall and cognitive work.
- In lists, use the more widely understood “etc.” rather than “i.e.” or “e.g.”
- Use second-person pronouns (you/your) rather than first-person (I/me). However, first-person pronouns are sometimes used in agree/disagree questions.
- When asking either/or bipolar questions, state both the positive and negative sides in the question stem and list the positive term first (To what extent do you agree or disagree with the following statements?).
References
Dillman, D. A., Smyth, J. D., & Christian, L. M. (2014). Internet, phone, mail, and mixed-mode surveys: The tailored design method. John Wiley & Sons.
Fowler Jr, F. J., & Fowler, F. J. (1995). Improving survey questions: Design and evaluation. Sage.
Groves, R. M., Fowler Jr, F. J., Couper, M. P., Lepkowski, J. M., Singer, E., & Tourangeau, R. (2011). Survey methodology. John Wiley & Sons.