Survey Best Practices Tips

SurveyMonkey has great resources on how to develop and design surveys. Beyond the tips we summarize below, we recommend reading this 10-15-minute-long article on survey best practices.

Know your goals

Before you start drafting survey questions, make sure you have clear goals for your survey.
You want to keep your goals specific and attainable, to avoid creating lengthy surveys, unless they are strictly necessary.
Your goals will also help you decide if you want to have open- or close-ended questions. Open-ended questions are great for capturing new ideas and giving respondents more freedom to answer how they want, but they are more time-consuming to answer and to analyze. Close-ended questions, on the other hand, are easier to answer and analyze, but might not give space for respondents to be creative and might cause bias in the answers. You could also combine both types and use open-ended questions to capture more explanations about specific answers in closed-ended questions, for example.

Language

Ask one question at a time instead of double-barreled questions that often contain the word “and.” A survey question should contain only a single concept. For example, the question “How satisfied are you with the price and quality of the product?” is confusing because the respondent might be satisfied with the quality but not the price (or vice versa). This would be better presented as two separate questions, one about the price and one about the quality.
Avoid jargon, acronyms, technical terms, and obscure phrases. Use simple words and specific language to create questions that yield precise answers.
Avoid leading and loaded questions (e.g., “Do you agree that…). These questions might bias the respondents to answer in a certain way, and hence will not provide you with the results you want.
Avoid using indefinite qualifiers (e.g., only, just, merely, many, few, or seldom).
This short guide has some good examples on how to rephrase questions to avoid key mistakes in survey design.

Question format

A 5-point Likert scale is recommended. The ‘distance’ between the values should be as ‘equal’ as possible. For example, responses to a question such as “How helpful was the content presented at the workshop?” could comprise 1. Extremely helpful 2. Very helpful 3. Somewhat helpful 4. Not so helpful 5. not at all helpful.”
Use matrix questions sparingly. These questions comprise a series of questions with the same set of answers on the same page. While this can be more efficient for respondents, it can potentially confuse respondents.
When asking respondents to rank options, keep the list to be ranked small, up to a maximum of 7 items.
For “Check all that apply” and “Select up to three responses” questions, keep the list from getting too long (e.g., a maximum of 10 answers). If there is a “does not apply” answer, this should be listed first.
Avoid forced-choice questions. Use options such as “don’t know” and “not applicable” sparingly; otherwise, it gives respondents an easy way to skip past a question.

Revise and pilot

Before sending your survey to respondents, make sure to revise the text with attention.
It is highly recommended that you first pilot test your survey with a small subset of people to identify areas for improvement. Ask respondents in your pilot study about the clarity of your survey questions.

Survey Best Practices Tips

Other useful links on survey best practices