10 Tips For Crafting Good Survey Questions

June 25, 2015

Online survey tools have made it easy for marketers to conduct their own research. But while it may be easy to create a survey, surveying requires careful planning if you want to collect meaningful results that you can act on.

When crafting your survey questions, consider these Ten Dos and Don’ts.

What Makes a Good Survey Question

Start with a clear survey goal.

DO: Stick to your goal

Only ask questions that pertain to your goal or an objective that will help you achieve your goal. No matter how nice it might be to know, don’t ask if it does not help you achieve your goal. Set a clear goal on what you want to achieve and don’t stray from it.

DO: Use the right survey question type

To get clean data, you need to use the right question type.

Qualitative questions are open-ended and are great for asking “why”. Use these when exploring an issue. Use them sparingly as they are fatiguing for respondents and subject to interpretation bias.

Quantitative questions are closed- ended. These are far less fatiguing and easy to measure. These offer easy answer options for answering how, what and when. They often appear as:

  • Radio buttons
  • Check boxes
  • Drop down menus
  • Rating scales (i.e. Likert scale or star rating)
  • Ranking scales (drag and drop or matrix table)

A mix of both qualitative and quantitative questions is often a good way to go. Follow up quantitative questions comment box to learn more about why the respondent answered the way they did. Or use and essay question type at the end of your quantitative survey.

DO: Offer mutually exclusive answer options

When using a range, whether it be a date, age, income, etc.) make sure your ranges are exclusive and that there is no overlap. For example:

Q: How long have you been a club member?
1-10 yrs
10-20 yrs
20- 30 yrs
30+ yrs

Instead, your answer range should be exclusive like this:

Q: How long have you been a club member?
1-10 yrs
11-20 yrs
21- 30 yrs
31+ yrs

Nonexclusive answer options not only confuse respondents, but result in bad data.

DO: Be specific and direct

General questions will not give you good results. You need to clearly specify what it is you want answered.

For example:

Q:Do you read the paper?

Instead, ask:

Q:Do you read the Boston Globe on Sundays?

Use short and concise terms. Wordiness adds to confusion. Unclear questions lead to unclear results.

DO: Use balanced scales

When asking survey questions using a scale, each point should hold the same amount of weight with the same number of options on either side of the middle point so as not to skew the data.

Things to Avoid While Crafting Good Survey Questions

DON’T: Be fatiguing

Questions that are ambiguous, lengthy, or tax the respondent’s memory are fatiguing. Limit the number of qualitative questions to 2 since they require more time and effort to answer.

DON’T: Ask sensitive questions

Personal questions can be uncomfortable to answer. Give respondents the option to opt out of these. If you make these questions required, they are likely to abandon the survey. Consider providing a Prefer Not to Answer (PNA) option to decrease survey abandonment while not forcing an answer that might skew your data.

DON’T: Beg the question

Leading questions create biased results that lead to poor data quality. Subtle wording differences can significantly impact results. “Could,” “should,” and “might” all sound about the same, but may produce different results.

For example:

Q: Folgers is a popular coffee brand. What brand of coffee do you drink?

This question might imply that if they do not drink Folgers they might not be popular.
Avoid any wording that introduces bias or impacts the results.

Instead, ask:

Q: What brand of coffee do you drink?

DON’T: Use double-barreled questions

Combining two questions into one confuses respondents and impacts data quality.

For example:

Q: How satisfied are you with our buffet food and drink selection?

The respondent might love the food selection but are dissatisfied with the drink selection. Splitting the question into two separate ones, will give you better results that let you know what you need to improve.

Instead, ask:

Q1: How satisfied are you with our buffet food selection?
Q2: How satisfied are you with the drink selection?

DON’T: Force the answer

Requiring answers will give you results but not necessarily the best results. If you are going to require an answer make sure that you provide a comprehensive answer list.

For example:

Q: What is your gender?
Male
Female

Instead, ask:

Q: What is your gender?
Male
Female
Transgender

Good Survey Questions Take Time

Surveying takes careful planning. Doing the hard work up front will save you time and effort later. Good survey questions lead to good results!

  • Get started with Alchemer today.
    Try Alchemer Start my free trial
  • See How Easy Alchemer Is to Use
    See Help Docs
  • Start making smarter decisions

    Start a free trial