Have you ever had to write survey questions? Anyone who has tackled survey copy understands how difficult it is. To get it just right, a survey should be written like a conversation with your customer: casual, engaging, and succinct.
Find out how to craft familiar and unbiased survey questions from an expert. Guest author Jon Gitlin, Content Manager at SurveyMonkey, (parent-company of Usabilla) shares advice on avoiding leading and loaded questions on your surveys.
The way you ask questions influences the quality of responses you receive.
Ask for information in a neutral, straightforward way, and you’re sure to get back honest, thoughtful, and valuable feedback. Inject an opinion or assumption into your survey question, and you’ll receive biased responses that offer little use.
Two common sources of biased survey questions are loaded and leading questions. Each sways respondents to answer in ways that don’t account for their true feelings and preferences.
So what exactly are leading and loaded questions? And how can you avoid them on your survey?
We’ll break down the answers to both of these questions! To start, let’s talk about leading questions.
The definition of leading questions and how to avoid using them
A leading question subtly guides respondents to answer a certain way. More often than not, leading questions influence respondents to provide feedback in a manner that aligns with the survey creator’s opinion.
Leading question examples can look as follows:
Our customer support team is rated as being the most responsive in the industry. How responsive—or unresponsive—do you think they are?
By establishing that your team is rated as very responsive, the respondent will feel more compelled to agree. You can easily fix this issue by taking out the first sentence and simply asking: How responsive—or unresponsive—do you think our customer support team is?
Was our customer service excellent?
This example is a bit more subtle, but it still qualifies as a leading question. By adding the word “excellent,” you’re hinting that the service was in fact excellent. Give the respondent space to come up with their own answer by removing this word and by providing a range of answer options, from poor to excellent.
To spot leading questions like the examples above, ask yourself the following questions:
Do any words reflect your own opinions and preferences? If so, remove them!
Is the prompt balanced? Does it account for both the negative and positive experiences the respondent might have? It’s critical that it does.
Are there unnecessary modifiers? Something as simple as calling a politician a conservative, for example, can influence how respondents choose to answer.
Did I test the question on friends, family members, and colleagues? We recommend that you do! After they answer the question, ask them if they felt compelled to provide a certain response.
Overview on loaded questions and ways to catch yourself asking them
A loaded question makes an assumption about the respondent that forces them to provide an answer on something they may not agree—or be familiar—with.
Here are a few examples of loaded questions:
How would you rate your experience in working with our customer support team?
If the respondent has never worked with your support team, they’d be left to answer at random.
Prevent these respondents from seeing this question by asking them at the beginning of the survey if they’ve worked with your support team. If their response is negative, use Question Skip Logic to hide this and other customer support-related questions.
Note: If you’re completely certain that the client has worked with your support team before receiving the survey, the prompt above doesn’t qualify as a loaded question.
What’s your favorite part about working with our support team?
Note: This is also a leading question, as it asks the respondent for their “favorite part” about working with the support team.
If there isn’t anything the respondent likes about working with your support team (which hopefully isn’t the case!), they won’t even know how to answer the question.
Solving this, again, comes down to using Question Skip Logic. Ask about their satisfaction from working with your support team earlier in the survey, and, only if the response is positive, program your Question Skip Logic to show the question above.
So how do you stay away from asking a loaded question? Here are some tips and tricks:
Think hard about the audience of your survey. Who’s taking your survey? How are you sending it to them? Why are they taking it? Making a survey that’s tailored to your audience will ensure you’re writing questions that they can actually answer.
For any question that makes an assumption, confirm that Question Skip Logic is built around it. As we saw in our examples, this step is paramount.
Preview your survey and share it with others around you. A quick check, both on your own and by others, can ensure that you’re not asking loaded questions.
At first it might be hard to avoid asking leading and loaded questions and that’s OK. Apply the guidance in this page to identify—and fix—them. Only then will you be able to keep your prompts impartial, giving respondents a better survey-taking experience, and leaving you with more reliable data for making decisions.