Questions, when used properly can yield a wealth of insights and help you separate assumptions from facts. Used incorrectly and they can lead you in the wrong direction.

There are countless types of questions that can be asked such as close ended, open-ended, and leading questions. In this guide, we’ll focus on the definition of leading questions, the different types available, and provide multiple examples. When you’re done, you’ll know whether your surveys are using them or not and what to do as a result.

What is a leading question?

A leading question can be defined as a question that pushes respondents to answer in a certain way because of the information given or the way they’re asked. A leading question is often phrased in such a way that it has the information the survey creator wants to confirm.

For example, ABC Inc. is considered one of the best places to work in New York, don’t you agree?

In the example above, the survey question has already told the respondent that ABC Inc. is considered one of the best places to work. With this information, responses are more likely to agree with the sentiment.

Leading questions can be intentional or unintentional but the outcome is the same – skewed data. Since the data is biased in one way or another, the final conclusions drawn may fail to have the desired impact. This can be especially damaging when making business or policy decisions.

Leading question vs loaded questions

A leading question is similar to a loaded question but there are differences. A loaded question is generally complex and it contains an assumption (that usually hasn’t been verified). It’s asked in such a way as to limit the range of responses that can be given.

For example, ‘Have you stopped wetting the bed at night?’ The way the question is structured assumes that the person wets the bed at night and the person asking is only seeking verification. If the respondent doesn’t wet the bed then none of the possible answers to the question are applicable.

Leading questions, on the other hand, don’t necessarily have an assumption contained within the question. Some leading questions do but it’s not the only way leading questions are structured. Leading questions tend to guide someone to a specific conclusion by including or omitting information. They can even be chained together to get the respondent to answer in certain ways.

For example, ‘You were working last weekend, correct?’  Yes. ‘You are in charge of shift rotations, right?’ ‘So why shouldn’t this issue be your responsibility?’

The example above is a series of leading questions ending with a loaded question. Let’s look at the different types of leading questions so you have a better grasp of when they’re being used.

Types of leading questions

Assumption based 

This type of leading question is often indistinguishable from a loaded question. The question is created with an inherent assumption. Many types of surveys are guilty of this but it’s most often found in feedback surveys. Instead of asking if someone felt one way or another, the question often assumes they feel one way and seeks to understand the degree of the feeling.

For example, how much did you like our services? The assumption is that they like your services at all. Instead, you should first ask whether they liked or disliked your services. After that, you can then ask how much they liked or disliked the services. A few more examples include:

  • How much do you dislike the new policies of the governor?
  • What’s your favorite part of the travel tour?

Direct implication

This type of leading question is more explicit and is designed to put respondents in a specific frame of mind. The goal is to understand how people would behave or react in specific situations if a specific event were to happen. It’s a classic if, then scenario. Though this is a leading question, when used in the right context, it’s not a bad or negative thing. It can help you predict the behavior of a population that shares certain characteristics.

  • If you’re happy with the service rendered, would you tell your friends about it?
  • If Governor Nix wins the election, would you still participate in politics?
  • If you lost fifty pounds, would you believe in our weight-loss regimen?

Coercive

These questions are structured with the intention of forcing respondents to answer in a specific way—usually in agreement with the question. They’re a large source of bias and should be avoided in almost every case. They’re much more direct and, at times, aggressive than most types of leading questions.

  • You had a great time, right?
  • We love that show, don’t you?
  • You’ll go to the next concert they’re having, won’t you?

Scale-based

Scale-based leading questions are more difficult to catch because they give the illusion of being objective. In reality, the options are tipped in the direction the research wants them to go. For example, the scale can have more options toward a negative stance or more extreme responses on the negative scale. Because there are more negative options, the respondent is more likely to choose a negative response.

  • How do you feel about our customer support rep’s performance today?
    • Extremely satisfied
    • Satisfied
    • Somewhat satisfied
    • Somewhat dissatisfied
    • Dissatisfied
  • In the example above, there are three options that express satisfaction while only two options express dissatisfaction and there is no neutral measure.

Interconnected leading questions

An interconnected leading question has two parts. The first part presents a statement that is designed to sway the respondent or introduce a bias. The second part of the question capitalizes on the bias and usually tries to seek agreement from the respondent. It’s subtle and is often used when the person asking the question wants to gain support for an idea that may receive resistance.

For example, ‘Atlanta is considered one of the best cities for growing tech startups, what do you think about Atlanta as a growing tech hub?’

Why you should avoid leading questions in surveys

Leading questions aren’t always a bad thing. For example, the direct implication question is trying to set up a situation to understand the behavior of respondents. It’s intentionally leading to get a certain type of information.

Most leading questions are not so benign. They’ll skew data in the direction the researcher wants in order to confirm their assumptions. This can be detrimental – especially if the original assumption is incorrect. For example, if there’s a real problem with a product but the research asks leading questions that end up masking these issues, the company will eventually lose market share.

Because of the possible drawbacks, it’s best to err on the side of caution and avoid leading questions during data collection. That way, you can get clean primary data with fewer biases. Of course, it’s not possible to remove biases completely

How to avoid leading questions

Few leading questions are obvious, that’s why they’re used so often. Before you can avoid them, you need to understand what they are and the different forms they can take. That’s what the majority of this guide has focused on.

When creating your surveys or preparing to administer questionnaires, you need to look critically at the questions you’ve asked. Review them with the following criteria:

  • Do any questions showcase your own opinions or biases? If so, rephrase them so they’re as objective as possible.
  • Does it leave room for the respondent to answer in unexpected ways? For example, adding words like right, won’t you, correct, etc. at the end of the question or statement tries to force agreement. Avoid this.
  • Does it have any inherent assumption in the question? For example, assuming they were satisfied with a product or assuming they’ve had a memorable experience.
  • Are you using certain adjectives or modifiers (for example, calling someone tall or calling a politician liberal)?
  • Have you tried using the questions on your inner circle and gauged their response? This is especially useful if you know people that have different opinions. If most of them answer the same way then it may be a sign of a leading question.
  • Ask for a peer review. Seek out the advice and perspective of colleagues to check if your questions may be leading respondents in a certain direction.

Conclusion

In the vast majority of cases, leading questions are to be avoided because they can reduce the accuracy of your data. Without accurate data, you can’t make accurate decisions and may invest resources in the wrong way. Nobody wants that.

Take a critical look at the questions you’re asking for data collection to ensure they’re not leading respondents in one direction or another. The goal is to get objective data – not to get data that confirms your own assumptions or biases.