
“Why are you dissatisfied with our customer service?”
At first glance, this sounds like a reasonable question for a customer satisfaction survey. It asks respondents to explain their experience, and the answer could provide useful feedback.
But there’s a problem.
What if the customer isn’t dissatisfied?
The question assumes dissatisfaction before the respondent has confirmed experiencing it. Someone who is satisfied may struggle to answer, skip the question, or provide an explanation that doesn’t reflect their actual experience.
This is a loaded question.
Loaded questions contain assumptions that respondents may not agree with, creating problems for both the people answering and the researchers interpreting the results.
They appear in customer satisfaction surveys, employee engagement questionnaires, political polling, product research, healthcare studies, and everyday interviews.
Sometimes they are deliberately designed to influence an answer. More often, they emerge unintentionally when researchers assume they already understand the respondent’s experience.
Either way, the consequences can be significant.
A survey question that assumes a problem exists may produce answers about that problem even when respondents would never have identified it themselves.
This guide explains what loaded questions are, how they differ from leading and double-barreled questions, and how to recognize and rewrite them. You’ll also find 15 practical examples, prevention techniques, and a checklist for reviewing questionnaires before launch.
What is a loaded question?
A loaded question is a question that contains an assumption or presupposition that has not been established and that the respondent may not accept.
The question is structured so that answering it directly can imply acceptance of that assumption.
For example:
Loaded question: “Why did you stop using our product?”
This assumes that the respondent previously used the product and has since stopped.
But what if the respondent never used it?
Or what if they still use it?
The question doesn’t provide an appropriate way to express either situation.
A better approach is to establish the relevant facts before asking for an explanation.
Step 1: “Have you ever used our product?”
Step 2: “Do you currently use our product?”
Step 3, if the respondent previously used it but no longer does: “What were your reasons for stopping?”
This sequence removes the unsupported assumption and allows the researcher to collect more accurate information.
Loaded questions are closely related to the complex question fallacy, sometimes called the loaded question fallacy, in which a question presupposes something that has not been demonstrated.
A classic example is:
“Have you stopped cheating on your exams?”
Both “yes” and “no” can appear to imply that cheating occurred.
The appropriate response may be to reject the assumption entirely.
In survey research, loaded questions are problematic because respondents may feel forced to accept assumptions that do not reflect their experiences.
Why loaded questions are a problem in survey research
Loaded questions can undermine survey quality in several ways.
-
They introduce unsupported assumptions
A good survey question should allow respondents to describe their own experiences.
Loaded questions instead present an experience or belief as if it has already been established.
For example:
“How much has our new platform improved your productivity?”
This assumes the platform improved productivity.
But some respondents may have experienced no improvement or even a decline.
The question can therefore produce a misleading picture of the platform’s impact.
-
They can pressure respondents into inaccurate answers
When a question assumes something untrue, respondents may choose the closest available answer rather than challenge the premise.
Consider:
“Which feature of our new product do you find most useful?”
A respondent who finds none of the features useful may still select one if the questionnaire offers no alternative.
The resulting data could exaggerate perceived product value.
-
They can increase survey abandonment
Respondents may become frustrated when questions do not reflect their experiences.
This is particularly relevant when questions concern sensitive subjects, personal circumstances, or controversial issues.
If someone feels a questionnaire makes assumptions about them, they may skip questions or abandon the survey.
-
They make data interpretation difficult
Suppose a survey asks:
“Why are employees unhappy with management?”
Respondents may provide answers about communication, leadership, or organizational culture.
But the researcher cannot determine whether respondents were actually unhappy with management before encountering the question.
The responses may reflect the question’s framing rather than independently expressed dissatisfaction.

-
They can reinforce researchers’ expectations
Researchers often begin a study with hypotheses about customer behavior, employee attitudes, or market conditions.
Those hypotheses are useful for planning research, but they should not automatically become assumptions embedded in the questionnaire.
Otherwise, the research risks confirming what the team already believes rather than testing whether those beliefs are correct.
The central problem with loaded questions is that they ask respondents to explain or evaluate something that has not yet been established as true.
15 loaded question examples and how to fix them
The following examples illustrate how loaded questions can appear in different research contexts.
| # | Loaded question | Better, neutral approach |
| 1 | Why are you dissatisfied with our customer service? | How satisfied or dissatisfied are you with our customer service? Ask for reasons based on the response. |
| 2 | Why did you stop buying our products? | Have you purchased our products before? Do you still purchase them? If not, why did you stop? |
| 3 | How much has our new website improved your experience? | How has our new website affected your experience, if at all? |
| 4 | Which of our innovative features do you like most? | Which features, if any, do you find useful? |
| 5 | Why do employees dislike the new policy? | What is your opinion of the new policy? |
| 6 | How often does your manager communicate unclear instructions? | How would you rate the clarity of your manager’s communication? |
| 7 | Why is our competitor’s product less reliable? | How would you compare the reliability of the two products? |
| 8 | How concerned are you about the harmful effects of this policy? | How concerned, if at all, are you about the possible effects of this policy? |
| 9 | Why did you choose our company over cheaper alternatives? | What factors influenced your decision to choose our company? |
| 10 | How much time do you waste using outdated systems? | How much time, if any, do you spend dealing with system-related inefficiencies? |
| 11 | Why do you distrust political institutions? | How much do you trust political institutions? |
| 12 | How has the training improved your performance? | What effect, if any, has the training had on your performance? |
| 13 | Why do you avoid contacting customer support? | Have you contacted customer support? If not, why? |
| 14 | How difficult was your recent hospital visit? | How easy or difficult was your recent hospital visit? |
| 15 | Why do customers prefer our new packaging? | Which packaging design do you prefer, if any? |
These examples share a common characteristic: the original questions assume attitudes, experiences, behaviors, or outcomes.
The revised versions either remove those assumptions or establish whether they apply before asking follow-up questions.
This is especially important when using conditional survey logic.
Instead of presenting the same question to every respondent, researchers can use screening questions to identify relevant experiences and then ask appropriate follow-ups.
Loaded questions vs. leading questions: what’s the difference?
Loaded and leading questions are closely related, but they are not identical.
A leading question encourages a particular response.
A loaded question contains an assumption that respondents may not accept.
Consider the difference.
Leading question: “Don’t you agree that our new product is better?”
This encourages respondents to agree with a positive evaluation.
Loaded question: “Why is our new product better?”
This assumes the product is better and asks respondents to explain that supposed advantage.
Both questions may introduce bias, but through different mechanisms.
| Feature | Loaded questions | Leading questions |
| Main problem | Contains an unsupported assumption | Encourages a particular answer |
| Typical mechanism | Presupposition | Suggestive wording or framing |
| Example | Why did you stop trusting our company? | Don’t you agree our company is trustworthy? |
| Main risk | Respondents implicitly accept a false premise | Respondents are steered toward a response |
| Best correction | Establish or remove the assumption | Use neutral, balanced wording |
A question can be both loaded and leading.
For example:
“How much do you appreciate the improvements we’ve made to our excellent customer service?”
This assumes improvements have occurred and uses positive language to encourage favorable responses.
For a deeper explanation, see Leading Questions: Definition, Examples & How to Avoid Them.
Loaded questions vs. double-barreled questions
Loaded questions should also be distinguished from double-barreled questions.
A double-barreled question asks respondents to evaluate two or more distinct concepts using one answer.
For example:
“How satisfied are you with our pricing and customer service?”
The problem is that respondents may feel differently about pricing and customer service.
A loaded question, by contrast, contains an assumption:
“Why are you dissatisfied with our pricing?”
The question assumes dissatisfaction.
These problems can overlap.
For example:
“Why are you dissatisfied with our high prices and unreliable customer service?”
This question is loaded because it assumes dissatisfaction and describes the service negatively.
It is also double-barreled because it combines pricing and customer service.
Correcting it requires separating the concepts and removing the assumptions.
For example:
- “How satisfied or dissatisfied are you with our pricing?”
- “How satisfied or dissatisfied are you with our customer service?”
Researchers can then ask appropriate follow-up questions based on the responses.
How to identify loaded questions in a questionnaire
Loaded questions are not always obvious.
Some contain strong emotional language, while others appear neutral but still make assumptions.
Here are five warning signs.
-
The question assumes a behavior occurred
Examples include:
- “Why did you cancel your subscription?”
- “When did you stop using the service?”
- “What made you switch providers?”
These questions are appropriate only if the researcher has already established that the respondent performed the behavior.
-
The question assumes an opinion exists
Consider:
“Why do you dislike our new pricing structure?”
This assumes the respondent dislikes the pricing.
A neutral question would first measure their opinion.
-
The question assumes an outcome
For example:
“How has our new software increased your productivity?”
This assumes a positive outcome.
A better question would allow respondents to report an increase, decrease, or no change.
-
The question asks for an explanation before establishing the facts
Questions beginning with “why” are not inherently loaded.
However, they can become loaded when they ask respondents to explain something that has not been confirmed.
“Why did you stop using our product?” is appropriate for respondents who have already indicated that they stopped using it.
It is inappropriate when presented to everyone.
-
The question makes disagreement difficult
Some loaded questions leave respondents without a clear way to reject the assumption.
For example:
“Which of our premium services provides the best value?”
Someone who believes none of the services provides good value may struggle to answer.
Adding “None of the above” can help, but a more neutral question structure may be preferable depending on the research objective.
How to avoid loaded questions: 8 best practices
-
Identify the assumptions behind each question
Before adding a question to a survey, ask what must be true for the question to make sense.
For example:
“Why did you change your insurance provider?”
This assumes the respondent changed providers.
If that behavior has not been established, introduce a screening question.
-
Separate screening questions from follow-ups
Use screening questions to determine whether respondents have relevant experiences.
For example:
Screening question: “Have you contacted customer support in the past six months?”
Follow-up for those who answered yes: “How satisfied or dissatisfied were you with your most recent interaction?”
Additional follow-up: “What contributed most to your rating?”
This structure collects information without assuming that everyone has contacted support or had a particular experience.
-
Use neutral language
Remove words that evaluate an issue before respondents do.
Terms such as “unfair,” “disappointing,” “successful,” and “improved” can introduce assumptions.
Instead of asking:
“How disappointing was the product launch?”
Ask:
“How would you evaluate the product launch?”
-
Allow respondents to reject the premise
When appropriate, include options such as:
- None of the above
- No change
- Not applicable
- I have not experienced this
- Don’t know
These options help prevent respondents from selecting inaccurate answers simply because the questionnaire offers no alternative.
However, response options should not substitute for fixing fundamentally problematic wording.
-
Avoid assuming positive or negative outcomes
Questions should allow respondents to describe the full range of possible experiences.
Instead of:
“How much did the training improve your performance?”
Ask:
“What effect, if any, did the training have on your performance?”
The response scale can then include improvement, no change, and deterioration.
-
Use survey logic carefully
Questionnaire routing can help prevent loaded questions by ensuring respondents receive only relevant follow-ups.
For example, a respondent who has never used a product should not receive questions asking why they stopped using it.
Similarly, someone who reports no difficulties should not automatically receive a question asking what caused those difficulties.
Well-designed skip logic helps preserve the relevance and meaning of each question.
-
Test questions with real respondents
Cognitive interviews and pilot surveys can reveal assumptions that researchers have overlooked.
Ask participants to explain what they think a question means.
If respondents say that a question assumes something untrue about them, revise it.
Testing is particularly useful for sensitive subjects, complex experiences, and questions involving personal circumstances.
-
Review the questionnaire from the respondent’s perspective
Researchers often know more about the study’s objectives than respondents do.
This familiarity can make assumptions seem obvious or harmless.
Before launch, review the questionnaire as if you knew nothing about the organization, product, or hypothesis.
Ask whether each question would make sense to someone with a completely different experience or opinion.
Can loaded questions ever be appropriate?
Yes, but only when their assumptions are justified by the research context.
For example:
“Why did you cancel your subscription?”
This question is appropriate when it is shown only to respondents whose cancellation has already been confirmed.
The assumption is no longer unsupported.
Similarly, a researcher conducting a follow-up interview with someone who previously reported dissatisfaction can reasonably ask:
“What contributed to your dissatisfaction?”
The distinction is between an assumption that has been established and one that has not.
Loaded questions can also appear deliberately in experiments studying framing, persuasion, or the effects of misleading information.
In those cases, the wording may be part of the experimental treatment.
However, researchers should distinguish between deliberately testing the effects of a frame and unintentionally introducing bias into an ordinary survey question.
A question is not problematic merely because it contains a premise. It becomes problematic when the premise is unsupported, disputed, or inappropriate for the respondent.

How loaded questions affect different types of research
Customer satisfaction research
Customer surveys sometimes assume that respondents have experienced a problem or improvement.
For example:
“Why was your delivery experience disappointing?”
A better approach is to measure delivery satisfaction first and ask follow-up questions based on the response.
This allows researchers to accurately identify both positive and negative experiences.
Employee engagement research
Employee questionnaires can contain assumptions about management, workplace culture, or organizational change.
For example:
“Why are employees unhappy with the new policy?”
This assumes dissatisfaction across the workforce.
Neutral questions can establish how employees actually perceive the policy before exploring their reasons.
Market research and product testing
Product research often evaluates new features, packaging, pricing, and brand perceptions.
Loaded questions can exaggerate product appeal by assuming respondents recognize certain benefits.
For example:
“Which of our innovative features would you use most?”
A more neutral question would first establish whether respondents consider any of the features useful.
Political polling and public opinion research
Political questions may contain assumptions about candidates, policies, or social issues.
For example:
“Why has the government failed to address this problem?”
This assumes failure before respondents have evaluated the government’s performance.
Neutral wording allows researchers to measure public opinion without embedding a predetermined conclusion.
Healthcare research
Healthcare questionnaires may involve sensitive experiences and personal circumstances.
Questions such as “Why did you ignore your doctor’s advice?” can be inappropriate if nonadherence has not been established.
They may also make respondents feel judged.
Neutral, behavior-focused wording and appropriate screening questions can produce more useful information.
How Survox supports accurate questionnaire execution
Avoiding loaded questions begins with good survey design.
Technology cannot automatically determine whether every assumption in a questionnaire is justified, but it can help research teams implement questions and routing consistently.
Survox by Enghouse Insights supports professional data collection across CATI, IVR, online, and multi-mode research.
Its questionnaire logic allows researchers to create conditional question paths based on respondents’ previous answers.
This is particularly useful when questions should only be shown to people with specific experiences.
For example, researchers can establish whether someone has used a service, reported a problem, or changed providers before asking follow-up questions.
Survox also supports centralized data collection, quota management, and fieldwork monitoring, helping teams manage complex studies while maintaining consistent questionnaire execution.
Explore how Survox supports professional questionnaire design and multi-mode survey research.
If your surveys use follow-up questions based on earlier responses, Survox’s questionnaire logic can help ensure those follow-ups only reach respondents for whom the underlying assumption is true.
Checklist: How to identify loaded questions before launch
Use this checklist when reviewing a questionnaire:
- Does the question assume a behavior occurred?
- Does it assume the respondent holds a particular opinion?
- Does it assume a positive or negative outcome?
- Has the assumption been established through an earlier question or reliable information?
- Can respondents reject the premise if it does not apply?
- Are response options neutral and sufficiently complete?
- Does the question ask for an explanation before establishing the relevant experience?
- Could the wording make respondents feel judged?
- Is the question shown only to respondents for whom it is relevant?
- Has the wording been tested with real respondents?
If a question contains an unsupported assumption, remove it or establish whether it applies before asking the follow-up.
Frequently asked questions about loaded questions
What is a loaded question?
A loaded question contains an assumption that has not been established and that respondents may not accept. For example, “Why did you stop using our product?” assumes that the respondent previously used the product and has since stopped.
What is an example of a loaded question?
“Have you stopped making mistakes at work?” is a loaded question because it assumes the respondent previously made mistakes. Answering yes or no may make it seem like you accept that premise.
Why are loaded questions bad in surveys?
Loaded questions can pressure respondents into accepting assumptions that do not reflect their experiences. This can produce misleading answers, increase frustration, and reduce the validity of survey results.
What is the difference between loaded and leading questions?
Loaded questions contain unsupported assumptions, while leading questions encourage a particular response. A question can be both loaded and leading, but the two problems are conceptually different.
Are loaded questions always biased?
Not necessarily. A question that contains a premise can be appropriate when that premise has already been established. The problem arises when respondents are asked to accept an unsupported or disputed assumption.
How do you fix a loaded question?
Identify the assumption, remove it where possible, or introduce a screening question to establish whether it applies. Then ask an appropriate follow-up question.
Are loaded questions the same as double-barreled questions?
No. Loaded questions contain assumptions, while double-barreled questions combine multiple concepts into a single response. Some questions can have both problems.
Can loaded questions affect survey response rates?
They can. Respondents may skip questions or abandon a survey when the wording feels inaccurate, judgmental, or impossible to answer appropriately.
What is the loaded question fallacy?
The loaded question fallacy occurs when a question contains an unproven assumption that a direct answer may appear to accept. It is also associated with the complex question fallacy.
Can skip logic prevent loaded questions?
Skip logic can help by showing follow-up questions only to respondents for whom the underlying assumptions have been established. However, researchers must still review the wording itself for bias and ambiguity.
Conclusion
Loaded questions can quietly undermine survey research by embedding assumptions that respondents may not share.
They often appear in otherwise well-written questionnaires, especially when researchers are familiar with the topic or expect a particular outcome.
The solution is to identify those assumptions before fieldwork begins.
Ask neutral screening questions, use appropriate follow-ups, provide relevant response options, and test the questionnaire with real respondents.
Good survey questions establish the facts before asking respondents to explain them.
That principle helps researchers collect answers that reflect genuine experiences rather than assumptions built into the questionnaire.