Leading questions in research: examples and how to fix them
By Juno · Published · Updated
A leading question is one that suggests its own answer. It might name the answer or quietly limit the options. Either way, the person answers the question in front of them rather than telling you what is true for them, and the finding becomes the interviewer's instead of the participant's.
Below are illustrative examples of common patterns, why each one can distort an answer, and possible rewrites. They demonstrate general interviewing principles; they are not participant quotations or findings from a customer study.
1. Naming the answer
Bad: "So price was the main issue, right?" Better: "What led you to that view of the price?"
The question states a conclusion and asks for agreement. Most people agree, so you learn what the interviewer thought, not what the participant did.
Bad: "Did you end up applying to that job?" Better: "Then what did you do?"
Naming one possible next step turns an open story into a yes or no. Anything else the person did never gets heard.
2. Putting a judgement in the question
Bad: "How do you make sense of him not having any coverage when he brings in most of the income?" Better: "Why do you have cover and he doesn't?"
The first version carries the interviewer's view that the situation is unreasonable. People respond to the judgement, defending or agreeing, instead of explaining.
3. Asking for only one side
Bad: "Thinking about your job search so far, what's been the easiest part?" Better: "How do you feel about the general experience of browsing for new jobs right now? What's it been like for you?"
Asking only about what was easy assumes something was, and collects half the experience. Either ask openly or ask for both sides together.
4. Embedding the 'right' answer
Bad: "Do you ever consult a health professional, such as a doctor or dietitian, for nutrition advice?" Better: "Have you talked to anyone else about nutrition and eating habits?"
Naming health professionals signals the expected answer. People who rely on friends, family or nobody at all are less likely to say so.
5. Offering a false either/or
Bad: "Do you shop more on your phone or computer?" Better: "What device do you use to shop with most?"
Two options become the whole list. A tablet, a work laptop or a mix of devices has nowhere to go, and the real answer is lost.
6. Feeding examples too early
Bad: "Who do you think decides if the sensor is really bad: the machine, the manual, or someone else?" Better: "Who do you think decides if the sensor is really bad?"
Examples narrow the answer before the person has thought about it, and people tend to confirm the ones listed. Ask openly first, and offer examples only if the person struggles.
7. Assuming something they didn't say
Bad: "What do you find most challenging about eating healthier?" Better: "How do you currently feel about your normal eating habits over a week?"
The question assumes the person is trying to eat healthier. Someone who isn't will often produce a challenge anyway, to be polite, and it reads as a finding.
Bad: "At the start of your day, do you talk with the crew about safety, or do they just know the rules?" Better: "How do you communicate with your crew about safety throughout a usual day?"
This one assumes when the conversation happens, then offers two versions of it. Every other way the crew handles safety is left out.
8. Asking two questions at once
Bad: "Can you tell me about your online application experience, including what worked, what didn't, and if there were any barriers and if there was anything you didn't understand in the process?" Better: "Tell me about your online application experience."
The person spends their effort working out what is being asked. One question at a time leaves that effort for remembering and reflecting.
Bad: "Have any candidates or campaigns contacted you directly this election, and how did you react?" Better: "Have any candidates or campaigns contacted you directly this election?" Then, if yes: "How did you react?"
The second half only makes sense if the answer to the first is yes. People answer one half and skip the other.
9. Moving to a premise they didn't raise
Bad: "What would make buying feel less like a hassle?" Better: "Why do the calls feel like a hassle for you?"
The person had said calling agents was a hassle, not buying. Jumping to an imagined fix for buying in general means the real reason never gets asked about.
10. Using jargon instead of their words
Bad: "Which e-commerce sites do you use most often?" Better: "Where do you shop from online?"
Industry words make an interview feel like a survey, and people answer in the same stiff way. Be plain and human.
Bad: "Oh, so did you say that you were on a small-format device in a web browser?" Better: "You said you were on your phone. Can you tell me more about that?"
Replacing the participant's word with the company's makes them feel corrected, and people who feel corrected say less. Mirror their language.
How Juno handles this
Juno checks a study's design for leading wording, built-in assumptions and either/or framing before anyone is interviewed, and pushes back on a leading brief. Interviews are led by research objectives rather than a fixed list of questions, so the interviewer can ask openly first and follow what each person says. See our methodology standards.
Frequently asked questions
What is a leading question in research? A question that suggests its own answer, either through its wording or by limiting the options. It pushes people towards a response, so the data reflects the question rather than the participant.
Are examples in a question always leading? Not always, but they should come second. Ask openly first. If the person struggles to answer, offer examples to help them.
Is an either/or question always a problem? Only when you don't know both options cover the person's situation. If screening has already confirmed it, a choice between two can be fine. If you don't know, it leads heavily.
What's the simplest fix for most leading questions? Remove anything that suggests an answer, and ask about what the person has already told you, in their words. "Then what did you do?" and "Tell me more about that" do more work than most carefully built questions.