You have one microphone, one minute and a question you actually care about. You ask it. The expert on stage smiles, says “That’s a great question,” and spends ninety seconds on something adjacent: the importance of the issue, their organisation’s commitment to it, a study that is related but not quite relevant. The moderator thanks them and moves on. You sit down with the vague sense of having been answered and the specific knowledge that you were not.
Most people blame the expert. Sometimes that is fair. But the more useful diagnosis is that most questions put to experts are designed, unintentionally, to be dodged. They are broad, they invite a statement rather than a commitment, and they come with no follow-up. An expert who has given the same talk forty times has a prepared answer for every broad question in their field. The talking point is not evasion so much as the path of least resistance, and the question laid the path.
The argument of this piece is that getting a real answer is mostly a design problem, and one that health professionals are unusually well trained to solve. A good clinical history is exactly this skill: asking specific questions, noticing when the answer doesn’t match the question and following up until you have something you can act on. We just don’t apply it when the person in front of us is on a stage.
Why dodges work
There is useful research on this. In a 2011 paper in the Journal of Experimental Psychology: Applied, titled “The Artful Dodger,” Todd Rogers and Michael Norton of Harvard studied what happens when speakers answer a different question from the one they were asked. They found that listeners often failed to notice the dodge when the answer addressed a question similar to the one posed, and that when they didn’t notice, they didn’t think any less of the speaker. Their explanation was that audiences are usually paying attention to whether they like the speaker, not to whether the question was answered.
That finding explains a great deal about conference Q&A. The audience is not auditing answers. It is forming impressions. An expert who responds fluently and warmly to a nearby question scores well on the thing the audience is actually tracking. The person who asked may notice, but the room usually does not.
There are several familiar ways the dodge happens:
- Answering the category, not the case. Asked whether a specific programme failed, the speaker explains how difficult programmes in that area are in general.
- Answering the intention, not the outcome. Asked whether a policy reduced out-of-pocket spending, the speaker explains that it was designed to.
- Answering the future, not the past. Asked what went wrong, the speaker describes the next phase.
- Answering with process. Asked for a view, the speaker describes the consultation that will produce one.
- Answering the easier half. Asked a two-part question, the speaker answers the part they prefer and the moderator never returns to the other.
None of these requires bad faith. Experts, like everyone, drift toward the answer they have given before. But the drift is predictable, which means the question can be designed to block it.
A question is a trap for the talking point
A good question for an expert has three properties. It is specific, it is falsifiable and it forces a trade-off.
Specific means it names a thing: a programme, a number, a year, a population, a decision. “What do you think about community health worker programmes?” is an invitation to a lecture. “In the programme you described, what share of community health workers were still in post after two years?” can only be answered with a number or with an admission that the number isn’t known. Both are informative.
Falsifiable means the answer could turn out to be wrong. Ask the expert to commit to a prediction or a condition. “What result from the next evaluation would lead you to recommend against scaling this?” If the expert can name one, you have learned what they are really betting on. If they can’t, you have learned that the recommendation doesn’t depend on evidence, which is also worth knowing.
Trade-off forcing means the question puts two good things in tension and asks the expert to choose. “If a district had to choose between hiring two more nurses or funding the digital reporting system you recommended, which should it choose?” Talking points are built to avoid choices, because every choice alienates someone. A question that makes the choice unavoidable defeats the prepared answer.
A useful test before you stand up: could the expert give the same answer to your question that they would have given to the question before yours? If yes, rewrite it.
The follow-up is where the answer lives
Most of the value in questioning an expert comes after the first answer, and most event formats make follow-up nearly impossible. One question per person, a queue of others waiting and a moderator anxious about time.
Where you do get a second chance, use it on the gap. Don’t ask a new question. Point out, briefly and politely, which part of the first one was not answered. “Thank you. I was asking specifically about the two-year retention figure. Do you know it?” That formulation matters: it is courteous, it restates the question in narrower form and it gives the expert an easy way to say “I don’t know,” which is often the true answer being avoided.
A few follow-ups that work across health topics:
- “What’s the number?” when an answer is qualitative.
- “Compared to what?” when an answer claims improvement.
- “Who paid for that?” when an answer describes a successful service.
- “What happened in the districts where it didn’t work?” when an answer describes the best case.
- “What would you tell a health officer who has to decide this on Monday?” when an answer stays abstract.
The single most useful follow-up is silence. After an evasive answer, if you still have the microphone, pause for a couple of seconds before saying anything. Speakers often fill the silence with the more candid version.
What moderators can do
Moderators hold most of the leverage, and most use very little of it. A moderator who treats their role as hosting will let every talking point pass. A moderator who treats it as questioning on the audience’s behalf can transform a session.
Three practices make the largest difference. First, prepare the pressure points, not just the introductions. Before the session, the moderator should know where the speakers’ published positions are weakest, where they disagree with each other and which questions they are likely to avoid.
Second, claim the right to follow up, and say so at the start. “I’m going to come back to people if I don’t think the question has been answered” sets an expectation that makes speakers more careful and audiences more attentive.
Third, collect audience questions in a way that filters for specificity. Written or app-based questions let the moderator combine similar ones, drop the speeches disguised as questions and choose the ones that force a commitment. It is less spontaneous. It produces better answers.
What audiences can do
For people in the audience, a few habits help.
Write your question down before you stand, and cut it to one sentence. Long preambles give the speaker time to choose which part to answer.
Ask one question, not two. Two-part questions are dodge invitations.
Say what decision you are trying to make. “I run a district hospital and we are deciding whether to adopt this. What’s the one thing that would make you tell me not to?” gives the speaker a concrete person to answer, and it is much harder to deliver a talking point to someone who has just described their Monday.
Coordinate. If you and a colleague are both interested in the same weak point, have one ask and the other follow up. Sessions where two questioners press the same issue are much harder to deflect than sessions of twenty unrelated questions.
In online formats, such as an “ask me anything” session, a live chat or a webinar, the rules are the same but the advantage shifts to you. Written questions can be precise. Others can upvote the ones that matter. And the record is permanent, so a dodge is visible to everyone who reads it later.
The uncomfortable incentive
It is worth being honest about why these practices are rare. Experts on stage are often there because they are useful to the organiser, the sponsor or a partner. A moderator who presses them hard may not be asked back. An audience member who presses hard may be seen as rude, especially if they are junior, female or from a less powerful institution, and professional hierarchies in health are steep. Many people who could ask the sharp question calculate, reasonably, that the cost to them outweighs the value to the room.
That calculation is why design matters more than courage. When the format expects follow-up, when the moderator has announced it and when questions are collected in writing, the individual asking doesn’t carry the social cost alone. The structure asks the hard question, not the person.
The strongest objection
The best counterargument is that experts dodge for good reasons. Many specific questions don’t have a reliable answer, and an expert who gives a precise figure they are unsure of misleads the room. Some questions touch confidential negotiations, unpublished data or colleagues’ work. And a culture of aggressive questioning can drive the most thoughtful experts away from public events, leaving the stage to the most media-trained.
All true, and it points to the right standard. The goal isn’t to force an expert to give a number they don’t have. It is to make “I don’t know” and “I can’t say” acceptable answers, and to make them more attractive than a fluent evasion. A question designed well gives the expert an honest exit. The moderator’s job is to make sure they take it, instead of the dishonest one.
A worked example (hypothetical)
A webinar on antimicrobial stewardship in primary care. An expert presents results from a stewardship programme showing a drop in antibiotic prescribing across participating clinics.
The typical audience question: “How can we make stewardship programmes more sustainable?” The expert responds with a thoughtful overview of leadership, training and data. Nothing in it is wrong. Nothing in it is new.
A redesigned question: “In the clinics where prescribing fell, was there any change in return visits or hospital referrals for the same infections? If so, how big, and if not, was it measured?” This is specific, it has a factual answer and it probes the obvious trade-off: did reducing prescriptions cause harm elsewhere? The expert either knows and says, or doesn’t know and says, or tries to answer about sustainability instead. In the third case, the follow-up is simple: “So it wasn’t measured?”
Whatever the answer, the audience now knows something useful about the strength of the evidence. That’s the point of asking.
The takeaway
Experts give talking points because broad questions invite them and formats let them pass. You can’t make an expert candid. What you can do is ask questions that can only be answered with a fact, a commitment or an admission, and build formats where the follow-up is expected rather than rude. Health professionals already know how to take a history from a patient who would rather not say. The same discipline works on a stage.
What is the best question you have ever heard asked at a health event, and what made it impossible to dodge? Share it in the comments.
