# What to Do When You Don't Know the Answer During a Presentation The most common situation behind a frozen moment in a scientific presentation is simple. Someone asks you a question that falls outside the boundaries of your own data, and you have never thought about it before. The first thing to check is whether the question is about your work or about work adjacent to yours. A question about your own result usually has an answer you can reason toward on the spot. A question about a neighboring field, a reagent you did not use, or a study you never read usually does not, and the correct response is different for each case. Not knowing an answer during a talk is not a failure of expertise. Every researcher, from a first-year graduate student to a senior principal investigator, gets questions they cannot answer. What separates a strong scientist from a struggling one is not encyclopedic memory. It is the ability to handle the gap gracefully, buy time, reason aloud, and close the loop afterward. This guide covers the causes of that stuck feeling, the responses that actually work, a diagnostic sequence you can practice, and the preparation habits that shrink the number of surprises at your next talk. ## Why This Happens in a Scientific Presentation The "what to do when you don't know the answer in presentation causes" question has a small set of answers that cover nearly every case. Understanding which cause you are facing tells you which fix to reach for. **The question is genuinely outside your study.** Someone asks about a mechanism you did not test, a cell type you did not measure, or a population you did not sample. No amount of preparation would have given you this answer, because the experiment was never done. This is the most common cause, and it is the one with the cleanest fix. **You knew it once but lost it under stress.** Public speaking raises arousal levels, and high arousal narrows working memory. The fact that you could recite an antibody clone number last week and cannot recall it in front of forty people is a memory retrieval problem, not a knowledge problem. Stress also distorts time, so a three-second silence feels like thirty seconds, which pushes people to blurt out a guess. **The question is ambiguous or bundles several questions together.** A long, layered question ("Given your dose response, and considering the field's recent shift toward in vivo models, and assuming the reviewer concern about batch effects, how do you interpret the endpoint?") overloads working memory. You do not actually lack the answer. You lack a parsed version of the question. **The question rests on an assumption you do not accept.** "Since your treatment clearly reduced proliferation, how did you control for the confound?" contains an embedded claim you may not agree with. Answering it directly accepts a false premise, and your hesitation is really your brain flagging the premise, not the answer. **The questioner wants something other than information.** Some questions are job talks in miniature, or turf disputes, or tests of whether you will cave. The content of the question matters less than its purpose, and an answer you cannot give may still be a conversation you can win. **You are being asked about your own data and you genuinely do not know.** This is the rarest and most uncomfortable cause. Perhaps a control failed, an outlier behaved strangely, or a collaborator ran an analysis you did not sit with long enough to own. It happens, and it has its own playbook below. ## The Troubleshooting Table This table is the core of the article. Find the situation you are in, confirm the likely cause, and use the matching fix. | Symptom in the room | Likely cause | How to confirm it | Fix | |--|--|--|--| | The question concerns a method, reagent, or topic you never touched | Question outside your study scope | You cannot recall any experiment in your design that addresses it | Say so plainly, offer to follow up, and point to the closest relevant finding you do have | | You knew the fact yesterday but it vanishes now | Stress-related retrieval failure | The information feels familiar and on the tip of your tongue | Buy time with a restatement, then reason aloud from what you can recall. Offer to verify the exact value in writing | | The question is long, layered, or hard to parse | Ambiguity or question overload | You cannot repeat the question back accurately | Ask a clarifying question or break it into parts and answer them one at a time | | You feel resistance before you start answering | False or contested premise inside the question | Repeating the question aloud makes the assumption audible | Name the premise politely, state what your data actually support, then address the rest | | The questioner seems more interested in the exchange than the answer | Status challenge or turf defense | The tone is pointed, the audience is watching, the questioner has history | Stay calm, acknowledge the point, answer the legitimate core, and offer a private follow-up | | You realize mid-answer that you do not have the data | Genuine gap in your own results | You start inventing a number and stop yourself | Stop, retract the half-answer, and switch to a deferral. Never leave a fabricated figure on the record | | Someone asks about a paper you have not read | Literature gap | The citation is specific and you have no memory of it | Say you have not read it, ask for its main claim, and respond to the claim rather than the paper | | You blank completely for several seconds | Total retrieval failure under arousal | Silence is stretching past comfortable | Use a scripted recovery line (below), move to the next tool, and keep your posture open | ## The Solutions That Work, in Order of Preference These are the "what to do when you don't know the answer in presentation solutions" that experienced presenters use. Practice them as spoken scripts, because under stress you will fall back on rehearsed language. ### 1. Defer honestly and commit to a follow-up The single most reliable response, and the one senior scientists use most often, is a clean deferral with a commitment. A working script: > "That is a good question and I do not have the data to answer it directly. My study did not test that endpoint. I would be happy to look into it and follow up with you after the session." Three things make this work. First, the admission is specific. "I do not have the data" is a statement about the experiment, not about you. Second, the follow-up commitment is concrete, and you must actually honor it, ideally with a short email within a day or two. Third, the whole exchange takes ten seconds, which keeps the session moving. Chairs and session organizers notice presenters who handle questions efficiently, and a crisp deferral reads as competence, not weakness. A deferral only works if you deliver afterward. Keep a notebook or a note on your phone during the question session, write down each question you deferred with the asker's name, and close every one of those loops within a week. An unanswered promised follow-up damages your reputation far more than the original "I don't know" ever could. ### 2. Restate the question to buy time and confirm understanding Repeating the question in your own words does three jobs at once. It buys five to ten seconds of thinking time. It confirms you understood correctly, which matters because question sessions are often noisy and questions are often mumbled. And it lets the questioner correct you if you misheard, which is far better than answering the wrong question for two minutes. > "So if I understand correctly, you are asking whether the effect we see in the knockout persists at the lower dose. Is that right?" If the question was long and layered, restate it as a single short question. Ask the questioner to confirm which part they most want answered. This is not stalling. It is standard practice in any rigorous scientific presentation. ### 3. Reason aloud from your own data Many questions you "do not know" are actually questions you can approach with the data in front of you. If someone asks whether a treatment effect would hold in an older cohort, and you only studied young adults, you cannot answer definitively. But you can reason transparently: > "We did not test that directly. What I can say is that the effect size was stable across the age range we did sample, which spans 18 to 35. Whether it holds in older cohorts is exactly the kind of follow-up study this result motivates." This response converts a gap into a roadmap. It shows the audience how you think, which is often what they are really evaluating. The key discipline is to label every step of the reasoning as inference. Use phrases like "based on," "I would expect," and "we did not test, but." Never let an inference drift into sounding like a result. ### 4. Ask a clarifying question before answering When a question is ambiguous, answering it is a gamble. A short clarifying question is cheaper: > "When you say the effect is inconsistent with the field, which papers are you thinking of?" or > "Are you asking about the [statistical significance](/blog/guides/statistical-significance) of the primary endpoint, or the biological magnitude of the change?" Clarifying questions also slow down hostile exchanges, because they force the questioner to state their point plainly, which often deflates it. ### 5. Separate the premise from the question If the question contains an assumption you do not accept, address the assumption first, politely and briefly. > "I would push back slightly on the premise. Our data show a reduction in one marker, which does not by itself establish reduced proliferation. Taking the question as about the marker specifically, the answer is that the change was consistent across all three replicates." If the premise is simply wrong, say so without heat, and offer the evidence. If the premise is contested in the field, say that too. Never accept a false premise just because it came embedded in a confident question. Audiences respect a calm, factual correction more than a fudged answer. ### 6. Admit the literature gap directly When someone cites a paper you have not read, honesty wins: > "I have not read that paper, so I cannot comment on it specifically. If their central claim is that X drives Y, that would be an interesting contrast with our model, and I would like to read it. Could you send me the reference?" This response does four things. It refuses to fake familiarity, which experienced audiences detect instantly. It extracts the substance of the question without the citation. It opens a follow-up channel. And it costs you almost nothing socially, because every scientist in the room has been in the same position. ## What Not to Do The failed responses follow recognizable patterns, and each has a specific mechanism behind why it backfires. **Guessing a number.** A confident fabricated statistic, dose, or p-value becomes part of the record. Someone in the audience writes it down, cites it, or repeats it. When the true value surfaces later, your credibility takes a hit far larger than the original question warranted. If you catch yourself mid-guess, stop and retract: "Actually, let me correct that. I should not quote a figure I am not certain of. I will follow up with the exact value." **Bluffing with jargon.** Long, vague, jargon-heavy answers to simple questions are transparent to expert audiences. The audience may not know the answer, but they know what a non-answer sounds like. **Deferring on everything.** If you defer four questions in a row, including ones you could reasonably attempt, the audience concludes you did not prepare. Defer when you truly lack the data. Attempt when you can reason from what you presented. **Getting defensive.** A pointed question is not automatically an attack. Treating it as one, with a stiff posture or an edge in your voice, reads as insecurity. The fix is behavioral, not intellectual: keep your stance open, thank the questioner, and answer the content rather than the tone. **Arguing with the questioner to win.** Winning an exchange while looking combative loses the room. The goal of a question session is to demonstrate scientific judgment, not to defeat an opponent. **Freezing in silence with closed body language.** A long silence feels unbearable to the presenter and far less awkward to the audience, but silence paired with a dropped gaze and a turned-away torso reads as panic. If you need a moment, narrate it: "Let me think about that for a second." Narrated silence is composure. ## A Diagnostic Sequence You Can Practice ![Decision path sorting a question into answer now, reason aloud, clarify, or defer with a follow-up commitment.](/article-images/illustrations/what-to-do-when-you-dont-know-the-answer-in-presentation-question-triage.webp) When a question lands and you feel that stuck feeling, run this sequence in order. With practice it compresses into a few seconds. 1. **Listen fully.** Do not start composing your answer while the question is half finished. Half-heard questions produce wrong answers. 2. **Classify it.** Is this about my data, my methods, my interpretation, the broader field, or something I did not test? Most questions fall into one of these five bins. 3. **Restate it.** One sentence, ending with a check: "Is that right?" 4. **Ask yourself: do I know this, can I reason to it, or is it genuinely outside my scope?** These three outcomes map to three responses. 5. **If you know it, answer it concisely.** Two or three sentences, then stop. Rambling answers invite follow-up attacks. 6. **If you can reason to it, reason aloud and label your inference.** Keep the chain short and flag every speculative step. 7. **If it is outside your scope, defer with a commitment.** Specific admission, concrete follow-up, and move on. 8. **Write down anything you deferred.** Name, question, and the date you will answer it by. ## Reading the Questioner Not every question wants the same thing, and tailoring your response to the questioner's purpose prevents most awkward exchanges. **The genuine curious colleague** wants to understand your work. Give them a direct, substantive answer, or a substantive reasoned attempt. Deferring to this person when you could attempt an answer wastes an opportunity. **The methodological examiner**, often a senior scientist or a reviewer, wants to know whether you understand the limits of your own approach. This person rewards candor. Naming your study's limitations before they do is one of the strongest moves available in a scientific presentation. **The turf defender** has a stake in an alternative approach. Acknowledge the legitimacy of their method, state why you chose yours, and find the shared ground. Avoid turning the session into a debate between two techniques. **The student or early-career scientist** may be asking a basic question and worried it sounds basic. Answer it generously and without condescension. The senior people in the room notice how you treat junior questioners. **The hostile questioner** wants a reaction. Deprive them of one. Calm, precise, brief answers, plus an offer to continue privately after the session, end most of these exchanges productively. ## Building Research Presentation Slides That Anticipate Questions Good research presentation slides reduce the number of unanswerable questions before the session even starts. Several practices do the heavy lifting. **Put your limitations on a slide.** A short, honest limitations slide answers the examiner's questions before they are asked, and signals that you understand your own study's boundaries. Presenters who hide limitations get them extracted in public. Presenters who state them control the framing. **Build backup slides.** Backup slides are slides placed after your closing slide, containing material you expect to be asked about. Typical backups include full statistical output, additional replicate data, the exact reagent and protocol details, an alternative analysis, and a timeline slide. When a question touches one, you say "I have a slide on that" and jump to it. This is the single highest-return preparation step for a question session, because it converts a likely stumble into a demonstration of foresight. Three to six well-chosen backups cover most sessions. **Know your own numbers cold.** Sample sizes, effect directions, p-values for your primary comparisons, and concentrations or doses you used should be recallable without the slide. If you cannot state your own n, no deferral script will save the impression. **Slide design also matters for question flow.** A slide crammed with dense text forces the audience to read instead of listen, and generates questions about details you cannot see from the podium. Keep one main message per slide, label axes and units on every figure, and make sure any number you show is one you can defend. **Rehearse the question session, not just the talk.** Ask a colleague to read your slides cold and fire five difficult questions at you. Practice the restatement, the reasoned attempt, and the deferral out loud. The first time you deliver these scripts should not be in front of a hiring committee. ## Recovering From a Full Blank ![Flowchart from a scripted recovery line through restating the question to a reasoned attempt or a clean deferral.](/article-images/illustrations/what-to-do-when-you-dont-know-the-answer-in-presentation-full-blank-recovery.webp) Occasionally the retrieval failure is total. Your mind goes empty, the slide looks unfamiliar, and the seconds stretch. A scripted recovery works here because it gives you something to say while your memory reboots. > "Give me one moment, I want to make sure I answer this properly." Then restate the question. Restating forces you to process the words, which usually unlocks the relevant memory. If it does not, drop to the next tool in your sequence, a reasoned attempt or a deferral, and keep going. The audience has almost no visibility into your internal state. A presenter who says "let me think about that" and then answers, or gracefully defers, appears composed. Only you know there was a blank. Physical levers help too. Take one slow breath, plant both feet, and keep your hands visible and relaxed. Slowing your speaking rate by even ten percent lowers arousal and improves recall. ## Common Mistakes and Limitations These are the errors that show up repeatedly, in lab meetings and at international conferences alike. - **Treating "I don't know" as a confession of incompetence.** It is not. It is a statement about the boundary of a dataset. Senior audiences evaluate the handling, not the gap. - **Overcommitting to follow-ups you will not deliver.** Promising to "look into it" for six separate people and answering none of them is worse than six clean deferrals with three real follow-ups. Prioritize the questioners who asked substantive questions. - **Answering the question you wish you had been asked.** Reframing is sometimes legitimate, but an audience notices when every hard question gets redirected to your favorite slide. Redirect sparingly and acknowledge the original question explicitly. - **Assuming a vague question is a dumb question.** Sometimes the questioner is struggling with English, or thinking out loud, or nervous. Restating their question generously often reveals a sharp point underneath. - **Treating every scripted response as foolproof.** Scripts are training wheels. If you deliver "that is an excellent question" mechanically before every answer, it becomes audible filler. Use the underlying moves, not the exact same words every time. - **Ignoring the chair.** In sessions with a moderator, the chair controls time and turn order. Fighting the chair for one more minute loses goodwill on both sides. If time runs out, offer the deferred follow-up and sit down. - **Assuming a smooth question session means a successful talk.** Some of the best talks generate sharp, uncomfortable questions, because sharp audiences engage with sharp work. A friendly, empty question session can mean a flat talk. One general limitation applies to all of this. Presentation dynamics vary by field, institution, and audience culture, and individual situations differ. The principles here are widely applicable, but the specific scripts that feel natural to you will come from practice in your own environment. ## Prevention Practices Prevention is where the biggest gains live. The habits below shrink the pool of surprises year over year. **Anticipate questions systematically.** After finishing your slide deck, read through it as a skeptical reviewer and write down every question you would ask. Sort those questions into three groups: ones you can answer, ones you can reason toward, and ones you cannot. Prepare answers for the first group, reasoning chains for the second, and deferrals plus follow-up plans for the third. **Update your backup slide bank after every talk.** Every question you deferred becomes a candidate backup slide for the next presentation of the same work. Over a conference season, your deck grows a robust question-answering appendix that no surprise can penetrate. **Maintain a personal fact sheet.** Keep a one-page summary of your project's key numbers, reagents, model details, and statistical choices. Review it before every talk. This turns many "I don't know" moments into "as I noted, the n was twelve per group" moments. **Close every deferred loop.** Follow-up answers delivered promptly build a reputation for reliability, and people remember how you handled the gap more than the gap itself. A short email with the answer, plus the reference or analysis you promised, takes fifteen minutes and buys years of credibility. **Debrief each session.** Within a day of every talk, note which questions stumped you and why. Was it scope, memory, ambiguity, or a false premise? Over time you will see your personal pattern, and you can prepare for the type of question that trips you most. **Present often in low-stakes settings.** Lab meetings, journal clubs, and practice sessions are where the recovery scripts become reflexes. The presenter who has deferred a question fifty times in practice delivers a clean deferral on stage without thinking about it. ## Frequently Asked Questions ### What should I say first when I do not know the answer? Restate the question in your own words and ask if you understood correctly. This buys thinking time, confirms the question, and lets the questioner correct you before you commit to an answer. ### Is it bad to admit you do not know in a scientific presentation? No. A specific, honest admission paired with a concrete follow-up commitment reads as integrity and precision. What damages credibility is guessing, bluffing, or being caught in an invented answer later. ### How do I handle a question about statistics I did not run myself? Defer to the analysis and commit to a written follow-up. Say that the analysis was performed by a collaborator or core facility, that you cannot speak to its details from the podium, and that you will connect the questioner with the right person or send the details yourself. ### What if the questioner keeps pushing after I defer? Acknowledge their persistence, restate what you can and cannot claim from your data, and offer to continue the discussion privately after the session. If the chair is present, let the chair manage time and turn order. Do not escalate. ### How should I follow up after the presentation? Send a short email within one or two days. Thank the person for the question, answer it directly or point to the reference or data that does, and keep the message brief. If you need time to check something, say when you will reply again and then do it. ### How many backup slides should I prepare? Three to six is a practical range for most talks. Include your full statistical output, additional replicates, protocol or reagent details, and any obvious alternative interpretation of your main figure. Reuse and update them across presentations of the same project. ### How do I recover if I blank completely during my talk? Say a brief scripted line such as "give me one moment, I want to answer this properly," take a slow breath, and restate the question out loud. Restating usually reopens memory. If it does not, move to a reasoned attempt or a clean deferral and continue. ### What do I do when the questioner's premise is wrong? Name the premise politely and correct it before answering. State what your data actually support, then address whatever remains of the question. A calm factual correction earns more respect than accepting a false starting point.