Questions to Ask Scientists
Questions for interviewing a researcher, hosting a science talk, or talking with a scientist at an outreach event, covering their current project, methods, uncertainty, funding, failure, and publishing.
The questions
Open any question for the note
What are you working on right now, and what question is it trying to answer?
Why ask it
Researchers often describe their field when asked about their work. A good answer names one question and what would count as an answer to it. A tour of the whole subfield usually means the specific project is stalled or belongs to someone else.
How did you end up in this particular corner of the field?
Why ask it
Scientific careers are shaped by who was hiring and which lab had bench space, not by a plan. The story surfaces the accidents, and it tells you which techniques they personally own rather than supervise.
What result would tell you your current explanation is wrong?
Why ask it
Being able to name the observation that would sink their model means they are still testing it. Someone who cannot, or who says the evidence is already overwhelming, has moved from investigating the idea to defending it.
What does a normal week in your lab actually look like?
Why ask it
Public accounts of science skip the week: the maintenance, the failed runs, the ordering, the meetings. The answer shows how much of research is craft and waiting, which is the part students most need to hear before they sign up.
What is the last result that surprised you?
Why ask it
Surprise marks the place where they learned something. If nothing recently surprised them, the work is either confirmatory or not being looked at closely, and both are worth knowing about.
Which part of your work is hardest to explain to people outside it?
Why ask it
The hardest thing to explain is usually the thing that most limits their conclusions: an instrument threshold, a statistical correction, a stand-in measurement. It is also where non-specialists misread the findings.
How long does it take you to go from an idea to knowing whether it worked?
Why ask it
The cycle length sets everything else: how much risk the lab can take, how long a student's project has to be, why certain questions go unasked. Long cycles explain scientific caution better than temperament does.
What does your data look like before it becomes a figure?
Why ask it
Figures sit at the end of a chain of cleaning and choices. Asking about the raw form surfaces what was excluded and why, which is usually where honest disagreement between labs lives.
How confident are you in the finding you are best known for?
Why ask it
Asking about their most cited result invites a real calibration. Strong scientists separate the part that has replicated from the part that has not. Uniform confidence across everything they have published is a warning sign.
Which claim in your field gets repeated in the press but is shakier than it sounds?
Why ask it
Researchers know which of their field's headlines are thin, and they will usually say so when asked directly. For a journalist or teacher this is the most useful thing in the conversation, and it never appears in a paper.
Who funds your work, and does that shape which questions you can ask?
Why ask it
Funding shapes questions long before it shapes conclusions: grant cycles favor projects that finish in three years and summarize well in a proposal. The answer explains gaps in the published literature.
What happens to a project that fails? Does anyone else hear about it?
Why ask it
Most experiments fail and almost none of that reaches print. Whether failures are logged, shared at group meeting, or quietly dropped tells you how much of the record you are not seeing.
How do you decide a result is ready to publish?
Why ask it
Publication timing is a judgment, not a rule. Some labs wait for a mechanism, others publish the observation and move on. The answer tells you how settled to treat their output.
What do reviewers usually push back on in your papers?
Why ask it
Reviewer objections point at the field's genuine soft spots: sample sizes, missing controls, alternative explanations. It gets you to the live argument faster than reading a discussion section.
Which tool or technique changed what your field could do in the last ten years?
Why ask it
Bursts of progress usually follow a new instrument or a price drop rather than a new idea. The answer dates the questions the field is currently able to ask.
What do you and your closest colleagues disagree about?
Why ask it
Disagreement between people who share a coffee room is the live edge of a field. Consensus statements hide it, and no review article will name it this plainly.
Money aside, what would let you answer your question faster?
Why ask it
Taking funding off the table forces a specific bottleneck into view: a reagent, a collaborator, an approval process, one person trained on one machine. That is the real constraint on the work.
What does your field still get wrong about training new researchers?
Why ask it
Training is where scientists are most candid, because most of them were trained badly in some respect. Answers name the parts of the job nobody teaches: writing, managing people, deciding when to stop.
If this work succeeds completely, what changes outside the lab, and when?
Why ask it
Real-world timelines are usually far longer than press releases imply. A researcher willing to say decades, or possibly never, is being straight with you about translation.
What question in your field do you expect to still be open when you retire?
Why ask it
Naming a problem that will outlast their career shows how they size questions. It also separates what they are genuinely pursuing from what they gesture at in grant summaries.
Interviewing a researcher
Practical guidance for the conversation itself
Before you talk to them
Read one paper's abstract and figures
You do not need the methods section. The abstract tells you the claim and the figures tell you what was measured. Twenty minutes here changes every question you ask afterwards.
Separate their work from their field
Most scientists contribute one narrow piece. Asking a climate modeler about ocean chemistry gets you a polite non-answer. Find the specific piece and ask about that.
Get the name and title right once
Confirm their role and institution at the start rather than guessing. Doctor, professor, postdoc, and research fellow are not interchangeable, and getting it wrong early costs you standing for the rest of the conversation.
During the conversation
- Ask for the number: how many samples, how many years, how large an effect.
- When a claim arrives, ask how they know rather than whether they are sure.
- Let silences run. Scientists often qualify their own statements if given three seconds.
- Ask them to say the same thing again in one sentence, then read it back and check it.
- Ask what else could explain the result before you ask what it means.
- Write down the caveats, since those are the first things lost in a summary.
Common mistakes
Asking for a verdict on a contested question
Pressing for a yes or no on something the field has not settled produces either a hedge or an overstatement. Ask instead where the disagreement sits and what evidence would resolve it.
Treating a preprint as a finished result
Preprints have not been reviewed. Ask whether the work has been through review, whether anything changed, and whether anyone else has reproduced it.
Asking for free consulting
There is a difference between asking about their research and asking them to analyze your data or vet your product. If that is what you want, say so and expect to pay for it.
If you are recording or writing it up
- Ask permission to record, and say what the material is for and where it will appear.
- Offer to check quotes for accuracy, not for approval of your conclusions.
- Keep the uncertainty in. Cutting may and suggests to tighten a sentence changes the claim.
- Name the funder if the work is paid for by an interested party.
- Send the finished piece afterwards, which is how you get a second interview later.