Habit, Not Hack: Saying I Don’t Know Early (Mentor)

A good mentor makes 'I don't know' cheaper to say than a three-week mistake.

The meeting had gone well.

At least, that's what Dr. Rao thought as he closed his laptop and glanced around the table. The data were moving. The timelines felt reasonable. Everyone had followed along — or appeared to. When he asked "any questions?" at the end, the room answered with silence.

That usually meant things were clear.

He packed up and moved on to the next thing.

Three Weeks Later

Maya knocked on his door holding her laptop the way people do when they're bracing for impact — too tight, like it might slip if she loosened her grip. The graph on her screen didn't make sense. Not because she hadn't worked hard — he could see immediately that she had, could see three weeks of careful, sustained effort in the data — but because something fundamental had gone wrong early, and everything built on top of it had been built on a foundation that wasn't there.

Dr. Rao scanned the protocol. The assay wasn't new. The lab had been running versions of it for years. He started to feel the familiar irritation of a setback that should have been preventable.

Then he saw it.

A single assumption, buried in one line of the protocol — something so second-nature to him that it had never occurred to him it needed explaining. The kind of knowledge that is invisible to the person who has it, precisely because they have had it long enough to have forgotten acquiring it. To him, it was obvious. To Maya, running this assay for the first time, it was a gap that nothing in the written procedure had prepared her for.

He looked at her.

"Why didn't you ask sooner?"

The words came out sharper than he intended.

She hesitated. "I thought I should be able to figure it out."

What That Answer Revealed

Dr. Rao sat with that sentence for longer than the moment required.

He had said, at the start of this project, that the assay was straightforward. He had said it was standard. He had asked "any questions?" at the end of the meeting and moved on when none came.

He started hearing those statements through a different filter.

When he said "this is straightforward," Maya had heard: you shouldn't need help with this. When he said "this is standard," she had heard: everyone who is competent already knows this. When he asked "any questions?" and the room was quiet, he had heard confirmation that everything was clear. She had heard that the meeting was over and questions were no longer on the table.

The silence hadn't been understanding. It had been a calculation — a risk assessment that had concluded, on the basis of his language and his manner and the ambient culture of the lab, that admitting confusion was more costly than attempting to work through it alone.

She had made that calculation for three weeks. She had been wrong about the cost, but she hadn't been wrong about the signals she was reading.

He had built those signals, without meaning to, one meeting at a time.

What He Did in That Moment

Dr. Rao pulled a chair closer to her laptop.

"This part," he said, pointing to the line in the protocol, "is on me. I skipped an explanation because it's second nature to me. That doesn't mean it should have been second nature to you."

They worked through it together. Ten minutes. That was what three weeks of failed experiments had cost — ten minutes of explanation that had never happened because neither of them had known it was missing.

After Maya left, Dr. Rao stayed in his office longer than usual.

He thought about the number of times he had mistaken silence for comprehension. How many meetings had ended with a round of "any questions?" met with silence that he had read as clarity and his trainees had experienced as a closed door. How often he had praised independent problem-solving without realizing he was, in the same gesture, signaling that asking for help was a lesser alternative. How the culture he had built — unintentionally, incrementally — had been taxing the most conscientious trainees most heavily, because the most conscientious ones were the most reluctant to appear as though they didn't know what they were doing.

The cost had not been Maya's. It had been his — distributed across three weeks of her time, the lab's reagents, and the timeline of a project that was now behind for reasons that had nothing to do with her capability.

The Habit: Making Confusion Safe To Admit Early

Dr. Rao made one change to how he opened presentations of new work. Before presenting, he started saying something he had never said before:

"If something here doesn't make sense as I explain it, that is not a reflection of you. It means I haven't explained it well enough yet. Confusion is information about my explanation, not about your ability."

He also changed the question he asked at the end.

He retired "any questions?" — which was a question that required trainees to volunteer vulnerability in a public setting, in front of their peers, with no guarantee of how the admission would be received. He replaced it with something that changed the cost-benefit calculation:

"What part of this would you feel least confident explaining to someone else?"

This question did something "any questions?" couldn't. It normalized partial understanding as the expected state rather than as an exception requiring disclosure. It made the answer something everyone in the room was likely to have, rather than something that distinguished the confused from the capable. And it gave trainees language for a specific kind of uncertainty — not "I don't understand" but "I understand most of it and am least sure about this part" — that felt more accurate and less exposing.

The first time he asked it, hands went up.

Not because his lab had become less capable. Because it had become safer to be honest early.

What He Also Started Modeling

Dr. Rao also changed what he did when he didn't know something himself.

Previously, in the way of most senior researchers who have learned to project authority, he would deflect or redirect when a question exceeded his immediate knowledge. Not dishonestly — but smoothly, in ways that maintained the impression that his uncertainty was temporary and minor.

He started doing the opposite. When he didn't know something, he said so directly, in front of his trainees, and then modeled what to do about it.

"I'm not sure about this — let's look it up together."

"I don't know the answer to that off the top of my head. Good question. Let me find out and come back to it."

"I'd have to check the protocol on that. I don't want to guess on something that matters."

Each of these was a demonstration that "I don't know" was not a statement that undermined credibility. It was a statement that preceded finding out — which is, in fact, how research works, and which his trainees needed to see modeled by someone with standing before they would believe it applied to them too.

What This Habit Asks of You

Three specific changes to how you create space for early disclosure of confusion.

Replace "any questions?" with a question that doesn't require volunteering weakness. The standard closing question puts the burden entirely on the trainee — they must identify their own confusion, assess whether it is significant enough to raise, and then disclose it publicly in a room where the power differential is real and visible. Questions that distribute the burden differently — "what would you want to double-check before starting?" or "what part would you feel least confident explaining?" — produce more honest answers because they require less of the person answering them.

Name assumed knowledge explicitly before it becomes a gap. When you are explaining something that involves context you have and your trainee may not, say so. "This assumes you've seen X before — have you?" or "There's a step here that's implicit in how we run this that I should walk through." These statements do not reveal weakness in your explanation. They reveal awareness of the gap between what you know and what you're assuming — which is exactly the awareness that prevents three-week misunderstandings.

Model uncertainty without deflection. When you don't know something in front of your trainees, say so and show them what you do about it. "I'm not sure — let's find out" is more instructive than a smooth redirect. It demonstrates that not knowing is the beginning of finding out, not a state to be concealed, and that the most experienced people in the room say it regularly.

A Note on What This Isn't

This is not a habit about eliminating the expectation of independent problem-solving or treating every moment of confusion as something requiring immediate escalation. Independent engagement with difficult material is a genuine research skill, and it should be developed.

The habit addresses the specific failure mode that sits on the other side of that expectation: the trainee who has been confused for longer than their effort justifies, whose confusion is not resolving, and who has not said anything because the culture they are working in has not made saying something feel safe.

That trainee exists in most labs. In many labs, they exist in significant numbers. The way to find out is to change the question you're asking — and to watch what happens when the answer is no longer riskier than the silence.

The meeting that ends with silence is not always a meeting that ended with clarity.

Sometimes it is a meeting that ended with everyone calculating that confusion was safer to hide than to name.

You can change that calculation. It starts with what you say next — and how you respond when someone finally answers honestly.

That's not a hack. That's a habit.

✨ Confusion addressed early prevents bigger problems later. Explore the Feedback Reflection Worksheet and Difficult Conversation Prep tools designed to support clearer communication, earlier clarification, and the confidence to say “I don’t know yet” before confusion turns into crisis.