A senior manager once told me, with complete confidence, that a project was on track. The spreadsheet said so. The status report said so. Every update for the preceding three months had said so. Six weeks later, the project was dead.

I have seen this pattern more times than I can count across 30 years of working in and around organisations that should have known better. The problem was never the data. The problem was how they were reading it.

Here is what I find interesting about that pattern: a trained pathologist would not have made the same mistake.

Not because pathologists are smarter than business managers — they are not, necessarily. But because pathology training instils a particular discipline around evidence that most professional education simply does not. Pathology teaches you that a result is not a conclusion. It is a data point. And the difference between those two things is everything.

When a sample comes into a pathology lab, the question is never "what does this number say?" The question is "what does this number mean, given everything else we know?" A single out-of-range result triggers a process — not a verdict. You check the sample integrity. You look at the clinical context. You consider what the result would mean if it were true, and what it would mean if it were an error. You form a hypothesis. You test it.

That is diagnostic reasoning. And it is a profoundly different cognitive process from the one most people use when they encounter information they want to be true.

Most people — most organisations — look at data the way a driver looks at a GPS. They follow it. They trust it. And when it tells them to turn right into a lake, they slow down slightly and turn right.

Diagnostic reasoning asks a different question entirely: what would have to be true for this to be wrong?

That is an uncomfortable question. It demands intellectual honesty about the limits of your information. It requires you to hold two possibilities simultaneously — the result might be correct, or the result might be an artefact of a flawed process — without rushing to close on either one. Most people find this difficult. Not because they are incapable of it, but because they have never been trained to do it habitually.

Pathology training makes it habitual.

This is not a minor professional skill. This is a fundamental cognitive upgrade. And it transfers — completely and directly — to any domain where decisions are made on the basis of imperfect information. Which is, without exception, every domain that matters.

A manager reading a project status report. An investor evaluating a business case. A policy maker interpreting population data. A doctor reading a scan. A hiring panel assessing a candidate. In every single one of those situations, the trained diagnostic thinker is asking a different set of questions than the untrained one. Not "does this evidence support what I hoped?" but "what does this evidence actually say, and what are the plausible ways it could be misleading me?"

Business schools teach strategy, finance, marketing, and leadership. Many of them teach data literacy in some form. What most of them do not teach — what is genuinely hard to teach without the experiential foundation of working with real specimens, real results, and real consequences — is the discipline of treating your own conclusions as a hypothesis until the evidence forces you to stop.

That discipline is not instinctive. It runs directly against the cognitive shortcuts the human brain prefers. We like to close. We like certainty. We like the result to mean what we want it to mean. Resisting that pull — systematically, reliably, even when the answer is inconvenient — is something you build through practice in environments where being wrong has visible, immediate consequences.

Pathology is one of those environments. So is any precision laboratory. So is any field where a flawed conclusion cannot be hidden behind a confident presentation and a well-formatted report.

This is worth sitting with if you are currently in training. Not because your qualification is going to teach you everything — it isn't — but because the underlying cognitive discipline it is building in you is genuinely rare and genuinely valuable well beyond the specific context in which you are learning it.

The question worth asking — not just once, but as a regular practice — is this: when was the last time I treated my own confident conclusion as a hypothesis that still needed testing?

Adam