Writing and communicating research · Lesson 3 of 3

Claims proportionate to evidence — and explaining it to anyone

Match the verb to the evidence, keep the scope at the sample, and lead with the finding for anyone outside the field.

9 min read · Free — no account needed to read

The last and most important discipline in writing is matching the strength of your language to the strength of your evidence.

Watch the verbs. "Suggests", "is associated with", "is consistent with", "indicates", "demonstrates", "shows", "proves" are not interchangeable, and the distance between the first and the last is the distance between honest and overstated. Correlational evidence supports "is associated with"; it does not support "causes", however strongly you suspect it.

Watch the scope. A finding about 200 university students in one country is not a finding about students, or about young people, or about people. The slide from sample to humanity usually happens between the discussion and the abstract, and it is almost always unintentional.

Watch the hedging in both directions. Under-hedging overstates; over-hedging — where everything "may possibly suggest" — makes work unusable, because a reader cannot tell what you actually think. State what you found and be clear about the confidence.

Then the other audience. Explaining a finding to someone outside your field is a distinct skill, and it is not dumbing down. What to drop: jargon, methodological detail that does not change the conclusion, and the history of the debate. What to keep: what you wanted to know, what you did in one sentence, what you found, how confident you are, and why it matters to them.

The most common mistake is leading with method — "we ran a cross-sectional survey using a validated instrument" — when the listener wants the finding. Lead with what you found, then say briefly how you know, then say what you are unsure about. Being honest about uncertainty to a general audience builds trust rather than undermining it; the people who never hedge are the ones who should be trusted least.

See it happen

You need to tell a non-specialist audience about a correlational finding on sleep and performance.

The read that arrives on its own

"Our cross-sectional study of 340 participants found a statistically significant negative correlation between sleep duration and error rate."

The same situation, with this idea

"We looked at whether people who sleep less make more mistakes at work. They do — the pattern is clear across 340 people. What we cannot tell from this is which way round it goes: less sleep may cause the mistakes, or the kind of stressful job that causes mistakes may also cost people sleep. To separate those you would need to follow people over time, which is what we would do next."

After this, you can

lead with the finding, then how you know, then what remains open.

The takeaway

Match the verb to the evidence, keep the scope at the sample, and lead with the finding for anyone outside the field.

Easy to take the wrong way

That careful language makes work sound weak. Precise claims are what make a reader trust the claims you do make strongly.

Try this

Take your main finding and write it three ways: as a sentence for your field, one for a manager, one for a friend. Notice what you dropped and whether it changed the claim.