How We Review

This page describes how Human Understood checks what it publishes. It is deliberately specific, because a vague standard is not a standard.

What we publish

Human Understood covers social reconnection, career reinvention, relationships and intimacy, cognitive vitality, financial psychology, and purpose and meaning, for adults roughly between 40 and 55.

We write about research. That means we are in the business of describing other people's findings accurately, which is harder than it looks and is where most health and psychology writing fails.

How a claim gets into an article

Every factual claim we publish has to pass four checks.

The source exists and says what we say it says. We read the paper, not the press release, not the news write-up, not another site's summary of it. Where a paper sits behind a paywall we say so rather than repeating a secondary account of it.
The numbers match. Authors, year, journal, volume and pages are checked against the publisher record. Effect sizes and sample sizes are taken from the paper.
The finding applies to the people it is being quoted at. A result from university students is not a result about people in their fifties. We say when a study's population differs from our reader, because it usually does.
Nothing newer contradicts it. A source that exists is not the same as a current one. We check whether a later and larger study has superseded the one everyone quotes.

A claim that fails any of these comes out of the article, even when it is the reason the article was interesting.

What our evidence ratings mean

Every recommendation carries a rating. It describes the strength of the evidence behind the thing, not how much we like it and not what it pays.

Strong. Multiple randomised trials, or a meta-analysis of them, in a population reasonably like our reader. Rare.
Moderate. Evidence with clear limits: smaller trials, mixed results, an evidence base rated low-confidence by the people who assembled it, or findings drawn from a population different from our reader.
Weak. Plausible mechanism, thin or purely observational support. We will still recommend something rated Weak if it is low cost and low risk, and we will say plainly that the evidence is thin.
Not recommended. Evidence is absent, or the product's claims exceed it.

Every recommendation also names who should skip it. If we cannot think of anyone who should skip it, we have not thought about it hard enough.

Corrections

We correct in public. A corrected article carries a dated note at the foot saying what was wrong and what the source actually says. We do not quietly edit a number and move on.

Several articles on this site carry correction notes. Those are not an embarrassment we are hiding; they are the process working. In September 2026 we rechecked every article on the site against its primary sources and found errors in our own work: a statistic attributed to an organisation that never published it, a widely repeated set of figures that came from a university press release rather than the study, and a headline claim that the research record contradicts. All of those have been corrected and the corrections say so.

If you think we have got something wrong, tell us: hello@humanunderstood.com. Point us at the source. We will check it and, if you are right, we will fix it and say that we did.

How we make money, and what it does not buy

Human Understood is funded by affiliate commissions. If you buy something through a link on this site, we may earn a commission at no extra cost to you.

Here is what that does and does not mean.

Commission never determines a recommendation. We decide what to recommend from the evidence, then find out whether a program exists. Not the other way round.
We recommend things that pay us nothing, and we turn down programs that pay well. We have declined several profitable programs because we would have had to rate the product Low to describe it accurately, or because the company was under regulatory enforcement.
Paid links are visually distinct from editorial content, always, on every page. You should never have to wonder which is which.
No advertiser sees an article before it is published, and no advertiser has any say in what we write about them or anyone else.
We say when we do not know. Where nobody has compared two products head to head, we say so rather than inventing a ranking.

If a recommendation on this site ever reads like it was bought, tell us at the address above. That is exactly the kind of error we want reported.

Who writes this

Articles are currently published under the Editorial Team byline. We think the right way to earn trust is a published, checkable method rather than a name you have not heard of, so the method is above and you can hold us to it.

What we are not

Human Understood is not a medical provider and nothing here is medical, psychological, legal or financial advice. Our articles are educational. They are not a substitute for a conversation with a professional who knows your situation. If you are in crisis, call or text 988 in the United States to reach the Suicide and Crisis Lifeline.

Last updated 26 September 2026.