Editorial Standards

Our editorial principles

• Accuracy. A headline never claims more than the evidence supports. Where research is thin or contested, the article says so rather than rounding up to confidence.
• Independence. No advertising, no sponsored content, no affiliate links. Nobody pays to appear on this site or to influence what it says.
• Transparency. We disclose how content is made, including where AI is used, and we name our sources so you can check them.
• Plain language. Wellness writing hides a lot behind jargon. We write so the point survives first reading.
• Accountability. When we get something wrong, we correct it visibly and say what changed.

How we research and source

Content is built from primary and authoritative sources wherever a factual claim is made:

• Peer-reviewed research, cited specifically rather than gestured at
• Reputable public health bodies and established mental health organisations
• Original studies in preference to press coverage of those studies

Cultural and folkloric material — much of what circulates about dream symbolism, for example — is clearly labelled as tradition rather than evidence. It appears because it is interesting, not because it is established fact.

Our review process

1. Research. Claims are gathered and traced back to their original source.
2. Drafting. The article is written, with AI assistance as described below.
3. Fact-checking. A person verifies every factual claim against the primary source before publication. Anything that cannot be verified is cut.
4. Plain-language edit. The draft is reviewed for clarity and for tone appropriate to a mental wellness context.
5. Publication. The article goes out with a byline and a publication date.
6. Review. Published articles are revisited periodically, and updated where the evidence has moved.

How we use AI

AI tools are used to assist with research and drafting. They are not used to publish unreviewed content.

A person verifies every factual claim against its primary source before an article is published. Editorial responsibility for everything on this site rests with Repattern, not with a tool. Where AI output could not be verified, it does not appear.

Editorial independence

Repattern carries no advertising, no sponsored posts, and no affiliate links. No third party reviews or approves content before publication, and no commercial relationship influences what we cover or what we conclude.

If this ever changes, any commercial relationship will be disclosed at the top of the article concerned — not buried in a footer.

Our sourcing standards

Not all sources are equal, and we weight them deliberately. When a claim rests on evidence, we reach for the strongest tier available and cite it at the exact sentence it supports.

• Tier 1 — primary and authoritative. Peer-reviewed studies and meta-analyses, and public health bodies such as the National Institutes of Health, the National Institute of Mental Health, and MedlinePlus.
• Tier 2 — established secondary. Reputable academic and clinical institutions summarising the evidence, used for definitions and context rather than as the origin of a claim.
• Clearly labelled cultural material. Folklore and tradition — much of dream symbolism, for example — presented as tradition, never as evidence.

We do not cite content farms, SEO blogs, AI-generated summaries, or any source that is monetising the claim we are quoting. Where a popular belief is unsupported, we say so and explain why, rather than repeating it.

Conflicts of interest

Repattern is an independent project and carries no advertising, sponsorship, or affiliate arrangements, so most conflicts of interest simply do not arise. Our own tools are the only product we point readers toward, and we say so openly every time. If a genuine conflict ever existed — a paid relationship, a reciprocal arrangement, a personal stake — we would disclose it on the page in question before you read the content, not afterwards.

How we handle uncertainty

A lot of what gets written about wellbeing is stated with more confidence than the evidence supports. We try to match our tone to the strength of the evidence:

• Where research is strong and consistent, we say so plainly.
• Where effects are small, mixed, or debated, we say that — including when a popular idea has thin or contested support.
• Where something is folklore or opinion, we label it as such rather than dressing it up as fact.

Naming the honest limitation of the evidence is a standard we hold ourselves to on purpose. It is also, not coincidentally, what most competing content leaves out.

Freshness and review cadence

Evidence moves, and stale content is a quiet way to mislead. Every article carries a published date and, where relevant, a last-updated date, so you can see how current it is.

We revisit published content periodically and update it when the evidence changes, when we find an error, or when readers surface a better question we should answer. Material updates are marked; minor fixes are made quietly.

Scope: reflection, not clinical care

Repattern publishes general information and self-reflection material. It is not therapy, diagnosis, or treatment, and nothing here is personalised medical advice. Articles do not tell anyone to start, stop, or change treatment.

Where a topic touches on distress, the article points towards appropriate professional help rather than positioning our tools as a substitute for it. The full position is set out in our Terms of Service.

Currency and corrections

Every article carries a publication date, and updated articles are marked as such. We correct material errors openly:

• Factual errors are corrected as soon as we can, with a dated note explaining what changed.
• Substantive rewrites that alter an article’s meaning are marked as updated.
• Typos and formatting are fixed without a note.
• Articles are never silently deleted to hide a mistake. If one is withdrawn, the page explains why.

Spotted an error? Report it via the contact page, or raise it formally through complaints. More detail on our process is in how we create content.