About Unmythed

Unmythed looks at the claims people repeat, whatever the subject, and asks one question: what does the evidence actually show?

Who writes Unmythed

I’m a researcher with a postgraduate degree in the life sciences and have authored multiple peer-reviewed publications.

I also work in AI training, checking how AI models answer scientific questions and flagging claims that sound confident but aren’t supported, and I’ve written about new research for a general audience.

Why I started it

Both research and AI work have taught me how easily a claim can sound convincing without any evidence behind it. I see the same thing everywhere outside work, too: nutrition advice shared as settled fact, and superstitions repeated so often that nobody questions them.

Most people don’t have the time or access to dig into the original studies. I do, and I enjoy it. So I started Unmythed to do that checking in the open: find the best evidence, explain it plainly, show every source, and let you judge for yourself. If you have a claim you’d like checked, send it my way.

Independence

Unmythed has no advertising, no sponsors, no affiliate links and no paid content. I don’t accept free products or payment in exchange for coverage, and no company, organisation or religious group sees an article before it’s published.

If that ever changes, it will be stated here and on every article affected. AI tools play a limited, supporting role, explained in the AI use policy.

How claims are checked

Each article starts with one specific claim, stated as fairly as possible. The evidence comes from peer-reviewed research, systematic reviews and reports from scientific bodies. Every source is listed at the end of the article so you can read it yourself.

The goal is to understand, not to mock. Many of these beliefs are old, widely shared and held by thoughtful people. They deserve a careful look.

Behind every article

Inside an investigation

5 stages · 15 checks · 1 verdict

  1. 1Frame
  2. 2Gather
  3. 3Weigh
  4. 4Decide
  5. 5Publish

Stage 1 of 5

Frame the claim

One claim, stated fairly.

  • Exact wordingAs people really say it
  • Trace the originWhere did it start?
  • Set the testWhat would prove it wrong?

Stage 2 of 5

Gather the evidence

Cast a wide, fair net.

  • Search researchJournals, databases, scientific bodies
  • Systematic reviews firstAll the studies, gathered by set rules
  • Seek the other sideInclude studies that disagree

Stage 3 of 5

Weigh the evidence

Not all studies count equally.

  • Size & designHow many people? Was there a control group?
  • Replicated?Found again by others
  • Who paid?Funding and conflicts checked

Stage 4 of 5

Decide the verdict

Evidence vs. the claim as stated.

  • CompareWhat was claimed vs. what was found
  • Check the sizeIs the effect as big as claimed?
  • Pick one verdictSee the scale below

Stage 5 of 5

Publish & revisit

Show the work. Keep it current.

  • Plain summaryThe answer, up front
  • Every source listedCheck it yourself
  • Open correctionsNew evidence, noted openly

Five possible outcomes

The verdict scale

  • Holds up

    Good evidence supports the claim.

    Evidence
    Strong
    Claim vs. findings
    Findings agree
    • Consistent results across studies
    • Effect as big as claimed
    • Plausible explanation
  • Misleading

    A grain of truth, stretched further than the evidence goes.

    Evidence
    Moderate
    Claim vs. findings
    Findings partly agree
    • Real effect, but smaller
    • True only for some people
    • Wording oversimplifies
  • No evidence

    Properly tested, and no reliable effect found.

    Evidence
    Good evidence of no effect
    Claim vs. findings
    No link either way
    • Well-designed studies find nothing
    • Effects vanish in larger studies
    • Results don’t replicate
  • Unproven

    Never properly tested. There are stories, but no credible research either way.

    Evidence
    Little or none
    Claim vs. findings
    Unknown
    • Only anecdotes or testimonials
    • Studies too small or poorly designed
    • Nobody has tested it properly
  • False

    Good evidence shows the claim is wrong.

    Evidence
    Strong
    Claim vs. findings
    Findings contradict
    • Findings point the other way
    • Strong, consistent data
    • Proposed cause ruled out

Corrections

If you spot a mistake or know of research that changes the picture, please get in touch. Corrections are made openly, with a note at the bottom of the article explaining what changed, and every one is listed in the corrections log.