Corrections

When an article changes a number, a claim or a verdict after publication, it is listed here with the date, what changed and why. The policy is on the method page.

Published corrections

None yet. The first three articles were published on 14 October 2026.

Fixed before publication

Each video script is fact-checked independently before it ships. These changes were made before anything here was published, and are listed because they show what the checks catch:

• 5G and cancer. The WHO-commissioned review included 63 studies "that met its criteria", not the "63 strongest". The highest exposure in ARPANSA's Melbourne survey is given as about 0.14% of the limit, which matches "more than 700 times below"; 0.15% would only be about 667 times.
• Phone in your pocket. The Swiss study's heavy-use figure is a median sperm concentration of 44.5 against 56.5 million per millilitre, about a fifth lower; the paper's adjusted "21%" is a higher risk of a low total count, not "21% fewer sperm". A "28,000 times" dose ratio on an older agency summary page is an arithmetic error (it is about 300 times) and is not used.