Corrections & Fact-Checking

Our working approach is:

Read → Understand → Verify → Attribute → Write → Publish

Each stage serves a purpose.

Read

Where possible, we go beyond summaries and examine the original material.

For research stories, this may mean reviewing the study itself rather than relying solely on a press release or social-media description.

Understand

Before explaining a scientific or medical development, we seek to understand:

• what was actually studied;
• who or what was studied;
• how the study was conducted;
• what the researchers found;
• what they did not find;
• what limitations exist.

Verify

Claims are checked against sources appropriate to the subject. These may include:

• peer-reviewed research;
• official government data;
• regulatory authorities;
• international health organisations;
• universities and research institutions;
• original documents;
• official statistics;
• credible original reporting.

No single source is automatically sufficient for every claim.

We pay attention to distinctions that can materially change the meaning of medical evidence. For example:

• laboratory research versus human research;
• animal studies versus clinical studies;
• observational studies versus controlled trials;
• association versus causation;
• preliminary findings versus established evidence;
• statistical significance versus clinical significance.

Experts can help explain complex evidence.

But expertise does not eliminate verification.

Where expert interpretation is used, we consider relevant qualifications, institutional affiliations and potential conflicts of interest.

Virality is not verification.

A health claim being repeated across social media does not increase its evidentiary value.

In some circumstances, widespread circulation is precisely why a claim deserves closer scrutiny.

Information generated by AI is not accepted as a substitute for primary evidence or authoritative sources.

AI can assist research. It cannot independently verify itself.

When information originates elsewhere, we seek to identify the original reporting or source.

Fact-checking is not simply about deciding whether something is true. It is also about understanding where the information came from and how confidently it can be presented.

Every newsroom wants to be accurate.

A trustworthy newsroom also knows what to do when it is not.

If JOSEPH MMWA publishes a material factual error, we believe the responsible response is to correct the record clearly and promptly.

We may issue a correction for material errors involving:

• names;
• dates;
• locations;
• numbers;
• medical findings;
• study details;
• quotations;
• attribution;
• institutions;
• descriptions of events;
• other facts that materially affect the reader's understanding.

Not every change is a correction.

An update adds new verified information that was not available when the article was originally published.

A correction addresses information that was previously published inaccurately.

Where appropriate, we distinguish between the two.

For material errors, we may include a correction notice explaining what was incorrect and what has been changed.

The purpose is not to draw attention away from the mistake. It is to give readers an accurate record.

Where an error substantially undermines an article, we may take additional action, including prominent correction, substantial revision or withdrawal of the article where necessary.

The response should be proportionate to the seriousness of the error.

Readers are encouraged to alert us to potential errors.

A useful correction request should identify:

• the article;
• the specific statement;
• why it may be inaccurate;
• evidence supporting the concern.

A correction request does not automatically result in a correction. We review claims against the available evidence.

We do not ask: "Is this interesting enough to publish?"

We ask: "Do we have enough evidence to present this as fact?"

We do not consider a correction an admission that journalism has failed. We consider refusing to correct a known error the greater failure.

We will not defend an error simply because it has already been published. When the evidence changes—or when we discover that we were wrong—we will act accordingly.