How fact checking works in broadcast news comes down to two layers of work: a reporter gathers the evidence and writes the script, then a separate person re-checks every name, date, number and quote in that finished script against the original evidence before a producer signs it off for air. Broadcast adds one hard constraint that print does not have. Once an item is transmitted it cannot be quietly amended, and the newsroom may have minutes rather than days.
That combination is why broadcast fact checking looks systematic on paper and slightly improvised in practice. Here is how it actually runs, who does what, and how you can judge a report for yourself the next time one reaches your radio.
Table of Contents
- What Fact Checking Means in Broadcast News
- How Fact Checking Works in Broadcast News
- What Sources Do Broadcasters Check?
- How Do Reporters Verify Time-Sensitive Claims?
- How Do Radio and TV Corrections Work?
- How Can Listeners Judge the Quality of a Broadcast News Claim?
- Frequently Asked Questions
- What does fact checking mean in broadcast news?
- Who fact checks a radio or television news report?
- How quickly should a breaking-news claim be verified?
- Does a journalist need two sources for every fact?
- What is the difference between a correction and a clarification?
- Can listeners verify a news report after it airs?
- Conclusion: Start With the Evidence
What Fact Checking Means in Broadcast News
Fact checking means independently confirming the factual claims in a finished piece of copy, rather than trusting that the person who gathered them got them right. In a broadcast newsroom it covers names, dates, figures, quotations, the wording of an official statement, the location of a recording, and whether a clip or picture shows what the script says it shows.
The discipline underneath it is verification. Fact checking is the newsroom’s application of that discipline, and it happens as a separate pass, after reporting is finished and before transmission. The reason for keeping it separate is simple: a reporter who has just spent two days chasing a council tax figure is not the right person to decide whether they have found it, and most workrooms will not let them try.
Three things sit outside this process. Opinion is a view, clearly labelled, and checking it is not possible or desirable. Analysis is interpretation built on checked facts. A developing story is different again, because the honest position is that some of it cannot be verified yet, and the script has to say so in as many words.
Speech makes the stakes higher than the same claim in a printed column. A presenter reading aloud gets one pass, at speed, with no chance for the listener to re-read a caveat. A misspoken name, a figure that drops a decimal, or a quote stitched from two sentences becomes the version people repeat by lunchtime. Broadcast checking spends so much of its effort on things that look trivial for that reason.
How Fact Checking Works in Broadcast News
The pipeline below is a generic model built from published newsroom practice. Larger organisations run dedicated fact-checking desks, while smaller stations fold the same steps into the producer’s normal read-through. The stages are the same either way.
| Stage | Who owns it | Evidence required |
|---|---|---|
| Claim identification | Reporter and producer | List of every checkable assertion in the script |
| Original source check | Producer | The document, recording or first-hand account the claim came from |
| Independent confirmation | Producer or fact-checker | A second, unrelated source supporting the same claim |
| Primary evidence | Fact-checker | Filings, transcripts, datasets, full recordings |
| Source clarification | Reporter | Written replies from everyone quoted or criticised |
| Documentation | Fact-checker | An annotated copy showing where each fact was confirmed |
| Editorial sign-off | Editor or head of news | A named person accepting responsibility for the item |
| Transmission | Presenter and producer | Final read-through against the checked copy |
| Correction or clarification | Editor | A logged decision and, where needed, an on-air notice |
How Fact Checking Works in Broadcast News: The Editorial Workflow
It starts with the reporter’s research package, which gathers everything collected during reporting: the annotated draft, transcripts, saved pages, photographs with their details intact, and audio with timecodes noted. If an item is handed over without that package, the checking stage stalls, and a producer who wants to broadcast will say so quickly.
The editor reads it first and decides whether it is a story at all, or a claim in search of one. That is a different question from whether it is true, and newsrooms get it wrong in both directions: killing accurate material because it is dull, and passing inaccurate material because it is dramatic.
The fact-checker then works line by line, taking each sentence that asserts something and tracing it back. The core rule is the two-source rule: a significant claim needs the original source plus an independent one that did not come from the same origin. Two outlets repeating a press release is one source wearing two hats.
Every person named in the piece gets contacted, including the people being criticised. That is the right of reply, and it is a legal requirement as much as an ethical one, because publishing an allegation without giving the subject a chance to respond is how libel claims are lost. A source who ignores a written question gets a second, recorded attempt by phone, and that attempt is logged.
Quotations are read back against the original recording before broadcast. The classic errors are joining two separate sentences into one quote, trimming the qualifier that changed the meaning, and using “says” for words that were never spoken.
Then comes sign-off. One named editor or head of news accepts responsibility for the item as it will be heard, and the checked copy goes to the presenter’s script. Any unresolved flag stops transmission until it is fixed, properly attributed, or cut.
What Sources Do Broadcasters Check?
Not all sources carry the same weight, and the useful question is what a given source can actually prove. A council minute cannot testify to a conversation. A witness can describe what they saw but not what a press release claims about it.
| Source type | What it proves well | What weakens it |
|---|---|---|
| Official documents and filings | The formal record: budgets, court papers, minutes, published figures | Prepared statements are claims, not findings, and get quoted as if they were both |
| Public records | Traces nobody can revise afterwards | Records reflect what was filed, not necessarily what happened |
| Direct witnesses | First-hand observation of an event | Memory under stress, partial views, personal stake |
| Named experts | Specialist interpretation, on the record | Unnamed “experts” quoted without affiliation |
| Data and statistics | Trends, when the method and sample are published | Figures quoted with no baseline, no period or no sample size |
| Wire services | Corroboration and a second newsroom’s reporting | Shared feeds mean many outlets are not independent |
| Social posts and user video | That someone said it, and roughly when | Altered timestamps, recycled old footage, cloned voices |
| Listener or reader submissions | Tips, local knowledge, phone evidence | Chain of custody, and whether the audio was edited |
The two-source rule does not mean two people who agree. Independence is the point, which is why a newsroom checks whether two “independent” confirmations both trace back to one press office.
For pictures and video the work is technical. Journalists search archive footage databases to date a clip, look for shadow and signage that place it somewhere, and run a reverse image search to find an earlier version. Metadata can confirm a device and a time, and it can equally be stripped in seconds, so it supports a conclusion rather than proving one.
With synthetic media now in routine circulation, a second round of checks has been added to broadcast desks. Cloned voices cut against old interviews to produce new sentences, and still images are generated to match a real event. The habits that catch most of it are unglamorous: ask for the raw file, insist on the original upload rather than a repost, check for inconsistent lip movement or breathing, and treat a very smooth voice on a shaky phone connection as suspect.
How Do Reporters Verify Time-Sensitive Claims?
Speed and accuracy are not opposing goals in a well-run newsroom. They conflict only in the first minutes, and the resolution is to be precise about what kind of claim is being aired rather than to hold everything until it is certain.
A script can carry four states, and the wording should make the difference audible in one sentence. A confirmed fact has been checked against primary evidence and is stated as it stands. An attributed claim is verified as something a named person said, with the attribution kept in the copy. A developing claim is flagged as provisional and attached to what prompted it. An unverifiable claim either waits or appears with an explicit note that it has not been independently confirmed.
The hold-or-publish call usually rests with the editor. Publishing a labelled, sourced claim carries less risk than publishing silence when a public is looking for information, and far less risk than publishing the claim bare.
This is also why live interviews are not fact-checked as they happen. A live-to-air window runs in seconds, and interrupting to query a figure would be clumsy and would surrender the point of the interview. Instead the checking happens beforehand through pre-briefing: the presenter and producer research the guest, prepare fact sheets on their record, and know in advance which claims to put to them. What reaches air is a claim tested against that preparation rather than in front of it.
How Do Radio and TV Corrections Work?
Errors get through. They travel further than the correction that follows them, because a correction airs in the same slot while the mistake is already being clipped and shared, and because most listeners simply never hear it.
A correction becomes necessary when a factual error could reasonably change how a listener understood the story. A wrong figure, a misnamed person, a false implication about who said something, or audio presented as live that was recorded earlier all qualify. Wording preferences, tone and the balance of a debate usually fall to a clarification, which is often published rather than broadcast.
The decision to correct rests with the editor, not with the person who made the mistake, which is the point of separating the layers. Good practice fixes three things: how soon the correction goes out, whether it names the original claim clearly enough that a listener who missed the first report still understands what changed, and whether it is logged. Many broadcasters keep a public corrections log; the honest ones make it easy to read and do not quietly remove entries.
A workable correction tells listeners what was wrong, what is right, and when it was wrong. Anything vaguer than that wastes airtime that was meant to be useful.
How Can Listeners Judge the Quality of a Broadcast News Claim?
You will rarely be able to audit a newsroom from home, but five questions get you most of the way.
First, is the claim attributed? A figure or assertion with a named person attached can be checked and challenged; the same words offered unattributed cannot. Second, is the source independent, or are two reports circling the same press release? Third, does the report distinguish what is known from what is developing, and does it use words like “says” where evidence has not been shown?
Fourth, are the unnamed experts identifiable? “A leading scientist” with no name, no field and no affiliation is a warning sign rather than a credential. Fifth, and most revealing: when a story does turn out to be wrong, does the station correct it, and can you find that correction afterwards?
For the bigger claims, independent checkers publish their evidence openly. Searching the exact wording of a headline usually surfaces the original document, the full statement, or an existing fact check with its sources attached. You rarely need to be an investigator, just willing to look past the summary.
Frequently Asked Questions
What does fact checking mean in broadcast news?
It means independently confirming the factual claims in a finished script rather than trusting the reporter who gathered them. Names, dates, figures, quotations, the wording of statements and the identity of recordings all get traced back to their original evidence, usually by someone who did not report the item.
Who fact checks a radio or television news report?
Usually a producer, editor or dedicated fact-checker, with one named editor taking final responsibility. At smaller stations the producer does it as part of the read-through, while larger organisations run separate checking desks reporting to an editor. In every case the person who wrote the item should not be the person who signs it off.
How quickly should a breaking-news claim be verified?
Before air, and as fast as the evidence allows. A claim that cannot be fully checked can still go out if it is attributed to a named source and described as provisional, but never as established fact. The editor decides whether waiting helps listeners more than publishing a clearly labelled, sourced claim.
Does a journalist need two sources for every fact?
Not for every detail. The two-source rule is generally reserved for significant or contested claims, where a single origin carries real weight. What matters more is independence: two confirmations that both trace back to one press office count as one source, however many outlets repeat the story.
What is the difference between a correction and a clarification?
A correction fixes a factual error that could change how a listener understood the report, and is usually broadcast. A clarification adds missing context or corrects wording without changing the substance, and is more often published online. Both should say plainly what was wrong and what is right.
Can listeners verify a news report after it airs?
Yes, in most cases. Search the headline wording to find the original document or statement, then check independent fact-checking sites that publish their sources. Ask whether the claim was attributed, whether the source is independent, and whether the report separated what is confirmed from what is still developing.
Conclusion: Start With the Evidence
Broadcast fact checking works because a second person traces every claim in a finished script back to its origin, and because one editor signs their name to it before transmission. When you are checking a report yourself, do the same three things: identify the claim, find the original source behind it, and look for a confirmation that does not come from the same origin.
