How to check a claim
Every topic on this site is built by the same procedure. It is not specialised and it is not owed to any particular subject. It works on a sermon, a headline, a study, or anything on these pages. Six steps.
1. Say what is actually being claimed
Most arguments die here, because most claims are never stated precisely enough to be wrong. “The earth is young” is not a claim you can check. “The earth is approximately six thousand years old” is. Pin it to a number, a date, an event, a mechanism. If a claim resists being made specific, that resistance is itself the finding.
State it in the form its own defenders would accept. Beating a version nobody holds tells you nothing, and it is the fastest way to lose an argument you were otherwise going to win.
2. Ask what would show it false
Before looking at any evidence, name the observation that would sink the claim. A young earth forbids rock that dates older, ice with more annual layers than years available, and starlight from beyond six thousand light-years. Those are falsifiers, and they are the reason the claim is checkable at all.
Run the same step on your own position. If you cannot say what would change your mind, you are not holding a conclusion. You are holding a commitment, and the rest of the procedure will not help you.
3. Separate evidence from testimony and authority
Evidence is an observation of the world that anyone with the right equipment can repeat. Testimony is a report that something was observed. Authority is someone stating a conclusion on the strength of their position. All three appear in real arguments; only the first settles anything on its own.
Authority is a useful shortcut, not a proof. When a scientific body states a figure, what makes it worth weight is the measurements underneath it, which is why a citation that leads to those measurements is worth more than one that leads to a press release.
4. Open the source
Follow the citation and confirm it says what it is cited as saying. This single habit catches more bad argument than any other, on every side of every question. Claims get passed along a chain of quotation until the thing at the end no longer resembles the paper at the start.
Prefer the primary source: the paper reporting the measurement, the excavation report, the dataset. Note who published it and whether the venue has review. Check the date, because fields move. A 1970 figure cited as current is not a lie, but it is not evidence about today either.
5. Look for independent agreement
One method can be flawed. The question is whether methods that share no assumptions land in the same place. Radiometric decay, ice layers, tree rings, coral bands and seafloor spreading rely on entirely different physics, and they agree. For that agreement to be an artefact, every one of them would have to fail in precisely the way needed to produce the same wrong answer.
That is the shape of a well-supported conclusion, and it is a different thing from many sources repeating one original claim. Count independent measurements, not citations.
6. Watch for the retreat
When a specific claim meets contrary evidence, it often does not get withdrawn. It gets modified into a version that no longer forbids anything: the world was made recently but built to appear ancient; the effect is real but only when conditions are not controlled; the plan is good but its workings cannot be known.
Each of these is unfalsifiable, and an unfalsifiable claim is compatible with every possible observation. That is usually presented as its strength. It is the opposite. A claim that cannot fail cannot be tested, cannot be corrected, and explains nothing, because it would have been equally at home in any universe at all.
Noting the retreat is not a refutation of the original claim. The evidence already did that. It is how you tell that the argument has ended.