Begin by identifying the claim
Before deciding whether a claim is true, we must know exactly what is being claimed. Vague statements are difficult to test because they can change meaning whenever evidence challenges them.
“This treatment works,” for example, might mean that it produces a measurable improvement, that one person felt better after using it, or merely that people have used it for a long time. Those are different claims and require different evidence.
Ask what evidence we should expect
A useful question is: if this claim were true, what should we expect to observe?
A claim about a historical event may require contemporary documents, physical evidence and independent accounts. A medical claim may require controlled research. A claim about a person’s motives may never be known with complete certainty, but available actions, statements and surrounding circumstances may support a cautious conclusion.
The kind of evidence required depends upon the claim. No single test applies equally to every question.
Examine the source
Evidence does not become reliable merely because it has been presented confidently. We should ask:
- Where did the information originate?
- Can the original source be examined?
- Does the source have relevant knowledge?
- Could the source benefit from persuading us?
- Has the information been independently confirmed?
- Is contrary evidence being omitted?
A source may be sincere and still mistaken. Expertise can increase reliability, but experts must still show the basis of their conclusions.
Consider alternative explanations
People often ask only whether some evidence is compatible with their preferred explanation. A stronger method asks whether another explanation fits the evidence equally well or better.
An apparent prediction might result from chance, vague wording or selective memory. An unusual experience might arise from perception, expectation, sleep disturbance or an unknown natural cause. A political event may have several contributing causes rather than one secret explanation.
Look for evidence against the claim
Honest investigation requires more than collecting support. We should actively search for information that could show that our belief is wrong.
This is difficult because beliefs can become connected to identity, community, fear, hope or pride. Nevertheless, a belief protected from every possible challenge cannot be meaningfully tested.
Match confidence to evidence
Many conclusions are not absolutely certain. That does not make every answer equally plausible.
We may conclude that a claim is well established, probably true, possible but unsupported, probably false, or presently unknown. The responsible position is the one justified by the available evidence, not the one that offers the strongest emotional satisfaction.
Evidence notes
The reliability of a conclusion depends on the quality, relevance, independence and completeness of the evidence. Repetition is not independent confirmation when every repetition can be traced to the same original source.
Absence of evidence is particularly significant when the claim predicts that clear evidence should exist. It is less significant when the expected evidence would be difficult to preserve or observe.
Ethical questions
The standard of evidence can have ethical consequences. Accepting weak accusations may injure innocent people. Rejecting strong evidence may protect harmful institutions. Demanding impossible certainty can also become a way of avoiding responsibility.
We should therefore ask not only whether a conclusion is justified, but what harm may result from accepting or rejecting it carelessly.
Conclusion
We cannot guarantee that every conclusion will be correct. We can, however, use a method that makes error less likely.
Define the claim, identify the expected evidence, examine the sources, test alternative explanations, search for contrary information and express confidence in proportion to what the evidence supports. Truth-seeking is not the elimination of uncertainty. It is the disciplined refusal to pretend that we know more than we do.