Evidence-Probing Question Ladder
Probe a claim with questions before accepting or rejecting it
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 98%
The Evidence-Probing Question Ladder is a conversational method for testing a claim without accepting or rejecting it outright. Tyson describes teaching his children to begin with what actually happened, then ask what a claimed object or mechanism consists of, what tests have been performed, and whether another explanation could account for the result. Each answer determines the next question. The approach separates curiosity from credulity: the speaker receives room to explain, but the claim does not gain credibility merely because it is confidently stated. The output may be support, rejection, or an honest unresolved state. Its key discipline is symmetrical skepticism—both automatic belief and automatic dismissal avoid the harder work of investigating how someone knows what they say they know.
Origin
Tyson says he and his wife raised their children to be scientifically literate and describes how they learned to probe claims about astrology and healing crystals with questions.
Core principles
- 01Immediate rejection can be as lazy as immediate acceptance
- 02Questions expose assumptions without requiring a fight
- 03Alternative explanations must be considered
- 04Confidence should follow the quality of evidence
How to run it
- 1
Pin down the claim
Ask the person to describe exactly what happened or what they believe the cause produced. Remove vague language before evaluating it.
Pro tip Start with a neutral question such as, 'What actually happened?'
Watch out Do not replace the speaker's claim with a stronger or easier-to-attack version.
- 2
Ask for the mechanism
Explore what the object, process, or influence is claimed to do and how it would produce the stated result. Clarify the materials and causal pathway where relevant.
- 3
Request the evidence
Ask what observations, comparisons, or tests support the claim. Separate a personal impression from a controlled test or repeatable pattern.
Pro tip Ask whether the result has occurred when the claimed cause was absent.
Watch out A sincere report is evidence of what someone experienced, not automatically evidence of their causal explanation.
- 4
Generate alternatives
Name other plausible causes and ask whether they could explain the same outcome. Include coincidence, prior conditions, measurement error, and ordinary mechanisms when relevant.
- 5
Find a discriminator
Ask what future observation or test would separate the competing explanations. Prefer evidence that could show the favored claim is wrong as well as right.
Watch out Do not move the standard of proof after an agreed test produces an inconvenient result.
- 6
Calibrate the conclusion
Accept, reject, or leave the claim unresolved in proportion to the evidence available. End the sentence at 'unidentified' when no identification has been established.
Watch out Unknown does not mean any preferred explanation becomes true.
In the wild
Tyson imagines his children asking what a healing crystal is made of, what tests were conducted, and whether the apparent healing could have another explanation. The questions investigate the claim without beginning with ridicule.
→ The claimant must supply a mechanism and evidence rather than rely on confidence alone.
When a video shows an unexplained object, Tyson accepts that it may be impressive but rejects the leap from 'unidentified' to 'alien.' The missing identification is treated as the endpoint until better evidence appears.
→ Uncertainty is preserved without turning it into a conclusion.
Common mistakes
Rejecting before investigating
Reflexive dismissal avoids the same evidentiary work as reflexive acceptance and rarely improves the conversation.
Treating unknown as proof
A lack of explanation does not establish aliens, a cover-up, or any other preferred cause.
Is it for you?
Best for
It is best for conversations about health products, astrology, paranormal claims, rumors, and other uncertain assertions.
Not ideal for
It is not ideal as a delaying tactic when established evidence already supports urgent safety action.
From the transcript
“If you just reject what someone says outright, that's as intellectually lazy as it is to accept what they say outright.”
“What tests have you made for this?”
From the episode
Neil deGrasse Tyson: The Brutal Truth About Astrology! Our Breath Contains Molecules Jesus Inhaled!