Learn area
Reasoning Errors
The topics change. The same reasoning errors recur. This directory connects individual claims to reusable errors in logic, evidence and inference.
Explore connections →Network view
See which reasoning errors appear together.
Each node is a reasoning error. A line connects two errors when they occur in the same visible investigation; thicker lines mean they co-occur more often. This is an empirical co-occurrence map, not a claim that one error causes another.
Directory view
Browse reasoning errors A–Z.
Absence-of-evidence inversion
Treating a lack of disproof as positive evidence that a claim is true.
perception patternsAnomaly hunting
Scanning noisy data for unusual details and treating the selected anomalies as evidence of a hidden explanation.
source evaluationAppeal to anecdote
Using one or a few vivid personal cases as if they establish a general pattern or causal effect.
rhetoric languageAppeal to nature
Treating something as safer, healthier or better merely because it is described as natural.
logicArgument from incredulity
Treating personal difficulty imagining an explanation as evidence that the explanation is false.
probability statisticsBase-rate neglect
Underweighting how common something is before considering case-specific evidence.
logicCategory error
Applying a property, comparison or explanatory rule to something that belongs to the wrong conceptual category.
source evaluationCherry-picking
Selecting supportive observations while ignoring relevant evidence that changes the overall picture.
perception patternsConfirmation bias
Favoring observations that fit an existing belief while discounting or reinterpreting evidence that does not.
causal reasoningCorrelation/causation error
Inferring that one variable causes another merely because they are associated.
scientific reasoningDose neglect
Reasoning about a substance or exposure without considering how much, how often, or by which route.
logicFalse dichotomy
Presenting two options as exhaustive when other possibilities or intermediate positions exist.
logicFalse equivalence
Treating two claims, sources or bodies of evidence as equally strong despite relevant asymmetries.
scientific reasoningHazard–risk confusion
Treating the ability to cause harm under some conditions as if it were the actual probability of harm at a specific exposure.
causal reasoningMechanism neglect
Ignoring whether a proposed causal pathway is coherent, evidenced and capable of connecting the claimed cause to the outcome.
rhetoric languageMisuse of scientific vocabulary
Using technical language to create an appearance of scientific support without a valid evidentiary connection.
source evaluationMotivated source rejection
Applying harsher credibility standards to sources that threaten a preferred belief than to sources that support it.
logicMoving the goalposts
Changing the evidentiary requirement after the original requirement has been met.
probability statisticsOvergeneralization
Extending a finding beyond the population, setting, dose or conditions that the evidence actually covers.
source evaluationPatent-as-proof error
Treating the existence or grant of a patent as independent proof that an invention works as claimed.
probability statisticsPossibility–probability confusion
Treating the fact that something is possible as if that makes it probable or evidentially competitive.
causal reasoningPost hoc reasoning
Inferring causation because one event happened before another.
conspiracy reasoningProportionality bias
Expecting major events to have correspondingly major, intentional or dramatic causes.
conspiracy reasoningSelf-sealing conspiracy reasoning
Interpreting counterevidence, missing evidence, and criticism as further evidence that the conspiracy is real.
scientific reasoningSurrogate endpoint error
Treating improvement in a proxy measure as if it automatically proves improvement in the patient-relevant outcome.
source evaluationTestimonial evidence
Treating personal endorsement or reported experience as sufficient evidence of general efficacy, prevalence or causation.
scientific reasoningTrace-amount fallacy
Treating the mere detectable presence of a substance as evidence of meaningful biological or toxicological effect.
rhetoric languageUndefined claim
Using a central term or outcome so vaguely that the claim cannot be measured, compared or decisively tested.
scientific reasoningUnfalsifiability
Framing a claim so that no conceivable observation is allowed to count against it.