Statistical and Scientific

Ignores prior probabilities when evaluating new evidence, leading to mistaken conclusions.

Selects only the data that support a claim while overlooking the full dataset.

Assumes that because one event follows another, the first caused the second.

Draws a conclusion from a comparison between things that are not sufficiently alike in relevant aspects.

Believes past independent random events change the odds of future independent events.

Assumes trends within observed data automatically hold between or beyond the data points without justification.

Misses an underlying relationship because another variable hides or distorts it.

Uses a striking anecdote or vivid event to outweigh statistical evidence or broader trends.

Confuses the probability of observing the evidence if someone is innocent with the probability of innocence given the evidence.

Mistakenly attributes a change after an extreme event to a specific cause, ignoring regression to the mean.

Distorts conclusions by using a non-random or non-representative selection of data or participants.

A trend appears in several groups of data but reverses or disappears when the groups are combined.

Focuses on successes that survived a process while ignoring failures, leading to wrong conclusions.

Draws a causal conclusion without sufficient evidence, often from mere correlation or sequence.

Assumes that because two things occur together, one causes the other.

Attributes a relationship between two variables to causation when both are driven by an unconsidered third factor.

Mistakes the direction of cause and effect between two correlated variables.

Two variables correlate by coincidence or external patterns, but no causal link exists.

Relies on personal stories or isolated examples instead of representative evidence.

Draws conclusions from a sample that does not represent the population of interest.

Evidence is distorted because studies with positive or exciting results are more likely to be published than null or negative ones.

Infers individual-level conclusions from group-level data, ignoring within-group variation.

Draws a broad conclusion from too little or unrepresentative evidence.

Assumes that because two things move together, one must cause the other.