Anecdotal evidence is basically saying "X happened to me/someone I know/someone I heard about" in a discussion and/or argument. If it's true that X happened to this person (which is not proven by the assertion alone), it proves only that X is not impossible. It most certainly does not establish that X has ever happened more than once, or that X is common or pervasive - and that's what this fallacy about. The anecdotal fallacy is using anecdotes instead of actual statistics and making general rules and grand claims off them, accurate picture of the reality of the matter be damned. Crops up often in heated arguments and amidst Hasty Generalisations, whereby an anecdote is not just used to illustrate or critique a general rule but as 'proof'note in itself. As they say, "The plural of 'anecdote' is not 'data'".
Anecdotal Evidence is extremely prone to Confirmation Bias; when it doesn't fit one's viewpoint, it can be very easily dismissed as this fallacy. If it does fit one's viewpoint, it's a perfect example of that viewpoint applying to real people in the real world.
- Seatbelt usage: many know of someone or have heard about someone who staunchly refuses to use seatbelts because they were involved or know someone who was involved in an accident where they got stuck because they couldn't undo the seatbelt, possibly suffering injuries or dying because of it, casually ignoring the many lives saved and injuries prevented by this device, and how seatbelt failures and defects are very rare.
- A belief that some event was caused by the supernatural usually falls into this category. For example, if you asked the Abrahamic-Monotheistic God to heal your grandma, and she got better, that wouldn't prove God did it. While such an episode might be powerful to the individual, it does not prove anything. A hundred other people might ask for the same thing and get dead grandmas. Indeed, this is a False Cause fallacy.
- Another instance of this is what is sometimes called "argument from mere analogy" — the mistaken belief that because analogy is often used to illustrate proof, analogy is proof.
Looks like this fallacy but isn't:
- When the question is whether people believe something is true rather than whether it is true.
- When the evidence presented is a thought experiment ("let's say I met someone who...") rather than a claim of proof.
- When the anecdote is used to argue that something is possible: If someone says "No one can live more than a hundred years", the answer "My grandmother did!" isn' t fallacious. Of course, one must consider whether the objector is telling the truth about their grandmother, but that becomes an issue of fact, not of the logic of the argument.
- Similarly, when the anecdote is used to counter a universal claim as a counter-example; using the example above, claiming that grandma is a long-lived heavy smoker with no health problems doesn't prove that smoking is healthy, but it does disprove the claim that "All smokers get cancer and die young". A universal claim (All A are B) is disproven by the existence of even one case where A is not B. That said, the maxim "extraordinary claims require extraordinary evidence" applies here, and if the only evidence that this old person who smokes exists is that the opponent asserts that she does, this is not a particularly compelling counterpoint.
- Eyewitness testimony will often fall into the anecdotal category, since it's generally non-repeatable and can't be scientifically verified (although not always). While of course this can be wrong and isn't the best, it's still evidence. Cross-examination and other mechanisms are used in an attempt to check its problems.
- An anecdote being used to illustrate an argument. If somebody can show legitimate evidence that something happens often, and uses an anecdote purely as an example, there is no fallacy.
- When a sufficient number are collected to be statistically significant, such as while testing new drugs.