Anecdotes and dozey doats and liddle lamzy divey

I'll bite.

An anecdote is a story about a person or something that happened -- typically a short, memorable one. (Anecdotal evidence is when you use such a story as evidence of the truth of a larger proposition.)

E.g., anecdotes:


I met a Trump supporter who insisted the infamous Access Hollywood "pussy grab" tape was fake because Trump said so. (anecdotal evidence if used as evidence that "Trump supporters believe everything Trump says, no matter how obvious it is that it's a lie.")

I read about a Black man who said he couldn't be friends with White people. (anecdotal evidence if used as evidence that "Black people are racist against White people.") 

I knew a college girl who falsely accused a boy of rape after she regretted having sex with him.  (anecdotal evidence if used as evidence that "girls commonly falsely accuse boys of rape."

When I was in Egypt, a Muslim tour guide implied American women were easy and inappropriately propositioned me.  (anecdotal evidence if used as evidence that "Muslim men are all sexual harassers.")

I heard about a guy who pretended to be homeless, came into town every day to beg, then went home to his nice house in the burbs with hundreds of dollars every evening.  (anecdotal evidence if used as evidence that that "homeless people are all lazy cheaters")

A story in the Washington Post turned out to be exaggerated.  (anecdotal evidence if used as evidence that "mainstream news is fake")



These anecdotes are all either mine or I've heard them, and I've heard them used as anecdotal evidence in the way I indicate.  

FWIW, in my opinion many are more prone to clutch anecdotal evidence even in the face of facts and evidence because (a) anecdotes tend to be more memorable, and (b) because people tend to clutch at things that reaffirm their own gut, regardless of whether they are demonstrably true, and happily seize on any excuse.

You can examine a bunch of facts and statistics about homeless people -- or you can remember that one story about the fake homeless guy begging for a living. If your own gut is that most homeless people are grifters looking to live off hard-working 'Mericans, you're more likely to point at the anecdote than bother with charts and numbers. If you're inclined to think they're mostly people screwed by a system gone wrong, you're more likely to dig into the evidence.

You can take a look at the actual percentage of false or exaggerated stories in the news versus true stories, or you can point to one or two false ones as proof it's all "FAKE NEWS!" If you want to believe it's all fake news, you cling to the anecdotes.

And so forth.

A fairly obvious, common one with vaccines:

"I got the flu shot a few years ago and immediately got sick. I'll never get it again -- the flu shot causes flu."

"My neighbor's kid is autistic. They started noticing signs after he got his vaccinations. Obviously, vaccinations cause autism."

You can tell these people a zillion times that you cannot get the flu from the flu shot, and that there is no link between autism and vaccines, show 'em all the studies you like -- what will stick in their minds is that one time they got sick the day they got the flu shot and their neighbor's autistic kid. I submit they're a little suspicious of vaccines (and likely modern medicine and Big Pharma) to begin with, and the memory of the illness and the autistic kid are more vivid than any stinkin' facts and studies, so the anecdotes stick.

I'm reminded of Bastiat's "That which is seen and that which is unseen," or seeing only the tip of an iceberg. The applicable denominator for any anecdote must be judiciously considered to avoid the Chicken Little Syndrome.

Anecdotes provide emotional support to give one instance of X a higher weight than it logically deserves.

Let me weigh in with this one as I had a discussion on it recently. So yes, that would be anecdotal. If you're looking to say this one case proves a pattern. It can also be considered an example of that it does happen. It is, however nowhere near proof that it happens all the time, or it's common. 

A person is accused of a crime. Rape, murder, assault, whatever. He's innocent. One story in anecdotal and doesn't prove a pattern. However, if they're convicted of that crime, it's natural to ask what went wrong. If it's because the evidence just wasn't available, then it just is one incident. But if it's that the evidence was there and was not given to the defense by the prosecutors, that's different. Just like one incident can prove someone is a killer, one case can show a flaw in a system. If there's no law that says the State must turn over any and all evidence (I know, here it's called discovery.) and that causes a miscarriage of justice, then it's more than anecdotal. 

So if you're having a discussion and someone uses a single case, it's anecdotal, but it's a single example of it. The question of if it points to a larger pattern is I think more complicated. 

Jack killed Simon. This is one case. If I use it to prove all men named Jack kill men named Simon, it's anecdotal. If I use it to prove that Jack is a bad guy, it's not.

Debby is arrested for killing Tom, however Debby was innocent. The investigation proves this. Again, it's anecdotal, but it's an example of how police might arrest the wrong person.

Tim is charged with killing Matt. He didn't, but a reasonable person would think he did. There was blood left behind from the killer, but it's 1978 and they can't get DNA. Decades later, someone extracts the DNA and they realize Matt as killed by Stan, who was also a suspect. It's one case, and it only proves DNA can exonerate an innocent man, not that all men convicted of murder without DNA are innocent. It depends on what you're trying to prove.

Dave is arrested for killing Kurt. He didn't, and the evidence proves that. However, the prosecutor seeing that charges her, goes through a trial and hiding the evidence convicts her. After 2 years in prison, she gets the proof and is exonerated. A review shows the DA was aware of the evidence and wasn't forthcoming with it when they were should have been. There's a loophole in the law that says they didn't break a rule. The one case can be seen as evidence that whatever loophole exists needs to be closed. 

To Vince's point:

If one is using an anecdote to disprove the notion of "X *never* happens," that's completely legitimate.

But if one is using an anecdote to attempt to prove that X always or often happens, it is not legitimate.

One might take a large number of similar anecdotes to support the idea that something frequently happens. E.g., I was very frequently sexually harassed in Egypt. I've been to some 40 countries now, and never did I encounter anything close to the amount of harassment I received there. Several women I know had the same experience--indeed, it is the experience of every single woman I've met who has traveled there--and I've read articles about other women who've had this experience. I might use all of these anecdotes to support the idea that sexual harassment seems to be a problem in Egypt, or at least more of a problem than it is in, say, Denmark or Australia. It's still not definitive PROOF of it, and there might be a great deal of evidence out there to disprove it (e.g., evidence showing that despite my personal experience, sexual harassment is equally prevalent in Denmark), but cumulatively, a number of similar anecdotes at least is some evidence that my hypothesis might be true.

However, a single instance of sexual harassment by my tour guide? That's not evidence of anything other than that one anecdote.

Poker is not gambling. It's a game of skill. Lots of people misunderstand that. The good players don't "gamble," they calculate the odds and use very mathematical reasoning (with a hefty dose of psychological and social analysis, because "reading" your opponent is also important). But bad players gamble, and will always tell you about the time they called an all-in with 9-5 offsuit and the other guy had Aces, but they hit two pair on the river... (In other words, for non-poker players: they made a bad call that would lose most of the time, but one time they got lucky and they use that to justify making what was still a bad call.)