Anecdotal Evidence

-thunks head on desk repeatedly-

Right, let's turn on my time travel device, and alter the course of history.

History changed. Now that the main (and largely irrelevant) offender has been removed, let's start this over again.

edited 3rd May '11 3:05:55 PM by annebeeche

I'd still say that I want to see evidence. I don't know the context. It seems fishy, especially since things like Conservapedia are just a click away.

Now, I have seen many internet arguments. Plenty abuse the anecdote fallacy. Some will be believed unjustly, others will be disbelieved unjustly. It's a bit of a fickle mistress, that way, much like the trust-worthiness of an anecdote, heh.

In a debate, it is best to use anecdotes as either a stepping stone or an add-on, not the main cource. Mostly because it comes down to he said vs. she said and what else have you.

Now, if someone does indeed disregard your argument, that does not mean what they did was a bad thing, even if you were sincere. They had their reasons, you had yours. The disaproval of one anecdote does not mean anecdotes as a whole are looked down upon irrationally. Or rationally, I suppose.

The main problem with anecdotes, as I have said before, is the execution. Giving an anecdote as fuel or to disprove a universal statement is great. Stating an anecdote, focusing the entire context of the debate around that particular incident, and treating it to be worth a hefty percentage is where the problem comes in. Many times, anecdotes are used this way, and it's a pity because that is not how an anecdote should be used in a debate. There are those who do use it correctly, and I am sure there are more than I give credit for.

To put it concisely, most of the time, I find that anecdotes are really more of a symptom of bad debating, which is why they tend to be given a bad reputation, and rightly so. Much like writing erotica, there are far more bad examples than good examples in the bunch, even if the good examples are really, really good.

edited 3rd May '11 3:14:31 PM by MrAHR

I see and understand all of your points.

I see using ancedotal evidence as a thin ledge or a slippery slope. There's a right way to do it, but there are a lot of mistakes you have to avoid making, treating your story as if it speaks for an entire unseen group being one of them.

Oooh. That's another good point. —steals it—

I have a question: What do you think of this kind of scenario?

1. Person A gives anecdote about how a pretty obscure, 'Insignificant' issue has affected him.
2. Person B is like, "Hey, that happened to me too!" and shares his story.
3. Persons C, D and E and others also share their stories, and pretty soon you have an army of anecdotal evidence.

Depends, really.

For instance, we're a website of a young nerds, yes? So we're going to have a biased demographic.

Most websites do. Even generic chat sites do, since the heavy personalities are going to either attract or repel certain people. For instance, I recall a website that had a lot of anecdotal evidence of the horrors of massachusetts legalizing gay marriage.

Here, we're more likely to have reclusive types, and what else have you.

Also, there is still a silent majority. If you and everyone else in the thread are talking about...I dunno, how you all had goldfish and they all died at age five, if I did not have a goldfish (thus irrelevant) or I had a different fish, or it died at a different age, if I were a less OMGMUSTPOSTATALLTIMES person, I might not feel so inclined to post about it, simply because it's nothing I feel I could contribute to.

Now, those are two factors that make them iffy.

A third is how much is an army of anecdotes? Ten? Twenty? Thirty? A thousand?

Lets say you do indeed have an "army" of anecdotes. What does that compare to the actual numbers? If there are millions of X type, and the army is a few hundred, what seemed like a lot suddenly seems a bit more fluke-ish.

edited 3rd May '11 3:44:44 PM by MrAHR

And that's why we need statistics.

Though personally, it doesn't really matter to me how big a minority is. It just matters to me that the minority exists, and that's what an anecdote can help prove.

edited 3rd May '11 3:46:47 PM by annebeeche

Fair enough. I guess it depends on whether you are trying to prove that someone is a fluke in the system (Like, say someone was in a restaurant, tripped over a curtain, slammed into the rock fountain, and now has a concussion. A serious event, possibly sue-able, but nothing that implies something wrong with the restaurant system), or a reoccurring issue that needs to be addressed as a general problem (a mass/minor outbreak of e coli, from Taco Bell).

Or, alternatively, if the minority is huge enough to affect your overall sales/effectiveness (medicine owns this section).

edited 3rd May '11 3:54:50 PM by MrAHR

Anecdotal evidence serves an important purpose in putting a face on statistics, which can otherwise be overlooked due to lack of emotional impact. That, and if anyone is serious about actually helping those whose story is told by said anecdote (a big if, unfortunately), that can help illustrate a solution.

edited 6th May '11 12:01:21 PM by Ratix

But at the same time, an emotional story can blind you from everything else.

Since I'm fond of anecdotal evidence, my two pennies...

As Madrugada said, it's great for disproving supposed "universals". It's also great to use in debate if the user recognizes that there's no proof that his/her anecdote is necessarily completely true for anyone else, and states so. Alternately, presenting it as "this is just my personal experience, but <counterargument>".

Finally, no person is unique; if one has had the experience, it is safe to say that many others have had something similar occur. I hold the "Rule of 34" up as an example we all know.

Besides, statistical evidence is subject to its own set of biases. For example, how was the data collected? Was proper scientific method followed? And who collected the data, and for what purpose? Plenty of partisan organizations fund "surveys" for the sole purpose of collecting "statistics" to throw at their enemies.

I'd say both have the same troubles, and the same basic rule; if used correctly they can bring order to chaos. If used wrongly, they make you look like a douchebag.

This. Just this. Like in the recent "oh noes vaccines cause autism" scare where the emotional nature of the anecdotes over-rode the science and statistics and led to an unfounded scare. Of course, you could argue that the medical sources ought to have better utilised emotional anecdotes in support of their position... but that seems a little bit like stooping to sensationalism.

TL;DR: Anecdotes can be great for providing a real and emotional face to statistics, but this can also lead to emotion-fed blindness to the facts and cause real harm.

If anything, I like ancedotal evidence, and I think it is quite good evidence or a good argument, IF one is willing to accept that it is ancedotal in the first place.
On the other hand, ancedotal evidence should never be dismissed, because it is still evidence, but it should be counted for what it is: Something that may or may not be relevant or large enough as evidence to be worth anything.
I agree with Mr AHR, BUT if one is willing to admit the bias in the first place, then ancedotal evidence is valid.