It certainly is data, if poor quality data. It seems like it is simply not rational not to consider anecdote, you just have to view it in the proper context. To simply dismiss it seems downright foolish, and I can’t believe anyone really does.
I mean someone could say that shortly after eating oysters their cousin died, that is a fairly useless anecdote and I’d pay it no attention.
Now if four people mention they knew someone who died shortly after eating oysters harvested from Hypothetical Bay with similar symptoms, I would not be prepared to proclaim oysters from this bay are poisonous but I’d sure think there was something there worth investigating AND I think I’ll pass on those oysters
Or is there an argument for ignoring all anecdotes?
The problem is usually you hear anecdotes such as “I got the flu after I got a flu shot”. When the person has no idea (usually) whether they actually got the flu - and there have been studies done that PROVE you don’t get sick from the shot. You can’t (or shouldn’t use) anecdotes when there are already well done studies on the matter.
Persoanlly I find them somewhat useful in assessing possible effects/side effects of anti-depressant medications. Reading these in Internet forums - when taken with huge grains of salt - can be useful. It is known that studies on these medications vastly under report incidents of sexual dysfunction - so in cases like this - anecdotes can be useful.
Generally anecdotes on vague symptoms like fatigue are much less useful than those on specific symptoms like sexual dysfunction, blurry vision and the like.
An anecdote is a datum. So yeah, ignoring all anecdotes as non-data is a matter of not understanding what data are. Some people may think data are magic. They’re really not. A data set is pretty much a big error-prone list. Science is messy.
But the more data you have to work with, the more you can get away from false predictions based on false expectations due to a narrow sample size. And we do want data that are selected by something other than the attention-getting of the odd sore thumb and squeaky wheel. Which statistical issues, taken together, more or less give us the point of the expression.
When I saw the thread title I did a double-take, since I’ve thought of starting a thread with similar title!
Obviously anecdotes are inferior to the best derived statistics but
[ul][li] The earliest civilizations were built before the science of statistics or experimental controls, and “anecdotal” information is still a major source for people’s practical knowledge.[/li][li] Good anecdotal data will beat bad statistical data anyday; there’s plenty of the latter in the Age of the Interwebs.[/li][li] For a good observer, personal anecdotes can be extrapolated.[/li][/ul]
There were very clear demonstrations of the point in threads about underage prostitution. Published estimates varied from absurdly low to absurdly high. lexi was able to provide reasonable estimates by extrapolating from her personal experience, yet was ridiculed for it in the thread (“Yadder-yadder-Yaddote, plur-eral of anecdote…”). Some Dopers seemed to think the absurd numbers were better because they were published by an “authority.” :smack: (Arrests were used as a proxy for totals despite the obvious absurdity: arrests were a “statistic.”)
Despite that she indicated how she extrapolated to get some numbers, she was answered with:
plural, anecdote, data, etc.
[/QUOTE]
I sympathise with lexi. I know little about Alcoholics Anonymous except anecdotes from my Mother, who worked as counselor with many AA people.
My vivid memory is walking downtown on Sundays with my Mom to drink tea or see a movie and have strangers walk up, introduce themselves and tell me “Your Mother saved my life!” This happened several times. I’m reluctant to mention it here, whether in support of AA or my Mother, since someone is sure to show up and show off their sophomore’s lesson “Yadder-yadder-Yaddote, plur-eral of anecdote…”
I wonder how many people who have the reaction** grude** and **septimus **complain of have ever done field research. Our ability to measure, count, and graph real life can be annoyingly imprecise when we’re trying to get our neat little charts accurate, even if we’re just counting wild onions.
Also, people have a natural filter to focus on what agrees with them and ignore what doesn’t. To say “the plural of anecdote is not data” depends entirely on how the anecdotes were collected - randomly (usually not) or in a controlled group (usually not). In other words, what are the the denominator (is it a statistically significant proportion) and the context of the collection?
When you say it’s an upgrade from “my cousin ate an oyster, and got sick the next day” to “…then I found FIVE OTHER people who have also eaten oysters and gotten sick the next day!”, it could still be meaningless because it could well be this person is mentally tallying 5 out of the N people he’s told the sick cousin story to (with an unknown, possibly very large value of N) who replied, “Yeah, that totally happened to me too!” The mistake is to think “my conclusion just got 500% more valid, that’s a big jump”.
Is it worth prompting you to look into it further? Maybe. It is “meaningful” on its own? Not really.
BUT, if you narrowed it down to “my cousin got sick from doing raw oyster shoots at Sam’s Clams last Friday night, and so did 5 more of the 8 people at his table”, that is significant. Or if it turned out that your cousin and those 5 people were among the first to ever eat a newly discovered variant, the Formica Oyster, found only in a bay downstream from both a nuclear power plant and furniture factory.
There are many categories of inconclusive evidence, anecdotal is one of them. I think “anecdotal” is more of a communications/linguistic term than a science or research term, and that’s where the confusion comes from.
I think it also matters whether you regard the anecdote in question as true. All too often, people rattle off what happened to them in a way that sounds unreliable, unlikely, or flat-out unbelievable. I think that contributes to the bad reputation anecdotal evidence has.
The situations in which it’s particularly dangerous to use anecdotes as data are those in which you’re talking about how common or how likely something is—when you’re dealing with percentages or probabilities. Like in the recent GQ thread Likelihood of same sex children?: several people responded with anecdotal evidence (“My brother has four daughters”), which doesn’t help in establishing how likely or how frequently something happens.
As a matter of fact, the original quote was “The plural of anecdote is data”. Of course, Wolfinger wasn’t saying that an anecdote or three trumps all other data, just that you can’t throw away data that doesn’t support your desired conclusion by labeling it “anecdote”.
I’m not sure why this zombie was bumped, but I’ll just add that I think the important point is that you cannot draw conclusions of any kind from anecdotes. To draw conclusions (i.e., the oysters are poisoned), you need to have carefully collected, unbiased data, controlling for every confounding variable you can think of (and possibly more). Anecdotes, almost by definition, are biased by self-reporting, if nothing else. I agree that they can indicate places where it would be valuable to do investigation, but conclusions require hard data.
For instance, the “five people ate oysters and died” might lead to a valid conclusion if, for instance, they were the only five people who ate those oysters and they had nothing at all else in common. But what if there were 3,000 other people who also ate those oysters without any problems, and it turned out that those five also happened to work in the same chemical plant that had a serious leak that day? Suddenly, with more data, the connection becomes far more tenuous.
There was a mention on Reddit not long ago, about how the original quote has been reversed, and it no doubt annoys people to find that something they have repeated for years is actually completely untrue. While doing a Google search I saw this thread, and rather than start a new one, about the exact same thing …
Question:
Their are comments of 5000 ( # pulled out of my ass ) people who have been declared dead by a medical doctor who later ( a short time more than 15 minutes) revived and stayed revived for at least 24 hours +
They were specifically asked if they saw or remembered anything of the occurrence. 50 of them told of unique things that their body was present for that only happened while they were dead.
Now, since there is no way this knowledge can be scientifically proved or disproved IMO, is it automatically untrue that there is some kind of consciousnesses after death because 4,950 of the people asked did not remember anything?
