I was in the odd position once of having someone here give me the old “anecdote is not data” line when the story that I told was about me and was being written up as a training case at Harvard Medical School. There is another case study being written about a deceased family member that will appear in the literature some time in the future.
I have a related academic background and I studied the cases for hours on end until I could discuss them almost as peers with any doctor they sent to see us.
Somehow these cases that I have close, personal experience with are good enough to help train doctors and push science forward but they are just anecdotes if I share any part about them here.
whole bean, I’ve seen, and used, the phrases you’ve pitted. The reason that I tend to think that the phrases, limited as they must be by common sense limits, are still useful is that often an observer may use their anecdotal evidence as proof of a thesis that is, in fact, only marginally supported by the account in question.
Forex, recently I found it suspicious that of the three persons in my family who vote, two of them were required, in a recent election, to vote via affadavit ballot. The fact that the two required to vote via affadavit ballot were both stubborn enough to remain unregistered with any political party is my anecdote. If, however, I start claiming that there is a conspiracy to keep independents from voting in Monroe County, and offering my anecdote as evidence, I would hope you’d throw it out.
I’ve always heard it as, “the plural of anecdote is not data.” The funny thing is, I think we could all agree that anecdote can mean:A short account of an incident (especially a biographical one) . We could also agree that data is the plural of datum, and datum can mean: A fact or proposition used to draw a conclusion or make a decision. If we take only those anecdotes that are based in fact (no embellishments), then I have no problem envisioning a scenario where the plural of anecdotes (which in themselves are datum) is indeed data.
I would question your conclusion but not because you had no evidence, only because you had very little. I would sugges that your conclusion was the product of flawed inductive reasoning.
Of course it all depends on context, and what me mean by “evidence”.
Joe: Do you have any evidence that golbalization is bad for the American economy?
Schmoe: My brother lost his job due to outsourcing!
Two problems here: Something can be bad for Schmoe’s brother and still be good for the economy overall. But even if we accept it as “evidence”, it’s not significant or convincing. This sort of stuff happens all the time in GD. I’ve got evidence to prove it!
It very much depends on what you’re talking about. I think the most famous use of the “plural of anecdote is not data” line is by Blake in the “how high can cats jump?” thread. In that thread, people were offering personal testimony that their cats could clear tall buildings in a single bound, and Blake was stubbornly refusing to believe it. In this case, if you ask me, anecdotes *were *data, because the subject wasn’t terribly complicated: either a cat could jump vertically however many feet, or it couldn’t.
But if, as in John Mace’s example, you’re trying to generalize from a tiny, tiny sample, then of course the plural of anecdotes is not data. Unless you can assemble ten thousand such anecdotes.
Yeah, Blake was the one who introduced me to the phrase.
As for John Mace’s example, in that case there is no plural – one bro, one job lost. But let’s say there were more – maybe 10,000 jobs lost – that’s data (of which each job lost is an anecdote for someone).
Nope; the nature of ‘anecdote’ is that it inherently isn’t rigorous. Proof requires rigor; no amount of anecdotes can impart it. After all, all 10,000 of those people might be mistaken.
(If you think I’m kidding, open up a GD thread asking for anecdotes that prove God exists. Then listen to the non-theists. Cardinality doesn’t matter.)
This proves nothing as you asked me to take anecdotes to prove something terribly difficult, if not impossible to prove, but that does not mean anecdotes, enough of them at least, cannot be assembled as proof of something simpler to prove. I guess a lot of it depends on the burden of proof we are assigning to the task – I am a lawyer, not a scientist, so I look at proof differently – I confess that “data” has scientific conotations, so I guess in a purely scientific sense, “the plural of anecdote is not proof” might be true. I am simply not in a position to say.
From a serious empirical standpoint, unverifiable anecdotes from anonymous sources on the internet are pretty much worthless as evidence. We’ve seen our share of bogus stories and even self-proclaimed identities exposed around here. It’s not unfair to say that, while a poster’s story may be plausible or may even be true, it doesn’t meet any real standard of evidence and it would be foolish to treat it as such. Personal anecdotes can influence a discussion, be taken seriously or even generally believed but it wouldn’t be very rigorous to say that they can be taken as dispositive or settle a debate. They certainly can’t prove anything.
That’s not to say that I’m in the habit of dismissing everyone’s anecdotal claims as bullshit. I tend to take people at face value, but only in a provisional way. I still keep it in mind that just about anybody could be lying so while I might (most of the time) lean towards accepting people’s stories as probably true, I never accept any of it as settled fact. To quote Jerry Seinfeld, I wouldn’t put anything past anybody.
It’s difficult to generate well-controlled personal anecdotes. It limits the value of personal anecdotal data in a rigorously scientific sense. “The plural of anecdote is not data” lacks nuance, but I always figured that’s what the adage was getting at.
This proves nothing as you asked me to take anecdotes to prove something terribly difficult, if not impossible to prove, but that does not mean anecdotes, enough of them at least, cannot be assembled as proof of something simpler to prove. I guess a lot of it depends on the burden of proof we are assigning to the task – I am a lawyer, not a scientist, so I look at proof differently – I confess that “data” has scientific conotations, so I guess in a purely scientific sense, “the plural of anecdote is not proof” might be true. I am simply not in a position to say.
Sure, but if your conclusion is ‘not all cats are black’ then your experience of seeing a white cat is, in fact, proof. It might not be well documented proof, mind you, so people still might not accept the story as true, but if it is true, it’s sufficient.
as you asked me to take anecdotes to prove something terribly difficult, if not impossible to prove, but that does not mean anecdotes, enough of them at least, cannot be assembled as proof of something simpler to prove. I guess a lot of it depends on the burden of proof we are assigning to the task – I am a lawyer, not a scientist, so I look at proof differently – I confess that “data” has scientific conotations, so I guess in a purely scientific sense, “the plural of anecdote is not proof” might be true. I am simply not in a position to say.