Nah, the plural of anecdote is confirmation bias. Why do people keep trotting out this "plural of anecdote is data" line, is it a quote from some high-profile idiot?
Of course data can also give you confirmation bias, maybe that's the point, maybe that's why people defend anecdotes. What you have to do to find out if something is really happening is reason about it, then test your theories by doing your best to knock them over. Often though we just end up testing for statistical correlation without any better theory than "these two things go together". In that case, the plural of anecdote is bad methodology. The best thing I can say about anecdotes is that they might give you ideas.
Edit: found it, it's due to one Ray Wolfinger, behavioral political scientist: When a student once categorized one of Wolfinger’s claims as “just anecdotal,” he paused for an expectant second, dropping a copy of Robert Dahl’s “Who Governs” onto his seminar table as he replied, “The plural of anecdote is data.” His quip, emphasizing that statistics represent human stories, would become a well-known aphorism throughout the field. Well this probably shouldn't be taken literally and I suspect the criticism of his claim was fair.
> Nah, the plural of anecdote is confirmation bias. Why do people keep trotting out this "plural of anecdote is data" line, is it a quote from some high-profile idiot?
It is standard Bayesian reasoning[1]. But it requires independence between observations, which many people forget!
The things people say aren't truly independent, even when they are. Like when Charles Sheffield and Arthur C. Clarke wrote novels with the same plots at the same time, they were working independently, but they weren't culturally independent.
Besides, even if several people independently assert "tying a ribbon to a wishing tree cured my warts", that's not an explanation of what actually took place. If repeated observations with unbiased instruments confirm this, then there's something wrong with the instruments (or something), until you have an explanation.
(But saying things like that usually prompts people to bring up cosmology or particle physics or other fields where we really do have to resort to saying "the measurements say it's happening, we'll have to assume it's happening" without understanding much.)
You might be missing the forest for the trees. Anecdotal evidence does not make a proper study, but, to quote myself "phenomena exists before the related studies do."
Oftentimes we (humans) use imperfect, but well-known, figures of speech to convey common ideas.
Of course data can also give you confirmation bias, maybe that's the point, maybe that's why people defend anecdotes. What you have to do to find out if something is really happening is reason about it, then test your theories by doing your best to knock them over. Often though we just end up testing for statistical correlation without any better theory than "these two things go together". In that case, the plural of anecdote is bad methodology. The best thing I can say about anecdotes is that they might give you ideas.
Edit: found it, it's due to one Ray Wolfinger, behavioral political scientist: When a student once categorized one of Wolfinger’s claims as “just anecdotal,” he paused for an expectant second, dropping a copy of Robert Dahl’s “Who Governs” onto his seminar table as he replied, “The plural of anecdote is data.” His quip, emphasizing that statistics represent human stories, would become a well-known aphorism throughout the field. Well this probably shouldn't be taken literally and I suspect the criticism of his claim was fair.