This one is probably the most difficult fallacies to fight. An anecdote is a brief story or vignette. Anecdotal evidence is the use of someone's story as evidence for a general hypothesis. It often starts with something like, "I knew a guy who," and then continues on with some story to provide evidence. "I knew a guy who smoked three packs of cigarettes a day and lived to the age of 94, which proves that smoking doesn't cause cancer." The problem is, statistics only work with large samples, and anecdotal evidence has very small samples, often sample sizes of one.
An awful lot of my own understanding of the world comes from my own experience, but I have to be careful because I understand my sample is just me. Just because some events happened in my life does not mean that they are generalizable to the whole population. I do write about my story here in my blog, but this does not mean that anybody else will have the same experiences.
One that always gets me is "most gay guys I know have relatives that are gay, too." This is used to explain that homosexuality is genetic. But that's anecdotal evidence. In reality, since the percentage of the population that seems to have gay orientation is around 6%, unless the friend has fewer than 16 people in his extended family (uncles, cousins, and second cousins often are called upon in these instances), you should expect a gay relative or two. In fact, even if there were only a dozen other people in a gay guys extended family, there is still a 52% chance that at least one of those relatives would also be gay, just using randomness.
Anecdotal evidence is almost always used in persuasive arguments rather than real research, so be wary of anyone using such evidence to make a point (or sell something "I used this product for three weeks and lost 45 pounds").
No comments:
Post a Comment