This site is not maintained. Click here for the new website of Richard Dawkins.


← Anecdote vs. fact

Schrodinger's Cat's Avatar Jump to comment 5 by Schrodinger's Cat

'Anecdotal' is not a swear word. The fact is that whilst one may not be able to discern anything from an individual anecdotal report, one can form a reasoned inductive probability based upon multiple cases.

That probability may not tell you anything about the phenomenon, but it will at least give you greater confidence that there is a phenomenon to be explained.

A good example would be ball lightning. If one person tells you he has seen a mysterious ball of light hover above the ground, pass through a window, and disappear in a shower of can perhaps dismiss this person as hallucinating, unreliable, and so on. But, if you get multiple reports.....all describing the phenomenon in a similar manner....then you can start to build up a possible scientific hypothesis and a far greater confidence that there is a real physical phenomenon.

The point being that many of the esoteric phenomenon where science can indeed make a statement of some sort about veracity are in fact based primarily on anecdotal evidence. Thus, even the most sceptical researcher of ghosts admits that people do experience what they perceive to be ghosts. There is a 'real' phenomenon, though scientifically the most convincing explanation is that there's a common psychological cause. Nevertheless, to argue that ghosts are simply 'all in the mind' and dismiss the anecdotal reports is precisely to miss the anecdotal evidence pointing to a common psychological cause.

Thus, anecdotal evidence is at the very least a source of valuable insights into human psychology. And given that people are not always mistaken about what they have experienced......for example ball lightning, sprites, earthquake lights, and a great many other anecdotal reports with real scientific has to wonder why the mere mention of the word 'anecdotal' gets such bad press among scientists.

Tue, 20 Sep 2011 19:46:13 UTC | #873254