Beth said:
Anecdotal evidence cannot distinguish between placebo and an actual effect, thus no anecdote can ever meet your requested standard of proof. That's fine and appropriate skeptical thinking. But anecdotal evidence cannot establish the cause of an effect, whether that effect is harm or benefit.
If you insist on thinking of causes as being singular (
the cause of an effect), then no evidence whatsoever will ever be able to establish the cause of an effect, because effects usually, if not always, have multiple causes. Evidence that can be used to establish a falsehood is no evidence.
If you accept that multiple contributory elements can all be causes, then it's easy for anecdotal evidence to establish the cause of an effect. Where the effect itself is anecdotal, then much of the cause will almost have to be anecdotal. Glasses break when dropped (generally) --- but the cause of that
this glass being now broken (an anecdotal effect) comes from
that person dropping it (an anecdotal cause). [And, of course, there are multiple causes. Another reason that the glass is broken is because the floor in
this room, where it was dropped, is hard tile as opposed to styrofoam.]
Loki had a pretty good analysis upthread -- let me briefly repeat (and expand). To establish a particular (anecdotal) cause, we need to establish two things : first, that a general cause is in play (glasses break when dropped), and second, that the specifics of this anecdote support the general cause (... and the glass was dropped, therefore breaking).
In the case of paranormal beliefs, I think the general statement that "people do stupid things when they believe things that aren't true" is another of those general causes that can easily be supported -- although, like breaking glasses, it's not universal.
So we have Loki's analysis of
belief -> action -> result
Anecdotal evidence cannot establish the "action -> result" link, especially when there are multiple potential causes for the particular result in play. But I think it's much easier to establish the link between belief and action; when someone claims to believe X and then acts as though X is true, I fail to see how you could not derive that the belief caused the action.