Aust. Skeptics updates

Richard said:
New free stuff to download this week from our site.

Calendar, title deed to the Sydney Harbour Bridge (Homoeopathic), more reports and more 'Skeptic tank' radio shows.

http://www.skeptics.com.au
Click to expand...


Hi Richard,

Just been browsing your site and came across the prize challenge. In the rules it says:

"The challenger must beat million to one odds against a chance result."

and later:

"It’s important to remember that as this is a scientific inquiry. A failure to successfully complete the test does not indicate that the power the challenger claims does not exist, it merely indicates that on this particular occasion it was not present, and that it remains unproved."


It could also be a possibility that a failure to successfully complete the test is a result of you setting an unrealistic alpha value. The challengers power may have been present but it may not have been strong enough to beat the odds.
 
Yeah... you might have a point. We are in fact working on a better set of rules etc. I think the wording is not what it could be.
 
Re: Re: Aust. Skeptics updates

davidsmith73 said:
It could also be a possibility that a failure to successfully complete the test is a result of you setting an unrealistic alpha value. The challengers power may have been present but it may not have been strong enough to beat the odds.
Click to expand...
However weak the challenger's "power" is, a sufficiently long series of trials would get you the one in a million, if the "power" really existed. Indeed, if the "psychic" could say in advance how much more accurate he or she was than chance, then it would be possible to arrange a test such that there was only a one in a million chance of failure --- if the "psychic" really did perform at the level claimed. That would be nice. Unfortunately, all claimants are people who have not tried to find out to what degree their abilities differ from chance. This is, I suppose, because people who do try to find this out find that the answer is 0, and so they don't become claimants.

Nonetheless, I make this suggestion to richard for what it's worth: if you can get them to state in advance how accurate they are, then it's fairly easy to do the maths to see how many tests they have to do so that they have, not just a one in a million chance of succeeding by luck alone, but also such that they have only a one in a million chance of failing --- if they really are as accurate as they claim.

Whether that would satisy davidsmith73, I cannot foretell: for someone who can complain about the alpha value of an experiment per se is clearly not mentally equipped to get a firm grip on a statistical argument.
 
Re: Re: Re: Aust. Skeptics updates

Dr Adequate said:
However weak the challenger's "power" is, a sufficiently long series of trials would get you the one in a million, if the "power" really existed. Indeed, if the "psychic" could say in advance how much more accurate he or she was than chance, then it would be possible to arrange a test such that there was only a one in a million chance of failure --- if the "psychic" really did perform at the level claimed. That would be nice. Unfortunately, all claimants are people who have not tried to find out to what degree their abilities differ from chance. This is, I suppose, because people who do try to find this out find that the answer is 0, and so they don't become claimants.

Nonetheless, I make this suggestion to richard for what it's worth: if you can get them to state in advance how accurate they are, then it's fairly easy to do the maths to see how many tests they have to do so that they have, not just a one in a million chance of succeeding by luck alone, but also such that they have only a one in a million chance of failing --- if they really are as accurate as they claim.

Whether that would satisy davidsmith73, I cannot foretell: for someone who can complain about the alpha value of an experiment per se is clearly not mentally equipped to get a firm grip on a statistical argument.
Click to expand...

The claimant could simply overestimate the strength of their "power" and there is still the chance that the pre-stated number of trials (and it has to be pre-stated) is not enough to reach a p-value of 0.000001. Thus a reason that the claimant fails to successfully complete the test is because their "power" is not strong enough. The claimant fails even though an effect was present. I'm just being objective about the level of significance that has been put forward in the test. I agree with the notion that extraordinary claims need extraordinary evidence, but you have to take into account the possibility that any real "powers" are quite weak. Performing more trials of course would solve this but you have to decide the number in advance which creates the problem of someone acurately predicting how well the claimant can perform. Thus there is the possibility that someone will overestimate. Does that satisfy Dr Adequate?
 
"one in a million chance" is just a saying, something whoever wrote that blurb put in. I'll take it out as soon as we update that section. Hell, people who win Lotto beat "one in a million chance"
 
You must log in or register to reply here.

Back
Top Bottom