At this point we are probably just arguing analysis styles and semantics, but I would say that unless the test was designed at the outset to handle test the lower level and it was stated before hand that was the goal I would be uncomfortable even looking at the resulting p-value.
I would say that it is incumbent on the petitioner to bring to Randi’s attention what exactly he says he is capable of doing and with what accuracy in the negotiation phase, rather than up to Randi to do a post-hoc analysis of his data to see if a previously undefined lower success rate was indicated. Again it comes down to whether Randi has rejected proposals on the basis that proper sizing would require too many replications, rather than methodological grounds (needing to touch the spoon with both hands). If so then you have a point, but if not then why hasn’t a petitioner brought in a statistician to help them in negotiations and create a test that is good enough to show the small effect with high probability.
Depends for what purpose. In terms of paying the guy, yes. In terms of what Randi holds out to the world, it makes no difference what the other guy should have done. (There have been conflicting claims in this thread as to whether or not Randi does in fact hold out to the world that his tests debunk the phenomena - my impression is that he does.)
I’m guessing that Randi wouldn’t agree to it. Because it would be hard to replicate the astronomical odds against a random chance success that he has at the levels he tests for, if he was testing for a lower level of effectiveness.
This is old news. It has only been possible to have a sensible discussion in the BBQ Pit about anything where the MDC has been a relevant datapoint for years because of this.
Sorry about that (though frankly I think your wording was a bit confusing on that score).
In that case, your point was off-target, though. We’re discussing how Randi represents the results of his tests and the implications of his challenge. Claiming that other tests by other people support the same claim doesn’t justify that.
Whether I was off target depends on what quotes exactly you can cite to show that Randi says that **his tests **show that dowsing does not work at any level. I must say over the last few days and two threads I have lost track a little of what has been cited but I don’t think you have found any such quote.
Until you do find that quote, you don’t get to say “what we are discussing”: the bounds of discussion are far less constrained than might suit your position.
I have no doubt that he has made very doubtful statements about dowsing, along the “reindeer off the roof” style, but I am not sure he has said categorically that dowsing is totally disproven. I am even more doubtful that he has said it is totally disproven and that he knows this by his tests and by his tests alone.
Until you find quotes to show that what Randi has proclaimed has been proclaimed on the basis of his tests alone and not his knowledge of the results of testing of dowsers generally, what other tests that Randi is likely to know about remain absolutely relevant.
We’re talking about testing supernatural powers. The only people who will apply sincerely believe they have magic abilities. This makes them more likely to be a “fringe character with bizarre claims”, it’s inherent to the process.
Sure, it’d be great to have someone well known for their supposed abilities to do the test (and I think that’s what Randi was always hoping), but they know they’re frauds and would be exposed. So the only people who are willing to do it are those who believe they have powers. And those people tend to be kooky.
It’s up to the applicant to describe their ability. If you have 10 buckets, one filled with water, and say “here, try your ability and see if it will work right now/in this room/using these containers/etc” and they say “yes, it works fine, we can start the test” then they can’t go claim “well the magic juju of the fluorescent lighting threw off my ability later” or something. Testing the claim is the most logical way to handle it.
I’m not sure how the JREF would handle someone who said they can call a coin flip 52% of the time. I don’t know if they’d outright reject them or design a high number of trials test for them. I suspect the issue has never came up, because if you’re going to invent yourself superpowers, they’re probably not going to be that subtle.
QUOTE=Fotheringay-Phipps;13530597]Depends for what purpose. In terms of paying the guy, yes. In terms of what Randi holds out to the world, it makes no difference what the other guy should have done. (There have been conflicting claims in this thread as to whether or not Randi does in fact hold out to the world that his tests debunk the phenomena - my impression is that he does.)
[/QUOTE]
Randi, like most skeptics, is careful when declaring a negative. Even Richard Dawkins says “god almost certainly does not exist” rather than “does not exist”.
I don’t think I’ve ever heard Randi say that any particular test completely disproves some type of magic, but that there’s a plausible alternative explanation and that it’s another piece of evidence.
Pedantically, sure, you can’t prove a negative. Practically you can come close with more and more confidence as your experimental data builds up. The million dollar challenge absolutely is a strong piece of evidence against the existance of magic, but it is not conclusive. But then neither is a hundred people checking under my bed for a pink unicorn completely conclusive that it isn’t there.
More than that, particularly as regards dowsing, they are not fringe characters with bizarre claims. In many country areas it is not doubted for a moment that dowsers can legitimately do what they say. These guys go out and tell you where to drill and you get water. They are respectable. They may be wrong, but they are not fringe characters with claims they or their community believes to be bizarre.
To digress slightly, on my father’s side I come to a signficant extent from a family of drillers. I didn’t know the great uncles involved but I understand from my father and grandfather that they (the great uncles) believed dowsing worked and dowsed to find where to drill. These weren’t kooks: they were successful working guys and businessmen. Later, they became sceptical and that developed into my grandfather’s and my father’s and my interest in scepticism.
Look at the big German study linked earlier and the Australian study in 1980 (see hereand further clarification of the stats here). These weren’t isolated studies of a few wackos. The German study was a big university study of the creme de la creme of German dowsers. The Australian one was smaller. Not only did the dowsers not achieve (in the latter) what they needed to achieve to win the prize, the results were null (the latter only after an additional study, sure, but that was the end result).
FP’s suggestion that all Randi can know is that “the claims of those tested” have not been found proven is just plain wrong. He needs to put up some strong cites showing that Randi goes further than the evidence, or shut up, frankly.
I have seen Randi ask someone to go fetch a spoon in the pantry, knock it hard against the table to show that it was perfectly solid and then suddenly it ends up being bent, so no preparation at all. I tried to not avert my eye from his hands (and I was sitting just a couple of metres away from him) but I have no idea about when in the process he actally bent it.
He probably switched it for a pre-bent one. For an experienced magician like him, it would be routine, even if you thought you were watching him carefully.
Not a good example (and forgive me if this point has been made before). Weathermen don’t make predictions. They make statements about the past, and (assuming they have accurate statistics about the past) those statements are correct. They’re just facts.
When the weatherman says “there’s a 50% chance of rain today,” what he’s really saying is “on 50% of days in the past with conditions like the conditions we’ve observed today, it rained.” And he’s probably right.
So, even if the weatherman says there’s a 99% chance of rain, and it doesn’t rain, he was right. He was still right even if this happens 100 times in a row.
I am far from being a magician but I learned one of Randi’s “psychic keybending” tricks (from one of his books) in about ten seconds and I’ve done it closeup for people. The basic setup is someone brings you a couple of keys (real unprepared keys), holds them in their own closed fist, you do a little patter about psychic vibrations, they open their hand and voila, there’s a sharply bent key.
If a rank amateur like me can do this with almost no practice right in front of people who are watching attentively imagine how good someone with Randi’s experience will be.
Valgard, which book was that in? I’ve seen closeup tricks where what’s in the person’s hand is not what they thought they had in there, and I would love to learn that trick.
Obviously you’re an untrained psychic using dowsing particles to weaken the* quantum-electro-love-field* of the key. No reason to resort to trickery to explain it when there is a perfectly adequate psuedo-scientific explanation.
That is precisely what I meant by Randi rejecting a test because proper sizing would require too many replications,. If we had complaints from petitioners saying that since their powers only work 30% of the time, they need more replicates and so the test should be 30 successes out of 150 instead of 14 out of 21 but that Randi refused, then I agree there would be a problem. But I haven’t seen any indication that Randi would refuse such a test. Again I think we mostly agree, and neither of us know for sure whether Randi would refuse such a test.
Fotheringay-Phipps, I understand the point you are trying to make, but I’m stymied as to why you think it’s really a very worthy point.
Can you give me an example of any human ability that we can typically do with a success rate slightly greater than chance? I’ve been trying to think of one, and I can’t.
So if there’s any value in arguing about Randi’s or anyone else’s tests, I can’t see it lying in the idea that perhaps people can do something at a rate just above chance.
Obviously there aren’t any known ones that involve supernatural forces, but I can throw a softball perpendicular to a bullets path and reduce the chances of it hitting a target by 0.002%
On the more general topic of disproving minor supernatural effects:
Here 's the link to FAQs about the Randi Challenge, and they say that if something is untestable then it can’t qualify for the prize.
Beyond that the claimants are required to pay for the tests expenses, and have some endorsements by legitimate local figures. So, if someone is willing to pay for the extensive testing required to (almost) prove a 2% supernatural effect, the Randi people seem entirely ready to put their money at risk.
Because if there really is something in any of these paranormal phenomena, it’s likely to be a rather weak effect. (Otherwise they would manifest themselves a whole lot more.) So for my own purposes, the entire question is not really being addressed by these tests, and people have the incorrect impression that it is.
It’s extremely common. Try this for example, or any number of similar matters that are studied and tested.