Psychic (Sylvia Browne) Who Said Amanda Berry Was Dead Silent After Berry Is Found Alive

[William Seger]

255. Read it again: It wan't intended as a rebuttal

Why should I rebut something after explaining why it's irrelevant? I see that you don't care to address why that's so, but the "issues raised in the link" can certainly be rebutted (at the risk of giving you an excuse to again ignore my point):

The Challenge begins with a red herring:

Randi boasts that the protocols of each test must be “mutually agreed upon.” But the only terms he agrees to insist that applicants obtain results beyond what would be demanded to determine scientific significance.

No, that's not a "red herring," but what Volk is trying to imply is that Randi will always insist on results that are impossible to produce. But just think about that for a minute: If we're talking about a phenomenon that's real, what's the reason that convincing results cannot be produced? The Challenge is for the claimant to provide convincing evidence of a paranormal claim in a series of only two tests, so yes, the terms should be tough to meet, to rule out chance and/or systematic procedural flaws. Otherwise, the results aren't convincing. If you want to win a million dollars by getting lucky, play the lottery. Actually, Volk appears to be trying to pull a fast one himself by confounding "statistical significance" with "scientific significance." In a single test, a result with 95% confidence of not being the result of chance -- i.e. odds of 1 in 20 of a chance result -- are sometimes (but not always) considered to be "statistically significant." However, such a result would never be considered "scientifically significant" because it simply doesn't rule out a chance result or a systematic flaw in the test. Such results would only be considered "scientifically significant" if they were repeatable to a degree that provides a far higher confidence level. And in fact, in recent years there has been quite a bit of controversy about scientists putting too much faith in statistical analysis, because statistical inferencing requires subtle assumptions that are difficult or impossible to verify.

The preliminary test, which must be passed before an applicant can try for the million, demands odds against chance of 1,000 to 1. The second test, to win the million, requires the applicant to show results at better than a million to one against chance.

Actually, there aren't any predefined "demands" for the odds in a test, but I believe odds of 1000 to 1 are generally used as a rule of thumb. Why is that a problem if the phenomenon is real? But anyway, if a particular test did have odds of 1000 to 1, then passing the same test again would be odds of a million to 1. However, since no one has ever passed a preliminary test, the odds for the second test have never mattered, so Volk's "million to one" point is moot.

The result is that an applicant can—and did—achieve statistically significant positive results, yet was deemed to “fail” the challenge

And again, a "statistically significant" result in a single test (which was 50 to 1 in his "and did" link) is not nearly the convincing evidence rightly required for the Challenge, for good reason.

“In the ganzfeld telepathy test the meta-analytic hit rate with unselected subjects is 32% where chance expectation is 25%. If that 32% hit rate is the ‘real’ telepathy effect, then for us to have a 99% chance of getting a significant effect at p < 0.005, we would need to run 989 trials. One ganzfeld session lasts about 1.5 hours, or about 1,483 total hours. Previous experiments show that it is not advisable to run more than one session per day. So we have to potentially recruit 989 x 2 people to participate, an experimenter who will spend 4+ years running these people day in and day out, and at the end we’ll end up with p < 0.005. Randi will say those results aren’t good enough, because you could get such a result by chance 5 in 1,000 times. Thus, he will require odds against chance of at least a million to 1 to pay out $1 million, and then the amount of time and money it would take to get a significant result would be far in excess of $1 million.”

What a shame that such an elusive phenomenon is unsuitable for the Challenge, but that simply allows me to redirect back to my point: Forget about Randi and the Challenge, then. What's the reason that Sheldrake can't convince other scientists that he's on to something? The reason is right there in the excuse for Randi-bashing: The claimed phenomenon is apparenlty so weak that it's virtually impossible to prove that it actually exists.

<Edit to add a good link that rebuts Volk's article>