At 07:29 PM 5/23/2005, you wrote:
I think I can answer to the whole message by saying "no way" isn't always "the way". The EPR paradox was supposed to prove quantum theory was wrong because it supposedly violated relativity. Alain Aspect proved that EPR actually worked as advertised, however it does so without violating relativity. Likewise I think there are ways that information, and perhaps other things, may be able to tunnel between worlds, despite the decoherence problem, of which I am well aware. Besides, Plaga has an experiment that is waiting to be tried that would prove other universes - http://arxiv.org/abs/quant-ph/9510007 . Time will tell, but I think history is on my side.
I remember Plaga's original post on the Los Alamos archives way back when the server there was a 386. Most of the methods I've seen--Plaga's, Fred Alan Wolf's, and others involve tweaking the mortar, so to speak---prying apart the wallboard to obtain evidence of the next room over.
Since all I'm interested in is whether behavior systems incorporate knowledge of clearly defined probabilities that may exist in the next lane over (so to speak)--I would like to make a modest proposal---
Assemble a hundred college students (a hundred will return a respectable Z score) in a double-blind experiment to determine their awareness of occult but clearly defined probabilities.
Here's how: set up a random number generator that will return a value on a screen--say 1 through 50 (or whatever object set you'd like). Tell the students it's a random number generator that will return a perfectly random result, and you'd like to see how good they are at guessing a value just before it appears. Pay the student a nominal sum each time she gets the value correct. Debit the student a small amount each time she gets it incorrect--so they'll have something invested in the outcome.
There's always a catch and here's this one: the values aren't really random, but are chosen (double-blind) to result in TWO randomly-chosen sets. These sets are transferred to a disc and placed in the "RNG" which then randomly picks which set to show--and which to keep in a state of unrealized probability. Of course, the researcher won't know either--until after the fact.
The experiment begins. One set of values gets shown to the student (immediately after they "guess" at the value). The other set remains as an unrealized probability.
If the student do not probe "probability space" then the number of "guessed" values from the unrealized set should not be significant. On the other hand, if the students guess by actually probing nearby probabilities (i.e. the next lane over), then the number of guessed values in the unrealized set should be significant. Given the nature of this experiment, I'd support a minimum z of 1.96 as a criteria---p<.05. And no meta-analysis allowed.
It seems to be a relatively easy experiment to try--RNG software is available (though some algorithms, I hear, are not as random as they should be.)