Quote:
Originally Posted by Venom  Sounds for me like saying: "be carefull, because you can have a few lucky trials, but after a while, of course, they will be a return toward the mean (or average)".
Well, that's kind of obvious. Alex should stop now that he's lucky, but with more and more trials and more and more dogs, things are going to look a lot less impressive.  Well, even now, he's just showing on YouTube his best trials, the lucky ones, but doesn't show the rest. It's kinda a filedrawer effect. He's goal is to make believers trough YouTube, not to really inform people... |
I can't dispute that it "sounds for [you] like saying" -- you are the only one who can speak to the truth of that statement. However, if this is what it sounded like to you then it is a purely a matter or you projecting your prior beliefs onto something that in no way resembles your beliefs.
What I said is that there is a mechanism that can cause a decrease in performance by any living or non-living incremental learning system over time of
any task where there is too much noise relative to the signal being discriminated (OK, that's a bit more specific than I was before). Charley presented evidence that this might be a factor in the psi decline effect.
Attributing the "decline effect" to classic "regression towards the mean" is a favorite cant of Skeptics, but that is complete nonsense, which has been clearly and unambiguously refuted time and again. One of the nonskeptical characteristics of Skeptics is that they persist in confidently asserting the truth of arguments that have been shown conclusively to be false. It is unclear in any particular case whether this is just a matter of them ignoring the facts or their being willing to make pronouncements despite being ignorant of the facts -- neither is consistent with the tenets of skepticism. Generally no effort is made to actually address the refutation.
The "decline effect" is only said to have occurred if it can be shown that the decline in question is not due to chance or "regression towards the mean." Whatever the cause of high scoring at the beginning of the tests in question it is not just a statistical fluke that naturally disappears at the end. There is, for example, no sign of an equal but opposite "incline effect" in some sets of experiments with fixed termination criteria. The decline effect, by the way, was also routinely found to occur within smaller units within experiments. So for example, it was fairly consistently found that there was more "hits" in the first quarter of a run of a deck of Zener cards than in the last quarter. Any "Skeptical" explanation of psi results must explain this and other patterns to psi hitting (scientific explanations are required to explain the actual observations), but they rarely made any attempt to.
Its also worth noting that the within subject "decline effect" has mostly disappeared in the last few decades. This is generally attributed to the whole thing having been primarily due to simple boredom. Modern tests are generally more interesting to the participants.
If I showed that some measure of sports performance (say batting average) declined over the course of a long, grueling period for the individual, recovering somewhat after brief rests, would you claim that this "sounds to you" that hitting the ball was simply a matter of chance all along? Or would you consider the possibility that it might be due to some factor such as fatigue on the part of the athlete?