Psience Quest

(2019-01-21, 11:59 AM)Laird Wrote: [ -> ]Can you think of any other than experimenter fraud/deception?

No.

(2019-01-21, 11:59 AM)Laird Wrote: [ -> ]Yes, which somewhat echoes malf's request a page or more back for participants in this thread to state their preferred hypothesis. It's a good question/request, though not a straightforward one - I think it invites discussion and further questions rather than firm conclusions, and I hope, after reviewing my notes, to be able to contribute some thoughts of my own.

My answer to that question from malf last October was:
"I suspend judgment on what they've measured/not measured. Peter Bancel has made some strong arguments in favour of experimenter psi rather than global consciousness, but he had previously made what appeared to be equally strong arguments pointing in the other direction. Perhaps it's not cut and dried."
https://psiencequest.net/forums/thread-t...8#pid21748

To clarify, when I say I can't think of any conventional explanation other than fraud/deception, I mean that while in principle there could be some kind of subtle error or artefact hidden in the statistical analysis, I can't think of a mechanism by which it could produce the observed results. Particularly bearing in mind the nature of the statistical measure being used (which reflects second-by-second correlations between different random number generators), the XOR processing and the fact that the different random number generators are not accurately synchronised, and the absence of a significant deviation from chance expectation in the control data.

(2019-01-21, 08:29 PM)Max_B Wrote: [ -> ]Here's a good study... the results are solid... and quite shocking... why don't people put some of the effort they spend dissecting the GCP into something like this...



Can I get any traction on here with this stuff... can I nelly...

Does this post need a link? And possibly its own thread?

(2019-01-21, 12:23 PM)Laird Wrote: [ -> ]As for how these are supposed to account for the results is left totally unspecified by you (and anybody else, including malf), which indeed does leave us in "the state we frequently find ourselves in": opponents making vague claims which they refuse to substantiate.

Bancel clearly thought the "experiment" could be tightened up and made some suggestions. They seem common sense and good practice, do you agree with them? They aren't vague in any way (much less vague than any hypothesis or conclusion associated with this project). 

We can't know for sure the finer details of procedure in any study let alone an open ended data collection exercise like this. I'm reminded of the Hennacy-Powell fiasco. Despite that being more recognisable as a piece of science, were it not for the video that appeared, some skeptics somewhere on some forum would be, to this day, getting it in the neck for not accepting the results and conclusion as presented in the paper.

(2019-01-22, 03:09 AM)malf Wrote: [ -> ]Bancel clearly thought the "experiment" could be tightened up and made some suggestions.

Why the scare-quotes around "experiment"?

Yes, he made some suggestions in the event there was a "future version of the experiment". What he didn't say is that the current lack of realisation of his suggestions could account for the results of the experiment as it is. If you think that the results can be explained via "flexible selection", then please make that case in detail. So far, you haven't seemed to be capable of doing that.

A quick (and very related) question for you: what did you understand the point of my magic show analogy to be?

(2019-01-22, 03:09 AM)malf Wrote: [ -> ]They seem common sense and good practice, do you agree with them?

They do seem sensible, although I find it hard to imagine what an algorithm for event selection would look like.

(2019-01-22, 03:22 AM)Laird Wrote: [ -> ]They do seem sensible, although I find it hard to imagine what an algorithm for event selection would look like.

There's an old Skeptiko podcast with Chris French in which he expressed an interest in doing some work on that, or getting someone to do some work on it. As far as I know, nothing came of it, though. He did say he was very busy.
https://skeptiko.com/83-chris-french-psi-claims/

(2019-01-22, 08:54 AM)Chris Wrote: [ -> ]There's an old Skeptiko podcast with Chris French in which he expressed an interest in doing some work on that, or getting someone to do some work on it. As far as I know, nothing came of it, though. He did say he was very busy.
https://skeptiko.com/83-chris-french-psi-claims/

Interesting, although I didn't see anything specific to developing algorithms in that transcript. That seems to me to be tricky: embedding the qualitative aspect of event selection - that which, from a human perspective, defines a "globally significant event" - into an essentially quantitative process, which is (wouldn't you agree?) what characterises an algorithm.

(2019-01-22, 10:16 AM)Laird Wrote: [ -> ]Interesting, although I didn't see anything specific to developing algorithms in that transcript. That seems to me to be tricky: embedding the qualitative aspect of event selection - that which, from a human perspective, defines a "globally significant event" - into an essentially quantitative process, which is (wouldn't you agree?) what characterises an algorithm.

I think it would be quite a difficult problem to design a really objective algorithm.

On the other hand I don't think it would be too hard to come up with a set of rules that left only relatively small scope for subjectivity. Of course, the people applying the rules would still have to be blind to the data. In those circumstances, if the original results were something like a psi selection effect (or a fraudulent selection effect), the effect size would presumably plummet, probably to something undetectable.

Unless the formulation of the rules was itself influenced by psi, of course ...

(2019-01-22, 03:22 AM)Laird Wrote: [ -> ]A quick (and very related) question for you: what did you understand the point of my magic show analogy to be?

I didn’t think your analogy worked, but I’ll grant you that it’s not easy to find a perfect one.  How about this: There’s rabbits all over the place and someone is throwing top hats over some of them. It turns out the rabbits covered are special ones. But what makes them special is related to slightly louder (or softer) squawking by chickens on a bunch of distant, arbitrary chicken farms at the moment a special rabbit was covered by the top hat.

Quote:They do seem sensible, although I find it hard to imagine what an algorithm for event selection would look like.

Well I find it hard to imagine how they’ve been doing it (so fortuitously) for the last couple of decades.

(2019-01-22, 05:21 PM)malf Wrote: [ -> ]How about this: There’s rabbits all over the place

But that avoids the key point which my analogy makes plain: that there shouldn't be any rabbits in these top hats in the first place! That is where the real "magic" is; not in the selection process (though there might be something anomalous going on there too, as Peter Bancel argues). I was hoping you'd get it but either you don't or you're pretending not to.