The Global Consciousness Project

well I’ve asked you several questions and had no replies!

Seriously, malf. You really can't expect people to go to the trouble of answering your questions when you waste so much of their time by behaving as you have today.

Seriously, malf. You really can't expect people to go to the trouble of answering your questions when you waste so much of their time by behaving as you have today.

Part of the problem here is when I said ‘event selection’ you thought I meant ‘timing selection’ or ‘duration selection’ or something. In fact I meant ‘event selection’...

My (ignored) questions:

1. What do you think is going on? Bancel’s ‘experimenter psi’, Radin’s ‘love’ affecting distant arbitrary electronic devices, or some sort of ‘p-hacking’?

2. Does anyone know if the protocols and hypothesis analysis have been through any sort of peer review?


3. Who chooses the ‘event’?

Only if you’ve enough time, obv.

malf

Just to be clear - I am not going to respond to your posts from now on. I said it last October, and I was a fool to go back on it.

malf

Just to be clear - I am not going to respond to your posts from now on. I said it last October, and I was a fool to go back on it.

Fair enough.

Jeffrey Mishlove has a one-hour interview with Roger Nelson about the Global Consciousness Project in his "New Thinking Allowed" series:

In case anyone is interested, the 1989 paper by Dean Radin mentioned by Mishlove, in which he claimed to have trained a neural network to recognise the "signatures" of difference experimental subjects - using data from random number generator studies done in the PEAR lab -  can be found here:
http://deanradin.com/articles/1989%20neu...rk%201.pdf

[Laird]

Ian and Laird had already decided to defray their past expenses with the first donations, I thought that was an unwise decision for themselves, as that was their original stake in the site.

when I later saw Malf and Linda’s contributions, I realised Laird might feel a little constrained by the donations.

clearly something had motivated such a long post.

Have responded to these three quotes in the "Donations are now possible" thread.

[Laird]

Linda, I don't think there's value for us in further arguing back-and-forth, so I'll leave it at that.

In terms of competing hypotheses, one must be that, once presented with colossal amounts of noisy data, mathematics can always produce some statistical significance. 

If Bancel’s comments can be trusted we know the one thing the GCP is not measuring is GC.

It might be helpful if others suggested their favoured hypotheses in the next few posts?

I've been thinking about this in relation to Gelman's and Loken's "Garden of Forking Paths".
http://www.stat.columbia.edu/~gelman/res...acking.pdf

They refer to 4 different scenarios:
1. There is no flexibility.
2. There is flexibility, but no use can be made of it because of pre-registration.
3. There is flexibility and use can be made of it, but there is no fishing.
4. There is flexibility and there is fishing.

They focus on #3 reported as though it is #2. The assumption is that there will be an inflation of falsely significant results under #3, but not under #2. This parallels the claim made by Nelson, et. al.

This leads to two questions with respect to the GCP:
Are they acting under #2 vs. #3?
Is there no difference in the number of falsely significant results under #2 vs. #1?

Bancel's and other analysis shows us that there is a difference between the results you would obtain under #1 and the results of the GCP. That is, there is a difference in the results obtained when flexibility is present and when it is absent. For example, while there was flexibility at the beginning of the process in how the New Year's Eve data would be blocked and analyzed, for a number of years there has been no flexibility, as it is necessarily analyzed the same way each time. And the cumulative results are definitively non-significant (http://noosphere.princeton.edu/events/newyear.2015.html). Bancel (https://www.researchgate.net/publication...xploration) also showed that the choices made in the presence of flexibility, but without the purported ability to use that flexibility (i.e. "formal hypotheses"), among choices of which identical events to include, test statistics, and blocking intervals gave a different result than that expected if flexibility truly could not be used. And he also demonstrated a correlation between the measured correlations and the timestamp errors, whereas there should be no correlation in the setting of random error unless there is a process which selects for fortuitous timestamp errors.

So either they are really acting under #3, or there is an interesting phenomenon where there are differences between #2 vs. #1. Probably most non-proponents just assume that they are acting under #3, given that there isn't anything to prevent it. As you mention in a later post, there are ways to tighten this up. It would be interesting to see what happens if they do that. If the effect remains, it still won't be Global Consciousness, but it would seem to be anomalous in some way.

Linda

I've been thinking about this in relation to Gelman's and Loken's "Garden of Forking Paths".
http://www.stat.columbia.edu/~gelman/res...acking.pdf

They refer to 4 different scenarios:
1. There is no flexibility.
2. There is flexibility, but no use can be made of it because of pre-registration.
3. There is flexibility and use can be made of it, but there is no fishing.
4. There is flexibility and there is fishing.

They focus on #3 reported as though it is #2. The assumption is that there will be an inflation of falsely significant results under #3, but not under #2. This parallels the claim made by Nelson, et. al.

This leads to two questions with respect to the GCP:
Are they acting under #2 vs. #3?
Is there no difference in the number of falsely significant results under #2 vs. #1?

Bancel's and other analysis shows us that there is a difference between the results you would obtain under #1 and the results of the GCP. That is, there is a difference in the results obtained when flexibility is present and when it is absent. For example, while there was flexibility at the beginning of the process in how the New Year's Eve data would be blocked and analyzed, for a number of years there has been no flexibility, as it is necessarily analyzed the same way each time. And the cumulative results are definitively non-significant (http://noosphere.princeton.edu/events/newyear.2015.html). Bancel also showed that the choices made in the presence of flexibility, but without the purported ability to use that flexibility (i.e. "formal hypotheses"), among choices of which identical events to include, test statistics, and blocking intervals gave a different result than that expected if flexibility truly could not be used. And he also demonstrated a correlation between the measured correlations and the timestamp errors, whereas there should be no correlation in the setting of random error unless there is a process which selects for fortuitous timestamp errors.

So either they are really acting under #3, or there is an interesting phenomenon where there are differences between #2 vs. #1. Probably most non-proponents just assume that they are acting under #3, given that there isn't anything to prevent it. As you mention in a later post, there are ways to tighten this up. It would be interesting to see what happens if they do that. If the effect remains, it still won't be Global Consciousness, but it would seem to be anomalous in some way.

Linda

What is the appetite for tightening things up? Given that Bancel suggested improvements in his 2014 paper, have any been implemented?

https://www.researchgate.net/publication...SS_PROJECT

A strong criticism  of the GCP is its reliance on an open-ended  protocol for deciding event parameters and this should be replaced with an algorithmic procedure in any future version of the experiment

In  practical  terms,  perhaps  the  most important consequence  of  the  analyses  is that  the  GCP effect may indeed  be  subject  to  signal-to-noise averaging.  If  this  is  so,  the  effect  can  be  studied  with  far  greater  statistical  power  by  increasing  the number of nodes in the network. A ten-fold increase in the number of RNGs would allow a full replication within 2 to 4 years. Augmenting the network 100-fold would allow  for the detection of single events in real  time.  The  detail  and  power  provided  by   vastly   increased  data  rates  would  also permit  the development of analyses to test models... 

I'm not sure whether there's any realistic possibility of getting the discussion back on track, but perhaps it's worth reposting the description of the project from my initial post.

It began as a kind of sequel to the microPK experiments conducted at the PEAR lab at Princeton. It consists of a worldwide network of several dozen random number generators. Essentially the idea behind it was that at the time of significant events - typically, events that engaged the attention of the whole world - the random number generators would exhibit unusual behaviour. Different measures of unusual behaviour were used at different times, but the commonest signified that the numbers produced by the different generators would tend to correlate with one another.

The network still exists, and continues to generate numbers. It has a Facebook page, where the latest post examines its response to Hurricane Irma:
https://www.facebook.com/EGGproject/
 
But for evidential purposes, the significant data are those produced by the "Registry of Formal Hypotheses and Specifications". According to the organisers of the project, for each of a sequence of 513 events in the period 1998-2015, a statistical hypothesis was specified before the data were examined, and was then tested. In subsequent analysis about a dozen of these events were excluded because the hypotheses were poorly defined, or not defined before any of the data were seen, but for the 500 classified as "rigorously defined", the cumulative Z value was 7.31, corresponding to a p value of 1.333 x 10^-13.
http://global-mind.org/results.html

As far as I'm aware, that result remains totally unexplained by sceptics. The hypotheses were stated to be pre-specified - that is, specified before the data were examined. The specification wasn't just a vague hypothesis - it was a specific statistical test that would yield a definite Z value for the event. And it was stated that all the pre-specified events would be included, so there would be no "publication bias" in the results.

Perhaps it's also worth reminding people that Peter Bancel's interpretation of what was going on changed quite radically over time. His most recent conclusions - which were published after the conclusion of the "Formal Hypothesis" series - can be read here:
https://www.researchgate.net/publication...xploration
And the accompanying reference material is here:
https://www.researchgate.net/publication...al_details