By George, I think he’s got it!
I wrote a long post about statistics and how a good test can be misapplied, but you seem to have the gist of it right there.
The EGG guys don’t even have a testable hypothesis yet, let alone an experiment to give significant results.
This is correct almost all of the time. As a brief counterexample, we can imagine an experiment intended to disprove a hypothesis that proves nothing if the experiment is a failure. An astrophysicist looking for the effect of gravitational waves caused by gravitons in general relativity would be a good example of this. If this person’s hypothesis is “General relativity is false”, then a failure on his part does not disprove his hypothesis, as there exist other ways thta general relativity could concievably be falsified (through the use of technology we do not have yet, for example).
In a general sense, though, you’re absolutely correct on this. Any testable hypothesis must be falsifiable for it to have any scientific meaning. An example of an untestable hypothesis was used in Carl Sagan’s “The Demon-Haunted World”, where he postulates that our entire universe exists as an analogue to a subatomic particle in another, larger universe. Such a suggestion, while interesting to contemplate, is unfalsifiable, and therefore, outside of the domain of scientific research.
You know, if I had a dime for every time I’ve used this line in a bar…
You’ve got to let us know if he writes back. Somehow, I’m not holding my breath.
This whole thing smacks of Bible Code to me, but not for the reasons one might think. Remember the Bible code? First you had group of Israeli mathematicians who demonstrated a series of what they called highly unlikely coincidences in the Hebrew text of Genesis. Then you had a journalist take their work and turn it into–surprise, surprise–a black book that predicts the future. Before anyone had time to evaluate the claims of the scientists, their work was lumped in with the journalist’s active imagination. All they (the scientists) were really saying was, ‘Wow, look at this. An interesting statistical puzzle.’ Maybe a statistical quirk. Something about math or very large numbers we don’t understand or that they neglected to consider. Maybe they were fooling themselves. But when three guys with doctorates in math fool themselves on such a scale, no matter what you think of it, it’s just plain weird.
But did we ever even find out conclusively what is was they were talking about in their original paper? (I know experiments were conducted with other novels and such). Regardless, it seems like there was never a chance to really consider it for what it was because of the hype surrounding the Bible code.
Same thing here: a group of scientists (sorry, mock them if you want, but they do work at Princeton and not Dunkin’ Donuts) saying, “Wow, look at this. We found something weird.” They don’t even really know what to think about it, yet. But we don’t hear about their work until some journalist needs a good story, and by the time he/she’s done with it, we have cell phones that can predict lottery numbers. That’s when it shows up in GQ, and before long, we have another Uri Geller / James Randi Celebrity Deathmatch scheduled.
I think the idea here is interesting because, it seems to me, it would not be hard to replicate this experiment, given a computer with a hardware RNG and some knowledge in programming and math. It’s not like we’re trying to slow down the speed of light or build a particle accelerator in the basement. RNG’s are practically plug-n-play in comparison. You shouldn’t be able to affect its output by just thinking about it, just like I can’t affect my pseudo-random NG by thinking about it. Not even a little tiny itsy bitsy bit. So we know there’s a set of results which is consistent with the theory being false–my pseudo-random NG can demonstrate this every day. It seems like the only question would be, given all the parameters, what is the threshold of success here? At what point could we say the data are no longer falsifying the hypothesis? And then at what point could we say this “shouldn’t” be happening?
I see your point, but in this case the scientists are being quite clear about what they want to believe this is. They’re not saying how, but they ARE saying they’re measuring global consciousness and perhaps predicting the future. They may have left themselves some wiggle room, but I don’t think the journalists are inserting things here.
What’s the hypothesis in the first place?
From the link in the OP:
My emphasis. “stabbing in the dark” != “quite clear about what they want to believe.”
The article draws from many sources, so a consistent tone from all of them would be surprising, don’t you think? The whole thing reeks of speculation.
I suppose stated affirmatively, it would be something like “operator intention can ‘significantly bias’ the output of a ‘true’ random number source”, however you want to define “significantly bias” and “true”
Here’s the way I’d form it, since it should be much easier to prove this way: “You shouldn’t be able to affect a true RNG by thinking about it, not even a little.”
What I see here are some hardcore skeptics trying to sterilize an environment of correlation.
This is their trademarked technique. They are Randimites. Nothing has meaning to these people, they’re approach is linear and out of date.
I don’t really think you have to be a “hardcore skeptic” at all to question this study. Just an average, healthy skepticism will do. For me, the most important question is whether the results of random number generators being skewed by thought can be scientifically reproduced. There are studies sited in one of the articles linked, and I’d like to know more about them before I come to any final conclusions. My gut feeling though is that such a feat is not possible in a statistically significant way. I would love to be proven wrong. I would love to think the world is not as random as we might think and that the power of thought can effect the physical universe. But, thus far, all the evidence points otherwise.
He has a doctorate in experimental cognitive psychology, and is technically retired.
I still maintain that this is the first thing that needs to be addressed. Are the numbers random? Because if not, then the whole hulabaloo is just that. I’ve been a programmer for going on ten years, and even done some low level crypto work. Random numbers are (as I said) very very hard to generate.
What magnifies the problem in this case is the size of the dataset. Let’s say those little black boxes are actually about 99.9999% random…but one bit in every 1,000,000 or even 10,000,000 is biased. Well, as the dataset grows, those seemingly rare bits are compounded and the effect magnified.
We shall see. He probably really is busy right now.
Applause
L. Ron Hubbard always used to talk about how easy it would be to start a religion when he was writing his science fiction novels. I wonder what he ever made of…
oh?
He did?
No kidding!
:smack:
Yes, you did a great job. This is exactly why the “Prophesies of Nostradamous” have never foretold anything- but dudes can have a lot of fun about interpreting them after the fact. :dubious:
1st you predict, then you examine the data. Not the other way around.
Umm, yes.
Granted, I’d rate the JSE as entirely speculative - though I’d regard that as no bad thing in and of itself, so long as one realises that that’s it market. Foundations of Physics is somewhat more mainstream, even if I wouldn’t want to be publishing there myself.
Sterlize = clean up improper methodology? Sounds about right.
I get paid to teach people math, including statistics. Pay me my hourly fee and I’ll gladly teach you everything you need to know about understanding statistical reasoning. Including why you can find a way to correlate things to look favorable doesn’t mean that the results actually are favorable. Until then, here’s my explanation.
Protocol before experiment = GOOD.
Experiment before protocol = BAD.
And finally, Correlation != Causation.
And, oh, apparently the scientific method is linear. That is sort of the definition of a “method”, a set of steps you follow to get from the beginning to the end.
Outdated? I missed the memo on that one. What was the expiration date on the scientific method? “Freshness guaranteed until 2004?”
If the scientific method is outdated and needs replacing, what method would you replace it with? The method that came before (praying about things, making wild guesses, and blaming magical entities for things we didn’t understand) did O.K. for a while, but it’s my opinion that the computer which I’m using to type this reply to you wouldn’t have come about if we still used that method for solving universal mysteries. But I’m reasonable, the scientific method is rather old and perhaps its time has passed. Please show us unelightened masses a better way.
I saw on his own homepage that this was his field, but did not see a mention of a doctoral degree, thus I didn’t include a “Dr.”. I saw that he considered his main job as being the head of the GCF, so I did not find “Professor” appropriate. From this point on, I will refer to him as Dr. Nelson, as is appropriate. My own ignorance has been fought today.
Notice how I predicted, among other things, the sudden hospitalization of one of the best known celebrities in the world, Michael Jackson!
I missed the earthquake in Japan by a few hours, although it was Tuesday here in Massachusetts when it happened. Maybe that counts?
Amazing! Getting any signals about tomorrow, Vern?!?
What I find interesting is that there are some folks in this thread who don’t bother to ask themselves why Nature and Science aren’t all over this. I’m not necessarily talking about you.
It does seem that they have mostly confined themselves to speculative and friendly journals and have not exposed their methodology to critical peer review (or if they have, it has not passed that peer review). Merely having publications (even in Nature and Science) demonstrates very little. The overlap between the society that publishes JSE and the people doing the research that Ferret Herder pointed out makes it very easy for them to publish.
Self-promotion is totally expected in science, which is why others need to see your methodology and data, hence the need for peer review. It’s not unforgivable. What is very hard to forgive is credulousness. Even harder to forgive is using statistical methods you don’t understand to “prove” something, which is what I think these folks are doing.
Never said it was, but even major universities have their share of pseudoscientists.
Although an internet messageboard in general enables you to make blind claims such as this, this forum does not. I am going to link you to a few formal publications Dr. Nelson has written his name on which are freely available and I expect a thorough presentation exemplifying the point that he does not understand statistical methods.
[ul]
[li]R. D. NELSON, G. J. BRADISH, R. G. JAHN, AND B. J. DUNNE (1994). A Linear Pendulum Experiment: Effects of Operator Intention on Damping Rate.[/li][li]RG Jahn, BJ Dunne, RD Nelson, YH Dobyns (1997). Correlations of Random Binary Sequences with Pre-Stated Operator Intention: A Review of a 12-year Program[/li][li]RD Nelson (2001). Correlation of global events with REG data: An Internet-based, nonlocal anomalies experiment.[/li][li]RD Nelson (1999). The Physical Basis of Intentional Healing Systems[/li][li]RD Nelson (2002). Coherent Consciousness and Reduced Randomness: Correlations on September 11, 2001[/li][li]RD Nelson. The Global Consciousness Project: Is there a Noosphere?[/li][li]RGJBJ DUNNE, RD NELSON. Engineering Anomalies Research[/li][li]RD Nelson (2002). CORRELATIONS OF CONTINUOUS RANDOM DATA WITH MAJOR WORLD EVENTS[/li][li]D Radin, R Nelson. Meta-analysis of mind-matter interaction experiments: 1959 to 2000[/li][li]YH Dobyns, RD Nelson (1998). Empirical Evidence Against Decision Augmentation Theory[/li][li]RG Jahn, BJ Dunne, YH Dobyns, RD Nelson(2000). ArtREG: A random event experiment utilizing picture-preference feedback[/li][li] RD Nelson, RG Jahn, BJ Dunne, YH Dobyns(1998). FieldREG II: Consciousness field effects: Replications and explorations[/li][li]RD Nelson, RG Jahn, YH Dobyns, BJ Dunne (2000). Contributions to variance in REG experiments: ANOVA models and specialized subsidiary analyses[/li][/ul]
The project is interesting, open, and the participants obviously express a genuine curiousity. Its essence does not narrow down to anyones knowledge of statistics, as those involved really aren’t all that complex. It narrows down to whether or not the data means anything, and how it should be read. I don’t see a giant flashing precedent. Many of the same hardcore skeptics we see here are quick to debunk all of psychology that can’t be measured with a stick (fMRI, PET, take your pick). The radical behaviorists of yesteryear who will claim up and down that their mind does not exist because they cannot see it are the same material skeptics inhabiting this thread. If it doesn’t have a number, it doesn’t have meaning. If it does have a number, lets make every possible attempt to completely devoid it of meaning and then hold the garbage data set above our head as our objective trophy of ego-validtion. The word Hope never comes to mind. Dr. Nelson is working in retirement on something he is truly interested in. It’s his little project. He makes the data available for anyone else who might be interested, gives good quote to feed those who have the same feeling that there is more than meets the eye, and is quite impervious to the (taken from above) Randimites, who, as exhibited in this thread really aren’t all that great at backing up their own “skepticism,” resorting instead to jokes, fallacies, and disrespectful and dishonest emails which do not express the true intention of the author who has already attempted to discredit the Dr. in a public venue and has every intention of using his words against him.
Probably, Sequent did a fair job of characterising why that which seems paranormal could well in a century’s turn of time be classified as physics.
That said, there have been independant analysis of one result, which found:
[ul]
[li]EC May, SJP Spottiswoode. Global Consciousness Project: An Independent Analysis of The 11 September 2001 Events.[/li][/ul]
Then to answer the OP,
No, it’s not outside the realm of possibility, and the data set as a whole has yet to be either found invalid or valid, and there are both staunch supporters and detractors. And I think, unfortunately, that the GQ just ended.
Depending on the exact numbers, yes. If I have a group of a thousand volunteers and they each flip a coin ten times, noting down the sequence of heads and tails, I wouldn’t be surprised if one of them gets 10 heads in a row. Assuming fair coins and no trick tosses, the odds of 10 heads are 1/1024. There’s no big mystery there.
Would you regard a sequence like HTTHHTHTTH as any more unusual than HHHHHHHHHH? Nobody gets all excited about the first sequence even though it’s exactly as likely as the second one which looks more unusual to us.
Note that this doesn’t mean that any given result MUST be from random chance, just that if it’s statistically likely given the number of trials I’ll be much less inclined to suspect cheating, unfair coins, psychic powers or other possibilities.
Now if you had ten people tossing coins a thousand times, and one person got a thousand heads in a row, I’d be much more interested.
The problem with looking at data after the fact is like the golf-ball problem - when you hit a golf ball you KNOW it’s going to land on a blade of grass. The odds of it landing on a particular blade of grass BEFORE you hit the ball are pretty small, but that doesn’t mean that after the golf ball has landed on one blade you can say “The chance of the ball landing on that blade was one in a jillion, let’s look for extraordinary powers at play here”. It’s only extraordinary if you named the target first.
This sounds like more of the “mainstream scientists are repressed [me/the data] because they’re closed-minded” that you often hear from people who are making outlandish claims.
No, it doesn’t. Why the hell should it? That, and your defense of the poor, victimized retired Doctor who’s just doing what he loves, seem irrelevant to this discussion.
I feel the likelihood of something happening is extremely high.
:eek:
When we base science on feelings instead of data, we end up with scientific non-events like the cold fusion fiasco. Science, like most other tools, works best when we are able to analyze things dispassionately. I would hate to lose an important part of my brain because a neurosurgeon had a “feeling” that there was something going on there.
Many of these are directed at me. I’ll take them one at a time.
If Dr. Nelson had seen fit to list the proper honorific either on the GCP page with him listed as a contact (link) or on his personal homepage (link ) then I would have used the proper title. Anyone who spends enough time in school to earn a PhD deserves the title, and I would absolutely have used it if I had seen it. I’m not here to throw insults at the guy, what purpose would that serve? We are both interested in scientific research to further mankind’s understanding of the universe. I belive that some of his conclusions are unjustified, he’d probably feel the same about a few of mine. That’s okay, we are both adults.
I feel that my email was not dishonest in any manner. If you feel differently, please elaborate. I am in fact interested in what criteria his data would need to satisfy to give a negative result. If this information didn’t interest me, I wouldn’t have bothered to request it.
My intention is to do just as I said, and post the information here. If Dr. Nelson wishes to know beforehand where his information will be going, I’ll gladly send him a link. If he wishes to know my current opinion on the data as currently posted, I will tell him. I left out my current opinion because it was irrelevant to the question I was asking, and because I didn’t wish to send a long letter, as I’m quite sure that he’s very very busy now with all the recent publicity.
I have not in any way attempted to discredit Dr. Nelson. I have posted some criticism of his data collection methods. This is part of science. I’m willing to be shown that I am in error, that his data collection methods are sound and proper follow scientific procedure, that I’ve misjudged Dr. Nelson’s experiment and his data, and that in fact the thoughts of a few random people in a completely different area will affect the results of a random number generator.
My expectation is that I will recieve no response, and, should this be the case, I will not hold it against him or make any form of derogatory comment here. I understand that running a large project such as his requires a lot of time, and I expect that he’s currently being flooded with calls and emails since the recent publicity has come his way. I don’t hate the Dr., I wish him nothing but success in his endeavours, and I expect he has more important things to do right now than talk to a mathematics major about his methodology.
Just because my first tendency is to be skeptical of earth-shaking claims doesn’t mean I’m for headhunting those that make them. Please, calm down and separate criticism of the experiment from criticism of the experimenter (he really does sound like a nice guy who believes in and is excited by what he is doing). I’ve got more in common with the guy than you’d think; in my less skeptical days, I was a firm believer in the idea of the Akashic records. We’d probably hit it off pretty well if we bumped into each other somewhere neutral.