Global Consciousness Project [Archive] - Page 2

T'ai Chi

12th April 2005, 10:57 AM

Originally posted by athon

I still can't figure it out. So far I feel that you are suggesting that in a trial using a RNG a result is produced that is not random. 'X' force is influencing these results.

Claus-minded generally attempts to try and make it look like I am saying a paranormal something exists.. but I'm not. I'm saying we should expect these RNGs to operate according to statistical 'laws'. We expect them to not go beyond a certain value, and we can calculate the probability of that happening. When it does happen, and when that probabiltiy is very very very very very very very very (get it?) tiny, we know that something, an anomoly, has occured.

Note, it could be an electrical shock, someone bumped the RNG, spilled soda on it, and so on. It could also be other explanations that we can't even imagine. But we should investigate those cases.

Taion

12th April 2005, 06:17 PM

Originally posted by jzs
We expect them to not go beyond a certain value, and we can calculate the probability of that happening. When it does happen, and when that probabiltiy is very very very very very very very very (get it?) tiny, we know that something, an anomoly, has occured.

How do we determine the likelyhood of this 'thing' happening? How do we know that the RNG is truely random normally?

If you say DIEHARD, I will cry. If it passes DIEHARD, it is random, but after it passes we analyze the data to see if its non-random, but if it's non-random it could not have passed DIEHARD. . .

T'ai Chi

12th April 2005, 06:39 PM

Originally posted by Taion
How do we determine the likelyhood of this 'thing' happening?

Standard probability theory.

How do we know that the RNG is truely random normally?

It passes the DIEHARD tests.

If it passes DIEHARD, it is random, but after it passes we analyze the data to see if its non-random, but if it's non-random it could not have passed DIEHARD. . .

It is always a RNG, that doesn't change. One looks at the RNG output on days specified in the formal hypothesis registry to see if the RNG output is odd on those days thought to be global events.

pmurray

12th April 2005, 07:30 PM

Originally posted by Luke T.
What the believers in these things don't realize is that I have neutralized the Planetary Meditation Grid (or GCP, if you wish) with a giant Negative Vibe Generator in my attic.

Ha! Nonsense! The NVG is simply television news. The PMG will never be big enough to counterct it. Watching TV is anti-meditation.

CFLarsen

12th April 2005, 11:45 PM

Originally posted by jzs
It passes the DIEHARD tests.

But the Orion RNGs do not pass the DIEHARD test. This is a fact, jzs.

And since you can't isolate the data from the Orion RNGs from the rest of the RNGs, you have to discard the whole dataset.

athon

13th April 2005, 01:21 AM

Originally posted by jzs

It is always a RNG, that doesn't change. One looks at the RNG output on days specified in the formal hypothesis registry to see if the RNG output is odd on those days thought to be global events. [/B]

Ok, I think I understand.

Let's assume that the results are significantly different on some days than others.

You are therefore suggesting that on the days a difference occurs, 'X' factor is at work, which is an undefined property. However, as it has been pointed out to you all along, without a second set of results on those same days which are not subect to 'X' factor, the results are meaningless. They could indeed belong to anything from a knocked machine to an electrical short. Every high school student knows you need a negative control for ANY experiment!

That's not even going into the whole 'randomness' issue.

We both agree on one thing; without a subsequent experiment with changes, this GC trial has offered us nothing.

Athon

CFLarsen

13th April 2005, 01:57 AM

Originally posted by athon
Every high school student knows you need a negative control for ANY experiment!

Indeed.

Jzs, do you acknowledge that the GCP does not have a negative control for their experiments?

If you don't, what is this negative control?

69dodge

13th April 2005, 10:50 AM

Originally posted by jzs
I'm saying we should expect these RNGs to operate according to statistical 'laws'.According to what statistical laws do you expect them to operate?

But before you answer too quickly, consider these questions: Why is the RNG output XORed with a mask containing equal numbers of 0s as 1s---shouldn't the output already contain an equal number of 0s and 1s? Why is there a need for the variance of the data subsequently to be normalized---shouldn't the variance already have the theoretically predicted value?

The reason is that the RNGs do not operate according to the statistical laws that one might naively expect them to. We expect them to not go beyond a certain value, and we can calculate the probability of that happening.If the calculation assumes that the RNGs follow statistical laws which, in reality, they do not follow, the calculated probability will be wrong.When it does happen, and when that probabiltiy is very very very very very very very very (get it?) tiny, we know that something, an anomoly, has occured.

Note, it could be an electrical shock, someone bumped the RNG, spilled soda on it, and so on. It could also be other explanations that we can't even imagine. But we should investigate those cases.One explanation that I can imagine is that the RNGs don't follow the statistical laws you expect them to, to begin with. Even in the absence of shocks, bumps, or soda.

T'ai Chi

13th April 2005, 06:13 PM

Originally posted by CFLarsen
But the Orion RNGs do not pass the DIEHARD test. This is a fact, jzs.

No, your "fact" is not a fact. It is a lie.

From the Orion site

http://www.randomnumbergenerator.nl/rng/home.html

"The Random Number Generator parallel port dongle(Mac/DOS/Win) is the first true RNG to pass Marsaglia's famous DIEHARD randomness test. "

As you well know, I emailed the contact email on that webpage to ask them about your 'critique' about the amount of bits, etc. As you well know, again, they responded back, and I posted their response, with permission, that the numbers in your 'critique' refer to an additional test they do on the RNGs prior to shipping to make sure they are functional prior to shipping.

Please, you have no excuse to be typing your lies.

T'ai Chi

13th April 2005, 06:15 PM

Originally posted by CFLarsen
Indeed.

Jzs, do you acknowledge that the GCP does not have a negative control for their experiments?

If you don't, what is this negative control?

Since athon mentioned "negative control", you should ask him. I've personally never heard those words used together like that.

T'ai Chi

13th April 2005, 06:18 PM

Originally posted by athon

You are therefore suggesting that on the days a difference occurs, 'X' factor is at work, which is an undefined property.

Well, I'm not suggesting it. That may be what the project creators are doing. Although, they seem to be exploring a correlation, not a causation.

However, as it has been pointed out to you all along, without a second set of results on those same days which are not subect to 'X' factor, the results are meaningless.

How do you know in advance what days will be "subject to X factor" or even if "X factor" exists.

T'ai Chi

13th April 2005, 06:24 PM

Originally posted by 69dodge
According to what statistical laws do you expect them to operate?

Approximately binomial, mean 100 and standard deviation ~ 7.07.

Why is the RNG output XORed

Because we are dealing with a real-life, complex electronic machine that is generating the data, that we know has a small amount of bias that can be measured and subtracted out.

---shouldn't the output already contain an equal number of 0s and 1s?

The output is ongoing.

But take a fair coin, flip it 10 times. You seem to saying if any number of heads is obtained that is not equal to 5, it is not random.

If the calculation assumes that the RNGs follow statistical laws which, in reality, they do not follow, the calculated probability will be wrong.

The probability is dependant on the bias. If the bias is large, yes, that is a real problem. Since they know the bias, can estimate it, they subtract it out for that very reason.

One explanation that I can imagine is that the RNGs don't follow the statistical laws you expect them to, to begin with.

Like Stimpy, you can imagine all the scenarios you'd like.

Walter Wayne

13th April 2005, 09:24 PM

Originally posted by jzs
Why is the RNG output XORed
Originally posted by jzs
Because we are dealing with a real-life, complex electronic machine that is generating the data, that we know has a small amount of bias that can be measured and subtracted out.You do realize that they justify the XOR mathmatically, as stimpy does?

If you don't think math applies to stochastic processes, remember that what the RNGs put out is a string of bits. Stimpy's math applies to any set of bits. Math is all that is required to justify it.

Walt

T'ai Chi

13th April 2005, 09:42 PM

Originally posted by Walter Wayne

If you don't think math applies to stochastic processes,

:rolleyes: Where did I say that??? Just show me. Please, please, PLEASE don't lie and say that I implied that. I've had just about enough of that for one day.

I am saying that if Stimpy has no measure of the bias, I could care less about his theoretical 'critique', whose conclusion of 'the analysis is meaningless' hinges on there being large bias.

Walter Wayne

13th April 2005, 10:09 PM

Originally posted by jzs

:rolleyes: Where did I say that??? Just show me. Please, please, PLEASE don't lie and say that I implied that. I've had just about enough of that for one day.

I am saying that if Stimpy has no measure of the bias, I could care less about his theoretical 'critique', whose conclusion of 'the analysis is meaningless' hinges on there being large bias. [/B] Stimpy doesn't need a measure of the bias, he has shown that their corrective methodology isn't corrective.

CFLarsen

13th April 2005, 11:36 PM

Originally posted by jzs
Please, you have no excuse to be typing your lies.

Please, have the Orion RNGs used in the PEAR experiments all been tested, or is it one example, at Orion, that was tested?

I'll answer for you: The latter.

You do understand the problem, don't you? It's like testing a prototype of a thermometer, and then producing a number of thermometers, without ever testing if just one of those thermometers are measuring temperature correctly.

You can't trust the data at GCP, jzs. It's that simple.

CFLarsen

13th April 2005, 11:38 PM

Originally posted by jzs
Since athon mentioned "negative control", you should ask him. I've personally never heard those words used together like that.

But you understand perfectly what they mean. What does GCP do to prevent electrical shocks, someone bumping the RNG, spilling soda on it, and so on?

DionysianSmile

14th April 2005, 12:29 AM

[QUOTE]Originally posted by jzs
[B]Claus-minded generally attempts to try and make it look like I am saying a paranormal something exists.. but I'm not. I'm saying we should expect these RNGs to operate according to statistical 'laws'. We expect them to not go beyond a certain value, and we can calculate the probability of that happening. When it does happen, and when that probabiltiy is very very very very very very very very (get it?) tiny, we know that something, an anomoly, has occured.

I'm not sure I know what is meant by "expect these RNGs to operate according to statistical laws". I thought a random number generator was an algorthim that's randomly seeded to produce a value. When a computer clock is used then if you can figure it's synced to the clock, it would not be hard to break down the algorithm. This device uses the parallel port so it's hardware removed from the clock and can use anything it pleases to seed? Am I read this stuff improperly? Isn't that why you'd want to use a rng say in casino so no one could use a computer to break down the algorithm. Is GCP normalize or make the RNG predictable? Like is it running the DIEBOLD (like the company?) test, passing nicely, and then suddenly starts to fail to generate random numbers? Or you are able to predict numbers numbers for a bit?
If the company has something used in gaming machines you would think they would not want to sell to psi researchers for fear of casinos getting smart assed GCPers in the casino making the slot machines bring 3 cherries just because they chinced out on ORION RNGs. I guess it's like an arms dealer selling to both sides.

athon

14th April 2005, 01:06 AM

Originally posted by jzs

Well, I'm not suggesting it. That may be what the project creators are doing. Although, they seem to be exploring a correlation, not a causation.

Huh? It's statements like these that just demonstrate how little of the scientific process you understand. I originally gave you the benefit of the doubt that maybe you were arguing something that had substance...now I see you don't.

They might not be testing the nature of the causation, but even a correlation assumes that a phenomena will have some sort of cause, obviously. That's the whole point of doing a correlation - to explore a direction in which a cause may lie.

They are looking at a correlation between 'non randomness' and 'select days of the year'. This is not causation, because a day of the year is not a phenomena, therefore cannot 'cause' non-randomness. However, the assumption therefore must be made that if the observation is real, a phenomena must be causing the non-randomness on these days.

By supporting the validity of the data, you are supporting the fact it has a real cause, even though that cause has not been speculated on (according to you).

How do you know in advance what days will be "subject to X factor" or even if "X factor" exists.

Until you've achieved a theory on this, you cannot hope to make valid predictions. But an 'x factor' must exist for the non-randomness to occur. That's the entire point! Maybe it is just random bat-farts in the ether, or bumped equipment, or God having a cosmic joke...if it is truly non-random, something is causing it.

Again, that's not even going into whether it really is non-random or not.

As for not knowing what a negative control is...I really do have to ask if you are serious. If you are, I'll happily explain it to you.

Athon

Stimpson J. Cat

14th April 2005, 04:12 AM

jzs,

Why is the RNG output XORed
Because we are dealing with a real-life, complex electronic machine that is generating the data, that we know has a small amount of bias that can be measured and subtracted out.
But as I have already analytically proven, applying an XOR mask does not "subtract out" the bias in the mean. It instead converts it into correlations in the data. Correlations which will then bias the variance.

This has been explained to you repeatedly. If you acknowledge that there is a bias in the mean of the data coming from these RNGs, then it mathematically follows that the XORed data will also be biased, just in a different way.

If the calculation assumes that the RNGs follow statistical laws which, in reality, they do not follow, the calculated probability will be wrong.
The probability is dependant on the bias. If the bias is large, yes, that is a real problem. Since they know the bias, can estimate it, they subtract it out for that very reason.
This is simply false. They do not "subtract it out". This is the entire point of my critique!

They incorrectly believe that by applying the XOR mask they will be subtracting out the bias in the mean, and you seem to believe them, but I have analytically shown that this is not the case! The XOR operation does not subtract out the bias in the mean. It produces a new signal whose mean is not biased, but which has correlations whose strengths depend on the initial bias. This is has been mathematically shown. How can you continue to ignore this fact?

I can only assume that for some reason you do not agree that XORing a signal with a bitmask will transform biases in the mean into correlations in the output. Is this the case? Or do you disagree with the claim that correlations in the XORed data can bias the variance of the resulting sums? If you agree that both of these things are true, then how can you believe that they are correctly subtracting out the bias of the mean?

Dr. Stupid

davidsmith73

14th April 2005, 04:52 AM

Originally posted by Stimpson J. Cat

But as I have already analytically proven, applying an XOR mask does not "subtract out" the bias in the mean. It instead converts it into correlations in the data. Correlations which will then bias the variance.

They fully acknowledge this on their site:

"After XOR'ing, the mean is guaranteed over the long run to fit theoretical expectation. The trial variances remain biased, however. The biases are small (about 1 part in 10,000) and generally stable on long timescales."

davidsmith73

14th April 2005, 06:01 AM

Originally posted by CFLarsen
But the Orion RNGs do not pass the DIEHARD test. This is a fact, jzs.

And since you can't isolate the data from the Orion RNGs from the rest of the RNGs, you have to discard the whole dataset.

Well, they can do this. Look on their site. They look at the bias in variance produced by the Orion REG and Mindsong REG separately

davidsmith73

14th April 2005, 06:06 AM

Originally posted by CFLarsen
Please, have the Orion RNGs used in the PEAR experiments all been tested, or is it one example, at Orion, that was tested?

I'll answer for you: The latter.

You do understand the problem, don't you? It's like testing a prototype of a thermometer, and then producing a number of thermometers, without ever testing if just one of those thermometers are measuring temperature correctly.

You can't trust the data at GCP, jzs. It's that simple.

They analyse the output of their REG's from 1999-2005 to look at the amount of bias in variance that is produced after XORing.

CFLarsen

14th April 2005, 06:17 AM

Originally posted by davidsmith73
Well, they can do this. Look on their site. They look at the bias in variance produced by the Orion REG and Mindsong REG separately

In the list of "Current results", do they separate the Orion data from the rest?

davidsmith73

14th April 2005, 06:18 AM

Originally posted by CFLarsen
In the list of "Current results", do they separate the Orion data from the rest?

Why should they?

athon

14th April 2005, 11:13 AM

Oh, God this is frustrating! The whole 'RNG' thing is a complete red herring in light of something more significant; it's all null and void without anything to compare it to!

We can apply all the maths in the world and argue until we're blue in the face that it was or was not calibrated. It's a simulacrum in numbers as far as I'm concerned, and we're debating whether the ruler used to measure it was the right size. ICAN's ol' demons in the static springs to mind; without a negative control, we have no reason to suspect that some sequences of numbers are anything more than faces in the snow.

Athon

CFLarsen

14th April 2005, 11:29 AM

Originally posted by davidsmith73
Why should they?

Because since the Orion RNGs are not calibrated (and I would love to see how the other two kinds were!), the data would pollute the whole dataset.

So, are they?

davidsmith73

15th April 2005, 06:06 AM

Originally posted by CFLarsen
Because since the Orion RNGs are not calibrated (and I would love to see how the other two kinds were!), the data would pollute the whole dataset.

So, are they?

As far as I can see, they perform the same analyses on control data for each type of REG. If they conclude that the output of each type is not producing any bias into the results (which they do claim, even though we have reason to believe they don't demonstrate it) then they have no reason to separate the data.

69dodge

16th April 2005, 07:50 AM

Originally posted by jzs
[I expect the output of the RNGs to be a]pproximately binomial, mean 100 and standard deviation ~ 7.07.Hey, that's what I expect too. And, hey, the output is approximately binomial with those parameters. So, what's the problem? Why should I think there's anything strange going on that needs to be explained?The probability is dependant on the bias. If the bias is large, yes, that is a real problem. Since they know the bias, can estimate it, they subtract it out for that very reason.So you expect the 200-bit sums, after XORing and normalization of the variance, to be exactly binomial and not just approximately binomial? And to be perfectly independent of each other too? I don't.

If you don't either, you should not defend statistical significance tests which assume that perfection.

If the Global Consciousness Project wants to demonstrate that certain global events affect its RNGs, it needs to compare the RNGs' behavior during the global events to their behavior the rest of the time. It is not sufficient merely to predict that the RNGs will act funny during a global event, and then to observe that they do act funny then. Maybe they act funny all the time.

(edited to correct spelling of "binomial")

T'ai Chi

16th April 2005, 08:12 AM

Originally posted by 69dodge
Hey, that's what I expect too. And, hey, the output is approximately binomial with those parameters. So, what's the problem? Why should I think there's anything strange going on that needs to be explained?

If there are more times when you compare that output to events specified in the formal hypothesis registry and they are beyond what you'd expect by chance.

So you expect the 200-bit sums, after XORing and normalization of the variance, to be exactly binominal and not just approximately binomial? And to be perfectly independent of each other too? I don't.

I don't either. "exactly" is where I disagree.

If you don't either, you should not defend statistical significance tests which assume that perfection.

All models are wrong, some are useful.

Walter Wayne

17th April 2005, 12:42 AM

Originally posted by athon
Oh, God this is frustrating! The whole 'RNG' thing is a complete red herring in light of something more significant; it's all null and void without anything to compare it to! I agree the debate is pointless because the data, good or bad, shows nothing. After several pages I almost forgot the original topic, and why I dismissed it in the first place.

Walt

CFLarsen

17th April 2005, 01:03 AM

Originally posted by davidsmith73
As far as I can see, they perform the same analyses on control data for each type of REG.
Where, exactly, do you see this?

Can you, in the datasets, see which data comes from which type of RNG?

69dodge

17th April 2005, 07:34 AM

Originally posted by CFLarsen
Can you, in the datasets, see which data comes from which type of RNG?The downloadable data identifies each RNG by its ID number. The GCP website has a page (http://noosphere.princeton.edu/egghosts.html) that lists all the RNGs with their ID numbers and says what type each is, along with some other information.

CFLarsen

17th April 2005, 10:07 AM

Originally posted by 69dodge
The downloadable data identifies each RNG by its ID number. The GCP website has a page (http://noosphere.princeton.edu/egghosts.html) that lists all the RNGs with their ID numbers and says what type each is, along with some other information.

Thank you. There's a problem, though.

Take a look at the 10th EGG in the list. It is referred to as "transferred", but installed on 2003-02-03. There are several cases of EGGs that are replaced, e.g. the 53rd EGG, that was once a MINDSONG, but was changed to an ORION. They are not even sure when the 72nd and 74th EGG was installed. And so on, and so on, and so on...

Do you see what's happening here? Their analyses are worthless (gee, we say that a lot, don't we?) because they are mixing the data from different EGGs.

T'ai Chi

17th April 2005, 10:58 AM

Originally posted by CFLarsen
Do you see what's happening here?

Certainly.

Out of a list that includes 946 pieces of data, 4 of those disturb you because they used to be one type of RNG, but now they are another.

An additional 2 out of 946 disturb you because they don't know the exact day they were installed.

So 6 out of 946 pieces of data disturb you.

Although, I'm surprised. You failed to mention the cases under 'Type' where they weren't sure of the operating system being used. That would have boosted it up to a whopping 10 out of 946.

You're slipping.

CFLarsen

17th April 2005, 11:11 AM

Originally posted by jzs
Although, I'm surprised. You failed to mention the cases under 'Type' where they weren't sure of the operating system being used. That would have boosted it up to a whopping 10 out of 946.

You're slipping.

Not at all. You, OTOH, have serious reading comprehension problems:

Originally posted by CFLarsen
And so on, and so on, and so on...

Do you understand these simple words? Please explain what they mean.

T'ai Chi

17th April 2005, 11:34 AM

Originally posted by CFLarsen
Not at all. You, OTOH, have serious reading comprehension problems: Do you understand these simple words? Please explain what they mean.

Aww, the patronizing approach. It is truly amazing how I've managed to get along this far, and this well, with the "serious reading comprehension problems" you need to believe I have.

So you find about .7% of the data in this table problematic.

Um, so...?

CFLarsen

17th April 2005, 12:27 PM

Originally posted by jzs
Aww, the patronizing approach. It is truly amazing how I've managed to get along this far, and this well, with the "serious reading comprehension problems" you need to believe I have.

I was throwing you a lifeline there. So, you deliberately misunderstand what I say?

Originally posted by jzs
So you find about .7% of the data in this table problematic.

No, that's not what I said. Since you won't admit to a reading problem, what is your excuse then?

Originally posted by jzs
Um, so...?

If they mix the RNGs, then the data is tainted. Do you understand?

T'ai Chi

17th April 2005, 12:38 PM

Originally posted by CFLarsen
No, that's not what I said. Since you won't admit to a reading problem, what is your excuse then?

You have a need to believe I have a reading problem. Whatever makes you happy.

If they mix the RNGs, then the data is tainted.

So you must believe the data from the RNGs is statistically different then? If you want to present your analysis, you are welcome to do so.

CFLarsen

17th April 2005, 12:51 PM

Originally posted by jzs
You have a need to believe I have a reading problem. Whatever makes you happy.

You skirt the issue.

If you don't have a reading problem, how do you explain that you constantly get things wrong?

Originally posted by jzs
So you must believe the data from the RNGs is statistically different then? If you want to present your analysis, you are welcome to do so.

Again, you are not able to understand what people are saying. No, I am not saying that. I am saying that, since one type of RNG is demonstrably not calibrated, and we have no idea what status the other types have, the data is inherently tainted. We simply cannot trust it.

T'ai Chi

17th April 2005, 01:22 PM

Originally posted by CFLarsen

If you don't have a reading problem, how do you explain that you constantly get things wrong?

If you have a strong need to believe that I have a reading problem, that is your issue.

I am saying that, since one type of RNG is demonstrably not calibrated, and we have no idea what status the other types have, the data is inherently tainted. We simply cannot trust it.

What do you mean by calibrated, and which RNG is not calibrated, according to you?

CFLarsen

17th April 2005, 01:43 PM

Originally posted by jzs
If you have a strong need to believe that I have a reading problem, that is your issue.

No, really, this is something you should address.

This isn't just the odd case of you getting a thing wrong here and there, it is disturbingly often that you do it. It hinders real discussion, because we have to go back and ensure that you get it right, over and over again. But we don't seem to succeed.

That you cannot seem to comprehend what people are saying is a problem. You need to realize this.

If you want to participate in a debate, then you have to be able to comprehend what people are saying. If you cannot, then how can we possibly engage in a debate with you?

Can you explain that?

If you can't, then you are simply a waste of time. Your points are automatically rendered worthless, because we cannot be sure that you understand what we are saying to you. By your refusal to acknowledge this as a problem, you cut yourself off from any debate here.

So, please explain how you can possibly get so many things wrong, if you insist that you don't have a reading problem.

What can explain it?

Originally posted by jzs
What do you mean by calibrated, and which RNG is not calibrated, according to you?

Why do you ask this? What are you trying to achieve here? Prove my point about you constantly misunderstanding what people are saying?

You know we've been over this before. So, why bring it up again, as if it has never been addressed?

Your questions have been addressed in detail, ad nauseam. You know it. You pretend it hasn't.

Can you explain that?

T'ai Chi

17th April 2005, 04:05 PM

Originally posted by CFLarsen

This isn't just the odd case of you getting a thing wrong here and there, it is disturbingly often that you do it.

Feel free to present an acutal % here. Oops, you can't possibly do that, yet you continue to spew your filth.

If you have the strong need to believe I have a reading comprehension problem, you'll just have to deal with that. But don't believe for a second that others don't see right through you.

Why do you ask this? What are you trying to achieve here?

Yawn. Forget me for once, if you can, and just focus on answering the question.

CFLarsen

17th April 2005, 04:16 PM

Originally posted by jzs
Feel free to present an acutal % here. Oops, you can't possibly do that, yet you continue to spew your filth.

"Filth"? Is that how you see it, when people point out that you have a problem?

Originally posted by jzs
If you have the strong need to believe I have a reading comprehension problem, you'll just have to deal with that. But don't believe for a second that others don't see right through you.

I am not worried about that at all. Your own posts clearly show that you have a major problem. Or, of course, you have some other agenda.

Whatever the reason, the problem is yours. You cannot understand what other people are saying, yet you deny that you have a problem.

You have to deal with this. Not me, not anyone else.

Originally posted by jzs
Yawn. Forget me for once, if you can, and just focus on answering the question.

Can you explain why anyone should engage in a debate with you, if you continuously misunderstand what people say, despite your claim that you do not have a reading problem?

Why should anyone answer any questions from you, if you misinterpret what they say?

Why do you ask a question that has been answered many times before? You know what I have said about the calibration issue, so why do you imagine that you can give the impression that it has never been addressed before?

Can you explain that?

rwguinn

17th April 2005, 04:32 PM

CF--
Leave him alone. He has a problem everyone but he can see.
Do you beat your head against a brick wall because it feels so good when you stop? The brick wall is never impressed, and won't move either.

Some battles just aren't worth fighting....

T'ai Chi

17th April 2005, 06:42 PM

Originally posted by CFLarsen
"Filth"? Is that how you see it, when people point out that you have a problem?

We've already established that you have a strong need to spread your lie that I have a reading comprehension problem.

You don't need to provide more evidence of your motives, really.

Why do you ask a question that has been answered many times before? You know what I have said about the calibration issue, so why do you imagine that you can give the impression that it has never been addressed before?

More evasion..

T'ai Chi

17th April 2005, 09:52 PM

I think we scared paintedbird away. :)

CFLarsen

17th April 2005, 11:13 PM

Originally posted by jzs
We've already established that you have a strong need to spread your lie that I have a reading comprehension problem.

Is it a lie? You don't seem to grasp that I have opened up for other possibilities in my previous posts (you can scroll up if you don't remember), so I honestly don't think you can argue that I don't have a case at all.

Originally posted by jzs
You don't need to provide more evidence of your motives, really.

I am only asking you to explain what it is, if not a reading problem. Because you demonstrably misunderstand - constantly - what people are saying to you.

Originally posted by jzs
More evasion..

Not at all. But we need to clear this up, before any debate with you can continue.

Can you explain why anyone should engage in a debate with you, if you continuously misunderstand what people say, despite your claim that you do not have a reading problem?

Why should anyone answer any questions from you, if you misinterpret what they say?

Why do you ask a question that has been answered many times before? You know what I have said about the calibration issue, so why do you imagine that you can give the impression that it has never been addressed before?

davidsmith73

18th April 2005, 03:42 AM

Originally posted by CFLarsen
Where, exactly, do you see this?

Can you, in the datasets, see which data comes from which type of RNG?

Go to the "normalising data" page. There's first a plot that shows the variance bias for the Orion and Mindsong REGs separately. Each type of REG produces a different degree of variance bias which is expected. It's actually the Mindsong REG's that are going to pollute the data set, not the Orion (but I guess you knew that already Claus ;) ). They then show the variance produced for all the devices and sure enough there is a large bias which they acknowledge is produced by the Mindsong REGs. They then show the variance for all devices after their normalisation procedure, and it falls within the boundaries expected from 0.05 probability. Therefore, if they think that their normalisation procedure is sufficiently correcting for the variance bias across all devices then I don't see a reason for them to separate the data into REG type for the main data set. Do you? (all this is assuming that their normalisation procedure is appropriate, I know there are reservations about it but that doesn't matter for the purpose of this argument).

CFLarsen

18th April 2005, 05:01 AM

Originally posted by davidsmith73
Go to the "normalising data" page. There's first a plot that shows the variance bias for the Orion and Mindsong REGs separately. Each type of REG produces a different degree of variance bias which is expected. It's actually the Mindsong REG's that are going to pollute the data set, not the Orion (but I guess you knew that already Claus ;) ). They then show the variance produced for all the devices and sure enough there is a large bias which they acknowledge is produced by the Mindsong REGs. They then show the variance for all devices after their normalisation procedure, and it falls within the boundaries expected from 0.05 probability. Therefore, if they think that their normalisation procedure is sufficiently correcting for the variance bias across all devices then I don't see a reason for them to separate the data into REG type for the main data set. Do you? (all this is assuming that their normalisation procedure is appropriate, I know there are reservations about it but that doesn't matter for the purpose of this argument).

It's very simple: As long as they cannot calibrate any of the RNGs, their data is worthless.

davidsmith73

18th April 2005, 06:09 AM

Originally posted by CFLarsen
It's very simple: As long as they cannot calibrate any of the RNGs, their data is worthless.

But I thought they claim that they do?

Fair enough, Simpy has pointed out some methodological errors but why should they separate their REGs if they think their normalisation has corrected for variance bias?

CFLarsen

18th April 2005, 06:50 AM

Originally posted by davidsmith73
But I thought they claim that they do?

What they claim they do and what they actually do are two separate things.

Originally posted by davidsmith73
Fair enough, Simpy has pointed out some methodological errors but why should they separate their REGs if they think their normalisation has corrected for variance bias?

Because if they later find an error in one of the types of RNGs, then they can isolate that data and not lose the rest. Or if they find one RNG that is faulty, they can isolate it. They don't store the RNG ID with the data, so they can't do that.

As it is, they don't have any data worth anything. They have a large database filled with 0's and 1's, signifying absolutely nothing.

Zep

18th April 2005, 07:46 AM

Originally posted by CFLarsen
As it is, they don't have any data worth anything. They have a large database filled with 0's and 1's, signifying absolutely nothing. Slight correction: They have a database filled with numeric values derived by means unknown from selected parts of an unknown number of series of raw data in the form of 0's and 1's that they have not kept.

Which is worse than what you said, Claus.

T'ai Chi

18th April 2005, 08:22 AM

Originally posted by CFLarsen
Is it a lie? You don't seem to grasp that I have opened up for other possibilities in my previous posts (you can scroll up if you don't remember), so I honestly don't think you can argue that I don't have a case at all.

I am only asking you to explain what it is, if not a reading problem. Because you demonstrably misunderstand - constantly - what people are saying to you.

Like I said, we don't need more evidence of your desparate need to believe I have a reading comprehension problem.

Why do you ask a question that has been answered many times before? You know what I have said about the calibration issue, so why do you imagine that you can give the impression that it has never been addressed before?

More evasion. Still!

T'ai Chi

18th April 2005, 08:27 AM

Originally posted by CFLarsen
What they claim they do and what they actually do are two separate things.

As is what you believe they do/claim and what they actually do/claim.

Because if they later find an error in one of the types of RNGs, then they can isolate that data and not lose the rest. Or if they find one RNG that is faulty, they can isolate it. They don't store the RNG ID with the data, so they can't do that.

If you read and understand their site, you'd find they have a section on rotten eggs and removing obviouslly erroneous data.

As it is, they don't have any data worth anything. They have a large database filled with 0's and 1's, signifying absolutely nothing.

You demonstrate, again, you have not read and/or understood what they are doing. Good job! :D

CFLarsen

18th April 2005, 08:41 AM

Originally posted by Zep
Slight correction: They have a database filled with numeric values derived by means unknown from selected parts of an unknown number of series of raw data in the form of 0's and 1's that they have not kept.

Which is worse than what you said, Claus.

Quite right.

CFLarsen

18th April 2005, 08:47 AM

Originally posted by jzs
Like I said, we don't need more evidence of your desparate need to believe I have a reading comprehension problem.

And yet, I have opened up for other possibilities, yet you ignore this. If you don't have a reading comprehension problem, what else could explain this peculiar trait of yours?

Is the explanation simply that you are "petty", as you have admitted to be? Why should anyone bother with you, then?

Originally posted by jzs
More evasion. Still!

Oh, well. How petty.

T'ai Chi

18th April 2005, 09:28 AM

Originally posted by CFLarsen
And yet, I have opened up for other possibilities, yet you ignore this. If you don't have a reading comprehension problem, what else could explain this peculiar trait of yours?

You have provided enough evidence of your need to believe I have a reading comprehension problem, and of your malicious intent to defame.

When your done, wake me.

CFLarsen

18th April 2005, 09:40 AM

Originally posted by jzs
You have provided enough evidence of your need to believe I have a reading comprehension problem, and of your malicious intent to defame.

When your done, wake me.

:hb:

T'ai Chi

18th April 2005, 10:03 AM

The picture comes to mind..

Have you ever had a thread about something paranormal, etc., where you didn't resort to personal attacks?

That would be a paranormal event I'm thinking.

Scoff away.

18th April 2005, 06:41 PM

Originally posted by jzs
The picture comes to mind..

Have you ever had a thread about something paranormal, etc., where you didn't resort to personal attacks?

That would be a paranormal event I'm thinking.

Scoff away.

When will you reply substantively to Stimpy's completely telling, absolute destruction of the PEAR methodology?

69dodge

19th April 2005, 01:34 PM

Originally posted by CFLarsen
They don't store the RNG ID with the dataThey do (http://noosphere.princeton.edu/data/basket_csv.html).

CFLarsen

19th April 2005, 02:07 PM

Originally posted by 69dodge
They do (http://noosphere.princeton.edu/data/basket_csv.html).

Correct. However, they don't isolate the data from a faulty EGG. If one EGG flunks during a day, the data is not included that day, but if it doesn't the next day, they simply include it again, without ever finding out why it malfunctioned.

Would you trust an RNG to provide real random numbers, if it had malfunctioned the previous day?

T'ai Chi

19th April 2005, 05:17 PM

Originally posted by CFLarsen
Correct.

However before you said

"They don't store the RNG ID with the data"

R E A D T H E I R S I T E P L E A S E.
(before criticizing their site)

Robin

27th April 2005, 11:12 PM

Originally posted by jzs

However before you said

"They don't store the RNG ID with the data"

R E A D T H E I R S I T E P L E A S E.
(before criticizing their site) [/B]
An interesting page to read on their site is:

http://noosphere.princeton.edu/results.html

Where they list 199 events and z scores for each. 19 out of 199 are identified as statistically significant and these have a z score of greater than 1.64 or less than -1.64.

If the data was purely random the proportion you would expect to have these scores is 10%. The proportion of events that have these score is 10%.

So this system does not detect global events, much less predict them - I am not sure why they are still persisting with this.

I wonder if the rest of Princeton are happy at having this site linked to their name.

Zep

28th April 2005, 12:00 AM

Originally posted by Robin
So this system does not detect global events, much less predict them - I am not sure why they are still persisting with this.

I wonder if the rest of Princeton are happy at having this site linked to their name.

PEAR research appears to have nothing much to do with anything - the whole setup appears to be simply a sinicure for the participants. Exceedingly wealthy but silent donors with rather odd views on the paranormal contribute greatly to its upkeep (or perhaps it's a massive tax-dodge, I don't know). So should the PEAR team ever actually say they can't find anything then the "game" is finished and the money stops. So, of course, they know which side their bread is buttered on so they never actually say that, do they! ;)

T'ai Chi

28th April 2005, 04:11 AM

Originally posted by Zep
So should the PEAR team ever actually say they can't find anything then the "game" is finished and the money stops. So, of course, they know which side their bread is buttered on so they never actually say that, do they! ;)
[/my_opinion]

Zep, feel free to reveal all of this money that you believe exists. Must be some conspiracy that you have some special knowledge on.

Personally, I see more "$" signs on Skeptic.com when I go to that page than I've ever seen on on any "woo" site.

T'ai Chi

28th April 2005, 04:14 AM

Originally posted by Robin

I wonder if the rest of Princeton are happy at having this site linked to their name.

I have never read anything suggesting they are unhappy. In places of learning, things tend to get investigated, no matter if they are not the current topics of science.

CFLarsen

28th April 2005, 04:50 AM

The University of Princeton is not associated with the Global Consciousness Project:
All members of the Planning Group participate in the EGG Project as individuals, and not as representatives of their affiliated institutions.
Source (http://noosphere.princeton.edu/planning5.html)

The Planning Group consists of:

Roger Nelson, Project Director, is Research Coordinator of Princeton Engineering Anomalies Research (PEAR) at Princeton University. He serves as producer, art director, and science manager for the GCP.
Dick Bierman, Professor of Psychology (albeit a physicist), University of Amsterdam, is a pioneer of World Wide Web experiments in anomalies. He is the editor of eJAP, the electronic Journal of Anomalous Phenomena.
Greg Nelson, American Nuclear Systems, Knoxville, TN, is a computer scientist (Artificial Intelligence) with major interest in systems integration. He is chief architect for the EGG network backbone.
John Walker, near Neuchatel, Switzerland, is founder and former CEO of Autodesk, Inc., developer of the AutoCAD computer aided design software. His current focus is a Web site (http://www.fourmilab.ch/) which hosts, among other resources, The RetroPsychoKinesis Project.
Rick Berger, Innovative Software Design/Innovative Product Marketing, San Antonio, Texas, has a long history and deep interest in anomalies research. He is chiefly responsible for the logic and aesthetics of GCP's website.
George deBeaumont, is a data analyst with a major utilities firm, and has a deep interest in consciousness and its possible effects in the world. He is the creator of most of the detailed graphical displays exploring the current results.
Charles Overby, Lifebridge Foundation, New York, is a computer engineer with experience in secure network communications. He is a shaman and scholar, investigating subtle energies that may emanate from humans.
Dean Radin, Interval Research, Palo Alto, CA, author of "The Conscious Universe", is an early explorer of effects of group consciousness. He has long-standing interest in integrating multivariate data complexes.
Marilyn Schlitz is Director of Research at the Institute of Noetic Sciences. She is a medical anthropologist, with major interests in research on consciousness and the interactions of intentionality and living systems.
Stephan Schwartz, various US locations, formerly Research Director of the Mobius Society, is an entrepreneurial pioneer in applications of psi to problem solving and discovery in archaeological, medical, and other fields.
Jiri Wackermann, has three hats: Neuroscience Technology Research, Prague; KEY Institute for Brain/Mind Research, Zurich; and Psychophysiology of Consciousness Laboratory, IGPP, Freiburg, Germany. He explores complex, hyperdimensional representations of states of mind.
Source (http://noosphere.princeton.edu/planning5.html)

Only one could be said to have been on the payroll of Princeton, Roger Nelson, and he retired in 2002. (http://www.princeton.edu/~rdnelson/)

The GCP does not operate under the aegis of the University of Princeton:

It operates under the aegis of the Institute of Noetic Sciences, a 501(c)(3) research entity, which will manage the funding for the project.
Source (http://noosphere.princeton.edu/planning3.html)

Zep

28th April 2005, 07:23 AM

Originally posted by jzs
Zep, feel free to reveal all of this money that you believe exists. Must be some conspiracy that you have some special knowledge on.

Personally, I see more "$" signs on Skeptic.com when I go to that page than I've ever seen on on any "woo" site. Oopsie.So let us look finally at PEAR again, and perhaps more specifically at their listed supporters: "Institut fÃ¼r Grenzgebiete der Psychologie und Psychohygiene, Lifebridge Foundation, Richard Adams, George Ohrstrom, Laurance S. Rockefeller, and other private donors."
The Institut fÃ¼r Grenzgebiete der Psychologie und Psychohygiene (IGPP) is based in Freiburg German, and states that it was founded in order to "achieve an improved understanding of mind-matter relationships from the perspectives of the humanities, social sciences and natural sciences". It also "offers information and counseling services for people with extraordinary experiences and entertains a special library and a research archive for parapsychology and border areas of psychology."
The Lifebridge Foundation "was established in 1992 for the purpose of supporting organizations and individuals who, through cultural, education, and/or scientific means, are dedicated to creating bridges of understanding among all people by bringing to realization the concepts of one humanity and the interconnectedness of all life."
Richard Adams I have not been able to accurately identify, although a particular candidate appears likely â€“ not the author of Watership Down, either.
George Ohrstrom is a highly reclusive person whose father was one of the darlings of Wall Street in the 1920â€™s. He is the chairman of G.L. Ohrstrom & Co., is on the boards of various large industrial corporations, is reputedly worth US$140 million per annum, and apparently contributes to a variety of charitable causes. His sister, Magalen Ohrstrom Bryant, seems to be an "eccentric millionairess" who is worth even more than George.
Laurance S. Rockefeller is a contributor to the Human Potential Foundation Inc., whose goals are "Exciting research was occurring in areas of inter-species communication, consciousness, life-after-death, non-human intelligence and alternative energy sources, all of which the Foundation wanted to examine and to then aid via scientific research in whatever ways seemed possible". Laurance appears to be the driving force behind HPFâ€™s UFO Matrix of "Belief". In this document, the HPF has encouraged the United States government to reveal all that it may know about, and technology supplied by, visiting UFOs.

In summary, it is probably fair to say that the supporters of PEAR are certainly wealthy, and not only sympathetic towards PEARâ€™s aims, but they also would likely be impressed by the "scientific" nature of PEARâ€™s published output. And they would certainly seem to be ideal "targets" for a summary of analyses of 25 years worth of remote-viewing work, with a metaphysical apologia of paranormal and pseudo-scientific buzzwords on the end.

Why try to impress them in this way? Let me posit a simple answer: the money. Consider the alternative. If PEAR had indeed published a summary of 25 years of remote viewing data and allowed the null conclusion to be clearly visible then their stream of support funding from their supporters becomes in jeopardy. However if this result could be buried in gobbledegook and made up to look like research was ongoing and yielding results then the income stream is protected. The impression on the rest of the scientific world is really secondary â€“ the Hansen Utts Markwick paper is not really a major issue. The sinecure of retaining PEARâ€™s continued existence and funding, not to mention prestige, would probably count more highly.
http://www.skepticreport.com/psychics/shapesintheclouds.htm

CFLarsen

28th April 2005, 08:48 AM

Originally posted by jzs
Zep, feel free to reveal all of this money that you believe exists. Must be some conspiracy that you have some special knowledge on.

I thought you had read the article that Zep referred to?

Guess not.

T'ai Chi

28th April 2005, 03:48 PM

Originally posted by CFLarsen
The University of Princeton is not associated with the Global Consciousness Project:

Good info, but we are talking about it being linked to their name (ie. the word "Princeton" is in the name), not that they are necessarily on Princeton's payroll.

Were they on Princeton's payroll at one time?

T'ai Chi

28th April 2005, 03:52 PM

Originally posted by Zep
Oopsie.
http://www.skepticreport.com/psychics/shapesintheclouds.htm

Zep, if you were paying attention (replying to me, even though in another thread you claim to have me on ignore?, :D ), I didn't ask you for a list of their supporters. I asked you to reveal all of this money that you believe exists. Like are they getting a lot, or a little, etc.

And like I said... as an exercise for you, go count the $ signs on the front page of Skeptic.com

I'm skeptical you can find any woo page with more. That is why the dollar argument that you are touting is pretty lame.

T'ai Chi

28th April 2005, 03:55 PM

Originally posted by CFLarsen
I thought you had read the article that Zep referred to?

Guess not.

I don't have much confidence anymore in non-peer-reviewed non-journal articles, many of which aren't even spell checked, hosted on a site beloning to a person known to be hostile to the subjects being written about, written mainly by anonymous people with no credentials in the fields they are writing about, and some deliberately poking fun at their subjects.

Guess you do.

T'ai Chi

28th April 2005, 04:06 PM

Originally posted by Robin
An interesting page to read on their site is:

http://noosphere.princeton.edu/results.html

Where they list 199 events and z scores for each. 19 out of 199 are identified as statistically significant and these have a z score of greater than 1.64 or less than -1.64.

If the data was purely random the proportion you would expect to have these scores is 10%. The proportion of events that have these score is 10%.

So this system does not detect global events, much less predict them - I am not sure why they are still persisting with this.

That's interesting. It seems we have somewhat of a paradox then, because the combined Z score is highly significant, but we'd expect it to not be.

In fact, about 60% of the z-scores are positive.

28th April 2005, 04:16 PM

Originally posted by jzs
I don't have much confidence anymore in non-peer-reviewed non-journal articles, many of which aren't even spell checked, hosted on a site beloning to a person known to be hostile to the subjects being written about, written mainly by anonymous people with no credentials in the fields they are writing about, and some deliberately poking fun at their subjects.

Guess you do. .

Have you, yet, any rebuttal to the total dismissal of the PEAR work offered by Stimpy? I have seen evasions and such, but no direct evidence that Stimpy is not right. Stimpy does not owe you proof, you owe him proof that he is wrong, since you are the one at least suggesting that the PEAR work is valid, which represents an extraordinary claim.

T'ai Chi

28th April 2005, 04:38 PM

Originally posted by jj
.
Have you, yet, any rebuttal to the total dismissal of the PEAR work offered by Stimpy? I have seen evasions and such, but no direct evidence that Stimpy is not right. Stimpy does not owe you proof, you owe him proof that he is wrong, since you are the one at least suggesting that the PEAR work is valid, which represents an extraordinary claim.

He couldn't calculate the variance of 10 whole numbers correctly, yet went out of his way to insult me for even suggesting that his expertise goofed. If he didn't even listen or understand in regards to a critique about somethig so simple, to talk about the much more complex would be an exercise in futility. Previously he made a comment that the GCP didn't even list the non-significant p-values, when a quick glance at their results page shows otherwise. Debating with someone like this is a waste of my time, and that is being said based on the evidence.

It is not up to me to show proof (and you mean "evidence", not "proof"; *sigh*, will you ever understand the difference?) that Stimpy is not right, as you stupidly believe. He has showed nothing convincing, except mistakes that convince me he doesn't know what he is talking about, and none of your bluster, or his repeated misunderstandings, will change that.

Robin

28th April 2005, 04:42 PM

Originally posted by jzs
That's interesting. It seems we have somewhat of a paradox then, because the combined Z score is highly significant, but we'd expect it to not be.

In fact, about 60% of the z-scores are positive.
And you would expect it to be 50%, but try this - generate 200 pseudo-random numbers in a spreadsheet and then calculate the z scores using the assumed mean and standard deviation (as GCP does). Then calculate the percentage of positive z scores and try hitting the recalculate key a few times. This will vary from about 45% to 60% so the results are not unusual for purely random data.

Additionally the analysis of the events is not consistent, there are 7 distinct resolutions, presumably the analyst has chosen the resolution that gives the most significant result.

Incidentally I made a mistake in my previous post - there are about 13% of events with an absolute z score of over 1.64. However a similar test of 200 z scores on assumed mean and variance shows that this is not unusual for random data, especially given the selection bias referred to earlier.

Robin

28th April 2005, 04:53 PM

Originally posted by CFLarsen
The University of Princeton is not associated with the Global Consciousness Project
It is hosted on the Princeton domain (and thus gains credibility). It is headed by Roger Nelson who your link says is still connected with PEAR, the project grew out of PEAR and the PEAR main page proudly displays cumulative deviation traces from random event generators. Certainly the article in the OP hammers the Princeton link for all it is worth. So it is certainly associated with Princeton, even if it is not actually run by Princeton.

I was just wondering if this is a sore point with Princeton faculty the way the ID link is a sore point with Baylor.
Originally posted by jzs
I have never read anything suggesting they are unhappy. In places of learning, things tend to get investigated, no matter if they are not the current topics of science.
Presumably they have some sort of maths department at Princeton, I wonder how happy they are at the statistical analysis of this and other PEAR projects. If I have time I may write to them and get their opinion of the 'cumulative deviation' methods used.

One of they key claims of the project is that cumulative z^2-1 from normally distributed random data will always do a random walk around zero. Well maybe, but how close to zero? In fact if you try this just on a spreadsheet you will find that cumulative z^2-1 from random data displays just the deviations that Nelson claims are evidence of a link with global consciousness.

T'ai Chi

28th April 2005, 05:00 PM

Originally posted by Robin
And you would expect it to be 50%, but try this - generate 200 random numbers in a spreadsheet and then calculate the z scores using the assumed mean and standard deviation (as GCP does). Then calculate the percentage of positive z scores and try hitting the recalculate key a few times. This will vary from about 45% to 60% so the results are not unusual for purely random data.

Feel free to post a link to your Excel file, or to the code you use for each column.

I agree that it would vary approximately over that range, and that is in fact what yields such a small p-value, because it is unlikely to be that far out.

T'ai Chi

28th April 2005, 05:06 PM

Originally posted by Robin

One of they key claims of the project is that cumulative z^2-1 from normally distributed random data will always do a random walk around zero. Well maybe, but how close to zero? In fact if you try this just on a spreadsheet you will find that cumulative z^2-1 from random data displays just the deviations that Nelson claims are evidence of a link with global consciousness.

Again, feel free to post a link to your Excel file, or to the code you use for each column.

Robin

28th April 2005, 05:06 PM

Originally posted by jzs

Feel free to post a link to your Excel file, or to the code you use for each column.

I agree that it would vary approximately over that range, and that is in fact what yields such a small p-value, because it is unlikely to be that far out. [/B]
I will post it over the weekend, but it would only take you a few minutes to put together such a spreadsheet.

In fact I find that about 1 in 5 recalcs gives me a value of 58% or greater, I have got 64% so it is not so very unlikely. And especially given the different resolutions under which the values were calculated.

Robin

28th April 2005, 05:37 PM

Originally posted by jzs
Again, feel free to post a link to your Excel file, or to the code you use for each column.
Put the formula =ROUND(RAND(),0) in cells A1:GR1, then =SUM(A1:GR1) in GS1. Then in GT1 put =STANDARDIZE(GS1,100,SQRT(50)). In GU1 put =GT1^2-1.

Now copy all this to as many rows as you like, then in GV1 put =GU1 and in GV2 put =GV1+GU2. Copy GV2 to all the rows that have values.

Now highlight all the figures in the GV column and click your graph tool, and select a line graph. Once the graph is created hit F9 a few times and you will see sometimes a random walk, sometimes a positive slope, sometimes a negative slope.

From the procedures page on the GCP site:
The 200-bit trial sums have expected mean = 100 and standard deviation = 7.071
I suggest copying to about 8,000 rows as this is about the same as the 6 days at 1 minute resolution in the Papal Visit to Israel example. They suggest that a continuous trend over 6 days is highly unlikely but it is not really. In fact they get a 7 day steeper positive slope in the same month from 9th March to 15th March.

T'ai Chi

28th April 2005, 06:10 PM

Thanks Robin for actualy showing something! It is hard to get that round these parts. :)

I think we'd need to also calculate the significance bands, like the GCP folk do, to interpret what is going on. Yes, the random walk will stray from 0 and that is not surprising, but how far from 0 until significance is declared?

Zep

28th April 2005, 09:31 PM

Originally posted by jzs
Zep, if you were paying attention (replying to me, even though in another thread you claim to have me on ignore?, :D ), I didn't ask you for a list of their supporters. I asked you to reveal all of this money that you believe exists. Like are they getting a lot, or a little, etc.

And like I said... as an exercise for you, go count the $ signs on the front page of Skeptic.com

I'm skeptical you can find any woo page with more. That is why the dollar argument that you are touting is pretty lame. I was interested to see what you had to say about PEAR and the money, so indeed I did read that particular post.

PEAR's list of supporters are on the PEAR website, if you care to look. Their details are readily available on the Internet. And I doubt that anyone could invent a more interesting set of sponsors for a research group...

So, are you suggesting that PEAR's "supporters" do NOT provide them with financial support? That they are simply a bunch of enigmatic billionaires and well-funded institutions who do little more than send them letters of encouragement and a woo club badge each year? That PEAR has managed to beg, borrow or steal premises for 25+ years in one of the USA's premier tertiary institutions (which, incidentally, I expect comes at an enormously expensive premium) purely on the good graces of the board of that university? When funds and premises are scarce enough for all the other more viable faculties?

Ahem. Excuse me while I laugh incedulously.

But perhaps we can all stand to be educated. Where do YOU think all the funding for PEAR has come from all these years?

Robin

28th April 2005, 09:34 PM

Originally posted by jzs
Thanks Robin for actualy showing something! It is hard to get that round these parts. :)

I think we'd need to also calculate the significance bands, like the GCP folk do, to interpret what is going on. Yes, the random walk will stray from 0 and that is not surprising, but how far from 0 until significance is declared?
I can't find how they calculate the significance bands though. Seems like it would be a horrendous calculation.

69dodge

29th April 2005, 01:14 AM

Originally posted by jzs
[...] the combined Z score is highly significant, but we'd expect it to not be.Who's "we"?

I would expect it not to be "significant," if physical RNGs were perfect. But they aren't. So I don't.

69dodge

29th April 2005, 02:30 AM

Originally posted by Robin
I can't find how they calculate the significance bands though. Seems like it would be a horrendous calculation.The sum of the squares of n independent standard normal variables has a chi-square distribution with n degrees of freedom.

But, of course, then you need to assume that their z scores really would be normal and independent in the absence of paranormal influences.

I'd much prefer a resampling type of analysis that doesn't make those assumptions. The GCP has a list of "global events," and their calculation of a p-value is based on the RNG data gathered during those events. (By definition, p-values less than, say, 0.01 would, if calculated correctly, happen only 1 time out of 100.) To see whether their calculated p-value is actually as rare as it claims to be, they should redo the same calculation repeatedly---but use, on each repetition, the RNG data from randomly chosen time intervals of the same number and lengths as the global events on their list. If, for example, their originally calculated "p-value" is 0.01, but it turns out that in a full tenth of the repetitions, rather than in just a hundredth, the similarly calculated "p-value" is less than 0.01, then the true p-value is 0.1 rather than 0.01.

CFLarsen

29th April 2005, 02:39 AM

Originally posted by jzs
I don't have much confidence anymore in non-peer-reviewed non-journal articles, many of which aren't even spell checked, hosted on a site beloning to a person known to be hostile to the subjects being written about, written mainly by anonymous people with no credentials in the fields they are writing about, and some deliberately poking fun at their subjects.

We are all aware of your hatred and envy of SkepticReport, but that doesn't change the fact: You have in the past claimed to have read the article, even discussed its contents, but can't remember anything about it now.

Or, perhaps you simply blotted it out? Wanted it to go away?

CFLarsen

29th April 2005, 02:46 AM

Originally posted by jzs
And like I said... as an exercise for you, go count the $ signs on the front page of Skeptic.com

I'm skeptical you can find any woo page with more. That is why the dollar argument that you are touting is pretty lame.

There is a lot more money in skepticism than in Woo-Land???

:dl:

CFLarsen

29th April 2005, 02:56 AM

Originally posted by jzs
Good info, but we are talking about it being linked to their name (ie. the word "Princeton" is in the name), not that they are necessarily on Princeton's payroll.

What is the GCP's official affiliation with Princeton University?

T'ai Chi

29th April 2005, 04:04 AM

Originally posted by CFLarsen
We are all aware of your hatred and envy of SkepticReport, but that doesn't change the fact: You have in the past claimed to have read the article, even discussed its contents, but can't remember anything about it now.

Or, perhaps you simply blotted it out? Wanted it to go away?

Nice attempt at villification. I point out flaws (no peer review (or as you claim 'the whole world' :rolleyes: ), little experts, not even corrected for typos, articles that poke fun, hosted by someone known to be hostile towards anomaly claims, no posting of reader corrections and comments, etc) and you emotionally claim I "hate" it.

It must really have hit home.

And like I have said before, some articles are good, and some are not.

Please, prove the person behind skepticreport.com is not really a pseudoskeptic in skeptic's clothing, and answer the question of 'what is the evidence of specific harm being done to that 5 year old girl that got handcuffed', unless you believe that others should do as you say but not as you do.

T'ai Chi

29th April 2005, 04:05 AM

Originally posted by CFLarsen
There is a lot more money in skepticism than in Woo-Land???

:dl:

You're changing the topic entirely. I specifically mentioned all the $ signs on skeptic.com, and wondered if anyone could find a "woo" site with more $ signs.

T'ai Chi

29th April 2005, 04:07 AM

Originally posted by CFLarsen
What is the GCP's official affiliation with Princeton University?

Why don't you tell us.

All I am saying is that if there is, or is not, or there was, an official connection with Princeton, the name "Princeton" is linked with PEAR.

T'ai Chi

29th April 2005, 04:24 AM

Originally posted by 69dodge
Who's "we"?

ie theory tells us.

I would expect it not to be "significant," if physical RNGs were perfect. But they aren't. So I don't.

Do the RNGs pass the DIEHARD tests?

CFLarsen

29th April 2005, 04:47 AM

Originally posted by jzs
You're changing the topic entirely. I specifically mentioned all the $ signs on skeptic.com, and wondered if anyone could find a "woo" site with more $ signs.

No, you did not. You clearly wanted to cast suspicion on a skeptical site. Something you do when you get a chance.

You are perfectly aware that there is a lot more money in woo sites. You are not fooling anyone.

CFLarsen

29th April 2005, 04:49 AM

Originally posted by jzs
Why don't you tell us.

I don't know. I can't find any. That's why I ask.

You claim to have studied these data. You believe that they are sound. You are the one who claims that they show something we can't explain. You are the one touting "Princeton", so you must be very interested in finding out. Why don't you tell us?

Originally posted by jzs
All I am saying is that if there is, or is not, or there was, an official connection with Princeton, the name "Princeton" is linked with PEAR.

Do you think that it in any way lends credibility to GCP?

29th April 2005, 04:51 AM

Originally posted by jzs
It is not up to me to show proof (and you mean "evidence", not "proof"; *sigh*, will you ever understand the difference?) that Stimpy is not right, as you stupidly believe.

In other words, you wish to assert that you're right, but you show no evidence, none at all, yet you demand evidence from the skeptic that points out your lack of evidence.

Until you produce evidence that Stimpy's trivial dismissal of the whole mess is wrong, you're the one shouldering the entire burden. Every time you refuse it, you show your contempt for both the rules of evidence and the scientific method.

Now, about your dishonest, malicious quibbling about "proof". Stimpy handed you an informal proof. That's "proof" as in mathematical, that's what Stimpy has given you. That's what you have ignored. Have any counter-proof of yours? Have any refutation? Nope, you don't. All you can do is maliciously quibble about the word "proof".

Got any evidence?

Got any proof?

(sound of one hand clapping)

CFLarsen

29th April 2005, 04:51 AM

Originally posted by jzs
Do the RNGs pass the DIEHARD tests?

Can the RNGs be shielded from the effect they are supposed to detect?

No?

Gee, you got a big problem...

T'ai Chi

29th April 2005, 05:22 AM

Originally posted by jj
In other words, you wish to assert that you're right, but you show no evidence, none at all, yet you demand evidence from the skeptic that points out your lack of evidence.

Nope, that's not in other words, that is in your words. Stimpy in fact showed no evidence. Neither have you. If you make a negative claim, that is a claim, and the burden is on you. His argument hinges on the level of bias, but he has made no attempts to even gauge the level of bias that he is saying is a big problem. He has provided evidence of several misunderstandings, and what he is currently talking about is no different.

Until you produce evidence that Stimpy's trivial dismissal of the whole mess is wrong,

Well I did show that his (mis)calculation of the variance of 10 whole numbers was wrong. :D But he still was trying to 'get' me, and wouldn't listen to me. His mindset makes this an exercise in futility.

All you can do is maliciously quibble about the word "proof".

Just pointing out your confusion. In science, providing evidence is talked about, not proving things, as you often demand.

69dodge

29th April 2005, 05:41 AM

Theory does not tell us that an RNG is guaranteed to generate six years' worth of perfectly random bits if a different RNG of the same design has once generated a few day's worth of bits that passed the Diehard tests.

69dodge

29th April 2005, 06:01 AM

Some versions of Excel have problems with their pseudo-random number generator.

http://support.microsoft.com/default.aspx?scid=kb;en-us;834520

http://support.microsoft.com/default.aspx?scid=kb;en-us;828795&Product=OFF2003

T'ai Chi

29th April 2005, 06:39 AM

Originally posted by 69dodge
Theory does not tell us that an RNG is guaranteed to generate six years' worth of perfectly random bits if a different RNG of the same design has once generated a few day's worth of bits that passed the Diehard tests.

Who said guaranteed? Theory is a model that tells you what is expected. If you get far from what is expected, then something is going on; where something could be a real phenomenon, or flaws, error, etc.

T'ai Chi

29th April 2005, 06:42 AM

Originally posted by 69dodge
Some versions of Excel have problems with their pseudo-random number generator.

http://support.microsoft.com/default.aspx?scid=kb;en-us;834520

http://support.microsoft.com/default.aspx?scid=kb;en-us;828795&Product=OFF2003

Well if one pressed F9 (to get new random numbers) many times, that is. The PRNG itself was perfectly fine; passed the DIEHARD tests.

People do look (I hope) at what their R or PRNGs are outputting. :)

CFLarsen

29th April 2005, 07:01 AM

Originally posted by jzs
Who said guaranteed? Theory is a model that tells you what is expected. If you get far from what is expected, then something is going on; where something could be a real phenomenon, or flaws, error, etc.

But that's not what the GCP gets, is it? Minuscule deviations, based on non-calibrated, non-shielded RNGs.

What do you think is going on?

A real phenomenon?
Flaws?
Errors?

I am not asking what other people think. I am asking what you think.

Stimpson J. Cat

29th April 2005, 07:56 AM

jzs,

He couldn't calculate the variance of 10 whole numbers correctly, yet went out of his way to insult me for even suggesting that his expertise goofed. If he didn't even listen or understand in regards to a critique about somethig so simple, to talk about the much more complex would be an exercise in futility. Previously he made a comment that the GCP didn't even list the non-significant p-values, when a quick glance at their results page shows otherwise. Debating with someone like this is a waste of my time, and that is being said based on the evidence.

It is not up to me to show proof (and you mean "evidence", not "proof"; *sigh*, will you ever understand the difference?) that Stimpy is not right, as you stupidly believe. He has showed nothing convincing, except mistakes that convince me he doesn't know what he is talking about, and none of your bluster, or his repeated misunderstandings, will change that.

Just as I predicted when it happened, you have taken a trivial and irrelevant arithmetic error, and are attempting to use it to discredit my arguments, rather than actually address them.

Never mind that you seem to be either unwilling or unable to point out what the errors in my argument are.

Never mind that when my arithmetic error, and my error regarding non-significant p-values in the PEAR analysis, were pointed out to me, I admitted my error.

Never mind that if you were to actually point out some error in my arguments, I would admit them as well.

Never mind that this arithmetic error which you now site as a reason for ignoring my arguments, was made long after I made the argument, and could not actually be your reason.

Never mind that the derivation which I provided to justify my arguments, and which you have ignored, was demanded by you.

You are pathetic. It is painfully obvious that you have no interest in whether the PEAR methodology is sound, or in having any kind of reasonable discussion of it. You are completely unable to make any sort of arguments about it at all. Instead, all you do is perpetually demand that anybody who thinks there is something wrong with the PEAR analysis prove it, and then ignore them when they do.

Nope, that's not in other words, that is in your words. Stimpy in fact showed no evidence. Neither have you. If you make a negative claim, that is a claim, and the burden is on you. His argument hinges on the level of bias, but he has made no attempts to even gauge the level of bias that he is saying is a big problem.

This is simply false, and why it is false has been repeatedly explained to you by both myself and others. My argument does not in any way hinge on the level of bias. The fact that you think it does simply proves that you not only do not understand my argument at all, but also that you do not understand even the basics of statistical analysis of experimental data.

Well I did show that his (mis)calculation of the variance of 10 whole numbers was wrong. But he still was trying to 'get' me, and wouldn't listen to me. His mindset makes this an exercise in futility.

No, you did not. What you did was simply assert that I was wrong, without any explanation. When another poster actually provided an explanation, I recognized and admitted the mistake.

Likewise, it is not sufficient for you to simply assert that there is something wrong with my arguments against the PEAR methodology. And it certainly is not sufficient for you to simply say that since I can make simple arithmetic errors, my entire argument concerning the PEAR methodology should be dismissed without even being addressed.

Point out the errors in my analysis. Provide arguments as to why you think my analysis is wrong. Explain why you think my arguments require me to prove that there is some huge bias in the PEAR data. Address my arguments for why I do not have to do so.

You have not even attempted to engage in any sort of actual discussion about it. All you have done is dismiss my argument with excuses which clearly demonstrate that you do not understand it.

Dr. Stupid

T'ai Chi

29th April 2005, 08:07 AM

Originally posted by CFLarsen

What do you think is going on?

A real phenomenon?
Flaws?
Errors?

I am not asking what other people think. I am asking what you think.

What I think is that I believe the results could be due to something you didn't list, and that is I think it could just be due to chance. I am also open-minded to state that I admit there could be some real phenomena there. But most likely, I believe it is from chance.

But luckily what I think isn't relevant. The data itself is what is important. Why don't you analyze some, instead of giving typical SR testimony?

CFLarsen

29th April 2005, 08:15 AM

Originally posted by jzs
What I think is that I believe the results could be due to something you didn't list, and that is I think it could just be due to chance. I am also open-minded to state that I admit there could be some real phenomena there. But most likely, I believe it is from chance.

OK, so, most likely, you believe it is from chance.

How do you reconcile that with your claim that the RNGs produce completely random numbers?

T'ai Chi

29th April 2005, 08:22 AM

Originally posted by Stimpson J. Cat

, you have taken a trivial and irrelevant arithmetic error,

Not trivial, as we are talking about VARIANCE here, stimpy. Morevoer, the incriminating part is, is that even after I pointed out your error, you still didn't believe me.

This evidence, and you not providing the evidence of the bias you claim is troublesome, tells me to not pay serious attention to your arguments.

You said
(bold mine)

"I am simply pointing out (1) that if there is any correlation in the bits, then mean of the XORed data may still be biased, and (2) that even if there is not, the initial bias in the mean will result in correlations in the XORed data. "

Your critique, for it to have any serious impact, hinges on their being correlation in the bits to cause the bias, and the initial bias in the mean. Your own words above show that. If these are small, your critique falls flat.

If you want to not present any measure of these, fine, but don't expect your merely raising the possibility of these things existing to be the same as exposing a serious flaw in their analysis.

Never mind that if you were to actually point out some error in my arguments, I would admit them as well.

I pointed out your error in calculating the variance, and you ignored and ridiculed me.

You are pathetic.

I can't help it if you haven't presented any convincing evidence. For you to resort to insults is quite telling.

but also that you do not understand even the basics of statistical analysis of experimental data.

:rolleyes: I guess you finally understood that bit about the variance?

Back on ignore for stimpee.

T'ai Chi

29th April 2005, 08:25 AM

Originally posted by CFLarsen
OK, so, most likely, you believe it is from chance.

How do you reconcile that with your claim that the RNGs produce completely random numbers?

They pass the DIEHARD tests, hence we can be assured of their randomness quality, but then when these random numbers are looked at on days specified in the formal hypothesis registry, there is statistical significance.

CFLarsen

29th April 2005, 08:25 AM

Originally posted by jzs
Back on ignore for stimpee.

So, you refuse to address Stimpy's points.

CFLarsen

29th April 2005, 08:48 AM

Originally posted by jzs
They pass the DIEHARD tests, hence we can be assured of their randomness quality

Which the manufacturers admit are not completely random. So, they are useless for the GCP experiments.

Originally posted by jzs
but then when these random numbers are looked at on days specified in the formal hypothesis registry, there is statistical significance.

No, you are, once again, seriously misrepresenting facts. On some days, hand-picked by the experimenters, the RNGs do not produce completely random numbers. There is no correlation between events and data.

How can the RNGs produce completely random numbers, if they don't produce results expected from chance?

Your conclusion must be that the whole idea of randomness is flawed. There is no such thing as coincidence. Is this correct?

T'ai Chi

29th April 2005, 11:20 AM

Originally posted by CFLarsen
So, you refuse to address Stimpy's points.

Only if you consder my saying his arguments aren't convincing because he hasn't presented evidence, as refusing to address.

Claus, will you EVER bring anything of SUBSTANCE to a thread? Your only focus is on bickering. Every thread.

Tsk.

T'ai Chi

29th April 2005, 11:29 AM

Originally posted by CFLarsen
Which the manufacturers admit are not completely random. So, they are useless for the GCP experiments.

'completely random' is your strawman, and moreover, is as moot as Stimpy saying that the bits can't be modelled exactly as a binomial. Duh! That is why they are called a model in the first place.

No, you are, once again, seriously misrepresenting facts

Not at all, but you do illustrate your vast misunderstanding of issues. On the days specified in the formal hypothesis registry, analysis shows it tends to be improbable.

Your conclusion must be that the whole idea of randomness is flawed. There is no such thing as coincidence. Is this correct?

No, again your take on things is skewed. I'm looking at the actual data, which shows that on days specified in the formal hypothesis registry, the data produced by RNGs that have passed the DIEHARD tests tends to be improbable.

But if you want to discard statistics when it is used to study topics you are cynical about, that is your issue.

Stimpson J. Cat

29th April 2005, 11:37 AM

jzs,

Not trivial, as we are talking about VARIANCE here, stimpy. Morevoer, the incriminating part is, is that even after I pointed out your error, you still didn't believe me.
You did not point out the error. You just said that I was wrong. Why should any of us simply believe each other's unsubstantiated claims? If you want me to believe you when you say that I have made an error, then you need to explain what the error was, and demonstrate that it was, in fact, and error.

You seem to be completely incapable of doing this.

You said
(bold mine)

"I am simply pointing out (1) that if there is any correlation in the bits, then mean of the XORed data may still be biased, and (2) that even if there is not, the initial bias in the mean will result in correlations in the XORed data. "

Your critique, for it to have any serious impact, hinges on their being correlation in the bits to cause the bias, and the initial bias in the mean. Your own words above show that. If these are small, your critique falls flat.
This has been explained to you several times already. My argument is not that if such biases are present, their analysis is invalid. My argument is that unless they demonstrate that such biases are not present (which they have not done), their analysis is meaningless, since positive results could simply be due to those biases.

This is really basic stuff, which anybody who has ever done any kind of data analysis should be familiar with.

If you want to not present any measure of these, fine, but don't expect your merely raising the possibility of these things existing to be the same as exposing a serious flaw in their analysis.
You've got it backwards.

It is the PEAR people who are "merely raising the possibility" that such biases are not present. That is the flaw in their analysis. I am sorry that you do not understand this. Why should we just accept on faith that such biases are not present? If we do not, then we cannot conclude from their results that anything unusual, or even interesting, has happened. Only by assuming that the REGs will, in the absence of external influences, be sufficiently unbiased, could we possibly conclude anything from their results.

I am not prepared to make such an assumption. Why do you think that we should?

I guess you finally understood that bit about the variance?
You really have no idea who you are talking to, do you?

I am a physicist. I do experimental data analysis for a living. I have several papers published in pear reviewed physics journals on the subject, and have given invited lectures on the subject at major physics conferences.

I "understand about the variance" quite well, thank you very much. But I am sure you would rather believe that I don't, and that my error was not a simple arithmetic error, but instead the result of me not understanding how variance works. You seem to be quite adept at believing whatever you want to believe is true, regardless of any evidence to the contrary.

Back on ignore for stimpee.
Boo hoo. It breaks my heart to know that you will now be passively ignoring my posts by having the board hide them from you, rather than actively ignoring them by simply refusing to respond to my arguments.

Please, do keep me on ignore. I am not posting for your benefit anyway. That would be a complete waste of time, since you have neither the basic understanding of statistical analysis needed to get anything from them, nor any interest in learning.

Dr. Stupid

Bronze Dog

29th April 2005, 12:55 PM

Originally posted by jzs
Back on ignore for stimpee.

Argument by covering your ears and going "LA-LA-LA-LA-LA!"

I've done some lurking in this topic, and you're just a troll, jzs.

jzs: "You're wrong!"

Stimpson: "Show me where the problems with my informal proof are."

CFLarson: "Show him the problems."

jzs: "You're wrong!"

Stimpson: "Show us where I'm wrong."

CFLarson: "What he said."

jzs: "You're wrong!"

Stimpson: "Show us where I'm wrong or shut up, you troll!"

jzs: "You're wrong! And you're insulting!"

CFLarsen

29th April 2005, 01:44 PM

Originally posted by jzs
'completely random' is your strawman

Huh?? What are you talking about? If the RNGs don't produce completely random numbers, how on Earth can you claim that there is "something" going on that is not explainable by science?

Originally posted by jzs
Not at all, but you do illustrate your vast misunderstanding of issues. On the days specified in the formal hypothesis registry, analysis shows it tends to be improbable.

Not all days, jzs. Not all days. That's what you are misrepresenting.

Originally posted by jzs
No, again your take on things is skewed. I'm looking at the actual data, which shows that on days specified in the formal hypothesis registry, the data produced by RNGs that have passed the DIEHARD tests tends to be improbable.

Man, you are a liar. You know that it isn't all days, and yet, you still perpetuate this lie of yours.

Originally posted by jzs
But if you want to discard statistics when it is used to study topics you are cynical about, that is your issue.

Oh, I am not discarding anything. Contrary to you, I am looking at all the data.

T'ai Chi

29th April 2005, 01:57 PM

Originally posted by BronzeDog
Argument by covering your ears and going "LA-LA-LA-LA-LA!"I've done some lurking in this topic, and you're just a troll, jzs.jzs: "You're wrong!"Stimpson: "Show me where the problems with my informal proof are."CFLarson: "Show him the problems."jzs: "You're wrong!"Stimpson: "Show us where I'm wrong."CFLarson: "What he said."jzs: "You're wrong!"Stimpson: "Show us where I'm wrong or shut up, you troll!"
jzs: "You're wrong! And you're insulting!"

Wow. Got anything to contribute?

T'ai Chi

29th April 2005, 03:55 PM

Originally posted by CFLarsen
Huh?? What are you talking about?

I suggest you look up the meaning of 'strawman'. No one is saying "completely random" as you claimed.

, how on Earth can you claim that there is "something" going on that is not explainable by science?

Feel free to show me where I claimed that it was not explainable by science.

Not all days, jzs. Not all days. That's what you are misrepresenting.

Read it again. I wrote "tends". Do you know what "tends" means?

Man, you are a liar. You know that it isn't all days, and yet, you still perpetuate this lie of yours.

You can't read and/or understand, so I'll help: I said "tends".

I, in fact, said

"On the days specified in the formal hypothesis registry, analysis shows it tends to be improbable."

and I said

"I'm looking at the actual data, which shows that on days specified in the formal hypothesis registry, the data produced by RNGs that have passed the DIEHARD tests tends to be improbable."

Which are both 100% true.

You won't apologize, because you are a cynical pseudoskeptic who must 'win' at all costs. That is also 100% true.

Oh, I am not discarding anything. Contrary to you, I am looking at all the data.

Feel free to present your analysis. I've asked you before. And you've chickened out before. All the "looking at all the data" we get from your website is your testimony which in terms of evidence, equates to dung, scientifically speaking.

T'ai Chi

29th April 2005, 04:23 PM

Originally posted by Stimpson J. Cat

You did not point out the error. You just said that I was wrong.

Fibber.

TheEternalVortex, pointed out that you goofed. I echoed it. You then said that you

"forgot to subtract off the mean"

which I pointed out was still(!) wrong, because if you subtract off the mean, you still don't get the correct variance. To which you replied

"Yes, it does. But why you would care is completely beyond me."

and

"Work it out yourself. I'm sure even you can figure out how to sum the squares of the numbers from 0 to 10, and then divide by 11."

So you see, you really are a cynic in skeptics clothing. A true skeptic would have re-worked it then and there without any glib comment. You? Nope.

My argument is that unless they demonstrate that such biases are not present (which they have not done),

What don't you understand about normalizing?

This is really basic stuff, which anybody who has ever done any kind of data analysis should be familiar with.

And they should know how to calculate the variance of 10 numbers.

Why should we just accept on faith that such biases are not present?

Why should we accept your argument on faith that bias (which you haven't shown to be an issue) is an issue even after normalizing?

For example, the bias that you are ranting about.. is it in the positive direction, negative, what? How big is it? I know, I know, you just throw out possibilities...it is not your job to provide any actual numbers.

You really have no idea who you are talking to, do you?

Here it comes..

I am a physicist.

So are Russel and Targ.

A few of the GCP and PEAR people are engineers.

What your day job is is moot, and a pathetic, desperate argument of pseudo-authority. Unless, of course, you work directly with the inner workings of RNGs, which you don't. My job is much closer to the topic than yours is, but you don't see me pulling that card. You must be desperate to think that passes in a debate on a skeptics board.

As BillyHoyt says, 'I am just a strip club bouncer' (even though he works with computers and is not a strip club bouncer). His point is that such arguments by authority don't hold up. Take notes.

I have several papers published in pear reviewed physics journals on the subject, and have given invited lectures on the subject at major physics conferences.

Any on RNGs? No?

I "understand about the variance" quite well, thank you very much.

Then you'll understand that subtracting out the mean still doesn't give you the correct variance as you claimed. I'm sure you understand.... now.

Oops, I didn't ignore. Must be failing memory. Oh well, I still remember how to calculate the variance of 10 numbers.

69dodge

29th April 2005, 05:06 PM

Originally posted by jzs
They pass the DIEHARD tests, hence we can be assured of their randomness qualityWe can?

Just above, you asked "Who said guaranteed?".

Is "assured" different from "guaranteed"?

Stimpson J. Cat

29th April 2005, 05:06 PM

jzs,

I thought you had me on ignore?

You did not point out the error. You just said that I was wrong.
Fibber.

TheEternalVortex, pointed out that you goofed. I echoed it. You then said that you

"forgot to subtract off the mean"

which I pointed out was still(!) wrong, because if you subtract off the mean, you still don't get the correct variance. To which you replied

"Yes, it does. But why you would care is completely beyond me."

and

"Work it out yourself. I'm sure even you can figure out how to sum the squares of the numbers from 0 to 10, and then divide by 11."
Like I said, you did not point out the error. You just said that I was wrong.

So you see, you really are a cynic in skeptics clothing. A true skeptic would have re-worked it then and there without any glib comment. You? Nope.
What on Earth does that have to do with skepticism?

My argument is that unless they demonstrate that such biases are not present (which they have not done),
What don't you understand about normalizing?
Good grief! Is it really possible that after all of this time you still don't understand that my argument was that their normalization procedure is flawed???

Why should we just accept on faith that such biases are not present?
Why should we accept your argument on faith that bias (which you haven't shown to be an issue) is an issue even after normalizing?
You shouldn't. I have explained with cold hard math exactly why it is. I cannot help it if you do not understand my explanation. I would have been happy to clarify any parts which are unclear, but you have not pointed out any.

For example, the bias that you are ranting about.. is it in the positive direction, negative, what? How big is it? I know, I know, you just throw out possibilities...it is not your job to provide any actual numbers.
That's right, it's not. I don't know what the bias is, or what direction it is in.

The point is that neither do the PEAR researchers!!!!

Without that information, the results of their analysis are utterly meaningless.

I am a physicist.
So are Russel and Targ.

A few of the GCP and PEAR people are engineers.

What your day job is is moot, and a pathetic, desperate argument of pseudo-authority. Unless, of course, you work directly with the inner workings of RNGs, which you don't. My job is much closer to the topic than yours is, but you don't see me pulling that card. You must be desperate to think that passes in a debate on a skeptics board.
I have not made any sort of argument from authority. I have never asked, nor expected, anybody to simply accept my arguments because I am a physicist. I pointed this out solely to refute your demonstrably false implication that I do not understand basic statistics. I do, and can easily prove that I do. Your representation of a trivial arithmetic error as evidence that I do not, is nothing more than a blatant attempt to discredit my arguments without actually addressing them. It is a classic example of the ad-hominem fallacy. Rather than actually address my arguments, you simply reject anything I say on the basis that if I can make a simple arithmetic error, my arguments can't possibly be valid.

You must have an incredibly low opinion of the people on this board if you think that such dishonest and underhanded tactics are going to distract them from the fact that you have completely failed to offer any substantial counter to my criticism of the PEAR methodology.

Dr. Stupid

T'ai Chi

29th April 2005, 05:26 PM

Originally posted by Stimpson J. Cat

I said that if you subtract off the mean, you still don't get the correct variance. That is pointing out an error. And you didn't check it before replying to me, saying I was wrong.

I have explained with cold hard math exactly why it is.

You didn't show the level of bias which is what your argument hinges upon. If it is tiny, as it is typically after normalizing, your poitns are not convincing.

That's right, it's not. I don't know what the bias is, or what direction it is in.

The point is that [b]neither do the PEAR researchers!!!!

Oh I see, so you are justified in saying they don't know, but no one dare say that you don't know? Is that how it works?

I pointed this out solely to refute your demonstrably false implication that I do not understand basic statistics.

Well you telling me that the variance calculation was correct if you subtract the mean out certainly demonstrated something.

I guess you don't have any publications relating to RNGs. Could you be mistaken? No, not possible!!

You must have an incredibly low opinion of the people on this board if you think that such dishonest and underhanded tactics are going to distract them from the fact that you have completely failed to offer any substantial counter to my criticism of the PEAR methodology.

Talk about dishonest. You still haven't really showed anything of substance, despite your belief and rantings.

Robin

29th April 2005, 10:11 PM

Originally posted by 69dodge
The sum of the squares of n independent standard normal variables has a chi-square distribution with n degrees of freedom.
The problem is that it is not a sum of squares, it is a sum of z^2-1, which is not a measure I have come across before and would be interested to know how they calculated the confidence bands.
But, of course, then you need to assume that their z scores really would be normal and independent in the absence of paranormal influences.
Which is why you would need a measure of expected values.
I'd much prefer a resampling type of analysis that doesn't make those assumptions. The GCP has a list of "global events," and their calculation of a p-value is based on the RNG data gathered during those events. (By definition, p-values less than, say, 0.01 would, if calculated correctly, happen only 1 time out of 100.) To see whether their calculated p-value is actually as rare as it claims to be, they should redo the same calculation repeatedly---but use, on each repetition, the RNG data from randomly chosen time intervals of the same number and lengths as the global events on their list. If, for example, their originally calculated "p-value" is 0.01, but it turns out that in a full tenth of the repetitions, rather than in just a hundredth, the similarly calculated "p-value" is less than 0.01, then the true p-value is 0.1 rather than 0.01.
It sounds like a good idea. First though, I would be interested in the p-values calculated from an equal playing field, rather than the 15-minute,10-minute,1 second etc switch that gets done here. If I have time I may do it.

In any case as I have said previously the values are not outside what you would expect from random data.

29th April 2005, 10:19 PM

Originally posted by jzs

Nope, that's not in other words, that is in your words. Stimpy in fact showed no evidence. Neither have you.

When you claim that we must, you lie. It's that simple. You are the one making tthe assertion. You must address criticisms. We don't have to do anything but point out difficulties. You are making the positive assertion, the burden of proof is entirely and completely yours.

If you make a negative claim, that is a claim, and the burden is on you.

Stuff and nonsense. You have assumed the burden, you stand and deliver.

Just pointing out your confusion. In science, providing evidence is talked about, not proving things, as you often demand. [/B]

A lie. You are dishonestly and maliciously confusing mathematical proof with scientific evidence. Your statement is an unproven, fraudulent misrepresentation.

T'ai Chi

29th April 2005, 10:41 PM

Originally posted by jj
When you claim that we must, you lie.

When you make a misinformed statement such as that, you lie. If you make a negative claim, that is a claim, and the burden of evidence (not 'proof'; take notes) is also on you.

And don't say "we", jj. You've done less than stimpee here. You may think you are his bulldog, but you come across more like a chihuahua. You don't even speak for most skeptics despite believing you do (ie. "we").

You are the one making tthe assertion. You must address criticisms. We don't have to do anything but point out difficulties.

And that is why you often are mistaken for a deluded cynic. If you make a negative claim, you have some burden. I don't care if you don't like that fact, but that is reality. Truzzi identifies the attitude of your type to a T (source:http://www.anomalist.com/commentaries/pseudo.html).

A lie. You are dishonestly and maliciously confusing mathematical proof with scientific evidence.

If arguments were tv shows, yours would be all laugh track.

Seek the attention you desperately need, elsewhere.

CFLarsen

30th April 2005, 12:09 AM

Originally posted by jzs
I suggest you look up the meaning of 'strawman'. No one is saying "completely random" as you claimed.

OK, so they are not completely random. That's your explanation. Gotcha.

Nothing more to talk about, is there?

Originally posted by jzs
Feel free to show me where I claimed that it was not explainable by science.

Isn't that your whole point??

Originally posted by jzs
Read it again. I wrote "tends". Do you know what "tends" means?

Yep. But when you don't mention that they are selecting the days, then you are misrepresenting the facts.

What does the analysis say, when they include the whole dataset?

Originally posted by jzs
"I'm looking at the actual data, which shows that on days specified in the formal hypothesis registry, the data produced by RNGs that have passed the DIEHARD tests tends to be improbable."

Can you make up your mind here? If the data tends to be improbable, then why are they simultaneously explainable by science?

Originally posted by jzs
Feel free to present your analysis. I've asked you before. And you've chickened out before. All the "looking at all the data" we get from your website is your testimony which in terms of evidence, equates to dung, scientifically speaking.

So, don't look at it. Can you do better?

CFLarsen

30th April 2005, 12:12 AM

Originally posted by jzs
So are Russel and Targ.

What the hell are you talking about, you idiot? "Russell" and "Targ" are not two persons. Russell Targ (http://www.parapsych.org/members/r_targ.html) is one person.

:hb:

Originally posted by jzs
A few of the GCP and PEAR people are engineers.

Emphasis on "a few".

Originally posted by jzs
Oops, I didn't ignore. Must be failing memory. Oh well, I still remember how to calculate the variance of 10 numbers.

Let it go. Stimpy has admitted his error, and it is nothing but petty of you to keep harping on this forever. You make mistakes too, but you don't see people use that as an excuse to address the points you are making.

Should we refuse to do that, given the fact that you don't even know who the people are you are so fond of referring to?

Stimpson J. Cat

30th April 2005, 01:47 AM

jzs,

I have explained with cold hard math exactly why it is.
You didn't show the level of bias which is what your argument hinges upon. If it is tiny, as it is typically after normalizing, your poitns are not convincing.
Please either explain why my argument hinges upon the level of bias, or respond to my explanation of why it does not. Simply repeating over and over again that it does, and ignoring my explanations of why it does not, is dishonest.

That's right, it's not. I don't know what the bias is, or what direction it is in.

The point is that neither do the PEAR researchers!!!!
Oh I see, so you are justified in saying they don't know, but no one dare say that you don't know? Is that how it works?
Huh?!!?! I just said that I don't know! Are you even reading what I post?

I pointed this out solely to refute your demonstrably false implication that I do not understand basic statistics.
Well you telling me that the variance calculation was correct if you subtract the mean out certainly demonstrated something.
Yes, it demonstrates that even I can make simple mistakes from time to time. Imagine that.

I guess you don't have any publications relating to RNGs. Could you be mistaken? No, not possible!!
What do my lack of publications regrding RNGs have to do with anything? I am not even making an argument about RNGs, other than that it is known that they are not perfect, which is trivial. My argument is about the analysis methodology. As I have explained before, my arguments would be applicable to any source of data, not just RNGs. Again my point is not that the biases in their data have had a significant effect on their results (thereby invalidating them). My argument is that they they have not shown that the biases in their data do not have a significant effect on their results, and that therefore they have shown that their results are valid.

And of course I could be mistaken. If I am, point out the mistake! Your attempts to paint me as some sort of elitist who demands that everybody have blind faith that everything I say about science is correct, are dishonest and completely out of line.

Dr. Stupid

69dodge

30th April 2005, 04:57 AM

Originally posted by Robin
The problem is that it is not a sum of squares, it is a sum of z^2-1, which is not a measure I have come across before and would be interested to know how they calculated the confidence bands.It's not a big problem, because 1 is a constant. If you have a sum of n independent (z2 - 1)'s, just add n to the sum, and you'll get a random variable that has a chi-square distribution with n degrees of freedom. I'm pretty sure this is essentially what they do.

T'ai Chi

30th April 2005, 08:39 AM

Originally posted by CFLarsen
What the hell are you talking about, you idiot? "Russell" and "Targ" are not two persons. Russell Targ (http://www.parapsych.org/members/r_targ.html) is one person.

:hb:

Yes I meant Targ and Puthoff(sp?).

CFLarsen

30th April 2005, 08:54 AM

Originally posted by jzs

Yes I meant Targ and Puthoff(sp?). [/B]

So, you admit to making an error. Should we now refuse to address your points from now on, because you made this error?

T'ai Chi

30th April 2005, 09:00 AM

Originally posted by Stimpson J. Cat

Please either explain why my argument hinges upon the level of bias, or respond to my explanation of why it does not.

You've stated it yourself. "Why should we just accept on faith that such biases are not present?" and that you don't know the level of bias or what direction it is in. So you're making a negative claim here. Why don't you analyze the data and answer both of your own questions? You won't, because you believe it is not your job to inquire.

You've said yourself that the frequency of 1's in the stream has probabiltiy .5, no matter what the starting p is. So where does the mean bias come in? What are all these higher level biases you talk about? Do you not know their measure too?

And you still have avoided talking about what happens with the other masks that are used. I wonder why? Oops, not your job to inquire.

You have still not talked about the meaning of a large autocorrelation at lag 1. If you feel this large autocorrelation is an issue, then please, you are invited to predict some time-series. Oops, not your job!

My argument is that they they have not shown that the biases in their data do not have a significant effect on their results, and that therefore they have shown that their results are valid.

You have not shown anything to demonstrate your negative claims.

And of course I could be mistaken. If I am, point out the mistake! Your attempts to paint me as some sort of elitist who demands that everybody have blind faith that everything I say about science is correct, are dishonest and completely out of line.

Yet you're the one who pulled the 'Do you know who you are talking to? I am a physicist.' card. Look, I still do not care if you look at rat neurons or whatever. It is your arguments I am concerned about, and the are simply raising the possibility of something, asserting a negative claim, not inquiring, and believing that is a hard-hitting skeptical analysis.

If you think this line of debate is worthless (and I agree), then you shouldn't have brought it up.

Expect chihuahua shortly.

T'ai Chi

30th April 2005, 09:01 AM

Originally posted by CFLarsen
So, you admit to making an error. Should we now refuse to address your points from now on, because you made this error?

My error didn't have to do with the specific mathematical details we are addressing, which is variance calculations. He goofed, twice, on calculating the variance of 10 numbers, and, more importantly, as I've said and you ignore, when I called him on it, he refused to believe me and ridiculed me.

You, stupidly and predictibly, will keep ignoring that last part.

CFLarsen

30th April 2005, 09:36 AM

Originally posted by jzs
My error didn't have to do with the specific mathematical details we are addressing, which is variance calculations. He goofed, twice, on calculating the variance of 10 numbers, and, more importantly, as I've said and you ignore, when I called him on it, he refused to believe me and ridiculed me.

You, stupidly and predictibly, will keep ignoring that last part.

Awwwwwwww.... Justin got a boo-boo? Justin was not treated right?

Justin stomps his foot and throws a tantrum! Justin runs off and tells!

http://www.skepticreport.com/resources/smilies/crybaby.jpg

Why don't you grow up, instead of collecting insults like other people collect stamps?

T'ai Chi

30th April 2005, 10:03 AM

Originally posted by CFLarsen
Awwwwwwww.... Justin got a boo-boo? Justin was not treated right?

Justin stomps his foot and throws a tantrum! Justin runs off and tells!

http://www.skepticreport.com/resources/smilies/crybaby.jpg

Why don't you grow up, instead of collecting insults like other people collect stamps?

Like I predicted, you ignored the point. :)

Thanks for being a bullying cynic. You annihilate your own arguments with such ease.

CFLarsen

30th April 2005, 10:13 AM

Originally posted by jzs
Like I predicted, you ignored the point. :)

Thanks for being a bullying cynic. You annihilate your own arguments with such ease.

Oh, I think I addressed why you are here. Please prove me wrong:

If the data tends to be improbable, then why are they simultaneously explainable by science?

What does the analysis say, when they include the whole dataset?

T'ai Chi

30th April 2005, 10:27 AM

Bully, seek the attention you desperately need, elsewhere.

CFLarsen

30th April 2005, 10:31 AM

Originally posted by jzs
Bully, seek the attention you desperately need, elsewhere.

Revealing that you find questions "bullying".

30th April 2005, 07:42 PM

Originally posted by jzs
Bully, seek the attention you desperately need, elsewhere.

WOLF! WOLF!

T'ai Chi

30th April 2005, 09:11 PM

And the chihuahua wets himself..

Still no data to back up his negative claims.

CFLarsen

1st May 2005, 12:59 AM

Originally posted by jzs
And the chihuahua wets himself..

Still no data to back up his negative claims.

If you think those who question your claims are "bullies" and are only seeking "attention", why do you pay attention to them?

And if you are so concerned about people insulting you, tot he extent that you collect insults, why do you insult them?

tim

1st May 2005, 04:36 AM

OK folks, calm it down. jzs, this was unneccessary - "And the chihuahua wets himself." - and it does not further your argument. Please desist.

Darat

1st May 2005, 04:40 AM

Several posts in this thread have been reported. I am not going to attempt to find out who "started it" rather I will state that if the trading of insults continues sanctions may be applied to individual Members. If you wish guidelines on what is acceptable I suggest you read your Membership Agreement (http://forums.randi.org/showthread.php?s=&threadid=42936). Whilst we accept that "...the nature of the forum inevitably involves strong emotions and opinions which can result in heated exchanges...." this does not mean we accept continuous petty bickering. Simply put: acting like mature adults debating an issue with an occasional heated remark is acceptable, acting like children in a playground isnâ€™t.

Zep

1st May 2005, 04:57 AM

THANK you, Darat. We seemed to have been going nowhere for a number of pages.

CFLarsen

1st May 2005, 06:43 AM

OK, let's get this thread back on track.

jzs,

If the GCP data tends to be improbable, then why are they simultaneously explainable by science?

What does the analysis say, when they include the whole dataset?

Is "assured" different from "guaranteed"?

Stimpson J. Cat

1st May 2005, 12:01 PM

jzs,

Please either explain why my argument hinges upon the level of bias, or respond to my explanation of why it does not.
You've stated it yourself. "Why should we just accept on faith that such biases are not present?" and that you don't know the level of bias or what direction it is in. So you're making a negative claim here.
What negative claim am I making? My claim is that if they do not establish that the output of the REGs will, in the absence of external influences, be sufficiently close to their model, that nothing can be inferred from statistically significant results of their analysis. This argument does not hinge on the data being too biased. It hinges on the fact that they have not shown that it is not.

Why don't you analyze the data and answer both of your own questions? You won't, because you believe it is not your job to inquire.
I can't do the analysis that would be required. I would have to look at long sequences of output from the REGs which were made during periods in which it is known that no external influences were present. How exactly do you propose that I do this?

Anyway, it is not my job to do their research for them. If they want their experiments to give meaningful results, they need to make sure the experiment is controlled properly. They have not.

My argument is that they they have not shown that the biases in their data do not have a significant effect on their results, and that therefore they have shown that their results are valid.
You have not shown anything to demonstrate your negative claims.
Stop attacking strawmen. I have just stated what my argument is, and I have shown that analytically that my argument is correct. If you think there is a mistake in the derivation where I showed this, then point it out. Repeating over and over again that I have claimed that the data is too biased, when I clearly have not, and demanding that I provide evidence to support a claim I have not made, is simply dishonest.

And of course I could be mistaken. If I am, point out the mistake! Your attempts to paint me as some sort of elitist who demands that everybody have blind faith that everything I say about science is correct, are dishonest and completely out of line.
Yet you're the one who pulled the 'Do you know who you are talking to? I am a physicist.' card.
I did not mention that as an attempt to back up my argument, and you know it. Again, the fact that you choose to misrepresent what I said in this way just demonstrates dishonesty on your part.

Look, I still do not care if you look at rat neurons or whatever. It is your arguments I am concerned about, and the are simply raising the possibility of something, asserting a negative claim, not inquiring, and believing that is a hard-hitting skeptical analysis.
If you are concerned about my arguments, then respond to them. Stop attacking a strawman. You know perfectly well that I am not simply "raising possibilities" and "making negative claims". I have explained this to you repeatedly, and you persist in simply ignoring my explanations and repeating claims which I have already responded to.

If you think this line of debate is worthless (and I agree), then you shouldn't have brought it up.
This is not a debate. It is more like arguing with a parrot. You just keep saying the same thing over and over again, even though it has no relevance to any argument I have actually made.

Let me know when you actually want to address my argument. If all you are going to do is continue to attack a strawman and throw around ad-hominem fallacies, you are just wasting everybody's time.

Dr. Stupid

CFLarsen

1st May 2005, 12:31 PM

jzs,

If the GCP data tends to be improbable, then why are they simultaneously explainable by science?

What does the analysis say, when they include the whole dataset?

Is "assured" different from "guaranteed"?

What negative claim is Stimpy making?

How exactly do you propose that Stimpy look at long sequences of output from the REGs which were made during periods in which it is known that no external influences were present?

Do you agree that if they want their experiments to give meaningful results, they need to make sure the experiment is controlled properly?

If you think there is a mistake in the derivation where Stimpy showed that is argument is analytically correct, can you point it out?

If you are concerned about Stimpy's arguments, could you respond to them?

I'll just keep summarizing for you, OK? Time will tell if you are interested in debating about the GCP, or proselytizing.

T'ai Chi

1st May 2005, 12:44 PM

Originally posted by Stimpson J. Cat
(more of the same)

If you don't care to provide evidence for your negative claims ("its not my job"), fine, that's your choice, just don't expect me to hold your argument in high regards.

CFLarsen

1st May 2005, 12:48 PM

Originally posted by jzs
If you don't care to provide evidence for your negative claims ("its not my job"), fine, that's your choice, just don't expect me to hold your argument in high regards.

jzs,

I'll repeat, in the hope that you are interested in debating about the GCP, and not proselytizing:

If the GCP data tends to be improbable, then why are they simultaneously explainable by science?

What does the analysis say, when they include the whole dataset?

Is "assured" different from "guaranteed"?

What negative claim is Stimpy making?

How exactly do you propose that Stimpy look at long sequences of output from the REGs which were made during periods in which it is known that no external influences were present?

Do you agree that if they want their experiments to give meaningful results, they need to make sure the experiment is controlled properly?

If you think there is a mistake in the derivation where Stimpy showed that is argument is analytically correct, can you point it out?

If you are concerned about Stimpy's arguments, could you respond to them?

T'ai Chi

1st May 2005, 12:51 PM

Originally posted by Darat
"...the nature of the forum inevitably involves strong emotions and opinions which can result in heated exchanges...." this does not mean we accept continuous petty bickering. Simply put: acting like mature adults debating an issue with an occasional heated remark is acceptable, acting like children in a playground isnâ€™t.

Where, in the membership agreement, does it state that you don't accept continuous petty bickering?

I've searched it, but "continuous" and "bickering" aren't words in the membership agreement.

You just made that up.

Please retract what you just made up, or adjust the membership agreement accordingly.

CFLarsen

1st May 2005, 12:56 PM

Originally posted by jzs
Where, in the membership agreement, does it state that you don't accept continuous petty bickering?

I've searched it, but "continuous" and "bickering" aren't words in the membership agreement.

You just made that up.

Please retract what you just made up, or adjust the membership agreement accordingly.

jzs,

I'll repeat, in the hope that you are interested in debating about the GCP, and not proselytizing:

If the GCP data tends to be improbable, then why are they simultaneously explainable by science?

What does the analysis say, when they include the whole dataset?

Is "assured" different from "guaranteed"?

What negative claim is Stimpy making?

How exactly do you propose that Stimpy look at long sequences of output from the REGs which were made during periods in which it is known that no external influences were present?

Do you agree that if they want their experiments to give meaningful results, they need to make sure the experiment is controlled properly?

If you think there is a mistake in the derivation where Stimpy showed that is argument is analytically correct, can you point it out?

If you are concerned about Stimpy's arguments, could you respond to them?

Robin

1st May 2005, 04:31 PM

Originally posted by 69dodge
It's not a big problem, because 1 is a constant. If you have a sum of n independent (z2 - 1)'s, just add n to the sum, and you'll get a random variable that has a chi-square distribution with n degrees of freedom. I'm pretty sure this is essentially what they do.
I was not so much worried about the -1, more about the z. Is the distribution of sum of z^2 the same as the distribution of the sum of squares? The sum of z is zero and the sum of z^2 is n so if you add n to the sum of z2 - 1 you just get n.

The problem is that the GCP analysts use the theoretic mean and variance to calculate z. This skews the data considerably, for example if you calculate each z using mean of 100 and variance of 50 and the mean is actually 100.1 and the variance is 51 then the distribution is very different.

And this is where I think it would be difficult to calculate confidence intervals - so I would still be interested in the calculation behind the confidence bands they use.

tim

1st May 2005, 04:59 PM

Darat

1st May 2005, 05:22 PM

Darat

1st May 2005, 05:25 PM

Originally posted by tim
jzs
We moderators do the best we can. If you don't agree you are free to leave. In fact, please do.

This post has been reported. Please ensure that if you are discussing an issue of Moderation your posts are clearly identifiable as either your personal views or your view as a member of the Mod Team.

T'ai Chi

1st May 2005, 05:57 PM

Originally posted by tim
jzs
We moderators do the best we can. If you don't agree you are free to leave. In fact, please do.

*sigh*

Thread posted in the Forum Management section as requested.

Darat

1st May 2005, 06:21 PM

(Miss posted - should have been posted in Forum Management.)

69dodge

1st May 2005, 08:41 PM

Originally posted by Robin
The sum of z is zero and the sum of z^2 is n so if you add n to the sum of z2 - 1 you just get n.I was using "z" in a different way, I think. I meant a standard normal variable. And by "sum," I meant a sum of independent such variables.

What you say about the constancy of the sums is true if we have a set of data points and we transform each into a corresponding "z-score" by subtracting the mean from it and then dividing by the standard deviation, where the mean and standard deviation used are calculated from that same set of data points. The z-scores are not independent in that case, and they are not necessarily normal: their distribution depends on that of the underlying data.The problem is that the GCP analysts use the theoretic mean and variance to calculate z. This skews the data considerably, for example if you calculate each z using mean of 100 and variance of 50 and the mean is actually 100.1 and the variance is 51 then the distribution is very different.

And this is where I think it would be difficult to calculate confidence intervals - so I would still be interested in the calculation behind the confidence bands they use.Confidence intervals are always calculated based on some null hypothesis or other. What null hypothesis do you want to test?

Using a mean of 100 and a variance of 50 to compute the z-scores tests the hypothesis that the underlying data is normally distributed with those parameters, because if it were, the sum of the squares of the z-scores so computed would have a chi-square distribution.

athon

3rd May 2005, 09:15 AM

No offence to anybody here, but I am surprised this is still going. No...literally none...nada...zip...progress has been made in pages. Justin cannot understand science, and evades the questions he has no means of answering. He has already said he does not understand what a negative control is or what it is needed for, and since he has demonstrated an unwillingness to understand it...well, as a teacher I can tell you it is futile trying to discuss an issue with somebody who has not the means to understand what it is you are discussing.

For the most part it's been educational, but now it is simply wasted pixels.

Athon

CFLarsen

3rd May 2005, 09:25 AM

Originally posted by athon
No offence to anybody here, but I am surprised this is still going. No...literally none...nada...zip...progress has been made in pages. Justin cannot understand science, and evades the questions he has no means of answering. He has already said he does not understand what a negative control is or what it is needed for, and since he has demonstrated an unwillingness to understand it...well, as a teacher I can tell you it is futile trying to discuss an issue with somebody who has not the means to understand what it is you are discussing.

For the most part it's been educational, but now it is simply wasted pixels.

Athon

Wise words.

T'ai Chi

3rd May 2005, 12:54 PM

Originally posted by athon
Justin cannot understand science,

People (like universities and employers and myself) disagree with your purposefully malicious statement.

He has already said he does not understand what a negative control is or what it is needed for,

I said I have never heard those terms used like that. That is a big difference between not understanding or not being able to understand.

, as a teacher

As a strip club bouncer, I'll say that is irrelevant.

CFLarsen

3rd May 2005, 12:58 PM

Originally posted by jzs
People (like universities and employers and myself) disagree with your purposefully malicious statement.

Do you take any criticism of you as a "purposefully malicious statement?

Are you really that thin-skinned?

athon

3rd May 2005, 02:26 PM

Originally posted by jzs

People (like universities and employers and myself) disagree with your purposefully malicious statement.

[/b]

I said I have never heard those terms used like that. That is a big difference between not understanding or not being able to understand.

As a strip club bouncer, I'll say that is irrelevant. [/B]

Wow, quick off the mark when it comes to somebody suggesting you lack the ability to understand science (BTW, do you really think it is purposefully malicious if it is true? I don't comprehend Japanese tea rituals. If somebody accused me of that, I don't think it is malicious. And I was making an observation based on your reluctance to answer certain questions).

So, why aren't you so quick to answer Claus' or Stimpy's questions? Easier to defend your personality than your claims, maybe?

And honestly, people have tried to explain negative controls to you, which are essential in any experiment. Yet you do not see this as relevant. Ergo, you do not understand them. That's ok, but you continue to ignore the help people offer. That's not ok if you wish to continue debating something you do not understand. I make this claim as a teacher due to experience helping others come to a conclusion by learning new skills.

WTF does being a bouncer have to do with anything?

Athon

T'ai Chi

3rd May 2005, 03:04 PM

Originally posted by CFLarsen
Do you take any criticism of you as a "purposefully malicious statement?

Are you really that thin-skinned?

Only the ones that are purposefully malicious statements.

T'ai Chi

3rd May 2005, 03:08 PM

Originally posted by athon
Wow, quick off the mark when it comes to somebody suggesting you lack the ability to understand science

Liar. You didn't "suggest" it, you stated

"Justin cannot understand science, "

If you state something purposefully malicious, where the information is available to all, including you, to show that this is patently false, expect the thug tactics to be pointed out.

athon

4th May 2005, 01:10 AM

Originally posted by jzs

Liar. You didn't "suggest" it, you stated

"Justin cannot understand science, "

If you state something purposefully malicious, where the information is available to all, including you, to show that this is patently false, expect the thug tactics to be pointed out. [/B]

:rolleyes:

Whatever. I could not be bothered having this sort of argument.

Now, answer Claus' questions or I think we can safely conclude that they are too difficult and this debate has run its course and it has been shown that this experiment's results have no merit.

Athon

CFLarsen

4th May 2005, 01:22 AM

Originally posted by athon
Now, answer Claus' questions or I think we can safely conclude that they are too difficult and this debate has run its course and it has been shown that this experiment's results have no merit.

Athon

jzs,

Here they are, in case you forgot:

If the GCP data tends to be improbable, then why are they simultaneously explainable by science?

What does the analysis say, when they include the whole dataset?

Is "assured" different from "guaranteed"?

What negative claim is Stimpy making?

How exactly do you propose that Stimpy look at long sequences of output from the REGs which were made during periods in which it is known that no external influences were present?

Do you agree that if they want their experiments to give meaningful results, they need to make sure the experiment is controlled properly?

If you think there is a mistake in the derivation where Stimpy showed that is argument is analytically correct, can you point it out?

If you are concerned about Stimpy's arguments, could you respond to them?

CFLarsen

5th May 2005, 09:19 AM

Bumped for jzs.

athon

5th May 2005, 09:28 AM

Originally posted by CFLarsen
Bumped for jzs.

Claus, you do realise that to make the djinni appear, you have to insult him three times. You know, insinuate that he doesn't understand something, or that he's stupid, or that science is as inaccessible to his pea brain as speech is to a goat's.

Otherwise, any calls into the ether quickly fade into silence. I'll bet all the lint in my left pocket that my post will get a response before yours.

Athon

T'ai Chi

5th May 2005, 03:59 PM

Originally posted by athon
:rolleyes:

Whatever. I could not be bothered having this sort of argument.

Oh well. I'm sure your rebuttal would have been marvelous! :)

T'ai Chi

5th May 2005, 04:05 PM

Originally posted by CFLarsen

If the GCP data tends to be improbable, then why are they simultaneously explainable by science?

Ooh! Switching into clausian mode to answer questions with twice as many questions. :)

Don't you think that is a meaningless question? Don't you agree that something can be improbable and also explainable by science? Why is this an issue for you?

What does the analysis say, when they include the whole dataset?

What is "the" analysis you are speaking of? The GCP results for the formal hypothesis registry? Or are you talking about another dataset? What do you mean by an analysis saying something? The statistics? The conclusion of the authors?

Is "assured" different from "guaranteed"?

Could you get a dictionary? And use it? Then report back?

How exactly do you propose that Stimpy look at long sequences of output from the REGs which were made during periods in which it is known that no external influences were present?

Do you believe it is up to me to figure out how to make his critique hard-hitting? Shouldn't it up to him?

Do you agree that if they want their experiments to give meaningful results, they need to make sure the experiment is controlled properly?

Will you ever define the phrase 'proper control'? Why do you think their results are not meaningful? Could you reveal your secret to make "sure" any experiment is properly controlled?

If you think there is a mistake in the derivation where Stimpy showed that is argument is analytically correct, can you point it out?

Don't you think he should point out any serious flaws first?

If you are concerned about Stimpy's arguments, could you respond to them?

Could you repeat your questions again?

:)

T'ai Chi

5th May 2005, 04:33 PM

Originally posted by Stimpson J. Cat

What negative claim am I making?

My claim is that if they do not establish that the output of the REGs will, in the absence of external influences, be sufficiently close to their model, that nothing can be inferred from statistically significant results of their analysis.

But forgot our main disagreements for now.

Could you just reveal to us how any experiment can be absent of external influences as you are demanding?

Thanks. :)

CFLarsen

5th May 2005, 11:24 PM

Originally posted by jzs
Ooh! Switching into clausian mode to answer questions with twice as many questions. :)

Don't you think that is a meaningless question? Don't you agree that something can be improbable and also explainable by science? Why is this an issue for you?

Question not answered.

Originally posted by jzs
What is "the" analysis you are speaking of? The GCP results for the formal hypothesis registry? Or are you talking about another dataset? What do you mean by an analysis saying something? The statistics? The conclusion of the authors?

Question not answered.

Originally posted by jzs
Could you get a dictionary? And use it? Then report back?

Question not answered.

Originally posted by jzs
Do you believe it is up to me to figure out how to make his critique hard-hitting? Shouldn't it up to him?

Question not answered.

Originally posted by jzs
Will you ever define the phrase 'proper control'? Why do you think their results are not meaningful? Could you reveal your secret to make "sure" any experiment is properly controlled?

Question not answered.

Originally posted by jzs
Don't you think he should point out any serious flaws first?

Question not answered.

Originally posted by jzs
Could you repeat your questions again?

Question not answered.

Athon, you are quite right. We can safely conclude that the questions are too difficult and this debate has run its course and it has been shown that this experiment's results have no merit.