New telepathy test, the sequel.

[Michel H]

You’re admitting that you considered results of your test as being correct without knowing for sure if it was a guess or telepathy!?

Yes, of course.

I had written and circled "4" on my paper, and gabeygoat correctly answered 4.

I knew that gabeygoat's answer was correct, but I couldn't know for sure the exact mechanism in gabeygoat's mind which led him/her to write the correct number (though I am fairly confident it was telepathy, I can't be 100% sure).

But that doesn't mean the test failed: I can examine some interesting comments, and I can see if more than 25% of credible answerers (after proper and necessary filtering) gave the correct answer.

 

[Gamolon]

(though I am fairly confident it was telepathy, I can't be 100% sure).

Explain how you know. How can you tell the difference between someone guessing and someone receiving your thoughts.

 

[JayUtah]

...though I am fairly confident it was telepathy, I can't be 100% sure.

What evidence and mathematics did you use to arrive at that confidence? If not 100% sure, then what percentage? And how did you arrive at that percentage?

But that doesn't mean the test failed: I can examine some interesting comments...

Comments are irrelevant. They should not be allowed in the protocol.

...and I can see if more than 25%...

How much more? Do the math for us, please.

...of credible answerers (after proper and necessary filtering) gave the correct answer.

Just no.

There is no "proper" filtering that allows subjective, unblinded decisions to be made about what data to accept. You were challenged to find examples of this anywhere in the psychology literature. If it's as "proper" as you say, then you can certainly find other examples of subjective, unblinded filtration. If you cannot, we have to presume you don't know what you're doing.

The filtration you suggest is "necessary" only because of other flaws in your experiment design. Since you have previously applied it in a clearly one-sided, cherry-picking fashion, we have to conclude that you designed the experiment purposefully to require a step that allowed you to manipulate the data to your advantage. The correct way to design the experiment is to disallow all input except for that which is a valid guess. The null hypothesis will take care of the rest.

Despite your publications in an unrelated field, there is no way you can continue to bluff your way past people who are quite professionally well acquainted with how to do this kind of research with proper scientific controls.

 

[Michel H]

Explain how you know. How can you tell the difference between someone guessing and someone receiving your thoughts.

From the statistical analysis, I knew that strong telepathy effects were apparently taking place during the first tests I conducted on this forum (back then, people were curious and interested, the forum was also more active, and I got many answers).

So it seems reasonable that gabeygoat's answer was obtained through telepathy, but still I can't be 100% sure about this.

 

[Michel H]

What evidence and mathematics did you use to arrive at that confidence? If not 100% sure, then what percentage? And how did you arrive at that percentage?



Comments are irrelevant. They should not be allowed in the protocol.



How much more? Do the math for us, please.



Just no.

There is no "proper" filtering that allows subjective, unblinded decisions to be made about what data to accept. You were challenged to find examples of this anywhere in the psychology literature. If it's as "proper" as you say, then you can certainly find other examples of subjective, unblinded filtration. If you cannot, we have to presume you don't know what you're doing.

The filtration you suggest is "necessary" only because of other flaws in your experiment design. Since you have previously applied it in a clearly one-sided, cherry-picking fashion, we have to conclude that you designed the experiment purposefully to require a step that allowed you to manipulate the data to your advantage. The correct way to design the experiment is to disallow all input except for that which is a valid guess. The null hypothesis will take care of the rest.

Despite your publications in an unrelated field, there is no way you can continue to bluff your way past people who are quite professionally well acquainted with how to do this kind of research with proper scientific controls.

What evidence and mathematics did you use to arrive at that confidence? If not 100% sure, then what percentage? And how did you arrive at that percentage?

This is explained at the end of the second test:

So, I received five answers which seem credible to me, the answers of GregInAustin, Tiktaalik, Loss Leader, gabeygoat and DuvalHMFIC. Out of these five answers, four are numerically correct, and one (the one by GregInAustin) was numerically incorrect. The {correct answer} rate for credible answers in this test is therefore equal to 80%, which is much higher than the "hit rate" from chance alone (about 25%).
In the previous test ( http://www.internationalskeptics.com/forums/showthread.php?p=8607740#post8607740 ), I found a {correct answer} rate equal to 100% for credible answers. In this previous test, I also obtained a {correct answer} rate (for all answers, credible or not) equal to 23.1% (the global hit rate is equal to 22.7% in this second test). These results (100% → 80%, and 23.1 → 22.7%) suggest a small decline in quality, from the first test on this forum to this second one. However, the number of valid numerical answers was much higher in this test (22, compared to 13 in the first test).

It may be interesting to introduce a credibility threshold, equal to CR=5, for exemple. Then, GregInAustin's answer (CR=2) is eliminated, and I obtain 3+4 = 7 ("strongly") credible answers for the two tests (on this forum, so far), all of which are numerically correct. The probability for this is equal to p = (1/4)⁷ = 6.10 x 10^-5 (assuming a 25% probability of answering correctly, for each answer).

.

There is no "proper" filtering that allows subjective, unblinded decisions to be made about what data to accept. You were challenged to find examples of this anywhere in the psychology literature. If it's as "proper" as you say, then you can certainly find other examples of subjective, unblinded filtration. If you cannot, we have to presume you don't know what you're doing.

You can't find exact examples of this filtering in the (para)psychology literature because I am doing new and original research on a very exceptional case.

The naive way of just counting the correct answers (and possibly doing a statistical analysis) promoted by (pseudo) skeptics just doesn't work (one may get about 25% of correct answers in four-choice tests): an additional effort is needed to extract the true parapsychological information (it's not as easy as skeptics imagine it is). Tests results must generally be analysed and interpreted.

 

[JayUtah]

This is explained at the end of the second test

What is the effect of degrees of freedom on your analysis? What was the rate before you edited the findings unblinded?

You can't find exact examples of this filtering in the (para)psychology literature because I am doing new and original research on a very exceptional case.

If this is what you believe, then you have no basis for claiming that your method is "proper." It either has precedent in the field, or you must make a new and original case for its propriety. Do so, please.

You can find many examples of proper data collection for this kind of experiment in the literature. You're by no means the first -- or even the hundredth -- researcher to attempt to prove telepathic ability by having subjects guess at information known only to another subject. Your belief that you're doing novel research that requires a novel method of collecting and "filtering" data is clearly countermanded by fact. There is ample precedent and amply validated methodology for collecting the kind of data you propose to use, and it has no room for "credibility" tests. Since your way do it is blatant cherry-picking, such methods are expressly disallowed in this kind of research.

And no, your credentials as a physicist do not qualify you to conduct this kind of research without input from peers.

The naive way of just counting the correct answers (and possibly doing a statistical analysis) promoted by (pseudo) skeptics just doesn't work (one may get about 25% of correct answers in four-choice tests)

Nonsense. Unless you can demonstrate that people guess the number you've circled at a rate statistically-significantly greater than chance, there is no effect to explain by any hypothesis except the null. This is basic experimental science.

an additional effort is needed to extract the true parapsychological information (it's not as easy as skeptics imagine it is). Tests results must generally be analysed and interpreted.

Try to set aside your hatred for skeptics and imagine that this conversation is how you convince skeptics that you have a method for "analysis" and "interpretation" that actually has an empirical test associated with it, isn't just cherry-picking, and has computable results. You're being asked by several people to describe, in suitable scientific terms, the method you use to determine that "hits" are due to telepathic ability and not chance. All you can tell us is that it's really hard, that your peers can't understand it, and that you're almost, but not quite 100%, sure that it's telepathy. That's not science. That's begging the question.

And you simply can't evade the significance issue. It seems like you're trying to tell us that even though the hit rate may not rise above chance, it's still due to telepathic ability and not chance. That's just special pleading.

 

[Michel H]

What is the effect of degrees of freedom on your analysis? What was the rate before you edited the findings unblinded?



If this is what you believe, then you have no basis for claiming that your method is "proper." It either has precedent in the field, or you must make a new and original case for its propriety. Do so, please.

You can find many examples of proper data collection for this kind of experiment in the literature. You're by no means the first -- or even the hundredth -- researcher to attempt to prove telepathic ability by having subjects guess at information known only to another subject. Your belief that you're doing novel research that requires a novel method of collecting and "filtering" data is clearly countermanded by fact. There is ample precedent and amply validated methodology for collecting the kind of data you propose to use, and it has no room for "credibility" tests. Since your way do it is blatant cherry-picking, such methods are expressly disallowed in this kind of research.

And no, your credentials as a physicist do not qualify you to conduct this kind of research without input from peers.



Nonsense. Unless you can demonstrate that people guess the number you've circled at a rate statistically-significantly greater than chance, there is no effect to explain by any hypothesis except the null. This is basic experimental science.



Try to set aside your hatred for skeptics and imagine that this conversation is how you convince skeptics that you have a method for "analysis" and "interpretation" that actually has an empirical test associated with it, isn't just cherry-picking, and has computable results. You're being asked by several people to describe, in suitable scientific terms, the method you use to determine that "hits" are due to telepathic ability and not chance. All you can tell us is that it's really hard, that your peers can't understand it, and that you're almost, but not quite 100%, sure that it's telepathy. That's not science. That's begging the question.

And you simply can't evade the significance issue. It seems like you're trying to tell us that even though the hit rate may not rise above chance, it's still due to telepathic ability and not chance. That's just special pleading.

I think that you are talking too much and not studying enough.

Read my analyses, for example: http://www.internationalskeptics.com/forums/showthread.php?p=9516155#post9516155
(and, by the way, there is no cherry-picking).

 

[JayUtah]

I think that you are talking too much and not studying enough.

You obviously have no answers for the points I raised. Sounds like I'm not the one who needs to do more studying. Come back when you have the answers, or have published your results. The link you posted exhibits the textbook definition of cherry-picking.

 

[abaddon]

I think that you are talking too much and not studying enough.

Read my analyses, for example: http://www.internationalskeptics.com/forums/showthread.php?p=9516155#post9516155
(and, by the way, there is no cherry-picking).

That link shows you intentionally cherry picking.

 

[Michel H]

That link shows you intentionally cherry picking.

No, I disagree.

By the way, you seemed to admit yourself (implicitly) that I am "telepathic":

Well, Michel's claim is that we we all hear his thoughts. If we all deny that we do, he claims we are lying about it.

How on on earth does he know in the first place that we hear his thoughts?

 

[Crossbow]

No, I disagree.

By the way, you seemed to admit yourself (implicitly) that I am "telepathic":

'Michel H', I do not want to sound mean or spiteful, but I just have to tell you that it is quite obvious that you ARE NOT telepathic.

However, it is quite obvious that you ARE delusional, and as such I urge that you seek professional help as soon as possible before something really serious happens to you.

Thanks much and I hope that you get well soon.

 

[JayUtah]

No, I disagree.

Good for you. The facts speak for themselves, regardless of your opinion. Now try to convince the larger scientific community that your method is not textbook cherry-picking. Put up or shut up.

By the way, you seemed to admit yourself (implicitly) that I am "telepathic":

Nope. He's explicitly asking what is the basis of your belief that you are telepathic.

 

[Michel H]

'Michel H', I do not want to sound mean or spiteful, but I just have to tell you that it is quite obvious that ARE NOT telepathic.
However, it is quite obvious that you ARE delusional, and as such I urge that you seek professional help as soon as possible before something really serious happens to you.

Thanks much and I hope that you get well soon.

You can't even express yourself in correct English, but nevertheless you want to give me mental health advice.

 

[Crossbow]

You can't even express yourself in correct English, but nevertheless you want to give me mental health advice.

Yes. You are correct.

I had a mistake in my posting and I have corrected it.

Thanks much for the correction.

In any event, it is still quite obvious that you ARE NOT telepathic and that you DO NEED help with your delusions.

 

[abaddon]

No, I disagree.

By the way, you seemed to admit yourself (implicitly) that I am "telepathic":

No Michel. I am directly asking you for your basis for the claim that everyone can hear your thoughts.

And your answer is to claim that I am hearing your thoughts. That is circular thinking. Or not even up to the level of thinking.

But the bottom line here is that you can provide no evidence that anyone hears your thoughts. None. Not a jot or tittle. Because you have none to offer.

 

[arthwollipot]

So far I think all current or previous "figures of authority" (Loss Leader and Arthwollipot) have a hit rate equal to 100% in my tests.

So keep it going. Pick another number.

 

[abaddon]

So keep it going. Pick another number.

Don't encourage him.

 

[jsfisher]

Michel H,

What is the end-game here?

Clearly, you are convinced of your telepathic abilities, so there is nothing more for you to accomplish in that regard. Clearly, too, all but one of the members here are equally convinced of your lack of telepathic abilities, so there is nothing more you are able to accomplish there.

What is it you actually expect to achieve?

 

[Gamolon]

From the statistical analysis,

Your data is based on your opinion regarding someone being sincere or not with their answer.

How is that science?

Why does your test have to have boundary parameters as to what your test subjects should be focusing on? If someone was truly receptive to your thought projection, you would be able to concentrate on anything and see if your test subjects got that particular projected thought. Any number, object, etc., would suffice, without telling them anything.

Your test is ridiculous at this point and proves nothing.

 

[steenkh]

So keep it going. Pick another number.

There is 25% chance that you will hit the right number again, which is way too much. Michel will not understand that with such probabilities he will need lots and lots of tests before it becomes reasonably valid, and in any case he will dismiss any guess that is not a hit as not credible.

So no, I don’t think continuing is a good idea.