Merged Telepathy test: which number did I write?

[IceSage]

No.

It proves that out of 20, 15 people randomly choose the number you picked. Increase the number of people by a 10 fold, a 100 fold or even a 1000 fold, the number of the answers would start to approach each other.

Roll a die six times, what results do you get? Roll a die a thousand times, what results do you get now?

Also, each die roll is independent of the previous roll, i.e. a die does't remember.

Whereas on a forum, the posters could be influenced (unconsciously) by posts made by previous posters.

Pretty much these answers.

Also, say for example everyone here ended up guessing your number. Everyone. Say we all just guessed 4. Nobody said anything else, we all just said 4.

Your number was 4.

That's 100% of the people here guessing your number right.

That would still not prove telepathy was involved.

Say we did it 5 more times, and we all guessed the same thing 5 more times.

Still wouldn't prove telepathy.

 

[Dave Rogers]

One disadvantage of polls, however, is that some people might answer in the poll without writing any comment, which would make it difficult (or even impossible)for me to assess the credibilities of their answers.

Well, IceSage, in order to be able to draw a conclusion, a statistical analysis has to be made. For example, if I say that I wrote one of the five numbers: 1, 2, 3, 4, 5 on my sheet of paper, and I wrote "5" (but this I don't say, you have to guess it), and I get 20 answers, out of which 15 are correct, when I would expect about 20/5 = 4 correct answers without telepathy, I would say that telepathy has been proved at a probability level given by the binomial distribution[...]

It should be obvious to you that there is a disconnect between these two statements. If you're basing your conclusion on statistical analysis, then you have no business assessing the credibilities of the answers in the first place. The big advantage of a poll is that it makes it difficult, or even impossible, to introduce your own personal bias by "assess[ing] the credibilities of the answers," because, by preventing the association of a comment with a particular choice, it reduces the answers to a pure choice with no confusing factors. Credibility is a purely subjective judgement, which can only skew the results. The whole point of double-blind studies is to eliminate that sort of subjectivity. This is all very basic stuff, which I think you need to understand a bit better before you can construct a valid test.

Dave

 

[dlorde]

... I would say that telepathy has been proved at a probability level given by the binomial distribution (see http://stattrek.com/online-calculator/binomial.aspx ) as p = 1.80 E-07.

You'd be wrong. It wouldn't prove telepathy, it would simply show the probability of getting that result by chance. That's not the same thing at all. It doesn't tell you that telepathy caused the result.

If you had done a well designed and controlled experiment with a large sample size and made a correct statistical analysis, you could consider it good supporting evidence.

 

[Ladewig]

I would say that telepathy has been proved at a probability level given by the binomial distribution (see http://stattrek.com/online-calculator/binomial.aspx ) as p = 1.80 E-07.

If I may be serious for a moment.

Even if something supernatural is at play, how do you know the results are from telepathy? Maybe the results are from precognition. Maybe the results are from your guardian angel talking to other guardian angels. Maybe it is mind control and these 15 people (consciously or unconsciously) made you pick that number.

Followup question: why did you decide to have only four numbers in your "test"?

 

[Ladewig]

I see a number between 1 and 4. Does that make sense?
The letter ''R'' is appearing too. Or ''M''.
The number is still living. I feel it is somewhere near water.
A man, an older man is coming through. Something wrong in his chest or his head. Do you understand?
The number wants you to know that it is happy and in peace.

I owe you an apology.

I did not see your post before I constructed post #47.


ETA: wait, maybe I have it backwards, maybe you read my mind before constructing your post and you are the plagiarist.

 

[Michel H]

One disadvantage of polls, however, is that some people might answer in the poll without writing any comment, which would make it difficult (or even impossible)for me to assess the credibilities of their answers.

I like this. So the only credible answers, as judged by you, will be the ones that got the right answer, therefore you will claim 100% success?

Norm

Not necessarily, fromdownunder, because my credibility assessment is based ONLY (at least in principle, it's the ideal situation) on the comment(s) (and, to a lesser extent, to some other elements, like the name, the avatar, even possibly previous contributions and so on), and everybody can verify it and is welcome to make constructive criticism.
An example of an answer I would consider credible (to a question: "did I write "1", "2", "3", "4" or "5" ? " ):
"Good question. I think I can tell you the answer is "5" , and I think I used telepathy. Good luck in your project ".
An answer I would regard as not-credible:
" 5, I don't believe in telepethy. Also, I think yu shoud larn more about experimentel design before yu do telepethy tests. hhhhhhhhhh."

 

[dlorde]

An example of an answer I would consider credible (to a question: "did I write "1", "2", "3", "4" or "5" ? " ):
"Good question. I think I can tell you the answer is "5" , and I think I used telepathy. Good luck in your project ".
An answer I would regard as not-credible:
" 5, I don't believe in telepethy. Also, I think yu shoud larn more about experimentel design before yu do telepethy tests. hhhhhhhhhh."

Why?

Is telepathy only credible in people who believe in it, can spell, think you're doing good work, and wish you well?

 

[Michel H]

An example of an answer I would consider credible (to a question: "did I write "1", "2", "3", "4" or "5" ? " ):
"Good question. I think I can tell you the answer is "5" , and I think I used telepathy. Good luck in your project ".
An answer I would regard as not-credible:
" 5, I don't believe in telepethy. Also, I think yu shoud larn more about experimentel design before yu do telepethy tests. hhhhhhhhhh."

Why?

Is telepathy only credible in people who believe in it, can spell, think you're doing good work, and wish you well?

Well, I think I have observed these people provide better (numerical) answers. See this analysis for examples: http://users.telenet.be/mhanck/analysis_jref.pdf , and also this thread: http://www.internationalskeptics.com/forums/showthread.php?t=237268 . It seems reasonable to assume that people who are aggressive and hostile in their comment will also be so in their (numerical) answer.

 

[Agatha]

I think you need to learn a great deal more about statistics and test design before you can draw any meaningful inference from the results of this (non) test.

 

[Agatha]

OK, I've read the pdf to which you linked. You need to learn more about test design and statistics, and also about the importance of not allowing subjectivity to skew your results.

Once you discard a 'wrong' answer - or a 'right' one - from your analysis, you are making your test meaningless. You have to take account of all the answers, and let the chips fall where they may.

Whether you are prepared to admit to it or not, by introducing so-called credibility you are falsifying the test results to get closer to the answer you want.

 

[Jodie]

three

 

[Ladewig]

It seems reasonable to assume that people who are aggressive and hostile in their comment will also be so in their (numerical) answer.

Even if that were true, why would predicting a number in a hostile manner preclude ESP?

 

[devnull]

I'm rather surprised so many people are playing along... Well, as far as even responding to this post goes.

Even if people did guess the number, what does this has to do with telepathy? Guessing a number would be just that... a guess.

But..... Im not guessing. Im psychic, and Im positive the number he wrote was 4.

 

[Michel H]

It seems reasonable to assume that people who are aggressive and hostile in their comment will also be so in their (numerical) answer.

Even if that were true, why would predicting a number in a hostile manner preclude ESP?

I don't think it precludes ESP, Ladewig. What I suppose is that these people knowingly answer incorrectly. This is very common ,I think. For some people, putting aggressive elements in their answer may also be a way to "send a warning signal" about the incorrectness of their answer, and therefore "a certain way" to be honest.

 

[Akhenaten]

I don't think it precludes ESP, Ladewig. What I suppose is that these people knowingly answer incorrectly.

This sounds a lot like you're saying that people know that they have ESP but they pretend not to.

Do you think they do this just to mess up your carefully thought-out test procedure or for some more nefarious purpose.

This is very common ,I think. For some people, putting aggressive elements in their answer may also be a way to "send a warning signal" about the incorrectness of their answer, and therefore "a certain way" to be honest.

Is "banana" an aggressive element? If someone wrote the correct answer but it was in a big, red, angry-looking font would it be rejected for being too aggressive? Are large numbers more aggressive than small numbers?

 

[alfaniner]

A sufficiently large value of 1.

 

[Ladewig]

I don't think it precludes ESP, Ladewig. What I suppose is that these people knowingly answer incorrectly. This is very common ,I think. For some people, putting aggressive elements in their answer may also be a way to "send a warning signal" about the incorrectness of their answer, and therefore "a certain way" to be honest.

I suppose that may be true, but isn't it unfair to discount their answers without knowing for certain that the hostility is evidence of an incorrect answer?


I fear you may not like hearing one more criticism of your test but your deciding which answers should be dismissed or disqualified while you, yourself, know the correct answer means your format is of very little value. If you are going to judge which answers should be included and which answers should be excluded, then someone else has to select and document the target number.


If a doctor proposes a new treatment and constructs a test where some patients are given the treatment and some are not, then it should be obvious that if the doctor knows who received the placebo and who did not then he is patently unqualified to evaluate the success rate of each individual patient. And he certainly should NOT be deciding which patients should be dropped from the study and which should be included.

 

[Andrew Wiggin]

I am lost at this point.
Are you saying that -7 and pi are not the same number?

They are, for certain values of pi.

 

[IceSage]

I'm sad, nobody at the JREF told me if my block of cheese answer was right.

I'm pretty sure it was though!

I swear today's answer was a block of cheese...

 

[Pixel42]

If a doctor proposes a new treatment and constructs a test where some patients are given the treatment and some are not, then it should be obvious that if the doctor knows who received the placebo and who did not then he is patently unqualified to evaluate the success rate of each individual patient. And he certainly should NOT be deciding which patients should be dropped from the study and which should be included.

I don't know exactly what Michel is trying to establish with her experiment, but certainly the poor design and small sample means she cannot hope to obtain meaningful results.

I'm also puzzled as to why she feels it necessary to do this experiment herself when the results of larger, much better designed experiments are freely available for examination. Not that I'm against doing your own experiments, quite the reverse, but they need to have at least some chance of producing meaningful results to be worth doing.