New telepathy test, the sequel.

Reformed Offlian said:
I have a better question: if everybody can hear Michel H's thoughts, why does he even post? He could just think what he wants to say and everybody will know it. The answer "for the convenience of other posters" doesn't work, because they all heard his thoughts already, too.
Click to expand...
I don't think Michel H has ever claimed that all people hear all his thoughts all the time.
 
arthwollipot said:
I don't think Michel H has ever claimed that all people hear all his thoughts all the time.
Click to expand...

On the contrary, I recall him saying the opposite. If the communication were synchronous, it could be tested with synchronicity as part of the protocol. He claims the communication is asynchronous so that he can expand the range of "hits" to those that happen anytime, anywhere. As far as I can recall, he has never proposed how that can happen, which seems to be something that a person who claims credentials in physics would know how to do.
 
JayUtah said:
On the contrary, I recall him saying the opposite.
Click to expand...
How? Telepathically?

But no, I don't think so. It seems to me that what arthwollipot said:
arthwollipot said:
I don't think Michel H has ever claimed that all people hear all his thoughts all the time.
Click to expand...
is reasonable. It seems obvious that most people need their own thoughts far more than mine, and I suppose there are some protective systems.

In my test, I have a simple message: one of the four words: "2022", "Russia", "attacked" and "Ukraine", that I repeat many times in various ways: I read it on my paper, I say it, I ask the mysterious (probably telepathic) voices in my head to say it (and they usually accept and cooperate on this).

Now, because this test is more than one week old, I repeat the right word less often.

So, I have a simple message that I repeat many times, as opposed to a test where I would send a complex message just once (it is likely that people can understand some complex thoughts that they receive telepathically too, but I don't use that much in tests).
 
Michel H said:
In my test, I have a simple message: one of the four words: "2022", "Russia", "attacked" and "Ukraine", that I repeat many times in various ways: I read it on my paper, I say it, I ask the mysterious (probably telepathic) voices in my head to say it (and they usually accept and cooperate on this).
Click to expand...
Please understand this. If telepathy was non-existent, and I guessed completely randomly, there is a 25% chance that I would be correct through nothing but chance. We require better odds in order to be convinced. This is why nobody is doing your test.
 
arthwollipot said:
Please understand this. If telepathy was non-existent, and I guessed completely randomly, there is a 25% chance that I would be correct through nothing but chance. We require better odds in order to be convinced. This is why nobody is doing your test.
Click to expand...
But many people can (potentially) answer, and many tests can be done.

In addition, I should pay attention to the various comments (especially if they are quality ones).

I have been doing many online telepathy tests for many years (I first reported on a telepathy test done elsewhere, on this forum, in 2012), and I generally did not find that increasing the number of choices led to better results.
 
Last edited:
Michel H said:
But many people can (potentially) answer, and many tests can be done.
Click to expand...
The problem is that if you've got more than 4 people answering, you've pretty much guaranteed at least one successful guess even if telepathy is not real. The more people take your test, the more false positives you'll get. Out of 100 people guessing purely at random, on average 25 of them will get it right by chance alone. Such a test is not more powerful the more people take it. It is only more misleading.

Michel H said:
In addition, I should pay attention to the various comments (especially if they are quality ones).

I have been doing many online telepathy tests for many years (I first reported on a telepathy test done elsewhere on this forum in 2012), and I generally did not find that increasing the number of choices led to better results.
Click to expand...
The purpose of a test is to find the truth, not to get "better" results. If the results of a reliable test do not show evidence of telepathy, we are obligated to accept that, not to develop a new test that does.
 
Michel H said:
But many people can (potentially) answer, and many tests can be done.

In addition, I should pay attention to the various comments (especially if they are quality ones).
Click to expand...
More tests only make for more accurate results if the test protocol is rigorous. Yours is not, precisely because you "pay attention to the comments". As soon as you introduce an element of subjective judgement you render you tests worthless. As far as I know you have only ever done a single valid test run.

I have been doing many online telepathy tests for many years (I first reported on a telepathy test done elsewhere, on this forum, in 2012), and I generally did not find that increasing the number of choices led to better results.
Click to expand...

Better (i.e. more reliably accurate) results come from a rigorous test protocol and more data. The latter can be achieved with either lots of runs with a few choices or fewer runs with more choices, it makes no difference whatsoever to the statistical significance of the result. It's the former where you fail abysmally.
 
Michel H said:
... I ask the mysterious (probably telepathic) voices in my head to say it (and they usually accept and cooperate on this).
[...]
(it is likely that people can understand some complex thoughts that they receive telepathically too, but I don't use that much in tests).
Click to expand...

You're speculating on the nature of phenomena which show every sign of being entirely imaginary. Since no type of telepathy has ever been shown to exist, despite all attempts to show they do, there is no justification for describing any aspect of telepathy as either "likely" or "probable".
 
arthwollipot said:
The problem is that if you've got more than 4 people answering, you've pretty much guaranteed at least one successful guess even if telepathy is not real. The more people take your test, the more false positives you'll get. Out of 100 people guessing purely at random, on average 25 of them will get it right by chance alone. Such a test is not more powerful the more people take it. It is only more misleading.
Click to expand...
If 100 people participate in one of my four-choice telepathy tests, and if 25 give the correct answer with no special (informative) comment, this may mean that no evidence for telepathy has been found in that specific test.

However, one has to be careful because it is also possible that (some) people are hell-bent on hiding the truth.

Some dishonest pseudo-skeptics for example might deliberately give wrong (or even right) answers so the final number of correct answers comes out exactly equal to 25. It is possible that a serious analyst would find that about 40% of the participants who are known for being honest, educated and reliable actually gave the right answer, even if they left no comment.

arthwollipot said:
If the results of a reliable test do not show evidence of telepathy, we are obligated to accept that, not to develop a new test that does.
Click to expand...
A test can fail for various reasons, for example because nobody wants to take part, or because people don't want to cooperate seriously and honestly, or (in a related way) because it's been tried on a skeptics' forum (where members, posing misleadingly as great and very rigorous scientists, are actually scared by the prospect of providing evidence for telepathy, or simply dislike the idea), or simply because extra-sensory perception does actually not exist.

Failure of one test does not prove that telepathy does not exist.

We don't live in a world of saints. If we did, people would probably not be spending so much time and energy trying to kill each other (like in wars).
 
Last edited:
Pixel42 said:
More tests only make for more accurate results if the test protocol is rigorous. Yours is not, precisely because you "pay attention to the comments". As soon as you introduce an element of subjective judgement you render you tests worthless. As far as I know you have only ever done a single valid test run.
Click to expand...
I see no reason why serious telepathy tests should be based only on statistical analyses, while disregarding all comments. If participants (especially those who gave the correct answer, which they couldn't know using "normal", sensory means) show an obvious tendency to make highly unusual comments, like "I knew it, I saw it, I heard it, the little voice told me and so on", I believe I am entitled to view this as evidence which confirms my hypothesis in this psychology research. Serious testimonies do matter, of course.
 
Michel H said:
A test can fail for various reasons ...
Click to expand...

One major reason is failure to exclude the tester's bias. You already know this. Cherry-picked results are completely worthless as you already know, even if you promise yourself you're going to try to be fair and rigorous.

That's why I previously suggested checking to see if people hear psychically-transmitted jokes. I was perfectly serious. Your fundamental problem in test design is the nagging belief that perhaps somehow literally everyone has agreed to pretend they can't hear you. So, use a test where their response is involuntary.

Watch a live event on TV with lots of crowd shots (tennis match? cycle race?) and watch a funny movie or standup comedian on a laptop at the same time. Do the crowd at the sports event spontaneously and inappropriately roar with laughter when you do? If not, why not?
 
Michel H said:
I see no reason why serious telepathy tests should be based only on statistical analyses, while disregarding all comments.
Click to expand...

Statistical analysis is absolutely vital for testing a claim that is intermittent. Chance is also intermittent, and what better way to separate chance results from meaningful results, that resists researcher bias?
 
Michel H said:
But many people can (potentially) answer, and many tests can be done.

In addition, I should pay attention to the various comments (especially if they are quality ones).

I have been doing many online telepathy tests for many years (I first reported on a telepathy test done elsewhere, on this forum, in 2012), and I generally did not find that increasing the number of choices led to better results.
Click to expand...

OK. So let's reduce the number of choices. I'm thinking of one word, WALRUS. What word am I thinking? "Walrus," you say. 100% proof of precognition . . . or is it telepathy? I'm scared to reduce the test to no words at all for fear the Universe might implode. :scared:
 
Michel H said:
I see no reason why serious telepathy tests should be based only on statistical analyses, while disregarding all comments.
Click to expand...
Because it's the only way to ensure you aren't fooling yourself. People manage to fool themselves into believing all sorts of things that aren't true - homeopathy, dowsing, astrology - by trusting their subjective judgement rather than objective data.

If participants (especially those who gave the correct answer, which they couldn't know using "normal", sensory means) show an obvious tendency to make highly unusual comments, like "I knew it, I saw it, I heard it, the little voice told me and so on", I believe I am entitled to view this as evidence which confirms my hypothesis in this psychology research.
Click to expand...
Then you believe wrongly. Your inability to tell when people are making fun of you is your problem, no on else's.

Serious testimonies do matter, of course.
Click to expand...
Rigorously obtained and analysed data is all that matters.

Note that this kind of test protocol is designed to test an hypothesis - in this case the hypothesis that one person's thoughts can telepathically affect another's - in a way which excludes known sources of error, such as confirmation bias. A higher hit rate than would be expected by chance would indicate respondents' guesses were not as random as they think they are, that their thoughts are somehow being unconsciously influenced by the sender.

If you seriously think there is a possibility that some of those responding are deliberately giving the wrong answer - that they actually know perfectly well what the correct answer is, they aren't simply guessing, and are choosing to give the wrong one - then this type of test protocol is simply not suitable.
 
You must log in or register to reply here.

Back
Top Bottom