Often we hear of stunning results from paranormal experiments, purportedly “proving” that paranormal phenomena exist. These experiments have a tendency to enter what could be called the Hodge Podge File Cabinet of Evidence of the Paranormal, this huge body of anecdotes, usually impossible to verify, and never repeated in a sound, scientific manner. Most often, these experiments are done by believers, eager to find evidence. One such example can be found in the Swedish Soul Travel Magazine, dedicated to Out-of-Body Experiences and the like, The magazine ran an article in 2003 about an informal experiment, where they tested two persons for telepathy.
There were two persons, lying on the floor, with one acting as sender and another as receiver. The roles switched between guesses. The sender “mentally projected” an image of a randomly chosen animal onto an imagined “screen” in the ceiling. The receiver tried to psychically “see” what animal it was.
It should be emphasized that the experiment is not claimed to be scientific, but the magazine does point to it as “private evidence”. But it is experiments such as this that are often referred to – glowingly – by believers and usually followed by a “Explain that, Skeptics!” comment.
But when skeptics investigate these experiments, they invariably find that the findings touted as “evidence”, be it of more of less “private” value, are nowhere near evidence.
Let’s take a look at this one…
The result was as follow:
|Sent image||Received image|
|Moose||Deer, owl (saw both)|
|Dragonfly (silhouette)||Albatross (silhouette seen from below)|
It was interpreted as a very successful experiment considering it was the first time it was tried and the participants had not had any training in sending images. With such a large number of animals in the world it was believed to be highly significant that three out of twelve tries were straight hits and three more were very close. Let’s remember that the test was not presented as a scientific test but the author (an experimentalist) wondered how it could be so difficult for scientists to design a good experiment for telepathy considering how easy it was to get good results.
One of the first things one may wonder about is how good is three hits? The number of animal species on the planet is huge but how many are the average person likely to think of? I have been told that one prominent biologist was once asked what question he would like to ask God, assuming that God existed. What he wanted to know from God was why he/she/it was so fond of beetles: There are an abundance of different species of beetles, by some estimates between 250,000 to 350,000. Should the three hits be counted against all different beetles or are beetles only one possible option? It is impossible to know how good a result is if we cannot calculate the probability of getting the same result by chance.
A scientist would probably not like to have this very large and undefined set of possible targets that the animal kingdom is. Instead, one would construct a set of say ten animals, fairly distinct and well known to both sender and receiver. This would make it possible to calculate how likely a certain number of hits would be. It would also make the result easier to interpret since a hit would be either correct or not, nothing could be almost a hit. This is because the receiver would know the ten possible animals the sender would concentrate on and he/she will just answer with one of them instead of some kind of description.
The problem with descriptions
The problem with descriptions is that they need to be interpreted and interpretation is something one should strive to avoid. One of the close hits above is “furry mammal” but if furry is to mean the same as in the case of a rat – an animal with fur – then no less than six of the twelve animals can be described with those words. This is the problem with a too general description.
To count albatross as a close hit for a dragonfly because of the likeness of their silhouettes would probably surprise a bird watcher since they are used to identify different birds in the sky from their silhouette. What is to be determined a close hit depends on the reference system of the participants. Using only a limited set of animals well known to the participants will remove these possible ambiguities, there would be no need for interpretation and to calculate the possibility for getting a hit by change would be easy. For ten different animals the possibility to get a hit by guessing is 1/10.
As can be seen from the table above the sender always thought of a new animal for each run in the test. It does seem to be the obvious way of doing things but there are a few things that complicate matters. If, as suggested above, only ten different animals are used the test can only be done ten times before there are no more animals left. Obviously one needs to use the same animal more than once. Everybody should understand that as sender it is then important to change the order of the animals in a random way.
The problem with randomness
The problem with randomness is that people are not very good at it. If a dice is thrown five times the result 2, 4, 5, 1, 3 is perceived as more random to us than 1, 1, 2, 2, 1 but the possibility for each of the results is exactly the same. Thus it would not be a good idea for the sender to choose what animal to send since the sender is likely to behave in a non-random way and the receiver is as well. It would be even worse if the two participants communicate while the test is running. A simple dialog could go as follows:
Sender: Are you ready?
Receiver: Sure go ahead.
Sender: Hear it comes…
Sender: Did you get that?
Receiver: Yes, it was one of the birds, the owl right?
Sender: No, it was the koala
This might not seem to be so serious, the sender never tells the receiver anything about the target before the test run is over and in this example the receiver does not even get the target right. But the information that has been transmitted might have an effect on the next run. Even if the choice of the target is done randomly, and not by the sender, it is unlikely that the receiver (if he/she is guessing) will choose the Koala in the next run. And if the selection of target is not random it is unlikely that the sender will choose either the Koala or any of the birds. It could be said that if telepathy really works it does not matter if the receiver is told about what is a hit and what is not. But the test is done to see if the receiver does better than chance or not, so it would be wrong to make any assumptions.
A better experiment
So far we have found that a practical experiment could be conducted as follows: The receiver and the sender agree on ten animals that they consider distinctly different so that they should have no problem to see which is which. Then they get a ten sided dice which the sender should roll each time to decide what target to send. Before sending
the target the sender makes a note about what target he/she is sending and the receiver than makes a note about what target is being received. They should not communicate in any way until they compare their notes after the experiment is finished. In fact they should not be able to communicate at all if the experiment is to be scientific.
People communicate in many different ways not only by talking or writing. Especially, if they know each other body language can communicate a lot. There are also other things that can influence the experiment. The time it takes the sender to write ‘hippopotamus’ is a lot longer than what it takes to write ‘bat’. It is not too bold to assume that the receiver might notice the difference, at least subconsciously, if the experiment is not conducted in a way to avoid all communication. But if the participants are not to communicate (besides by telepathy) it will be a problem for them to know when to send and when to expect to receive the images.
This can be solved either by limited communication, as a button at the sender side and a lamp with the receiver that the sender can turn on when he is sending, or by a third party, which will tell the participants what to do and when to do it. But both techniques are open for mistakes and actual cheating. For example if sender and receiver synchronise their watches the sender can turn on the lamp for the receiver when the clock show a time that corresponds to one of the ten targets.
Of course it is not so that everyone participating in a test of telepathy is cheating, it might even be that no one ever is cheating in these tests, but still it is best to remove all possibilities just to be sure. To solve the problem of knowing when to send and when to receive it is actually possible to use the same synchronised watches as the cheaters in the example above used. If for example the sender is to send every even minute the receiver knows when to concentrate and when to relax. There are many other ways to solve the problem but the most important is to be aware of it.
It is safe to assume that if two people want to test their ability to send and receive telepathic messages for their own amusement they will not try to cheat. Still they may be deceived by their subconsciously communication or failure to avoid the other pitfalls described above.
Humans: Good at fooling, better at fooling ourselves
The fact that we humans are so good at fooling ourselves is unfortunate; it makes it very complicated to perform good experiments without understanding all the different ways the experiment can go wrong. The example above was not a scientific experiment and gave a positive result, but as the gaps in the protocol is plugged by scientific methodology the positive results will get less and less significant until it will finally disappear. At least that is what I predict. For most sciences it works the other way around, the more precise you make your measurements the more significant will the result be. A skeptic will not rule out the possibility that telepathy works but it must be shown in such a way that all other explanations are less likely. Soul Travel Magazine