Low Contrast High Contrast

Paranormal Statistical Results


One of the things that makes scientists less receptive to supposed evidence of paranormal phenomena is the statistical form of the evidence.  Suppose we lived in a world where sometimes things fell up and sometimes down.  Imagine trying to prove a theory of gravity by saying things should have a 50% chance of falling up and a 50% chance of falling down, but some experiments seem to say things fall down 60% of the time.  There's nothing wrong with trying to understand why the experiments don't seem to show things falling down 50% of the time.  But it's hard to draw real conclusions, especially when some experiments say things fall down 40%, some say 55%, some say 70%...  All the more so if there had been a long history of con artists using sleight of hand to convince people things fall down rather than up.

Statistics certainly have their uses.  In a case like the one above, their main use is to tell us we don't know enough and there is some factor(s) (either in the outside world or in our testing procedures) that is unclear.  If we could identify some factor so that experiments showed that [for instance] when the temperature was over 70 degrees or when it's daylight that things fall down 99% of the time - then we'd have a theory.  If after 100 years of investigation we didn't find any factor that would give us such high consistency, it would be time to stop and wait until somebody had a whole new idea of something to look for.

What can we make of this kind of statistical results?  If I had a theory that green leaves get energy from light and my study of the subject showed that 80% of plants with green leaves are in places where they can get light, it would be an interesting statistic.  But if 20% of plants with green leaves were in caves and other lightless places, it might be reasonable to be skeptical about my theory.


Partly, I wonder what is the point of mind reading, remote viewing and other such things if they only work for some people some of the time.  One would wonder about a species whose eyes or ears only worked 70% of the time.  You'd wonder even more about a species if 80% of its individuals had eyes that more or less never saw, 15% of its members had eyes that saw limited amounts 30% of the time, and 5% of the species' members saw fairly well 80% of the time.  This kind of situation might be understandable if we were talking about a species hundreds of millions of years ago when eyes were first evolving.  And it might make sense if a species had moved to an unlighted environment and its eyes were devolving.  But humans don't seem to be in a comparable situation.  Also, if one believes having more or less paranormal ability today gives an individual a natural selection advantage (more surviving children), then let us see studies supporting that.  Otherwise, this doesn't seem to make sense.


Many of the experiments that supposedly indicate a statistical leaning towards paranormal effects are based on trying to identify one of a small number of choices and using interpretation to decide whether or not the right choice was selected.  I'd suggest something more like the following:

Make a larger number of choices, with well defined differences between them.  This will decrease the number of times the right choice is picked by random means.  By making the different alternatives distinctive, we reduce the chances a wrong answer is caused by ambiguity - a person with paranormal ability has no excuse not to get it right.

We might use photos of individual humans.  Identifying one person from another is a common and important skill.  We should be sure our test subjects have at least normal skill levels in doing this using non-paranormal means.  If paranormal abilities exist, this should be a fairly easy task to perform paranormally.  To make sure the individuals won't be confused, we'll use photos of people who are obviously classifiable into unique groups.  Their ages will be either 5, 15, 30 or 70.  Their races will be Scandinavian, dark African, clearly Middle Eastern or quintessentially Chinese.  Their gender will be clear based on physical features, grooming and clothing.

If there are four age groups, four racial groups and two genders, we can have 32 distinctive pictures of varied people none of whom share the same combination of age, race and gender.  We should not have to show all 32 pictures to the test subject afterward and have him choose.  He should be able to describe the race, gender and age to us.  I'm not asking them to distinguish two teenage Caucasian girls - even though that would not be such a difficult task using non-paranormal means.  If paranormal means can't help us do such a common and simple task of correctly identifying age, race and gender 2/3 of the time what is the point of this supposed ability?  What useful information can be passed paranormally if this cannot?


Suppose I tell you I'm selling telephones that wouldn't let you tell the person on the other end any clear, specific details of things you were talking about.  If the other person got any impression at all of what you were talking about, it would only be indistinct dream-like impressions.  Afterward, the other person would have to be shown a few pictures and try to figure out which one represented what you were trying to describe.  The other person is unable to pick the right picture as much as 90% of the time, but they do pick the right picture a bit more often than you'd expect by random chance.  Can I sell you one of these phones?  Well, maybe that's why it's hard to sell scientists on the idea of paranormal phenomena.

Suppose a person had the ability to get these vague impressions.  In the real world, nobody is going to show him five pictures and tell him that one of them shows what his vague impressions were about.  All he's going to have is vague impressions.  What is he going to do with that?

Let us assume we have a test subject with the intellectual skills so that if he did get vague impressions paranormally he could draw correct and useful conclusions from those vague impressions.  Those same intellectual skills might also allow that test subject to draw conclusions from subtle details in the testing environment that would let him guess the right choice more often than by pure chance.  If the subject does not have such intellectual skills, the kind of nonspecific impressions that most supposed paranormal experiences involve would not be of much use to him in the outside world.