Risk and the Paranormal

Do you believe in the paranormal? Are you a good judge of risk?

My guess is that if you answered ‘yes’ to the first question you should answer ‘no’ to the second. And in a recent study that’s just what we found.

The idea is simple and goes back to experiments I did many years ago when I was a parapsychologist investigating telepathy, clairvoyance and other paranormal claims. The more research I did the less convincing the evidence became. I began to think that the reason so many people believe in the paranormal is that they misjudge the probability of ordinary events and end up misinterpreting coincidences as psychic phenomena.

Here’s a common example; you wake up in the morning from dreaming you were driving a red sports car across water. Later that day you see exactly the same kind of car driving through a puddle in the road. Or, more seriously, you dream of a friend who has died and the very next day you hear that it was true. Few people can resist jumping to the conclusion that the dream was a precognition. Yet we know from simple probability calculations that death-dream coincidences are bound to happen by chance to someone, somewhere, many times a day.

So, with my late husband Tom Troscianko, we ran experiments to test the idea that believers are worse at probability judgements, and according to the results, they are.

This is why I was so excited when I heard about Dylan Evans’ new test for measuring risk intelligence (RI). I thought — yes — this is another, better, way to test my hypothesis. If I am right then believers will score lower, and sceptics higher, on his new test.

Dylan defines RI as “the ability to estimate probabilities accurately." His test is online and you can easily try it yourself. It is not a test of how much you know, but how good you are at estimating whether you do know something or not.

This is how it works: there are 50 statements such as “Cats are not mentioned in the Bible," “More than 8 out 10 victims infected by the Ebola virus will die in two days” and “There are more people in the world than chickens." You have to say not just whether they are true, but how confident you are that they are true. So if you are absolutely sure a statement is true you click on 100%, absolutely sure it’s false click on 0%, and no idea at all click on 50% — but the interesting numbers are the rest. If you think it’s probably true you’d click on 70%, 80% or 90% and so on.

If your risk intelligence is high then of the statements you gave 70% to, 70% of them will be true; of the ones you estimated at 20%, 20% will be true and so on. Plotted on a graph these figures give a calibration curve, and this is a straight line from 0 to 100% for someone with very high RI. The figure shows a very good calibration curve and a bad one. Then a special algorithm is applied to calculate the risk intelligence quotient (RQ), which lies between 0 and 100.

I was horrified when I first did the test. Like many people I was over-confident about what I knew and tended to go for extremes. My graph was very far from a straight line and my RQ was far below 100. Maybe that’s why I so strongly believed in the paranormal when I was young! Whether it is or not I’ve learned a lot about myself by doing the test and I hope I’ve improved my RI since then.

But back to the experiment — I told Dylan about my hypothesis and he agreed to include a question about paranormal belief in his massive online survey of roughly 40,000 people. His colleague Benjamin Jakobus wrote the code for the test and others helped with the analysis, and so we were able to find out the relationship between paranormal belief and RI for over 10,000 participants of all ages and from a wide variety of countries.

The results were clear. A quarter of the women believed in the paranormal (25%) and their average RQ was 56. The other three quarters were sceptical and their RQ was higher at 62. The same pattern was found in men. There were fewer believers (13%) and more skeptics (87%) but the believers had an average RQ of 57 and the skeptics 64. These differences were significant for both men and women. So we can safely conclude that paranormal believers have lower risk intelligence.

This leaves many questions unanswered. For example, if we trained people to improve their RI, would they lose their belief? We simply do not know — yet.