A new edition of the beautifully produced RadioLab has just hit the airwires with an excellent programme on the science of randomness.
The hour long science trip largely focusses on how we make sense of random or unpredictable events, from coincidences to statistical white noise.
There’s a wonderful part where the presenters visit statistician Deborah Nolan who has a neat party trick to demonstrate the properties of random sequences to her students.
She asks one group of students to write down the results of 100 coin flips, and another to write a list of imaginary coin flips. She then leaves the room, waits while each sequence is written down, returns, and tells the students which sequence was imaginary.
It works because humans are bad random number generators. Nolan looks for longer runs of heads or tails which are not included in imaginary sequences because we underestimate the variation in randomness.
In fact, there’s been quite a bit of research on how we generate ‘random’ number sequences, and it turns out that far from being a messy and effortless function of the brain, it requires some heavyweight intervention of the frontal lobes.
Brains are very good at stereotyped routines but it’s breaking these learned patterns which takes the real effort. To generate ‘random’ sequences, we need to check we’re not repeating ourselves and match the sequence against a model of randomness in our heads.
In fact, asking people to generate a sequence of random numbers is a good test of frontal lobe function, the more mathematically random it is, the better functioning the frontal cortex. And if we dampen down frontal cortex function using electromagnets, we see a drop in actual randomness of the numbers.
There are plenty more fantastic insights into the science of the unpredictable in the programme with the constantly surprising RadioLab team.
Link to RadioLab on randomness.