Tuesday, May 30, 2006

The Probability of Towers of Terror

The RMIT Brain Tumour Inducing Towers received mention on Ockham's Razor, with a discussion by Professor Simon Gandevia, a neurologist from the Prince of Wales Medical Research Institute. Professor Gandevia discusses just how difficult assessing probabilties is.
We thus readily attach a cause to a sequence of random events. Technically, we 'see' or actually 'invent' a positive serial correlation (indicating that one successful shot predicts another).

The cognitive illusion in assessing randomness is that we believe a small sample is as reliable as a large one. This has consequences not just for gamblers. The failure to note the true randomness accompanying a short sequence of outcomes means that scientists are often too easily swayed by the results of small numbers of 'experiments'. As a counter to this cognitive bias, in medicine, clinical trials must include large numbers of participants, alternatively, the results of several 'similar' trials are pooled in what is known as a meta-analysis. This pooling of numbers of participants is a deliberate attempt to counter or dilute the superficial attraction of the initial results of an experiment with small numbers.

Would EoR be surprising anyone here by momentarily reminding everyone of the alternatistas' preference for small trials?
Why do these illusions exist? In the evolutionary world of predator and prey, snap decisions are quite literally vital. It has been argued that because we need time to evaluate probabilities before making a decision, a default system has evolved that rapidly evaluates choices. The Nobel laureate, Francis Crick, is well known for his discoveries about the double helix of our genes, but he later worked in the field of neuroscience. He and his colleagues postulated that humans needed to develop what he termed 'zombie thinking' in order to deal efficiently with the massive sensory input we continuously receive about the external world. This mode of thinking is thus necessary to allow us to react rapidly to external events, so that these cognitive illusions are 'built in' to us, almost certainly for evolutionary reasons. None of us is immune to them, not even those trained as scientists or judges. Our capacity for rational thinking is limited. Propagandists and advertisers are all too well aware of this.

Would EoR be surprising anyone here by momentarily reminding everyone of the alternatistas' 'proof' that 'it works for me'?

Professor Gandevia concludes
The purpose of my talk has been to make you conscientiously cautious in evaluating probabilities. I have also introduced the sometimes harrowing world of cognitive illusions and how we really think. So, while William of Ockham reminds us that conclusions should not (in his words) be 'multiplied beyond necessity', I am reminding you not to 'jump' to them if you have the opportunity for considered reflection.

Would EoR be surprising anyone here by momentarily reminding everyone of the alternatistas' frenzied jumping?


  1. And this guy works for the Prince of Wales Medical Research Institute - the irony.

  2. Actually it's located at the University of New South Wales (http://www.powmri.edu.au/history.htm) and has nothing to do with woo research at all. A sad name however it's part of the Prince of Wales Hospital which has been in existence for far longer than the current homeopathic incumbent.


Note: only a member of this blog may post a comment.