Thursday, December 25, 2008

Neuroscience: Unconscious brain makes best possible decisions

According to a recent Physorg story, the 1970s claim that humans rarely make rational decisions is false:
Pouget analyzed the data from a test performed in the laboratory of Michael Shadlen, a professor of physiology and biophysics at the University of Washington. Shadlen's team watched the activity of a pair of neurons that normally respond to the sight of things moving to the left or right. For instance, when the test consisted of a few dots moving to the right within the jumble of other random dots, the neuron coding for "rightward movement" would occasionally fire. As the test continued, the neuron would fire more and more frequently until it reached a certain threshold, triggering a flurry of activity in the brain and a response from the subject of "rightward."

Pouget says a probabilistic decision-making system like this has several advantages. The most important is that it allows us to reach a reasonable decision in a reasonable amount of time. If we had to wait until we're 99 percent sure before we make a decision, Pouget says, then we would waste time accumulating data unnecessarily. If we only required a 51 percent certainty, then we might reach a decision before enough data has been collected.
I would be interested to know whether this is in part learned behaviour - as I suspect it is. That is, determining how much information is necessary to make a decision is probably partly a matter of experience. It would be interesting to test small children and see.

Labels: ,