baby as natural-born statisticians!

A good article-- by Roger Highfield (of the UK Telegraph)-- on interesting research about babies doing statistical work...

Compelling stuff, but it begs the question that will surely cross the minds of those who try to talk math with students: so when do they lose it (or why do they fail to develop it)?

Actually, this is not so much about math-- as what economists call "statistical discrimination" or what is commonly called stereotyping. In a word, people must make important decisions with limited and costly-to-obtain information. Statistical discrimination occurs everyday-- a type of discrimination in which ALL of us participate-- but some of its most interesting applications occur in labor markets, as people try to figure out who to hire [or not].

Many adults may have a deep set fear of mathematics but it turns out that babies are natural born statisticians, say scientists.

A test of eight month olds has shown how they are able to work out the likelihood of an event occurring, based on their knowledge of what has come before, showing they have a working knowledge of probability and statistics years before they even go to school.

The study, by Prof Fei Xu and Vashti Garcia at the University of British Columbia, did six experiments in which coloured ping pong balls were drawn from a box.

They found that the babies can intuitively predict the colour of ping-pong balls pulled out of a box, based on what they have seen before. "The infants' performance in these studies is impressive," says the team.

The team measured this by studying how long the babies looked at the red and white ping-pong balls as they were taken out of the box: If the box contains mostly red ping-pong balls, the infants looked longer if a mixture of mostly white ping-pong balls were pulled out compared to a mixture of mostly red ping-pong balls. Conversely, if the infants were shown a mixture of mostly red ping-pong balls being pull out, they expected to see the big box containing mostly red ping-pong balls.

The psychologists conclude in the Proceedings of the National Academy of Sciences that "infants possess a powerful mechanism for inductive learning, either using heuristics or basic principles of probability."

As for why many adults feel inadequate about their mathematical abilities, Prof Xu says: "They are bad at probability judgements, but recent work suggests that the story is complicated. Adults can certainly do the simple tasks we present to infants, and there is new evidence suggesting that adults use statistical information to make inferences."...