Mind over Computer 27 RAAF a | «| *& IQ Test Humans are intelligent. (If you disagree with this premise then you're going to have a problem with this argument!) If you talk to a computer and cannot tell it from a human, it must also be intelligent: QED. The logic is sound but somehow feels wrong. It neatly, but irritatingly, sidesteps the whole problem of defining intelligence. In 1912, William Stern devised a method for measuring intelligence in children. He named it ‘IQ’ from the German Intelligenz-Quotient. You may have taken one of these tests at school. The tests consist of a series of abstract reasoning problems that minimize cultural references. For example, you might be asked to look at a set of blocks with dots on them and identify which is the odd one out. Numerous versions of the test have been developed over the years, but nowadays we mostly use one of three standard tests, Wechsler being the most common. Measuring intelligence is complicated. Culture and language play a big part. If we take a tribe of Amazonian Indians and ask them to list the presidents of the United States, they will fail. That does not mean they’re stupid. Drop me into the Amazon Rainforest and I will probably starve to death; they, on the other hand, can live off the land as hunter-gatherers with only a few hours work per day. Who is more intelligent? One problem with IQ is that individual candidate scores can differ wildly from test to test, sometimes by as much as 20 points. That’s huge. At the high end of the scale it can be the difference between being classified as smart or as a genius; and, at the low end, between being average or mentally subnormal. These variations don’t usually matter and most universities and colleges take IQ with a pinch of salt, preferring more specific tests such as SATs in America, the Baccalaureate in Europe or A levels in the UK. IQ can be very important; and is sometimes a matter of HOUSE_OVERSIGHT_015717