[fauigerzigerk]

There's something I keep wondering about whenever I see people draw any conclusions whatsoever from IQ alone: How can it not matter whether a particular score was achieved by being equally good at all tasks or by being very good at some and very bad at other tasks?

I would expect a totally different outcome to many IQ related studies if that were taken into account. Or maybe it is actually taken into account but never reported, I'm not sure.

[Daniel_Newby]

Empirically, the score on the whole IQ test is highly correlated with the score on any subset of the questions. There is a difference between scores on verbal/symbolic and visual/spatial questions that depends on the testee, but it is usually modest except in cases of severe brain damage or malformation.