Maggie's Farm

IQ? What's that?

No, seriously... what is it a measure of?

Is it conformable across societies or only within context of a single society? Can it be given equally to, say, a Yanamamo and Tunisian and someone in East LA and record a result that accurately describes anything across all of them?

Does it measure 'street smarts' or the ability to analyze situations on-the-fly in complex social settings? Is it truly just 'book learning' or is it amenable to deductive, inductive and synthetic logic applied across multiple realms? Does it measure how to find the right solution to a given situation and stick to it even if the given solution does not yield immediate results?

I really am not joking on these things, as there are very few tests and systemic views of human personality, capability, attitude and outlook that have measurements that can be applied against them that are meaningful across humanity. Without firm and fast descriptive foundation that can have common measurement that is also well known and repeatable across all of humanity, with expected and known deviations via culture which would also be systemically described and measured, our ability to actually create a working science in this field of human personality that is commonly called 'human nature'.

If we can't accurately describe and measure the underlying capability or this thing known as 'IQ' and have repeatable measurement standards that are not only invariant across cultures but time as well, then exactly what does it measure?

Because, in science, if you do not have all the parameters nailed down in an experiment and you conduct the experiment and then find something you did not compensate for in the way of starting parameters you will then have many measurements. That is often not data, with confined knowns and unknowns to be built upon, but a series of readings that may or may not have any meaning at all. Sometimes one can compensate for an uncalibrated parameter. Often one needs to repeat an experiment over again that can take minutes, hours, days, weeks, years or longer.... do a sieve analysis of sediment data, let the sieves filter it out then do a water settling analysis and use Stoke's Law to then measure those and you have data, unless you left a sieve out of the initial screening and you had unwisely dumped them out after measuring them all. Then you have a re-do of that portion... unless you dumped the sample out entirely... then you go back in the field and start over from scratch and compare the two to ensure you actually did get a similar sample and not one miles from where you took the first one...

I looked at this a bit more thoroughly here: http://ajacksonian.blogspot.com/2006/04/when-science-isnt-one-and-why-it.html