Debunking the IQ myth

[Jim__]

From MedicalXpress:

...

The study included more than 100,000 participants from around the globe, asking them to complete 12 cognitive tests looking at their memory, reasoning, attention and planning abilities. It found a simple IQ score is misleading when assessing one's intellectual capacity.

...

"In the past, when people tried to examine how intelligence is related to the brain, they generally approached it with an assumption that there is one dominant form of intelligence which is sub-served by a specific system in the brain. What we found is that the brain regions associated with whatever the 'G Factor' is – what general intelligence is – actually housed more specialized systems, not just one," he explained.

"What we did in our study, that's been different than what's been done before, is to try and understand what the structure of intelligence is by considering the way in which the brain is organized into specialized functional systems – that is, when you look at the brain and you see there are different areas that form networks and support different types of functions," he explained.

...

Results from the study found that given a broader range of cognitive tasks, the differences in ability relate to at least three components of intelligence – short-term memory, reasoning and verbal aptitude. These three components combined create an intelligence, or "cognitive profile."

more ...

[Lionel Mandrake]

1. Reading between the lines ...

the IQ myth seems to be that there is a single number (g, or IQ) which underlies all mental aptitudes. This was never very plausible and has frequently been debunked. It is possible for a given test and a given population to apply a mathematical technique called factor analysis to identify a linear combination of the test items that contributes more to the overall variability of results than any other linear combination. This can be called "g", or the principal factor. But this result is valid only for the population tested.

It is also possible to show that IQ (as measured by a standard test, like Stanford-Binet) is very heritable in one population. But this is only a statement about the population, not about the individuals within the population. The conclusion that any test score is an overall measurement of mental aptitude or that it is primarily the result of nature (as opposed to nurture) would be unwarranted.

[napoleon_in_rags]

2. I do a little math.

And I can tell you a thing or two about it.

Suppose you have a 6-side dice. in the language of probability theory, what you have is "random variables" 1-6, (numbers on the sides of the dice) associated with a "probability density function", which ascribes a 1/6 probability to every variable, so a flat line if you plotted it.

But now suppose you have 2 6 sided dice, and your counting what they add up to when you roll them. Now random variables 2-12, and a distribution that looks like a triangle: low values for 2 and 12 (only one way to get each of those, snake eyes or double 6s') but a lot of different ways 2 die can add up to 7, so a peak there for probability.

Now suppose you have 3 six sided die. Random variables 3-18, starting to round up near the peak, flatten near the extremes.

And now suppose you have more. The shape you end up with as you add more die, converges to the "normal distribution", also known as the "bell curve". That's the shape that describes the distributions of IQ's. Yet by scaling the model and calculating the variance, you can always see any bell shaped curve as the sum of a number of independent variables, like counting the number of die that make it up. Thus there is a natural dimension of any bell -shaped distribution. what it represents its the n-dimensional space, where each test subject is evenly distributed in an n-dimensional space.

So what does that mean in practice?

I believe it means there is no such thing as IQ as a singular entity. It means what we call IQ is actually the sum of a number of totally independent variables. What could they be? Nutrition? Childhood inspiration? Its definitely worth finding them out. But until we know those independent dimensions for sure, IQ will always be a weak abstraction.

Peace!