1
Individual iq subtest scores are a more accurate representative of specific cognitive abilities than a full scale iq score is a representative of general intelligence. That is because intelligence is a more abstract concept than say verbal working memory ability so it is more difficult to quantify.
However even the individual sub-test have there limitations because of the nature of the data.
Height, for instance, has a fairly complex genetic basis yet due to its physical nature it can easily be measured for what it is-continuous. However, intelligence cannot be measured continuously even though that would be logical (assuming there are infinite variations of cognitive ability) because it has not been reduced to continous physical units like height has.
Iq is a discrete system as with any written test and that carries with it certain limitations. The bell curve of iq is really an approximation not an actual bell curve from a continuous distribution.
In addition to the concern about representation of data, the weighting of the test can vary significantly. Some tests weight more for processing speed and working memory, disadvantaging individuals with attention deficits who preform as least as well as others on other sub-tests.
As you pointed out the type of cognitive skills valued varies between different cultures and this could be responsible for some of the disparities between groups.
For instance, Inuits have a lower iq than Europeans (in the low 90s if I am not mistaken) but compete well in certain spatial tests. This may be because there is little lead in their barren environment to aid in tracking and direction and those with acute spatial memory would have an advantage.
In certain tests of memory Africans and African-Descended populations are not at a disadvantage to Europeans, but in Spatial tests they tend to underprefrom. East-Asians on the other hand have a strong inclination towards spatial reasoning.
Ashkenazi Jews have average spatio-visual iq, but elevated verbal reasoning abilities.
The variation between sub-tests may very well be greater than the variation between full-scale iq between different groups.
The bottom-line is that iq carries with it inherent flaws, but was intended to be an index that represents intelligence and may be one of the best ways to do so short of basing cognitive ability on a relationship of physical quantities, which may very well be unrealistic.
Bookmarks