It's important to note here that IQ may be fairly immutable...
but IQ isn't "performance."
Tools which measure IQ are inherently flawed for that purpose since they by definition measure some form of "performance."
Because of that inherent factor, there will always be some degree of bias no matter what sort of tool is used.
One could design a perfectly valid tool for assessing 'street smarts' in urban, minority children from low-income homes. Chances are very good that my child, being suburban and fairly sheltered, would not perform well on such a test.
On the other hand, it is very obvious that measuring cognitive ability using test items which rely upon situations that my daughter is familiar with (such as purchasing produce, or a lengthy car trip) would be relatively unfamiliar to that hypothetical inner-city child.
