The term 'intelligence' implies the ability to form concepts, reason abstractly, and notice complex patterns and behaviour. Intelligence Quotient (IQ) tests were devised in 1913 and used to screen out 'educationally subnormal' children. It would appear the reason for IQ tests to diagnose young children as having special needs remains the same today; only the terminology differs.
What exactly do the tests measure? Are they fair? What kind of people do they promote or exclude? Do these tests reflect other abilities, or exclude them?
The testing environment and atmosphere can unfairly alter the results. If a child is tested in an unfamiliar room, as is often the case, their concentration will inevitably be affected. If the tester is unknown to the child, this can affect their disposition. Psychological factors such as a fear of failure, as well as such physical factors as whether the child is cold, has not eaten or is wearing ill-fitting clothes can all have a profound effect.
Then there is the 'focus' of IQ tests. A common theoretical question is, 'If you saw a homeless person and had 1, would you give it to the person or to a charity?' Studies reveal most children would approach the homeless person and give them the money, as it is a more social interpersonal act of kindness. But these children would have 'failed' the question. The correct answer is to donate to charity, as this is a more practical solution. Scores are only achieved for rational answers, not those indicating an emotional response. Empathy is treated with disdain.
A question's wording also matters. Asking a child which piece of string is longest serves no purpose, nor does placing an object out of sight. But add a 'naughty little girl' to the equation and you are likely to receive a more realistic answer, because the child may empathise with being 'naughty'.
If more meaningful problems were used we would get different answers. But not with the way IQ tests are devised.