Another factor is that it's a bit easier to get a high score if you pick your best score out of FSIQ, GAI, VCI, PRI, etc.
and
If you're going to report the highest number achieved out of FSIQ, GAI, VCI, and PRI (those are the Davidson criteria, I think), then you have a greater chance of one of them being over a certain threshold than the chance that just the GAI will be over a certain threshold. It is also complicated by the fact that not everyone gets a GAI calculation - it is properly only added to a test report if there is a significant discrepancy in subscores, which probably means that the reported GAIs skew high. So there are probably more people entitled to run around saying, "I got 160 on my IQ test"* than you would think just on a straight normal-curve calculation.
* This should not be taken as an endorsement of bragging on IQ scores for yourself or anyone else socially.
Agreed. This may skew high. Some may say this entails a bit of cherry-picking through the child's scores to present the highest scores, which may not be the most representative. I'm linking to a related post,
How much error can there be in test scores? While some may prefer the GAI as they believe it to be more descriptive of a child's
intellectual ability, a child's Working Memory and Processing Speed (whose scores are more apparent in the FSIQ) may greatly impact the child's daily
classroom performance and achievement.
Linking to an old thread,
FSIQ vs GAI.
