I misspoke. I meant to say that most professionals think that GAI is a better indicator of LOG than FSIQ. Both measure potential ability, but ability of different things. It's apparently very common to find gifted kids with high VCI and PRI but average to low average PSI. If you need to use one measure as a "cutoff" for gifted programs, it's recommended to use GAI.

That's not to say the other subtests aren't important to consider. I was reading an interesting study (I forget where but I could find the link) that essentially demonstrated that a higher WMI provides advantages in school, whereas they weren't really able to correlate PSI to advantages.



This topic seems to cover the difference very well: http://giftedissues.davidsongifted.org/BB/ubbthreads.php/topics/112015/FSIQ_vs_GAI.html