Remember also that the OLSAT is a group ability test not an IQ test. I don't have it handy right now, but there was a report out a bit back to the effect of what Dottie was saying, that the scores on those types of tests vary a lot in elementary school children who take them more than one. I think that is was called something like "Gifted today, but not tomorrow."
eta: here's a link to that study: http://faculty.education.uiowa.edu/dlohman/pdf/Gifted_Today.pdf Hoagie's also has a link to an older study which indicated that that particular test didn't correlate well with IQ for gifted kids although it did for average kids.
Both the CogAT and the OLSAT publishers state that these tests were designed to measure "developed abilities" not innate ability or intelligence. Despite the amount of stock put in them by schools as an indicator of who is and isn't gifted, I think that distinction is more significant than it is often considered to be. The scores can be significantly impacted by academic placement and what the child has been taught.
For instance, the kids who are high achievers in our local schools get GT pull-out classes for the early years of elementary in which they work on critical thinking skills and puzzles. When those same kids are tested on these group tests along with everyone else who wasn't ided as possibly gifted early on, they have "developed" a lot of the skills tested on these tests and usually score a lot better.
