The book "Nurture Shock" has a chapter on this and it's very interesting. Basically the research shows that IQ scores are not stable in children in preschool/early elementary. So a child who is tested at 5 could earn a gifted score, but 3 years later they may be tested again and not score gifted anymore. Or vice versa.
Did they include data on the score variability? Because this makes total sense in a lot of cases, but for extreme outliers on the scale it would seem to be a poor excuse for not testing.
I'm thinking here of children with LDs that would benefit from early intervention and HG/PG children who are waaaaay out of band and not benefiting from school. It seems cruel to make them wait. Not to mention that I always wonder if high LOG kids drop (or "even out" as the premise goes) because they have to spent 4 years of school (K-3) being completely underestimated and not learning anything.
Our previous district tests in like 2nd or 3rd (I don't remember) but they don't do anything with the data anyway, so I don't see why they bother.
