True (test behavior). But avoiding overexposure would require giving a different measure of cognitive ability. Some examiners also want to compare cognitive and academic ability statistically, which you can only do with some validity if they are administered within six months of each other.
I would echo the comment about using data from any curriculum-based assessment, progress monitoring probes, etc. that the school may already be collecting. The primary flaw with these is that they usually target grade-level skills (as they are mainly intended to track the growth of below-grade level skills), so they tend to have rather low ceilings, unless out-of-grade-level probes are administered. The advantage is that they generally are better matched to the curriculum of the school, which can be useful when you are discussing grade/subject acceleration.
