As a pure example: A sub-section may have 40 progressively harder questions, an eight year old may hit the ceiling after getting 23 correct generating a scaled score of 19 (which is 3 standard deviations above the mean.)
A fourteen year old may get a scaled score of 19 with 38 correct.
It is also possible for an eight year old to get 38 correct. And in this case, neither kid may have reached the discontinue criterion (the point where several items are answered incorrectly and the subtest is stopped). However, the impact of this type of ‘ceiling’ (running out of questions) increases as a child gets older. If the extended norms are used, the standard score for the eight year old may increase from a 19 to a 28 (the highest possible standard score using the extended norms), but the standard score for the fourteen year old may only be able to go up to a 20 with the extended norms and may even stay at a 19 on certain subtests. There just aren’t enough questions on the WISC to extend the norms above a 19 after a certain age. So while the IQ scores of both the 8 and 14 year olds would be considered underestimates if the discontinue criterion were not met (both children may have been able to answer more questions had there been more available), you’d expect the underestimate to be much more severe for the fourteen year old.
I think this is one reason Linda Silverman from the Gifted Development Center believes that ages 4 - 9 is the best time to test gifted children (see
http://www.gifteddevelopment.com/PDF_files/problems.pdf). The extended norms can be used to differentiate HG, EG, and maybe even PG kids in this age range. But once kids are older, there just aren’t enough questions on the WISC to measure the full range of abilities.
