The thing is, I don't think you can easily draw a bright line with some of these issues. How much teaching/enrichment/stimulation is unfair hothousing? How much, exactly, would taught material have to resemble the specific format and/or content of items on a particular test, to even constitute specific test prep?
And is it really desirable that kids not be taught, just to try to get a more valid benchmark on a test, especially knowing that some parents would always cheat the system? Or should the solution be to try to improve the imperfect assessment tools, rely on more than test numbers from a single test for identification, etc.?
Testing is inherently imperfect, and you can't control what a parent will do in her own home. It would be more workable and fair in my opinion to expose all children to some standardized test prep before a test, rather than try to enforce a lack of prep.
I would agree with your points here, although I would hate to see yet more time spent on test prep in school: it's hard enough to find the time to teach everything....
One of the more interesting ideas I've heard suggests looking at "x" percentage of each demographic tested. That makes sense to me, although the follow up would have to be differentiated with current achievement levels in mind. In my fantasy approach, schools identify and form several types of gifted clusters (each in a different classroom). One cluster would be students who need fully individualized learning opportunities/instruction. These students likely top out both aptitude and achievement testing. Another cluster would be kids who can be instructed as a group, but with an approach and materials that are actually different in both form and content from general education practices/curriculum. These students are near or at the top of achievement and aptitude testing, but aren't quite as far out there and/or are not the kinds of learners who want a completely individualized approach. I'm thinking that the fourth grader who has not been grade accelerated but tops 20 on everything on the EXPLORE test would be typical of the first group; the 16-20's or scattered scores would be typical of this second group. A third cluster would be a combination of bright/mildly gifted students. These students will excel given curriculum that is a 1-2 years ahead and/or with differentiated materials and assignments. However, they can succeed and have needs met with a regular approach, as long as the material is sufficiently advanced. Finally(?), I would envision a cluster that includes kids with high ability indicators relative to their demographic group, but lower than expected achievement. Depending on the specific students, school, etc. this might be two clusters--one in which kids are underachieving due to intrinsic differences which need to be addressed, and one in which kids are underachieving due to lack of an enriched lifestyle. In my fantasy world this also allows teachers at each grade level to learn and specialize in different types of gifted education. Sometimes it seems that even where schools use clustering, they lack an appreciation of the differences between gifted students and create clusters that don't work well together.
I wouldn't want parents to withold what they think will meet the needs of their children, and I recognize that the types of problems on CogAT and other instruments are just plain fun for some of our children. If there was a variety of options for meeting the range of needs on the gifted end of the spectrum, maybe parents wouldn't feel a need to be so cautious about what they share with schools, and prepping wouldn't be an issue at all--it would just be part of the overall picture a school had of a child when trying to figure out which placement was the best fit. I guess what I want is an all cards honestly on the table approach to understanding typicality/atypicality. So many of us have had experiences in at least one subject area where we hear, "oh we have lots of students who....your child will fit right in here". With some of the picture obscured by circumstances that are not readily shared, and with other parts of the picture obscured by instruments or learning opportunities with inadequate ceilings, it is no wonder that this is too often a false statement.
