Yet I've met many educators (not my friends; people on many committees and panels I've been on) who DON'T think the way you mentioned. I've been horrified at some of the stuff I hear, such as:
I agree this is a serious problem.
It is just the kind of argument that you make above that leads to racial stereotyping in schools that discriminates against individual kids. There is a well documented phenomenon of average characteristics of a group being applied inappropriately to individuals. The strategies schools use for identifying gifted kids are far from perfect in this regard.
I'm interested; can you provide references for this?
There is a large body of scholarly research on this - you can find some of it by doing a search on "unconscious bias."
AFAIK, all kids get the same test for giftedness.
Systematic testing may happen in some school districts, but not ours. Some (very little) testing happens, but only as a result of parent or teacher recommendation. Being a squeaky wheel matters.
It seems reasonable to set the cutoff as a percentile rank for an area (say, per school or small group of neighborhood schools where results are similar, rather than all of New York City). Everyone who scores above some cutoff (~97th percentile?) is admitted to a local gifted program. The ones who are past the 99th should get services that are different from the ones in the 97th-99th.
Here are the arguments I've heard that say this strategy is unfair: the tests are not a perfect measure for what they're being used for. They can be gamed -- test preparation by a family with resources may get one child over the cutoff, while another is just under due to the environmental factors described above by Grinity. What if a kid is sick the day he/she takes the test? And a corollary to this is that the cutoffs are arbitrary from the perspective of which kids might benefit from the program. Is the kid in the 96th %ile really better off without the services those in the 97 %ile get? Not necessarily, it depends on the individual kids, i.e. factors beyond what the test measures....