3 ms·
Leta Hollingworth agreed with you that 1) is the optimal approach. Unfortunately for the exceptionally gifted, they are rare enough (by definition—4 standard d
by waterhouse 7y ago
Leta Hollingworth agreed with you that 1) is the optimal approach. Unfortunately for the exceptionally gifted, they are rare enough (by definition—4 standard deviations is less than 1 in 10k—although I've read that it's flatter than a normal distribution) that it is difficult to get enough of them into one place to make a whole class out of them, unless you run a boarding school or have a very dense population (Leta Hollingworth worked out of New York).
She also agreed, at least implicitly (I'm struggling to find the best quote), about the importance of starting early. "Such struggles as these [describing a 7-year-old who tried to show his favorite books to the other kids, who "resisted his efforts, made fun of him, threw the treasures on the floor, and finally pulled his hair"], if they continue without directing the child's insight, may lead to complete alienation from his contemporaries in childhood, and to misanthropy in adolescence and adulthood."
- nopinsight 7y agoWhat I’ve seen working in practice is grouping the highly gifted with older ‘merely’ gifted kids (combining 1) and 2)). If the latter group is 1 in 100, there would be enough of them to form a class even in a smaller city (Say a city with pop 200,000 & 2000 births per year means 20 gifted kids per cohort). The gifted ones tend to focus on learning and have similar inclinations to the highly gifted so it works out socially, as long as overt competition is kept to a minimum.
- sokoloff 7y agoThat probably works better at the elementary and middle-school-age ranges (I'm thinking pre-calculus, pre-high-school work). Once you get to traditional high-school level, I think even the 2 years older "merely gifted" kids will be holding back the younger truly exceptional. 100% agree that it's way, way better than the current state of the art which seems to me to basically be "as long as they aren't causing trouble, we're happy to have them bored in class and pulling up our standardized testing results..."
- nopinsight 7y agoYes, agree about the increasing difference. At the high school level, it’s often better to group more highly gifted students together. They are generally able to live further away from their parents so population size is less of a problem by then. This is actually a common practice in several Asian countries. There are a couple national-level magnet schools and (highly) gifted classes within those schools where these kids are taught and trained by special teachers and curriculum specialized to their talent—math, science, language. It’s likely that excessive desire for educational “equality” despite adverse consequences in some other countries squanders the potentials of many such kids.
- someguydave 7y agoYeah I wonder how much "grading schools & teachers by standardized testing" has discouraged advanced learning programs (because they want to juice their scores). The future historians will condemn this age for holding back talented kids.