In thinking further, the range might not change much because the people at the far left tail of the distribution are severely impaired, and many of them already receive specialized services. With improved nutrition and so on, there might be fewer people at the lowest end of the ability distribution, but some of the most impaired people would still anchor that end of the range. I would predict that mean scores for the entire population would rise, as well as mean scores for the ethnic subgroups. This would shift the highest part of the bell shape to the right, indicating actual improvement in ability population-wide. If this were to occur, psychometricians would probably adjust the tests, so that the average score would be 100 again, but this number would be the equivalent of 110 (or whatever) on the first version of the test.
In thinking further, the range might not change much because the people at the far left tail of the distribution are severely impaired, and many of them already receive specialized services. With improved nutrition and so on, there might be fewer people at the lowest end of the ability distribution, but some of the most impaired people would still anchor that end of the range. I would predict that mean scores for the entire population would rise, as well as mean scores for the ethnic subgroups. This would shift the highest part of the bell shape to the right, indicating actual improvement in ability population-wide. If this were to occur, psychometricians would probably adjust the tests, so that the average score would be 100 again, but this number would be the equivalent of 110 (or whatever) on the first version of the test.
Thanks for the dialogue, Richard.