As a layperson reading through that, I found it really informative, but also it makes me wonder. How do you use that info, that under 60% of top performers may still be in top 3% percentiles from year to year, with respect to grade acceleration? It just makes me see why schools are so hesitant to accelerate, if they are uncomfortable with such predictions. Especially if you're testing kids at grade-level instead of above-grade-level, so you're not comparing them to the correct set.
And is there always regression to the mean downward? Or do you see children who score high on an IQ test at 7 score similar or higher at 9? It seems like the 95% confidence range would be better suited to a .8 prediction, if this data is accurate. I might be misunderstanding!
