<p>@hinsdale</p>
<p>Uhh, according to AP statistics, you dont need a very large amount of the population to be representative. Around 50% of kids report their rank in class. Laws of statistics holds that the reported rank holds for the rest of the 50%. </p>
<p>Furthermore, the ranking criteria is fine. I dont see any glaring flaws with using % of kids hailing from top 10%. Higher ranked kids = more selectivity. Sounds reasonable.</p>