"which connected China’s best teachers "... I wonder why they choose the best teachers for this type study? Wouldn't a randomized, average sample produce more objective results?