My totally unscientific approach would be to build some sort of "curiosity index" and the higher that is for a child up to a certain age the more likely they ar

My totally unscientific approach would be to build some sort of "curiosity index" and the higher that is for a child up to a certain age the more likely they are to be successful. For extra credit calculate a delta to see if the curiosity stays at the same level (or increases/decreases) with age.