so how then, do you understand the very small sample sizes? Perhaps as a severe limitation to interpretation of the initial graphs you posted?