What is it, then, when a scientist formulates a hypothesis that growth patterns follow temperature, and tests the hypothesis against data only to find that growth patterns do not follow temperature at all for 30% of the data and only partially for the rest?