In parametric ANOVA, you could overlay a significance interval on the straight (or curved) line throughout the regression. Looked at this way, it doesn't really matter whether or not the Modulus is ...