Fig. 2.From: Divide and conquer? Size adjustment with allometry and intermediate outcomesTwo scenarios of the relationship among an experimental treatment, a trait of interest (focal variable, y) and an intermediate outcome (x). a The treatment affects both x and y, and therefore x and y are correlated (dotted line with a double-headed arrow) but x does not affect y. b The treatment affects both x and y, and then x also affects y Back to article page