The paper addresses Kilman's criticisms of the observed correlations between prices and labour values. It argues that the notion of spurious correlation is not relevant in this case. It examines Kliman's own simulations and shows that his statistical correction techniques involve dividing through by the signal to leave the noise.