When people have to make predictions and diagnosis they make use of their causal knowledge. This knowledge refers to two constituting aspects of causality: sufficiency and necessity. In standard theories both aspects are considered as being independent from each other. The present research tests this assumption. In an experiment we examined how peoples confidence in one of both aspects is affected, if they receive negative evidence for the complementary aspect. The presented data show that peoples confidence related to the aspect that has not been challenged by negative evidence decreases under such conditions. This devaluation effect is not predicted by standard theories.