Inductive reasoning allows us to go beyond the target hypothesis and capitalize on prior knowledge. Past research has shown that both the similarity of categories and specific knowledge about causal relations affect inductive plausibility. We go one step further and focus on the role of abstract causal schemas about main effects and interactions. Two experiments show that both prior assumptions about abstract causal schemas and the similarity of the corresponding causal effects affect inductive judgments. Reasoners have different prior beliefs about the likelihood of main-effect versus interactive schemas, and rationally combine these prior beliefs with new evidence in a way that can be modeled as Bayesian belief updating.