I have examined young children's ability to engage in counterfactual reasoning and its relation to causal reasoning. Counterfactuals are interesting because the causal graphical model formalism does not treat them as a special kind of reasoning, and suggests that if a reasoner has a representation of the causal structure, then counterfactual reasoning should be no different from predictions about future hypotheticals.
This is a different approach than many developmentalists, who suggest counterfactual reasoning is influenced by domain general abilities, such as the development of mental representation. I believe that counterfactual reasoning is not a domain general ability, but rather dependent on a child's developing theories of the world. I have shown that children's ability to make counterfactual inferences does not differ from their ability to make predictions about future events, and is related to their ability generate explanations of past events in the same domain. Further, I have shown that there isn't a general relationship between counterfactual reasoning and the child's mental representation abilities. Instead, there is a specific relationship between children's developing mental representation abilities and their ability to engage in counterfactual reasoning about the same causal domain (i.e., false belief).