Modelling Causal Reasoning under Ambiguity

Causal reasoning under ambiguity requires subjects to estimate and evaluate ambiguous observations. This paper proposes a hierarchical model that accounts for the uncertainty of both the distribution of the functional form selection and distribution of the ambiguity treatment selection. The posterior distribution of the causal estimates is determined by both the functional form and the ambiguity processing strategy adopted by the reasoner. A model is tested in a simulation study for its ability to recover the strategy and functional form adopted by subjects across a range of hypothetical conditions. The model is further applied to the results of an experimental study.