[How does model- and situation-specific lack of knowledge affect causal inferences?]

A model proposed by Thüring (1991) for inferences based on causal knowledge was empirically tested. According to this model, two variables affect the certainty with which a causal inference is concluded: insufficiency (model-specific uncertainty) and ambiguity (situation-specific uncertainty). Within an experiment these two variables were manipulated. Both had a very significant (p < .01) influence on causal inferences. In respect to its quantity, variation of ambiguity had the effect predicted in the model. Concerning insufficiency, distinct differences between predicted and empirical ratings were found. Reasons for these deviations and model modifications resulting therefrom are discussed.