Causal and diagnostic reasoning - solution to example 1


Most people answer correctly (c). In an experiment conducted by Tversky and Kahneman, 75 out of 165 subjects chose this answer. However, nearly as many 69 -chose (a) and only 21 chose (b). This indicates the expected asymmetry of inference. We generally perceive the daughter as more similar to the mother than vice versa, and we attribute properties of the daughter to the mother with greater confidence than vice versa.

Conclusion: the impact of causal data on the judged probability of a consequence is greater than the impact of diagnostic data on the judged probability of a cause.


It seems it is easier and more natural for us to reason from causes to consequences than to reason from consequences to causes. (See availability - biases of imaginability ).