When mental models go wrong: co-occurrences in dynamic, critical systems

This paper highlights a psychological phenomenon affecting the accuracy of mental models. It occurs when two consecutive events happen as expected by an operator. Typically, such a situation reinforces the confidence in one’s mental model. However, consecutive events can happen as a random co-occurrence, for reasons that actually differ from the ones believed by the operator. Nonetheless, because of the consistency between the environmental data and the operator’s expectations, one event can be taken to be the cause of the other. When this false belief happens, the mental model is erroneously assumed to be valid. We discuss this phenomenon and its potential disastrous consequences using the example of a real commercial air crash. We finally address some implications for systems’ design and support tools.