Double-loop learning

Involves not only recognizing the mismatch between actual and desired states, but also using past experience of unsuccessful attempts to meet the goals to evaluate and modify mental models and rules involved in determining an action to respond to the mismatch.