How to achieve sustainable growth in a feedback-rich system? The development of mental models and decision rules of naïve decision-makers in a dynamic task experiment.

Project proposal Fondecyt ID# 447625.

Researchers: Martin Schaffernicht, Miguel López-Astorga, Ramón Castillo, Cristian Rojas-Barahona (all University of Talca, Chile) and Joel Fuentes (University Bio Bio, Chile)

In previous work, we have shown that the mental models of dynamic systems and the theory of mental models (Johnson-Laird) can be combined. This combination led to the definition of two types of mental model errors.

Type 1 error: relevant features of the situation are not accounted for.

Type 2 error: relevant possibilities are not recognized.

We took a simple dynamic decision task, defined three intuitive decision policies any naïve decision-maker might think of and analyzed the underlying mental models of dynamic systems and mental models of possibilities. All three policies are based on mental models with bot types of errors. When simulated, the two simpler policies generate undesired behaviors and poor performance; the third policy achieves a good performance but still generates undesired behaviors.

In this project, we will empirically test several hypotheses derived from that theoretical study. Besides the experimental design, we also plan to develop a special computational platform for these experiments. The general design is explained in the short note Design for a dynamic decision experiment platform to capture and analyze mental models .