Models and Modeling

The Modeling Process

Human beings implicitly model all the time. Routine decisions in everyday life are guided by elaborate and yet incomplete mental models of the tasks at hand. For example, a person driving a car associates a yellow light with the need to slow down (or hurry through the intersection). In making this association, the person's mental model simulates, or translates the perception of a yellow signal into, a decision of whether to slow down or hurry up, depending on rapid evaluation of one's position relative to the intersection. This decision is carried into action by an adjustment of foot pressure on the accelerator and brake pedals. This sequence of information, perceived and acted on by a person's mental model, is one experienced many times over in a given day. In this way, mental models may be largely subconscious but yet represent intricate webs of understanding that guide (and misguide) human behavior.

Formal modeling enables explicit testing of underlying assumptions that humans carry in the form of mental models. One strategy for making models explicit is to slow down one's thought to become aware of the implicit mental models that simulate both immediate and longer-term scenarios of the system of concern. Group modeling experiences benefit from the process of slowing, exposing, and challenging implicit mental models.

The process of making mental models at least partially explicit—explaining ideas to others in conversation or writing, drawing, or encoding them in computer models—exposes the limitations of their scope, based as they are in an individual's personal experience. As mental models are formalized, the explicit model becomes an artifact readily shared with others, while the implicit mental models continue to update based on information feedback from both the model and the real worlds.

Figure 1 The modeling process in context. The iterative process of modeling creates an experimental learning feedback loop in the context of the real world.

Source: Adapted from Sterman, J. (2000). Business dynamics: Systems thinking and modeling for a complex world. New York: McGraw-Hill.

In a dynamic modeling process (Figure 1), observations and decision rules may be input as data and assumptions into computer models, and simulated output may inform policies for making decisions. As mental models are updated with observations of both the real and the model worlds, they translate into updated decision rules or heuristics that shape the decisions that manifest as actions in the real world. The reflexive relationship between mental models and the process of modeling reveals it to be at once an art and a science. The art of modeling includes the role of intuition in problem formulation and hypothesis development, as revealed by the selection of relevant variables for the system of interest. The science of modeling centers on a consistent, reproducible methodology for testing theory by comparing simulated output with observed data. Steps of the modeling process generally include the

...