Vocabulary (a common vocabulary for teaching modeling

Conceptual model – a simplified version of reality designed to achieve a specific purpose expressed qualitatively, often in pictures or words; a purposeful simplification of reality. One goal of a concept model is to identify the main components of a process and how these components relate to each other.

Computational model – a purposeful simplification of reality in which the components and the

relationships between components are expressed quantitatively and so the predicted outputs of the system are quantitative as well.

“Model World” (vs Real World) – a world populated only by those simplest components and

relationships needed to create a model to meet the specified purpose; creating a model world is the first step in building a model. Many assumptions are made when determining the minimal necessary components and rules (and it is often useful to explicitly identify important assumptions at this firststep).

Rapid prototyping – an approach to modeling that attempts to create the simplest possible model that will serve the identified purpose as rapidly as possible; it encourages an iterative modeling process.

Verification - examines whether the computational model is implemented correctly, tests are performed to look for bugs, errors and oversights; does not determine if the model’s objective is met.

Validation - examines how well the model’s predictions match experimental results and addresses whether the model is useful in the sense that the model addresses the right problem and provides useful information about the system being modeled.

Reality check – identifies justifications for the structure and assumptions of the model as well as evaluates the level of confidence in the outputs of the model; a weaker form of validation for when validation is not practical.

Sensitivity analysis – examines how strong of an effect there is to the model’s output when uncertain values vary in their value. It can tell you where good data are needed because small differences have large effects, which components may not need precision as they have small effects, and the range within which the model produces useful results.

Assumption analysis – articulates explicit assumptions of the model and examines how robust the model is to deviations from the assumptions.

Modeling process – the iteration between 1) simplifying the real world to a model world designed to meet a specific purpose, 2) creating a rapid prototype conceptual model that addresses the objective creating and implementing a rapid prototype computational model of the conceptual model, 4) verify the computational model is implemented correctly, 5) validating how well the computational model serves the original purpose using both sensitivity and assumption analyses, 6) determining if altering either the conceptual or computational model is needed to achieve their purpose.