In my Master’s Research Project, I developed a very large model of a complex system. In so doing, I identified multiple ways to expand or contract the scope and size of a modelling project: breadth, depth, density, detail, evidence and quantification.
The “scope” of a model refers to how much complexity of the real-world system will be expressed in the model’s simplification. The scope determines the model’s size. Modelling projects can balance their scope requirements with the time and other resources available.
These six dimensions were defined for causal models (an extension of causal loop diagrams) but similar scoping decisions would arise for other types of system maps & models with nodes and links. This extract from the Poverty Reduction Model (PRM) will illustrate some dimensions of scope:
- Breadth means the diversity of topics to be covered. This example has 9 colour-coded subject areas from the Poverty Reduction Model’s very broad scope, including employment, housing, health, finances and social dignity. Narrower scopes are easier to tackle but may miss connections outside the scope boundary.
- Depth refers to the quantity of elements (nodes, such as variable factors) identified about each topic. The PRM integrates deeper knowledge about employment (dark green) and housing (apple green). Its shallower subject areas include a few important & generalized elements such as “Newcomer Settlement”.
- Density is the quantity of connections (such as cause-and-effect relationships) between elements. If density is low, only a few important connections are shown. A denser model contains more knowledge but the diagrams are busier and harder to read, like this example.
- Detail indicates the amount of description and other information captured about each element or connection. Description may be useful for decision support, but take time to research, write and read. There are descriptions for some of the PRM elements, captured from the source documents or expert discussions.
- Quantification indicates whether there are measures, ratings or statistics about the elements (such as costs or benefits) or of the connections (such as strength of causality or length of delay). The PRM database contains impact ratings for the elements.
- Evidence refers to justification by citing studies or recounting personal narratives. Capturing more evidence increases the validity and reliability of the information conveyed by the model. This research has not yet been done for the PRM.
I defined a comprehensive model as one that thoroughly represents the breadth of a system, and captures rich (deep, dense and detailed) information about its complexity. You can read more about scoping a system model in section 6.1 of my research report.