���������������������������������������������������������������
quantitative or qualitative . Qualitative models are typically diagrams or possibly a game . The method may also include formal steps in the modeling process so that the model conveys a particular perspective of the target system . For example , Soft Systems Methodology has several steps , including a formal definition of the target system that must meet specific criteria ( referred to as CATWOE ) and creation of ‘ rich text pictures ’ of how the target system works ( Checkland & Poulter , 2006 ). Another example is that of Causal Loop Diagrams , that depict how changes in one aspect of a system impact on changes in some other aspect ( Sterman , 2000 ).
The advantage of qualitative methods is in their flexibility . They can be used to model relationships between issues , ideas , people , or any other entities . Relationships could be expressed as comparisons ( such as A is larger than B ) or as influences ( such as A impacts on B ) or as descriptors ( such as A is related to B ). As well as being used as the implemented model , qualitative methods can be used to assist with the design of other models , to explore the problem and encourage all relevant relationships to be considered .
While modeling methods cannot be neatly partitioned into qualitative or quantitative , selecting an appropriate method requires the policy analyst and modeler to consider the type of available data and how the relationships should be expressed . Some methods will be more or less suitable in situations where , for example , important detail would be lost by quantifying qualitative information .
��������������������������
Modeling methods that capture the
situation at a particular moment in time are referred to as static .
The relevant moment in time could be the present , and the model describes some aspect of the current situation . Alternatively , the model could be set in the past or in a hypothetical future .
In contrast , dynamic modeling techniques explicitly refer to the passage of time , representing the processes or activities within a system . For example , dynamic models may describe how a change in rainfall leads to a change in river flow , or how the parts of car move through an assembly plant and the time taken at each station . Dynamic models that reproduce changes rather than simply describe how they occur may also be referred to as ‘ simulations ’ because the model represents the process of change in a system through time .
Causal Loop Diagrams is a popular modeling diagram technique for dynamic systems , but the design documentation of any simulation would be considered dynamic . For example , UML diagrams that describe how an Agent Based Model is to be implemented describe how to update each agent each time step and are therefore also dynamic . More usually , modeling techniques for dynamic systems are those that implement the passage of time , such as games , State Transition Models , Agent Based Modeling and System Dynamics .
Whether a system is to be modeled for a specific point in time or over the passage of time also impacts on whether other characteristics are relevant to selecting a modeling technique . For dynamic models particularly , there are several characteristics that affect how the system changes over time and must therefore be represented in the model . The most important are the variability or heterogeneity of the model entities , and whether there is interaction between these entities . Heterogeneity and level of aggregation are an important topic for discussion between the modeler and the policy analyst concerns the amount of heterogeneity . For indistinguish-
78