Decision Making / 6 / 6.9

6.9 ANALYSIS OF DYNAMIC BEHAVIOUR USING CAUSAL LOOP OR INFLUENCE DIAGRAMMING

Our investigative efforts so far (whether concept mapping, SSM or other knowledge elicitation techniques are used) have been directed at coming to grips with the nature of problems at hand and appreciating the various views that exist. These views are critical and need to be accommodated in the interventions we design and implement. Without the accommodation of these views or perspectives, we could find outright rejection of management strategies, opposition to them or resistance to change.

Whilst the Simplified Executive Summary Concept Map, Figure 6-7 provides a consolidation of the views of the various stakeholder groups, it does not provide a sufficiently stringent depiction of causal structure upon which to build models of the dynamic behaviour of ThreeTwoOne’s business. The Simplified Executive Summary Concept Map is valuable as an aid to identifying in broad terms where intervention efforts may be directed. By itself it is of little use as mentioned earlier: it only becomes valuable when considered in the light of the maps from which it was derived. The next step involves building a set of directed causal graphs providing a sound basis for analysis of causes of dynamic behaviour of ThreeTwoOne’s business. At this point the task is to build dynamic hypotheses for further analysis. Concept maps, or other outputs from knowledge elicitation, have to be converted into directed causal graphs as the next step in IISD. When built, these causal graphs hold the secrets about where the pressure points are, that is, where best to expend effort. There are two alternate techniques. The choice is left to the reader:

• Causal loop diagrams. A causal loop diagram is a convenient way of representing the principal feedback loops and related causal relationships relevant to a particular problem situation, without distinguishing between the nature of the interconnected variables. Causal loop diagrams serve as preliminary sketches of causal hypotheses. Later, they enable the essential components and interrelationships to be summarised. The main purpose of building causal loop diagrams is to permit diagrammatic representation of dynamic hypotheses (Sterman 2000: 137-190). Basic casual loop diagramming skills are quicker to acquire and causal loop diagrams are easy to use. Causal loop diagrams are relative intuitive in their reading and can be highly valuable. Unfortunately, they can be ambiguous if the conventions for drawing them are not rigorously observed.
• Influence diagram. An influence diagram is a list of factors in a problem, together with arrows and signs showing the relationship between them (Coyle, 1996: 31). Influence diagrams involve greater rigour than casual loop diagrams and have superior functionality as tools for stringently depicting the causal structure of a problem. Unfortunately, influence diagrams are more difficult to learn, but the effort is very worthwhile. An influence diagram is a form of causality-directed graph having much more rigorous rules than a causal loop diagram, in that it pays strict attention to differentiating the rate and level variables used in system dynamics modelling. In some cases, the use of influence diagrams can mean that more sophisticated quantitative models are not required (Coyle, 2000).

The choice between the two causal diagramming techniques might be guided by the time available to stakeholder groups to build familiarity. Whether they may have already been exposed to one or other form is also a consideration.