2.7. Causal Loop Diagrams – Basic Tips
In this chapter, several causal loop diagrams have been presented. Despite their variety they all follow the basic tips of good practice.
2.7.1. Picking and Naming Variables
The choice of words is vital. Each variable must be a noun. Avoid the use of verbs or directional adjectives. For example, a causal diagram can use the word 'sales', but not 'sales planning' or 'increased sales'. Simple nouns like 'accounts' or 'staff' can be augmented with adjectives to give phrases like large accounts' or experienced staff'. Sticking to these basic naming rules helps when assigning polarity to causal links and explaining how changes propagate around loops.
Words are versatile, but they should also be grounded in facts. The range of concepts that can be included in causal loop diagrams extends from the hard and easily measureable, such as 'new products' and 'recruits', to the soft and intangible such as 'morale' or 'customer perceived quality'. A powerful feature of feedback systems thinking and system dynamics is its ability to incorporate both tangible and intangible factors. However, for any variable no matter how soft, you should always have in mind a specific unit of measure, a way in which the variable might be quantified, even if formal recorded data do not exist. Hence, you might imagine morale on a scale from 0 (low) to 1 (high) or product quality on a scale from 1 (low) to 5 (high). Be sure to pick words that imply measureability, such as delivery lead time' thought of in weeks or months, rather than a vague concept like 'delivery performance'.
2.7.2. Meaning of Arrows and Link Polarity
Arrows show the influence of one variable on another - a change in the cause leads to a change in the effect. The assignment of link polarity (+) or (−) makes the direction of change clear. In Figure 2.13, an increase in marketing budget leads to an increase in sales, which is a positive link. A useful refinement in polarity assignment is to note whether the effect of a given change is an increase (or decrease) more than it would otherwise have been. The use of this extra phrase avoids ambiguity in situations where the effect is cumulative. Hence, an increase in industry reputation leads to an increase in customers interested; but since customers were accumulating anyway, we really mean more customers than there would otherwise have been. Polarity assignment works equally well for intangible variables. Industry reputation here is viewed as an intangible concept measured on a scale from 0 to 1. In the particular case of the materials handling business described earlier, industry reputation depends on the bidding power of new competitors. An increase in bidding power (measured as the fraction of bids won by new firms in the industry) leads to a decrease in industry reputation, below what it would otherwise have been, a negative link. As reputation takes a long time to change, a delay is shown on the link.
Figure 2.13. Arrows and link polarity
2.7.3. Drawing, Identifying and Naming Feedback Loops
For the systems thinker, feedback loops are the equivalent of the sketches created by political cartoonists. They capture something important about the situation or object of interest. Just as a few bold pen lines on a canvas can characterise George Bush, Osama bin Laden or Margaret Thatcher, so a few feedback loops on a whiteboard can characterise an organisation. Like celebrity sketches feedback loops should be drawn clearly to identify the dominant features, in this case important loops. Sterman 2000 identifies five tips for visual layout:
Use curved lines to help the reader visualise the feedback loops
Make important loops follow circular or oval paths
Organise diagrams to minimise crossed lines
Don't put circles, hexagons, or other symbols around the variables in causal diagrams. Symbols without meaning are chart junk' and serve only to clutter and distract.
Iterate. Since you often won't know what all the variables and loops will be when you start, you will have to redraw your diagrams, often many times, to find the best layout. (Sterman, J.D., Business Dynamics: Systems Thinking and Modeling for a Complex World, © 2000, Irwin McGraw-Hill, Boston, MA. Reproduced with permission of the McGraw-Hill Companies.)
As we have already seen, for the hot water shower and drug-related crime, there are two main loop types, balancing and reinforcing. A loop type is identified by imagining the effect of a change as it propagates link-by-link around the loop. A reinforcing loop is one where an increase in a variable, when traced around the loop, leads to a further increase in itself. Such an outcome requires an even number (or zero) of negative links. A balancing loop is one where an increase in a variable, when traced around the loop, leads to a counterbalancing decrease in itself. Such an outcome requires an odd number of negative links. Once you have identified loop types it is good practice to label them R for reinforcing and B for balancing, the letter encircled by a small curved arrow shown clockwise for clockwise loops (and vice versa).
By following these tips and by studying the examples in the chapter you should be able to create, label and interpret your own causal loop diagrams.