Drawing a diagram from scratch

When you're drawing a multiple cause diagram you often start with one event or state you want to explain. Perhaps because it's something you want to achieve, or maybe something you want to avoid. You ask yourself the question 'what causes this?'

These may be the causes that first come to mind, but before moving on, it’s often helpful to ask ‘what else causes this?’

Next you ask ‘and what causes each of these?’

And again ‘what else causes each of these?’

And for each of these factors you can ask the same questions again... ... and so on.

In practice, it’s not usually as straightforward as this, and there are two or three further questions you have to ask.

Firstly, for each link, you need to ask is this a direct link or are there any intermediate factors it might be helpful to identify?

This is important because each of these intermediate factors may itself have multiple causes, and one of these may provide the most promising way to intervene in the situation.

Here’s an example of the sort of thing I mean.

Drought can cause famine, but opening out the chain of causal factors between the two can help you to see where action might be taken.

Secondly, you need to look for causal connections between the factors you've identified.

These interconnections can be the most important ones to identify, because some of them may form feedback loops where the arrows follow each other around a closed loop. Feedback loops play an important role in governing the behaviour of the system as a whole. So they can help you identify the points where intervention might be most effective.

If you're working on paper, you're likely to go through quite a few sheets at this stage. Some people find it helpful to write the factors onto post-it notes, so that they can move them about the page without having to re-write them each time.

An alternative way to draw a multiple cause diagram might be to start with an event you want to explore the possible consequences of, and to proceed in a similar way, asking ‘what might be the effect of this?’ ‘what other effects might there be?’ and ‘what might be the effects of these?’

This sort of approach could be useful for exploring the likely consequences both intended and unintended of initiatives you might take in an existing situation. In this case you'd be likely to have already analysed the situation, using the first approach, so you'd be more likely to be adding to an existing diagram than starting one from scratch.