The Psychology of Quality and More
Cause-Effect (Fishbone) Diagram
What's it for?
The Fishbone Diagram, traditionally used for finding causes of problems, can be attached to anything to show a hierarchical set of factors that affect the item to which it is attached. In this way, it can be used for annotation, exploration and expansion of a wide range of diagrams, leading to an improvement in the understanding and communication of the diagram.
How does it work?
A Fishbone Diagram, more properly called a Cause-Effect Diagram (and also sometimes called an Ishikawa Diagram, after its originator) is basically a simple way of showing a hierarchy, or tree. The standard use of the Fishbone Diagram is to show causes of a known effect, as in Fig. 1.
Fig. 1. Standard Cause-Effect diagram
What makes a fishbone a fishbone is the angling of the arrows which gives the diagram a 'fishy' appearance. This is simply a pragmatic ploy, used to help fit new text in around existing text.
The use of a hierarchy is an important consideration, as it acknowledges that causes are not simple, with each cause often being caused by one or more other causes. It is often those lower level causes (the 'root' causes) which are the real culprits. Fixing the root cause is almost always better than fixing the symptom.
These principles of hierarchy and angled arrows need not be the sole preserve of the Cause-Effect Diagram – they can be exported for use with other diagrams. This is illustrated in Fig. 2., where a Flowchart is annotated with potential problem areas.
Fig. 2. Potential problems and causes in a process Flowchart
How to do it
1. When drawing any diagram, from Flowcharts to Force-Field Diagrams to Graphs and Charts, consider why you are doing it. Are you looking for current or potential problems? It is to communicate specific points to other people? If you want to do more than simply draw the diagram, a Fishbone annotation may be useful.
2. Identify the elements of the diagram which you are going to investigate. Thus, in Fig. 2, both the actions and the outputs of each step are being considered.
3. Identify a specific question to ask of each element identified in step 2. This will be translated into arrows in the annotation. For example, in the Force-Field Diagram of Fig. 3, the question for the angled arrows might be, 'What contributes to this force?'
Sometimes, two different questions are useful. For example, in Fig. 2, the first question is 'What could go wrong here?', and subsequent questions are 'What causes this?'.
4. Ask the question from step 3 of each of the diagram elements identified in step 2, drawing in arrows to show the answer. Also find sub-elements by repeatedly asking questions from step 3.
In Fig. 2, asking 'What could go wrong?' with the engineer visit (the last step), gives one answer that he or she might be 'unable fix the problem'. The subsequent question asked about causes indicates that the fix may not be possible because of bad parts or insufficient knowledge.
5. Lastly, and most importantly, make use of the information gained. Thus from the questions in Fig. 2, you might ensure only qualified engineers go out on service calls.
Next time: KJ and the Affinity Diagram
This article first appeared in Quality World, the journal of the Institute for Quality Assurance
And the big