Fishbone diagram is known by different names: Ishikawa diagram, cause and effect analysis or cause and effect diagram. Fishbone diagrams are popular on shop floors as they are easy to understand visual tools.
The principle of a fishbone diagram is to list all possible causes leading to an effect (usually the problem to solve) and to sort them into “families” sharing commonalities.
These families are the main bones of the diagram. In order to get a fishbone diagram started a common way is to go for the famous “5Ms”, standing for Man, Machine, Material, Method, Mother Nature.
The five Ms are common families of causes of trouble experienced in industry and became a kind of standard to start a cause and effect analysis. Typically, one draws a triangle (the fish’s head holding the effect to analyse), a horizontal line which is the fish’s backbone and five slant bones from the horizontal one with the 5M’s at their end.
Note that a fishbone diagram must not be limited to these five Ms and it is not mandatory to use them. The 5Ms are just a convenient mnemotechnic.
A fishbone diagram is an appropriate support for brainstorming and ordering the outputs. The different named main bones will lead to explore the relative families of causes and propose ideas of potential causes. Then, the various ideas can be sorted according to the family each main bone stands for.
Yet too often I see fishbone diagrams cluttered with listed causes that are not likely to influence the effect. This is maybe where nicknaming the cause-effect diagram / analysis “fishbone” leads to forget what should be looked for: cause-to-effect relationship.
Now a true cause-effect relationship should respond to the “if [cause] then [effect]” sentence. If the proposal makes no sense, the cause-to-effect relationship is questionable.
A variant may be “if [cause] then [effect] because [justification]”
If the issue to solve is a frequent drive motor overload on a conveyor belt causing a circuit breaker to shut off. The proposal “rain” as such will not be valid: “if (it) rain(s) then drive motor gets overloaded” is not logically sound, as nobody can understand a direct link between rain and and circuit breaking.
The proposal may be refined and rephrased. It could be that the person thinking about rain was actually considering the supplement of weight of transported goods when they get soaked by rain: “if it rains, then goods get heavier when soaked and if goods are heavier, the power required to move the belt may cause the breaker to shut off”.
If no logical relationship can be established between the proposed cause and the effect under study, the proposal may be discarded for the sake of efficiency.
Using the “if [cause] then [effect]” sentence will filter out irrelevant proposals and get better diagrams.
A valid proposal would be “worn out ball bearings”, expressed in full sentence like: “If ball bearings are worn out, then the drive motor gets overloaded because the belt dragging increases”.
Improving the cause-effect analysis
When using a cause-effect diagram, it can be improved in two ways:
- Ask participants to bring ideas compliant to the cause-effect relationship, which somewhat constrains the brainstorming
- If the brainstorming phase should be kept uncensored, collect all proposals and then check each of them with the “if [cause] then [effect]” sentence before placing it onto the diagram
An additional benefit of using the “if [cause] then [effect]” sentence is to improve the proposals’ statement. Too often elder cause-effect diagrams are undecipherable because of the way the proposed causes have been stated. Using full (short) sentences with sound logic will help anyone to read the diagram, even if not involved in its construction. Such diagrams will still be understandable long after they’ve been drawn.
Having nice logically sound causes in a cause-effect diagram does not insure the causes actually exist. Therefore after populating such a diagram with possible causes, it is mandatory to check their existence and reality on the spot. In Lean-aware companies, we would say go to the gemba!