Ishikawa's quality tools: how to apply the fishbone and quality circle

Ishikawa's two primary quality tools are the fishbone diagram (cause-and-effect diagram) and the quality circle. The fishbone diagram organizes potential root causes of a quality problem into six categories — known as the 6Ms: Man (people), Machine (equipment), Material (inputs), Method (process), Measurement (data systems), and Mother Nature (environment). To construct a fishbone diagram: write the problem statement at the head, draw the six M branches, brainstorm potential causes under each branch with the team, identify the most probable causes using data — not consensus — and verify each suspected root cause with a scatter diagram or designed experiment before implementing a countermeasure. The quality circle applies the fishbone in a team context: a small group (5–10) of frontline employees from the same work area meets regularly to identify, analyze with the fishbone, and solve quality problems within their direct sphere of control.

Ishikawa fishbone diagram guide showing six M categories on the left and six-step construction process on the right with quality circle session structure below.

Ishikawa's quality tools are the most widely used in quality management — and the most frequently misused. The fishbone diagram is often used as a brainstorming recording tool rather than a root cause analysis tool, producing a list of possible causes without the data verification step that turns a cause hypothesis into a confirmed root cause. Ishikawa's design intent was clear: the fishbone organizes the hypotheses, and data analysis verifies which hypothesis is the actual cause.

How to Construct a Fishbone Diagram

  1. Write the problem statement at the head of the fish — be specific: 'Dimension X out of tolerance on 3.2% of parts at Station 4' not 'quality problem.'.
  2. Draw six M branches from the spine — Man, Machine, Material, Method, Measurement, Mother Nature.
  3. Brainstorm potential causes under each branch — use '5 Whys' under each branch to go deeper than the surface symptom.
  4. Identify the most probable causes — not by team vote but by asking 'what data would confirm or deny this cause?'.
  5. Collect data to verify — scatter diagrams, stratified Pareto charts, or a designed experiment.
  6. Confirm the root cause — only a cause that data confirms is a root cause. Everything else is a hypothesis.

The Six M Categories — What to Look for in Each

Category 

What to Investigate 

Common Causes Found Here

Man (People).

Training adequacy, procedure adherence, fatigue, skill variation. 

Operator-to-operator variation, untrained substitutes, workarounds.