What are the 7 QC tools?
The 7 QC tools are seven simple methods for finding and solving quality problems with data. They are the check sheet, stratification, the Pareto chart, the fishbone diagram, the histogram, the scatter diagram and the control chart. Together they take a team from "we have rejections" to "this is the cause".
Many plants say "our rejection is high" and stop there. People then argue about whose fault it is. One person blames the material, another the machine, a third the operator.
The seven tools replace the argument with data. Each tool answers one question.
| Tool | The question it answers |
|---|---|
| Check sheet | How often does each defect happen? |
| Stratification | Is it one machine, shift, lot or supplier? |
| Pareto chart | Which problem do we take first? |
| Fishbone diagram | What could be causing it? |
| Histogram | Is the size centred and inside the limits? |
| Scatter diagram | Are two things linked? |
| Control chart | Is the process stable over time? Should we act? |
One example, followed through all seven tools
This example uses made-up figures from a small shop that turns and grinds shafts.
1. Check sheet. A ready-made form counts each defect where and when it happens. In one month the team counts 48 defects in 600 shafts. They write the machine, the shift and the date on every tick, even before they suspect anything.
2. Pareto chart. Bars in order of size, biggest first. The biggest bar is "journal diameter oversize": 18 of the 48 defects, or 37.5%. Work on this first. The other defects can wait.
3. Stratification. Split the 18 by source before adding them up. A total hides the story. A split shows where the problem lives. Here, 15 of the 18 came from shafts turned on one of the two lathes. This is possible only because the check sheet carried the machine from the start.
4. Fishbone diagram. List the possible causes of the problem under headings such as man, machine, material, method, measurement and environment. List first, argue later. The team writes 12 possible causes. Two are circled: the grinding stock left, and the route card.
5. Histogram. A bar chart of measured readings in equal classes. It has two humps. One lathe's output sits against the upper limit. Two humps usually mean two streams mixed together.
6. Scatter diagram. Plot pairs of numbers. More stock left by the lathe means a bigger diameter after grinding. The points rise in a line. This suggests a link. But a strong link is not yet proof, so you must still check it on the machine.
7. Control chart. A time chart with a centre line and limits calculated from the process itself. These are not the drawing limits. The chart shows the process is not stable: several signals, and all the high points are from the same lathe's stock.
Now the team knows where to look. The next step is to find the real reason, with a 5 Why.
The tools do not solve the problem. They show you where to look.
Which tool is for which stage?
- To see the problem: check sheet, stratification, Pareto.
- To find the cause: fishbone, scatter, histogram.
- To hold the gain: control chart.
A department head should know the tools of his own area. A quality head should know these seven. A production head should know the Lean tools. When a manager does not, he cannot admit it to his team, and he lives with stress.
Five mistakes to avoid
- Collecting data without a purpose. Ask first: what decision will this data support?
- Leaving out the machine and the lot. If the check sheet does not carry them, you cannot split the data later.
- Judging by the total. A total hides which machine or shift is the problem.
- Treating a link as proof. A scatter diagram suggests. The machine confirms.
- Mixing up control limits and drawing limits. Control limits describe what the process does. Drawing limits describe what the customer needs.
How to start in your plant
- Pick one product with a known rejection problem.
- Make a check sheet that records the defect, machine, shift and date.
- Collect data for a month, or a lot of 200 to 300 pieces.
- Draw a Pareto chart. Choose the biggest bar.
- Split that bar by machine, shift, lot and supplier.
- Hold a short fishbone session with the operator, setter and supervisor.
- Move to the 5 Why to find the real cause, then act and watch the chart next month.
Keep the first round small. Use a pencil and paper. Fix the input not the output: if the counting on the shop floor is careless, no chart will be right.
Key points
- The seven tools replace arguments with data.
- Start with the check sheet and the Pareto chart.
- Record the machine, shift and lot from the start, so you can split the data later.
- A scatter diagram suggests a link; check it on the machine.
- Tools show where to look. A method like 5 Why finds the cause.
