The truncated distribution (a smooth curve with an abrupt ending on one side):
This is often caused by external forces to the process, such as screening, 100%
inspection, or a review process. Since truncating unusually indicates added
costs, it is a good candidate for improvement.
The isolated peak distribution (a small, separate, group of data to one side
of the parent group): Look for poor inspection, measurement errors or data
Edge-peaked distribution (a peak right at the edge): This usually means that
data from an otherwise long tail have been lumped together at one point (data
from outside the limit have been recorded as being inside the limit).
2.7.4 Pareto Analysis
It uses a specially organised histogram (the Pareto chart) to provide a picture that
instantly identifies those problems of the greatest concern, those problems that
should be addressed first.
2.7.5 Cause and Effect Diagram
As the name implies, this tool is just a group of causes and effects diagrammed
to show the interrelationships. The diagram is a form of tree diagram on its side
so that it looks like a fishbone. The name Ishikawa refers to the Japanese man
who conceived the chart.
The diagram is formed by first stating the problem in the form of an effect. In fig.
2.7, the problem is to maximise hardness, within limits. Hardness, then, is the main
effect. The main causes are then placed at the end of the lines extending up and
down at an angle from the centre line. These main causes are usually some form
of the following manufacturing factors like materials, equipment, work methods,
operators/workers, processes, tooling, management (policies), measurement,
environment, etc. the first three usually account for 80% of all problems.
Each of these main causes is now treated as effects and causes determined for
each of them, and then each of the secondary causes is treated as effects and
causes found for each of them, and so on. The procedure is continued until all
causes are identified.