Excerpted from the book “Statistical Quality Control,” translated and edited by Cheng-Hsien Chang, published by Hwa Tai Publishing
With respect to quality, there are two general dimensions:
quality of design and quality of conformance.

 
All products and services are made in many different ways, and people often use this term without clarifying whether we are talking about quality of design or quality of conformance.
Achieving quality of design requires clear decisions during the product or process design stage to ensure that certain specific functional requirements can be satisfied.

 
For example, the designer of an office copier might design a circuit component with more parts because they know this will improve the product’s reliability, increase the copier’s mean time between failures, and thereby reduce after-sales service requirements, so that consumers will be more satisfied with the product’s performance.
Products whose quality of design is established in this way often have higher production costs; however, these added costs are in fact prevention costs, incurred to prevent quality problems from arising in the late stages of the product life cycle.
Quality of conformance is often improved by changing certain aspects of the quality assurance system, such as using statistical process control methods or changing the type of inspection procedures.
Higher quality of conformance often means lower total cost, because it leads to fewer scrap and rework units, a lower defect rate, and less service.
Quality control is an engineering and management activity through which the quality characteristics of a product are measured, compared against specifications and requirements, and once a difference between actual performance and the standard is found, appropriate corrective measures are taken.

 
Most enterprises believe that they cannot provide customers with flawless products.
The biggest reason lies in variability, which is precisely why no two products are exactly alike.
For example, the blades of a jet turbine engine impeller vary in thickness even within the same impeller, let alone between different impellers; if this difference is very small, it does not affect the customer, but if it is too large, it will be unacceptable to the customer.
The causes of product variation come from differences in raw materials, differences in the operation and performance of process equipment, and even differences in the performance of operators in their work.
We therefore define quality improvement as the reduction of variability in products and processes.
Since only statistics is the best tool for expressing variability, statistical methods have become an indispensable tool in quality improvement engineering.