Mean and variance rules are more powerful or selective than quality control rules based on individual values

Area of Science:

• Quality Control
• Statistical Process Control
• Analytical Chemistry

Background:

• Traditional quality control (QC) methods rely on individual values or simple statistical measures.
• Complex multi-rules have been proposed for computerized QC systems.
• The efficacy of different QC rule types requires comparative analysis.

Purpose of the Study:

• To compare the performance of individual value QC rules against mean and variance rules.
• To evaluate the power and selectivity of different QC rules under various conditions.
• To determine the most effective QC rules for computerized systems.

Main Methods:

• Theoretical computations and simulations were employed.
• Comparison of simple (1(3)s) and combined individual value rules with mean and variance rules.
• Evaluation of performance based on type I errors and detection of location shifts or scatter.

Main Results:

• Mean rules demonstrated greater power in detecting location shifts compared to individual value rules at identical type I error rates.
• Mean rules exhibited better robustness against non-normal data distributions.
• Variance rules showed higher power for detecting increased scatter and superior selectivity over individual value rules.

Conclusions:

• Simple mean and variance rules are computationally efficient and offer increased power or selectivity.
• Traditional mean and variance QC rules are preferable to complex multi-rules in computerized quality control.
• The findings support the continued use and implementation of established QC methodologies.