Measurement Variation

Measuring performance is the foundation of all Six Sigma improvement projects. Just like any other process, the measurement process has a certain degree of variation and it is important to recognize the potential implications of measurement error for decision making and ongoing process control. Measurement error is the extent to which the indication of a measuring instrument differs from a true value. There are many factors that can result in measurement error including operator variation, operator to operator variation, procedural variation, equipment variation, material variation, calibration, temperature, humidity, vibration, and others. Measurement error is mathematically described by the equation shown below.

At its most basic level, every measurement system includes measurement instruments, operators or appraisers, measurement methods, and work environment. Having capable measurement systems, properly maintained calibration systems and clear traceability to reference standards can prevent questionable analysis, erroneous conclusions and ineffective solution to the problem. A deliberate decision should be made on how sophisticated the measurement system has to be in order to get a reasonably accurate picture of reality. As a rule of thumb, the measurement system may already need improvement if more than 10% of the total variation is due to the measurement system. Measurement System Analysis (MSA) is performed to evaluate capability of the measurement system, identify sources of measurement variation, quantify measurement variance, and take necessary steps to ensure the accuracy and precision of measurement methods. A number of analytical methods are used to investigate key characteristics that contribute to the effectiveness of a measurement system including bias, linearity, stability, repeatability and reproducibility. Resolution or discrimination is another important parameter that shows ability of the measurement system to detect small differences or changes in the characteristic measured. The fundamental measurement system concepts and definitions are briefly outlined below.