A novel evaluation model for measurement system analysis

Six-Sigma management has been considered as a powerful business strategy that employs a well-structured continuous improvement methodology to reduce process variability and to raise quality within the business process effectively using statistical tools and techniques. Many quality improvement programs employ measurement system analysis (MSA) in measure step of Six-Sigma to ensure the reliability of measurement results, which are the basis for decisions regarding the behaviour of critical quality characteristics. The repeatability and reproducibility (RR) study is an important approach for evaluating the precision of MSA. Although many measurement capability studies exist and are common in industry, few have discussed cases with attribute data, especially on the case related to two or more related quality characteristics simultaneously. This study presents a novel model for evaluating RR variance for bivariate attribute data. An example is utilised to illustrate the application process of the proposed model. The proposed model is illustrated to be capable to assess and to improve measurement systems with bivariate attribute data.