Comparison of Evaluation Criteria in the Use of Measurement System Based on Regression with Gauge R&R Study

Abstract

This project involves two systems of measurement evaluation: a Method based on Regression [1] and a Gauge R&R study [2]. There are different tools for the analysis, control and improvement of processes but this paper will only address those that involve the data of the realized measurements. These tools will be defined by a Method based on Regression and the Method of the ANOVA in the Gauge R&R Study [3]. Besides using different criteria to accept or to reject measurements realized under certain conditions, the approach used a number of significant numerical models that meet certain statistical conditions to be evaluated using both tools. This way it will be possible to compare the results shown by the Regression and a Gauge R&R Study by method ANOVA. By obtaining the number of different categories [4] and the relation between the projections in the measurements, it is possible to know if what is reliable for one system is also suitable for the other. Finally, this thesis proposes a comparative table of criterion of evaluation of both systems, showing where the line between acceptance and rejection is broken.