▪Linearity: It is the study of variation in results of measurements along the gage scale. It aims to ensure that the measurement results are reliable at any point in the range of the gage scale. This study only applies to gages with a scale and should be performed by a single operator.
|
▪Range: It is a study of the measurement gage that quickly provides an approximate value for the measurement divergence. Operators measure each part only once. This method uses two operators and five samples of part to be measured.
|
▪RR (average and range/ANOVA): This is the mathematical method to estimate the combined variation of repeatability and reproducibility of measurement systems by variable. This process can be accomplished in the following ways:
|
oAverage and Range: It is an approach that provides an estimate of the repeatability and reproducibility of a measurement system, allowing the decomposition of the variation. oANOVA: It is a standard statistical technique that can be used to analyze the measurement error and other variability sources of the data relevant to a study of the measurement system. |
|
▪Bias: The difference between the observed average of measurements and a reference value. Bias can be determined by the following methods:
|
oIndependent sample method: Bias is calculated based on the average of readings performed on the part. oChart/Stability method: Bias is calculated based on a control chart. |
|
▪Cross tabulation: This method is used when the team does not know the nominal decisions of reference for the parts. For this reason, comparative tables were developed between the appraisers, with the purpose of determining the extent of the existing agreement between them.
|
▪Signal Detection: As an alternative approach one can use this method to determine approximately the magnitude of a given region, and from there determine the Repeatability and Reproducibility of the measuring device.
|
▪Details: In this method, the stability of the process must be checked and monitored, if necessary. To do that, the most common way to verify the stability is to plot the results of a constant sample over time on a graph. |
