Gage Repeatability and Reproducibility
Below is a step by step example of how to determine repeatability and
reproducibility. The example utilizes the
software package Measurement Assurance. Click here to download a free demo version of
Click here for the technical details of the determining repeatability and reproducibility in Adobe Acrobat format. For the same article in Microsoft Word for Windows 95 format click here.
Select "Procedures" then “Repeatability & Reproducibility” from the menu and enter your data. An example Repeatability & Reproducibility screen is shown below. In the “Method” frame, the “ANOVA Method” or the “Range & Average Method” can be selected for analysis. The ANOVA method is used as the default because it is more accurate. In the “Based On” frame, the analysis can be conducted using the process variation or the process tolerance; process variation is the default. The default output is the 99% R&R index. The level of the index can be changed from 99% to any desired level using the “Index Level” box on the toolbar.
To obtain the results, and to display other options, click on the “Compute” button. When this is done, the “Charts” button is enabled, and if the “ANOVA” method is used, the user is shown the R&R table with an option of viewing the ANOVA table. This is shown in the figure below.
To view the ANOVA table click the "ANOVA Table" button in the "Output Type" box. The figure below shows the ANOVA table that accompanies the results shown above.
The ANOVA table is useful in determining which sources of variability are significant statistically. In general, if the “Significance” value is less than 0.05, than the source of variability is significant. In the figure above, the variability due to the appraisers and the variability due to the parts is significant, while the variability due to the interaction between the appraisers and the parts is not significant. The closer the significance level is to 0, the more likely that the source of variability is significant, and the closer the significance level is to 1, the more likely that the source of variability is not significant. Standard practice is to use a significance level of 5% as the decision point, with anything above 5% not considered significant.
The figure below shows the same analysis using the range and average method. Notice the results differ from the results using the ANOVA method.
Click the chart button to gain access to a variety of charts to aid in interpreting the r&r data. The figure below shows the charting screen that accompanies the repeatability and reproducibility screen. The range control chart is the default, and all appraisers can be displayed or a specific appraiser can be displayed by selecting from the “Choose Appraiser” frame.
Clicking the “Edit Graph” button or right clicking the mouse while the mouse is over the graph brings up the screen shown
This dialog box can be used to edit the graph color, background, titles, markers, scale, and other properties. It can also be used to export the graph in a variety of formats. A detailed description of the functions are found by selecting the “Help” button at the bottom of the dialog box.
Engineered Software, Inc.