The Attribute Gage Study examines the precision of a measurement system when measurement is binary. It calculates the amount of bias and repeatability of the measurement system. This study only give indicative information on whether a part is good or bad based on a set of predefined standard. To use this tool, you need to select parts of different values and you should know the standard value of each part. If you are using the AIAG method, you must at exactly 20 trials per part. If you are using the regression method, the minimum is 15 trials per part. However, to obtain good estimates of bias and repeatability, 20 trials per part is recommended.
Addition requirements are as follows:
Data collected for at least eight different parts
Smallest part are all bad
Largest part are all good
For AIAG method, exactly six parts have a mixture of good and bad units. For regresion method, more than six parts are allowed
If a lower tolerance limit is specified, the part with the smallest standard value are all bad and the part with the largest standard value are all good. Also the number of good parts should increase as the standard value increases
If a upper tolerance limit is specified, the opposite must be specified
At the Excel Menu (For Excel 2007, go to Add-ins first)
Choose ProcessMA > Quality Tools > Gage R and R > Attribute Gage Study
In Part, select the column containing part indicators (>=2 distinct values)
In Reference, select the column containing reference values or known standards (Numeric)
In Acceptance, select the column containing number of accepted trials (Integer, >=0)
In Trial, select the column containing number of trials (Integer, >0)
In Constant No. of Trial, enter the number of trials if it is constant (Integer, >0)
In Tolerance Limit to Use, choose if you want to use the lower limit or upper limit
In Tolerance Limit, enter a value for the tolerance level (Numeric)
In Method, choose the method use to test the bias of the gage
Click OK
You want to determine the bias and repeatability of an attribute (pass/fail) measurement system. You measured 10 parts of different reference values, 15 times each. You know that the system has a lower tolerance of 0.2 and an upper tolerance of 1.0.
Open data worksheet by choosing ProcessMA > Tools > Data
Choose ProcessMA > Quality Tools > Gage R and R > Attribute Gage Study
In Part, select AW - Part_Att
In Reference, select AX - Reference_Att
In Acceptance, select AY - Acceptance_Att
In Constant No. of Trial, enter 20
In Tolerance Limit, enter 0.2
Click OK
The bias and repeatability is shown in the output. The p-value for the test for bias is 0, suggesting that bias is present in the measurement system.
Copyright © ProcessMA