Example of Type 1 Gage Study
main topic      interpreting results      session command      see also 

An electrode manufacturer wants to evaluate the measurement system that measures the outer diameter of electrode shafts to determine whether the system accurately measures the shaft within the 0.05 mm tolerance.

An operator measures a reference shaft with a known outer diameter of 12.305 mm 50 times.

1    Open the worksheet SHAFT.MTW.

2    Choose Stat > Quality Tools > Gage Study > Type 1 Gage Study.

3    In Measurement data, enter Diameter.

4    In Reference, type 12.305.

5    Under Tolerance, choose Upper spec - lower spec and type 0.05. Click OK.

Graph window output

Interpreting the results

The bias in the measurement system is -0.00231. The test for bias indicates a significant difference from 0 (T = 4.506892, p-value = 0.000), suggesting that bias is present in the measurement system. Also, notice that many of the observations plotted on the run chart are lower than the reference value of 12.305.

Cg compares tolerance with measurement variation, while Cgk compares tolerance with both measurement variation and bias. Large Cg and Cgk indicate that variation due to the measurement system is small compared with tolerance range. A typical threshold for both these values is 1.33.

In this example, the capability indices are Cg = 0.46 and Cgk = 0.25, which are much lower than 1.33. These values suggest that variation due to the measurement system is large.

%Var (repeatability) is determined by Cg and %Var (repeatability and bias) is determined by Cgk. Small %Var values indicate small measurement variation compared to the tolerance. The capability indices threshold of 1.33 correspond to a % Var threshold of 15%.

In this example, %Var (repeatability) = 43.49% and %Var (repeatability and bias) = 81%. These values are much larger than 15%, which again suggests that the variation due to the measurement system is large.

These results indicate that the measurement system can not measure parts consistently and accurately and therefore should be improved.