1.

Open the Wafer.jmp sample data table.

2.

Select Analyze > Quality and Process > Measurement Systems Analysis.

3.

4.

5.

Notice that the MSA Method is set to EMP, the Chart Dispersion Type is set to Range, and the Model Type is set to Crossed.
6.

Click OK.

Take a closer look for interactions between operators and parts. From the red triangle menu next to Measurement Systems Analysis, select Parallelism Plots.
Take a closer look at the variance between operators. From the red triangle menu next to Measurement Systems Analysis, select TestRetest Error Comparison.
Use the Shift Detection Profiler to see how the probability of detecting a shift in your process changes when you alter the parameters or add tests. From the red triangle menu next to Measurement Systems Analysis, select Shift Detection Profiler.
Change the parameters to see what your chances are of detecting a 3 sigma shift when only one test is selected. For the Part Mean Shift value, change 2.1701 to 6.51 (2.17 multiplied by 3). Your probability of detecting a 3 sigma shift goes up to almost 96%.
Next, change the parameters to eliminate bias and see how that affects your chances of detecting warnings. For the Bias Factors Std Dev value, change 1.1256 to 0. Your probability of detecting warnings goes up to almost 99%.