You can use a multiple variable analysis to compare the capability for several variables or to compare the capability for several groups of the same variable.

For instance, you want to compare the process capability before and after a process improvement. After you perform an initial capability analysis, identify possible process improvements that you can implement to improve performance. After you change the process, perform another capability analysis to determine how much the capability has improved.

If the change has improved the capability, it is critical to implement control mechanisms on the process to sustain the process improvements. Without control mechanisms, the process may shift back to its initial capability.

In Minitab, each column in the worksheet typically represents a different variable. If you want to perform capability analysis on each of the variables contained in several different columns without having to run a separate analysis for each one, you can use the following procedures.

If you have continuous data, use the Capability Analysis - Multiple Variables commands to perform capability analysis on multiple columns, even if each column has different subgrouping or specification limits. The multiple variables commands can perform capability analysis on normal or nonnormal data, and also include options to analyze between/within capability.

For example, suppose you want to perform normal capability analysis on each of the columns C1, C2, C5, C10, and C15. The subgroup sizes are 2, 3, 5, 5, 5, the lower specification limits are −3, −3, −2, −2,−1, and the upper specification limits are 3, 3, 2, 2, 1, respectively.

- Choose .
- In Variables, enter
`C1 C2 C5 C10 C15`. - In Subgroup sizes, enter
`2 3 5 5 5`. - In Lower spec, enter
`-3 -3 -2 -2 -1`. - In Upper spec, enter
`3 3 2 2 1`. - Click OK.

You can use command language to perform

or on multiple columns individually, if they have the same sample size.Suppose you want to run

on each of C1, C2, C5, C10, and C15. The sample size is 500 for each sample.- Choose .
- Type the command
`PCAPA C1 C2 C5 C10 C15 500`.