When you perform a nonnormal capability analysis in Minitab, the overall capability indices are calculated by default using the Z-score method.
To calculate the Z-score, Minitab first determines proportions of observations that are outside the specification limits based on the nonnormal distribution that you specified for the analysis.
The probabilities P1 and P2 are then used to calculate the corresponding Z-values for a standard normal distribution. In this way, the original specification limits can be represented by corresponding Z values on the standard normal scale.
To determine the specification spread in the standard normal distribution, the value of Z.LSL is subtracted from the value of Z.USL.
The overall capability indices are then calculated based on the relationship between the standard normal specification spread and a specific spread of the standard normal distribution.
Pp is the ratio of the standard normal specification spread to the 6-sigma spread (6), which captures 99.74% of the process measurements from the standard normal distribution.
PPL is the ratio of the one-sided specification spread relative to the lower spec limit (Z.LSL) over the one-half the 6-sigma spread (3).
PPU is the ratio of the one-sided specification spread relative to the upper spec limit (Z.USL) over one-half the 6-sigma spread (3).
Ppk is the minimum of PPU and PPL. Therefore, in this example, Ppk equals PPU ≈ 0.43.