You can use Individual Distribution Identification functionality to perform capability analysis when the distribution of your data is unknown. The validity of statistics depends on the validity of the assumed distribution.

Use Individual Distribution Identification to compare the fit of up to 14 distributions and 2 transformations: normal, Box-Cox transformation, lognormal, 3-parameter lognormal, gamma, 3-parameter gamma, exponential, 2-parameter exponential, smallest extreme value, largest extreme value, logistic, loglogistic, 3-parameter loglogistic, Weibull, 3-parameter Weibull and Johnson transformation. The two transformations can be used to correct nonnormality in data

Minitab provides both graphical (probability plots) and numerical (goodness-of-fit statistics) output for comparison. The quantitative output also includes the estimates of the distribution parameters and percentiles for each distribution.

A company wants to assess the calcium content in vitamin capsules. You take a random sample of 50 capsules and record their calcium content. The distribution of the data is not known.

Data: Capsules.MTW (available in the Sample Data folder).