Minitab displays information about the method used to analyze the PLS model. The output depends on which method you choose to select the components and on whether you use cross-validation.

If you do not use cross-validation, Minitab displays:

·    Whether the components to calculate are set or user-specified (when you specify the maximum number of components).

·    The number of components to calculate, which is, by default, 10 or the number of predictors, whichever is less.

If you use cross-validation, Minitab displays:

·    How cross-validation was performed.

·    Whether the components to evaluate are set or user-specified (when you specify the maximum number of components).

·    The number of components to evaluate. This equals the maximum number of components you specify or, by default, 10 components or the number of predictors, whichever is less.

·    The number of components selected by cross-validation, based on the model with the highest predicted R-squared.

If categorical predictors are in the model, the method table specifies the type of coding used.

The components selected by Minitab form the model that is used in the PLS analyses and plots.

In this example, the scientists chose to use cross-validation, leaving out one observation at a time. They did not specify the maximum number of components to validate. The output shows that:

·      Minitab left out 1 observation at a time during cross-validation.

·      Minitab evaluated a total of 10 components, which is the default.

·      Minitab selected 10 components for the PLS model.