Setting reference levels in General Linear Models
main topic       

When you include a categorical predictor in your general linear model, Minitab assigns one categorical level as the reference level. Minitab designates the reference level based on the data type and coding scheme. You can choose the coding scheme in the Coding subdialog box.

For -1, 0, 1 coding:

·    For numeric categorical predictors, the reference level is the level with the largest numeric value.

·    For date/time categorical predictors, the reference level is the level with the latest date/time.

·    For text categorical predictors, the reference level is the level that is last in alphabetical order.

For 1, 0 coding:

·    For numeric categorical predictors, the reference level is the level with the least numeric value.

·    For date/time categorical predictors, the reference level is the level with the earliest date/time.

·    For text categorical predictors, the reference level is the level that is first in alphabetical order. If you defined a value order for a text categorical predictor, this rule does not apply. Minitab designates the first value in the defined order as the reference. See Ordering Text Categories.

If you choose 1, 0 coding, you can change the default reference level in the Coding subdialog box.

Regression creates a set of design variables for each categorical predictor in the model. See Coding factors in General Linear Models for examples.