In this tab, you will specify the measured variables (sometimes called predictor variables or just X variables) to use for PCA. The analysis is performed completely on the measured variables, allowing you to determine the underlying structure of the variables, identify clusters of variables or rows, and visualize your data.

Variables for analysis

Choose at least two continuous variables to include in the PCA. Categorical variables cannot be analyzed using PCA. Remember that with PCA you don’t need to designate a response, or Y, variable. If you plan to run Principal Components Regression, do NOT select the dependent (outcome) variable when choosing variables for analysis.

Perform Principal Component Regression (PCR)

If you wish to run Principal Component Regression (PCR) on the dimension reduced Principal Component Scores, check this option and choose the dependent (outcome) variable.  The dependent variable must be continuous and cannot also be a selected variable, so you can only choose from continuous variables that you did not include as variables for analysis.