If an observation in the input data set contains a missing value for any of the regressors or dependent values, that observation is dropped from the analysis.
Note: This procedure is experimental.