If there is a missing value for any model variable in an observation, the entire observation is dropped from the analysis.
