Need clarification on a question in SAS exam preparation

[Siuli]

Hi,

Please find the attached document. As per my understanding option a is not appropriate as it increases the size of the training data. But option C was highlighted by SAS as correct option.

Please can someone help me to understand.

Thanks,

Siuli Basu

[Tom]

I assume what they are talking about in A is to add the observations with missing values for the outcome variable. Because they are missing the outcome variable they will be excluded from the logistic regression, but they will have the dependent variable values available for the procedure to predict the score.

[Siuli]

Hi,

Thanks for your reply. Even if we add new data with missing values of Y variable to the training data set but the logistic model needs to rerun to score the new observations, right? In that case also it increase the size of the data, given that in predictive modeling the input data is huge.

In the explanation written in blue states the same if I understand correctly and that is why I thought option A is correct.

However, I am not sure why option C is selected as not an appropriate way as that is a one of the standard procedures to save the estimates from logistic and use that estimate to score new observations.

Kindly let me know what I am missing here.

Thanks,

Siuli Basu