A marketing manager attempts to determine those customers most likely to purchase additional products as the result of a nation-wide marketing campaign. The manager possesses a historical dataset (CAM

Question

Options

• Aproc logistic data=MYDIR.CAMPAIGN descending; class Homeowner; model Respond = Income Homeowner; run;
• Bproc logistic data = MYDIR.CAMPAIGN descending; by Homeowner; model Respond = Income Homeowner; run;
• Cproc logistic data = MYDIR.CAMPAIGN descending; model Respond = Income Homeowner; run;
• Dproc logistic data = MYDIR.CAMPAIGN descending; class Homeowner; model Respond = Income Homeowner; run;

• A
  80% (20)
• B
  12% (3)
• C
  4% (1)
• D
  4% (1)

Explanation

Option A is correct because it includes both the class Homeowner; statement - required to tell SAS that Homeowner is a categorical (nominal) variable rather than continuous - and the descending option, which instructs PROC LOGISTIC to model the probability of Respond=1 (a purchase) instead of the default Respond=0.

Option B is wrong because by Homeowner; runs separate logistic models for each level of Homeowner (Y and N), rather than including it as a predictor variable in a single model - this does not answer the manager's question.

Option C is wrong because it omits the class statement; without it, SAS treats Homeowner as a numeric/continuous variable, which is inappropriate for a Y/N categorical predictor and will produce misleading results.

Option D appears structurally identical to A in this rendering, but in the original exam context likely contains a subtle error (e.g., a different dataset path, missing library reference, or omitted descending) - always read option details character by character on exams.

Memory tip: Think "Categorical needs Class" - any time a predictor has character/text values (like Y/N), you must declare it with a class statement, and pair descending with binary targets whenever you want to model the event (1) rather than the non-event (0).

No community discussion yet for this question.