A telecommunication company wants to predict their customers' purchase decisions from four service options (categorical value). Using IBM SPSS Modeler Professional, which modeling node should be used?

Question

Options

• ALinear
• BLogistic
• CSuperNode
• DRegression

• A
  3% (1)
• B
  7% (2)
• C
  13% (4)
• D
  77% (23)

Explanation

Predicting a categorical outcome with four classes requires a modeling approach that handles multi-class classification, and in IBM SPSS Modeler Professional, the Regression node (D) covers this through multinomial logistic regression - making it the correct tool when the target variable is a nominal category (e.g., which of four service plans a customer chooses).

Linear (A) is wrong because it predicts continuous numeric outcomes (e.g., revenue), not discrete categories. Logistic (B) is a tempting distractor, but in SPSS Modeler's node architecture, a standalone "Logistic" option handles simpler binary problems, whereas the full Regression node encompasses the multinomial logistic capability needed for four-category prediction. SuperNode (C) is not a modeling algorithm at all - it's a container used to group and encapsulate a set of nodes within a stream for organizational purposes, unrelated to prediction.

Memory tip: Think "Regression rules them all" - when you see a categorical target in SPSS Modeler, reach for the Regression node because it subsumes logistic modeling. If the answer choices include both "Logistic" and "Regression," the exam is testing whether you know that Regression is the broader, more capable node for multi-class categorical targets.

No community discussion yet for this question.