True or false: The values true/false or 0/1 would both be valid combinations for the dependent variable in a logistic regression model.

• A. False
• B. True
• C. Not Sure
• D. Depends on Data

Answer: (B)

Binary outcomes are what logistic regression is built to predict. The dependent variable should have two states, representing the event of interest. These two states can be represented in data as 0/1 numeric values or as boolean values true/false. In practice, true maps to 1 and false maps to 0, so the model estimates P(Y = 1 | X). As long as there are exactly two categories and the mapping is consistent (which is generally handled by the software), either encoding works. This is why the statement is true. If you had more than two classes, you’d use a different setup, but for a binary outcome, 0/1 and true/false are both valid representations.