The data types of all independent variables in linear regression must be numeric.

• A. Numeric
• B. Categorical
• C. Boolean
• D. Text

Answer: (A)

In linear regression, the predicted value is formed as a weighted sum of the predictor values: ŷ = β0 + β1x1 + β2x2 + ... . This math only makes sense if each predictor xj is numeric, so you can multiply it by its coefficient βj and add the results. Non-numeric predictors can’t be used directly in this equation. Categorical variables must be converted into numeric features (for example, via one-hot encoding) before they can be included. Text data likewise needs to be transformed into numeric representations (such as bag-of-words or embeddings) before fitting the model. Boolean values can be represented as 0/1, which are numeric, but the underlying requirement is that the predictors fed into the model are numeric. Therefore, numeric is the appropriate choice.