Which description best distinguishes parametric, non-parametric, and semi-parametric approaches?

• A. Non-parametric methods use a fixed distribution form.
• B. Parametric assumes a fixed distribution form; non-parametric uses data-driven shapes; semi-parametric blends parametric tails with nonparametric components.
• C. Semi-parametric models ignore distribution tails.
• D. Parametric methods require no data.

Answer: (B)