Which statistic is based on the maximum vertical distance between the primary event EDF and the secondary event EDF?

Question

Options

• AKS
• BSBC
• CMax EDF
• DBrier Score

• A
  88% (30)
• C
  9% (3)
• D
  3% (1)

Explanation

KS (Kolmogorov-Smirnov) is correct because it is defined precisely as the maximum vertical distance between two empirical distribution functions (EDFs) - in a two-sample context, the "primary event" and "secondary event" EDFs. This maximum gap measures how differently two groups are distributed across the outcome range.

Why the distractors are wrong:

• SBC (Schwarz Bayesian Criterion) is a model selection penalty criterion based on log-likelihood and parameter count - it has nothing to do with EDFs.
• Max EDF is a plausible-sounding distractor but is not a standard named statistic; the actual statistic that uses the EDF maximum distance is called the KS statistic.
• Brier Score measures the accuracy of probabilistic predictions (mean squared difference between predicted probabilities and actual outcomes) - unrelated to EDF distances.

Memory tip: Think "KS = Keep the Suprémum" - the KS statistic is mathematically the supremum (maximum) of the absolute difference between two CDFs/EDFs. If you see "maximum vertical distance" + "EDF," that's always KS.

No community discussion yet for this question.