Which of the following describes a concordant pair of observations in the LOGISTIC procedure?

Question

Options

• AAn observation with the event has an equal probability as another observation with the event.
• BAn observation with the event has a lower predicted probability than the observation without the event.
• CAn observation with the event has an equal predicted probability as the observation without the event.
• DAn observation with the event has a higher predicted probability than the observation without the event.

• A
  3% (1)
• C
  5% (2)
• D
  92% (35)

Explanation

In logistic regression, a concordant pair is formed by taking one observation where the event occurred and one where it didn't, and checking whether the model ranked them correctly. Option D is correct because "concordant" means the model assigned a higher predicted probability to the observation that actually had the event - the model's ordering agrees with reality.

Why the distractors are wrong:

• A is irrelevant - comparing two event observations to each other has nothing to do with concordance, which always involves one event and one non-event observation.
• B describes the opposite: when the event observation has a lower predicted probability than the non-event observation, that is a discordant pair (the model got the ranking backwards).
• C describes a tied pair - equal predicted probabilities for an event and non-event observation are neither concordant nor discordant.

Memory tip: Link "concordant" to "correct" - a concordant pair is one the model got right, meaning the event observation scores higher than the non-event one. Discordant = the model got it backwards, Tied = the model couldn't decide.

No community discussion yet for this question.