Skip to content

IAI Actuarial Core Principles · Risk Modelling and Survival Analysis · Estimation procedures for lifetime distributions

Which statement about non-informative censoring in likelihood construction for survival data is correct?

Non-informative censoring means the censoring mechanism is unrelated to the lifetime parameters, so its terms factor out and can be ignored when maximising. Censored lives still contribute survival probabilities; removing them would bias the estimates.

  1. ACensoring times must be independent of the lifetimes' parameters, so the censoring mechanism factors out of the likelihood for the parameter of interestCorrect
  2. BCensored lives must be removed from the likelihood to avoid bias
  3. CCensoring is non-informative only if every life is censored at the same time
  4. DCensored lives contribute the density at the censoring time
  5. Non-informative censoring means no deaths occur after the censoring time

Explanation

Under non-informative censoring the mechanism does not depend on the lifetime parameters, so its terms are constant factors in the likelihood and can be ignored when maximising. Removing censored lives would discard information and bias estimates.

Did you get it right without looking?

One question tells you little. A timed set on Estimation procedures for lifetime distributions shows your real accuracy, how long you take and where you lose marks.

Practise Estimation procedures for lifetime distributionsFree diagnostic mock

More Estimation procedures for lifetime distributions questions