All MCQs

Tick marks indicate times when observation ended without the event.
They help viewers assess where the curve is supported by censored data.

For example, death from another cause can prevent observing disease recurrence.
Ordinary survival methods may misstate event probabilities when competing risks are ignored.

Kaplan-Meier calculations use those who could still experience the event.
Events and censoring reduce the risk set over time.

Median survival divides the estimated survival distribution in half.
It may be undefined if the curve never falls to 0.5.

Hazard concerns the immediate rate of event occurrence conditional on survival so far.
It is related to, but not identical with, event probability.

S(t) equals the probability that event time T is greater than t.
It starts near one and does not increase over time.

Kaplan-Meier uses observed events and risk sets to estimate survival probabilities.
The resulting curve is a descending step function.

Right-censored observations provide a lower bound on event time.
They remain part of the risk information up to their censoring time.

Censoring preserves partial information about event timing.
A censored subject may be known to survive beyond a recorded time.

Survival analysis models durations such as time to failure, recovery, or death.
It is designed to handle incomplete event times.

Forecasting should mimic using past data to predict later periods.
Training on future observations would produce an unrealistically optimistic evaluation.

Future outcomes are uncertain even when a central forecast is available.
Prediction intervals communicate how uncertainty typically grows with horizon.