Introduced on day 6
MNAR (missing not at random)
Missingness that depends on the unobserved value itself, so no analysis of the observed rows alone can remove the bias.
In detail
Example: fares withheld when they were disputed or unusually low. Handling requires extra data on why values are missing, explicit assumptions, or a sensitivity analysis over plausible values.