How does the assumption of MAR differ from MCAR in terms of data missingness?

  • MAR assumes the missingness is only related to the observed data
  • MAR assumes the missingness is related to the unobserved data
  • MAR assumes the missingness is unrelated to any variable
  • There's no difference between MAR and MCAR
In MCAR, the missingness is completely random and doesn't depend on any variable. In MAR, the missingness is not random but is related only to the observed data, not the unobserved (missing) data.
Add your answer
Loading...

Leave a comment

Your email address will not be published. Required fields are marked *