The DHS Program User Forum: Weighting data » Shall I drop observations with missing value Before or After weighting?

Home » Data » Weighting data » Shall I drop observations with missing value Before or After weighting?

Show: Today's Messages :: Show Polls :: Message Navigator

Shall I drop observations with missing value Before or After weighting? [message #12194]

Fri, 07 April 2017 04:04

AmsP
Messages: 24
Registered: April 2016

Member

My variable of interest is v133, but in DHS database there may be some observations whose v133 is missing (e.g. blank or 999). I am pooling databases of various survey rounds for a particular country into a single database. So I drop observations with missing values AFTER de-normalisation, but BEFORE weighting. Is this correct?

Thank you very much!

Report message to a moderator