A household survey is being conducted in which a selected primary sampling unit (PSU) could not be included because of operational inconvenience due to difficulty in reaching the unit and also lack of limited time for survey. Due to this an additionally selected (randomly selected) unit was surveyed. How can this change be addressed in the sample design with justification?
Since the added units were randomly selected, you are good. Re-designing would have been called for if the units were conviniently added, in which case you sampling would be non-probablistic.
Systematic, stratified, cluster and multi-stage samplings all fall under probabilistic sampling!
Like Adrian said, just highlight the addition and protocol of the random selection, but not necessarily as a limitation.
I agree with Adrian and Mohammed, but to be very accurate there is one more issue to consider: Why exactly the other unit dropped out and whether there is reason to believe that the drop-out is in some form correlated with variables of interest. But assuming that you have other cases of unit non-response and that the overall sample is rather large, this would probably mean being unduly meticulous.
The unit was dropped since it was in a very remote area and needed to be reached by walking for hours. The survey had to be completed within the alloted time.
- Badges
- Science topic
- Similar topics
- Political Science