Dear all stat member,
I have been asked to provide some tables accompanied by information that
will allow for the calculation of sample/non-response weights and complex
standard errors.
As I didn't dare to ask what she meant by
'information that will allow for the calculation of sample/non-response
weights and complex standard errors'
Would anyone on the list be kind enough to explain me how to calculate non
response weight. I assume the outcome is to try to correct the biases
created by non response for every variable, but how can I calculate this if
by definition we don't know what would have been their answers ?
What is the meaning of 'complex' into complex standard error ?
Looking forward to learn something from all of you guys,
Best wishes,
Olivier
|