I have survey responses from a survey where the sampling plan was stratification with sampling rates proportional to the square root of the number of households.
The plan created 12 strata, and the base weight for each stratum was to be computed as the ratio of the proportion of the total households for stratum h to the proportion of total returns for stratum h. The first answer to the question asked here describes how the base weights were calculated.
However, once the survey responses came in 6 of the 12 strata had such a low number of responses that it was necessary to pool the responses from the most similar strata. This resulted in 4 of the original strata being left as is (they had enough responses), while the other 8 were grouped into 3 new consolidated strata.
Can I use the same base weights as before, or do the base weights for the 7 redefined strata need to be recomputed?
If the base weights need to be recomputed, how would I do that?
