Dear all,
I received only two answers to my question regarding sampling for a study to detect the difference of prevalence between two populations regarding a chronic disorder.
I thank E.Corda and Cor Stolk for that.
The summary is:
1. If the two populations are similar except for geographical location then we may take
two independent samples of 1311 each through simple random sampling from the population
stratified according to the geographical location. If, on the contrary, the two populations have little in common, then these are two different populations, and one needs to estimate the prevalence in each population with the same accuracy and confidence level.
2.Finite Population Correction is to be incorporated in the results if the sample is >= 5% of the population. If we consider the sample as a stratified sample from one population then heterogeneity of sampling probabilities as well as FPC become irrelevant.
Regards
Basharat
[log in to unmask]
|