I am performing stratified random sampling on a population (country) containing ~20 strata (states). I need to calculate mean of the population. I also know the proportion of the strata. My question is: Should I have a minimum sample size (say >=30) in each strata/state? I have greater than 30 samples for the country, but I do not have greater than 30 samples for each state. My assumption is since I am calculating mean for the population (country), at a state level, I can as well have 1 or 2 samples.

Will this be an issue in estimating the means? Please let me know.



1 Answer 1


Minimum sample size is useful once you consider that units might not respond. However, if you make it too large your design is going to be less efficient. If you aim for about $5$ reponding units should be enough. Bump this up with an assumed response rate $5/rr $. $30$ is more than enough - but will probably mean you are putting sample where it isn't needed for a good estimate of the total


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.