here is a question one teammate asked. Please help to answer it.
The number of verbatim comments from the survey for X is in the table below.
Is there a sampling plan we can use for reading and categorization of these
that will get us to a 95-99% confidence level?
verbatim X
CA 181
TX 1254
FL 82
GA 697
OH 234
We want to ensure adequate representation of verbatims from each state. Do
we need to sample from each state specifically? Or can I mix all of the
states together and apply a random sample to the whole to achieve the same
thing?
F****n 发帖数: 3271
2
You need to sample each state if you want to capture state-based variation.
Just treat it as a stratified sampling. You can over-sample states with
small number of cases.
below.
these
【在 j******w 的大作中提到】 : here is a question one teammate asked. Please help to answer it. : The number of verbatim comments from the survey for X is in the table below. : Is there a sampling plan we can use for reading and categorization of these : that will get us to a 95-99% confidence level? : verbatim X : CA 181 : TX 1254 : FL 82 : GA 697 : OH 234