Good point. On the other hand, the confidence interval I provided is correct as long as the original survey and the repeat are sampled in the same way (regardle

Good point. On the other hand, the confidence interval I provided is correct as long as the original survey and the repeat are sampled in the same way (regardless whether it is un-biased sampling or not)