Some notes on the probability space of statistical surveys