Callistemon, having expressed deep suspicions about this sample, I am now going to say that you can have a sample of 4103 that is representative of the total population.
The ability to do this depends on how your sampling frame is set up. Somewhere I still have the set of statistical tables that shows me the probability that a random sample of a given size taken from a large population will be a cross section of said population. It is disconcerting how small a sample can be, to give you a 95% certainity that your sample is representative of the whole population.
However the secret is exactly who makes up the population you are drawing your sample from and, as I have said, the phrase politically representative people is a very imprecise term and could hide a multitude of biasses and exclusions.
ALPHABETICAL FOOD AND DRINK (Jan 26)
🦞 The Lockdown Gang still chatting 🦞


