It is common for non-statisticians to set the sample size for sampling from
a finite population of n items to sqrt(n)+1. I have never found a
statistical or non-statistical justification for this and would appreciate
justifications or equally to destructions of this rule.
Thanks and best wishes .... Max
Max Porter
PPT Manager- Statistics, Primary Supply, GMS
Phone Ulverston (7-736 or 01229-48) 2482
Fax (7-736 or 01229-48) 2257
|