On Tue, 7 Aug 2007, Philip Taylor wrote: > > That rounding seems quite misleading - if I haven't forgotten how to > do statistics, and if the details I am forgetting are not critical > ones, and if I'm not misinterpreting how you collected the data, then > the samples are independent [...]
The samples are not independent. In fact they have a high degree of overlap. > and from a binomial distribution that can be approximated as a normal > distribution with standard deviation sqrt(n*p*(1-p)), and if assuming > n=100,000 [...] Sorry, I meant in the order of 100 million, not 100 thousand. My bad. > and guessing p from the data then the 95%-confidence (+/- 2 s.d.) ranges > are something like: I don't know if the rest of it is correct or not (what distribution the data follows, etc). The fact that I haven't done these analyses is one reason why I'm skeptical about publishing this data more "officially". -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
