Thalatte said:Oh wow didn't hink of using such large samples...that changes what I can do. I can provide some money towards it then.
Yeah...the reality of these foods, is however hard a company tries, no two lots will be the same...almost like dying yarn. So, you could test a small sample, but the reality is there is no guarantee your next lot will be the same. That's why even the data available on packaging is "min" or "max". Once you get up to 30+ samples, the mean data from the sample then is approximately distributed normal (...add some more statistical terminology here most people won't understand...), etc. At that point, one can then truly estimate what the typical content of the food is.