| http://www.w3.org/ns/prov#value | - Small sample sizes are a problem when a) the effect size is small, and a large sample is needed to obtain sufficient statistical power to separate signal from noise or b) there is a possibility of selecting a small number of people who fall on the extremes of a distribution - oops, our mean income was off by a few million because of the chance inclusion of Bill Gates in the sample.In this case, st
|
