Assuming the set is homogeneous! Which in this case, is not. A sample extracted from a more or less financially stable region of the country will have drastically different output than a sample extracted from a poorer area. Your assumption about 'the soup well mixed' is less probable to be true, given the lack of any evidence supporting the claim that this study is either meaningful or properly executed.
Re:Which country read the most
by shalmanese on Monday March 03, 2014 @04:40AM
(#7332845)
Right, if there's a flaw in the study, it's in the representativeness, not the sample size.
An analogy I've found useful thinking about this is, if you have a pot of soup, tasting a teaspoon is enough to figure out whether the soup is oversalted or not, regardless of the size of the soup pot, providing the soup is well mixed.