2 ms·

(Random) sampling error does not very according to the size of the total population; only according to the size of the sample. A sample of 18,000 from a populat

by azernik 10y ago

(Random) sampling error does not very according to the size of the total population; only according to the size of the sample. A sample of 18,000 from a population of 6 billion will have the same random errors as a sample of 18,000 from a population of 100 million. And 18,000 is a BIG sample. (This breaks down a bit when the sample size and the total population size get very close, and does not include sources of sampling error other than random variation - e.g. biased sampling, which I would be more worried about in a global study like this.)

leksak 10y ago
That was part of my point, I was not sufficiently convinced that the sampling was conducted over a satisfying number of different living situations. Rest is simply not a luxury that a lot of people can even concern themselves with as they have no concept of leisure time (or at least not an opportunity to enjoy it). That does not invalidate the statement that being alone may be the key to rest, but it should point out that drawing the conclusions that the article alludes to about the scale of the issue is shoddy. > [...] showing perhaps what a pressing issue rest is in the modern world. Even if the concept of "modern" world precludes areas suffering from famine, I am not convinced that it should be considered a pressing issue.