2
$\begingroup$

I want to compute the standard deviation of some data points that I obtain during four series of experiments.

For the first three experiments that I have conducted, the number of data points that I obtain is quite limited and I am able to compute the "classic" population standard deviation. For the fourth experiment however, I obtain a very large number of data points (more than 2^48) for which I can not compute the population standard deviation. I can nethertheless compute the sample standard deviation on say, the first 2^32 data points.

I am then wondering, is it right to compare the population standard deviation of the first three experiments with the sample standard deviation of the fourth experiment ?

Thanks

$\endgroup$
0

1 Answer 1

2
$\begingroup$

Wikipedia has a discussion of this. For a large sample, your sample standard deviation is very close to the population standard deviation. I would be more concerned about bias from taking the first samples. It would be better to take a random sample in case the first data is somehow different from the rest.

$\endgroup$
0

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .