standard error (se) and sample standard deviation

[sleepybird]

David,
You have the same formula for se and sample standard deviation on page 24 and 25. So are you saying se = sample standard deviation? or should se = sample standard deviation divided by √n?

Since we're dealing with sampling distribution here, why the denominator is not n-1 here?

Thanks.

 

[David Harper CFA FRM]

sleepybird, i don't see where we define the sample standard deviation on 24 or 25.
SE is a standard deviation but of the random variable that is the sample mean. So variance(sample mean) = sample variance/n, since population variance not available; and, stddev(sample mean) = SQRT[sample variance/n] = sample StdDev/SQRT(n).

Per the note on 25, you can include an (n-1) adjustment for small samples or without replacement, but it's esoteric (and not exam worthy). The exam loves to test the CLT idea that the sample mean tends to be normal with SE = StdDev (sample mean) = sample StdDev/SQRT(n)

David