置信区间里到底是SD还是SE?

NetDagger

Common Probability Distributions那章的LOS: Construct and explain confidence intervals for a normally distributed random variable里面说:
the 90 percent condidence interval for X is X-bar -1.65s to X-bar +1.65s.
(这里用的是Standard deviation)

Sampling and estimation那章的LOS: Identify and describe the desirable properties of an estimator里面说:
a confidence interval for the population mean can be calculated as:
x-bar +/- z*(stand error of the mean)

印象中,上学时构建confidence interval时,一直就使用的是standard error, 那上面使用standard deviation的公式是怎么来的呢?

Augu

the fisrt one is one discriptive statistics
the 2nd one is one referential statistics.
Both right

NetDagger

but which stats i need to put into the formula if i meet this kind of question in the exam?

niles

认真看题就好了，如果题目说的是对总体进行描述性统计，比如一个总体，均值方差告诉，问你置信区间，就用sd。相反抽样统计，则用se，比如一个样本，均值和方差已知，就用se。关键是分清总体还是样本。

NetDagger

是不是这样呢:

在Common Probability Distributions那个Session中,有这样一个例题:
Using a 20-year sample, the average return of a mutual fund has been 10.5 percent per year with a standard deviation of 18 percent. what is the 95 percent confidence interval for the mutual fund return next year?
因为题目给的是样本的均值和SD,所以answer中用的是SD.

相反,在Sampling and estimation中,confidence interval那个例题中:
Consider a practive esam that was administered to 100 Level 1 candidates...Assuming a population standard deviation equal to 15...
因为这里讲的是population SD,所以要用SE?