cathyz · 2020年01月12日
问题如下：
We use interval estimation to study the mean of populations. If the population is normally distributed but the variance is unknown, How can we construct confidence intervals?
选项：
A.Both The t-statist and the Z-statistic
B.The Z-statistic at α/2 with n - 1 degrees of freedom.
C.The t-statistic at α/2 with n - 1 degrees of freedom.
解释：
C is correct.
Because the sample size is unavailable, the z-statisticcan not be used under this case.
为什么是N-1 的自由度？
NO.PZ2015120604000127 问题如下 We use intervestimation to stu the meof populations. If the population is normally stributebut the varianis unknown, how cwe construconfinintervals? A.Both The t-statist anthe Z-statisti B.The Z-statistic α/2 with n - 1 grees of freem. C.The t-statistic α/2 with n - 1 grees of freem. C is correct.解析（总体）方差已知用z，（总体）方差未知用t，非正态小样本不可估计。 The t-statistic α/2 with n - 1 grees of freem.
请问在correlation的假设检验中，说用到的是t分布，因为是当总体方差未知时。那这里当样本足够大时（n大于等于30），可以用z分布吗？
为什么是α/2 而不是α
