开发者:上海品职教育科技有限公司 隐私政策详情

应用版本:4.2.11(IOS)|3.2.5(安卓)APP下载

tongmopwt · 2019年11月15日

问一道题:NO.PZ2015120604000130

问题如下:

We want to use z-statistic to construct confidence interval for a normally distribution. Assume the sample size is 100, sample mean is 15% and the standard deviation of sample is 25%. The significance level is supposed to be 10% , the confidence interval is  :

选项:

A.

- 10.88% to 19.13%.

B.

10.88% to 19.13%.

C.

10.88% to 20.57%.

解释:

B is correct.

In this case, we can know that the Confidence Interval=[Point Estimate +/- (reliability factor)*Standard error]=

x¯±zα/2σn=15%±1.65×25%100=10.88%,19.13%

请问25%不是已经是样本的标准差了吗?为什么还要再除以根号下100?

1 个答案
已采纳答案

星星_品职助教 · 2019年11月16日

同学你好,

这道题涉及到对于中心极限定理的理解。中心极限定理说的是对于样本均值X bar来说,X bar服从正态分布(均值就是总体均值μ,标准差为总体标准差/根号N,N>30)。这里面要注意两点,首先,X bar本身也是一个随机变量,所以就有分布。第二,X bar的标准差就叫做标准误。形式就是总体标准差/根号N

而在估计中,总体标准差往往是不知道,所以这个时候我们就会用抽出来的那一个样本的标准差来代替总体方差。也就是S/根号N。这个是标准误的概念,并不是样本标准差的概念。可以理解为标准误是X bar的标准差(样本均值的标准差),而并不是“样本标准差”。

所以对于X bar的置信区间来说,后面±的也是X bar的标准差,也就是标准误。而并不是样本标准差。可以这么记忆,前面是X bar,后面对应的就是一致的X bar的标准差。也就要在样本标准差的基础上在除以根号N。

这个知识点老师基础班和强化班都强调过。等听到的时候可以留意一下,加油~

tongmopwt · 2019年11月20日

明白了,谢谢老师

  • 1

    回答
  • 0

    关注
  • 653

    浏览
相关问题

NO.PZ2015120604000130问题如下We want to use z-statistito construconfinintervfor a normally stribution. Assume the sample size is 100, sample meis 15% anthe stanrviation of sample is 25%. The significanlevel is supposeto 10% , the confinintervis :A.- 10.88% to 19.13%.B.10.88% to 19.13%.C.10.88% to 20.57%. B is correct.In this case, we cknow ththe ConfinInterval=[Point Estimate +/- (reliability factor)*Stanrerror]= 什么时候会考这个区间,一般问这个区间的问题是怎么问的这个区间的标准差指的是总体标准差吗

2023-01-18 11:01 1 · 回答

NO.PZ2015120604000130 问题如下 We want to use z-statistito construconfinintervfor a normally stribution. Assume the sample size is 100, sample meis 15% anthe stanrviation of sample is 25%. The significanlevel is supposeto 10% , the confinintervis : A.- 10.88% to 19.13%. B.10.88% to 19.13%. C.10.88% to 20.57%. B is correct.In this case, we cknow ththe ConfinInterval=[Point Estimate +/- (reliability factor)*Stanrerror]= SignificanLevel都是要除以2吗?

2022-11-01 18:08 1 · 回答

NO.PZ2015120604000130 问题如下 We want to use z-statistito construconfinintervfor a normally stribution. Assume the sample size is 100, sample meis 15% anthe stanrviation of sample is 25%. The significanlevel is supposeto 10% , the confinintervis : A.- 10.88% to 19.13%. B.10.88% to 19.13%. C.10.88% to 20.57%. B is correct.In this case, we cknow ththe ConfinInterval=[Point Estimate +/- (reliability factor)*Stanrerror]=x¯±zα/2σn=15%±1.65×25%100=10.88%,19.13% 请题干以及该题涉及的原理所对应知识点

2022-10-13 23:10 1 · 回答

NO.PZ2015120604000130 问题如下 We want to use z-statistito construconfinintervfor a normally stribution. Assume the sample size is 100, sample meis 15% anthe stanrviation of sample is 25%. The significanlevel is supposeto 10% , the confinintervis : A.- 10.88% to 19.13%. B.10.88% to 19.13%. C.10.88% to 20.57%. B is correct.In this case, we cknow ththe ConfinInterval=[Point Estimate +/- (reliability factor)*Stanrerror]=x¯±zα/2σn=15%±1.65×25%100=10.88%,19.13% 这里的Stanrerror 不应该是25%/ √100 吗?解析里的式子是什么意思

2022-10-09 10:47 1 · 回答

NO.PZ2015120604000130问题如下We want to use z-statistito construconfinintervfor a normally stribution. Assume the sample size is 100, sample meis 15% anthe stanrviation of sample is 25%. The significanlevel is supposeto 10% , the confinintervis :A.- 10.88% to 19.13%.B.10.88% to 19.13%.C.10.88% to 20.57%.B is correct.In this case, we cknow ththe ConfinInterval=[Point Estimate +/- (reliability factor)*Stanrerror]=x¯±zα/2σn=15%±1.65×25%100=10.88%,19.13%老师,既然题目说是正态分布,而且是选择题。想问下能否利用正态分布的特性,代入解题?正太分布中轴是均值,题干给出的置信区间应该也是轴对称。直接带入,分位点-均值19.13-15=4.13均值-分位点到均值的距离15-4.13=10.87然后选出最符合的的

2022-09-27 15:46 1 · 回答