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

大蒙 · 2018年11月04日

问一道题:NO.PZ2017092702000133 [ CFA I ]

问题如下图:

老师你好,看了之前其他同学问题下关于A选项的解答,其中两个解答中分别对p-value和显著性水平的解释

有一个解答说:“可以理解成P-value是最小的能拒绝原假设的显著性水平。

另一个说:“所以P-value 不属于significant level中的最低标准,这是两个独立的概念。


那到底是什么样的呢?


    

选项:

A.

B.

C.

解释:



1 个答案
已采纳答案

菲菲_品职助教 · 2018年11月05日

同学你好,这两句话并不矛盾。

首先,“可以理解成P-value是最小的能拒绝原假设的显著性水平。”这句话是正确的。这也是原版书的原话。我们怎么来理解这句话呢?因为P值越小越拒绝原假设。当p值小于α时,我们拒绝原假设。就如何老师在基础班中讲的一样,举个例子,已知p值=6%,问α怎么设定才能拒绝原假设。因为当p值小于α时,我们拒绝原假设,所以α至少要大于等于6%,我们才能拒绝原假设,也就可以理解成P-value是最小的能拒绝原假设的显著性水平。

第二个说法,“所以P-value 不属于significant level中的最低标准,这是两个独立的概念。” 这句话也没有错,这确实是两个独立的概念。P-value是最小的能拒绝原假设的显著性水平,但不代表着显著性水平的最低标准就是P-value。显著性水平本身是没有什么最低标准的,可以是1%,也可以是5%,是人为规定的。这句话的关键是“拒绝原假设”,在这个前提下,P-value才是最小的显著性水平。

  • 1

    回答
  • 3

    关注
  • 454

    浏览
相关问题

NO.PZ2017092702000133 问题如下 Whiof the following statements is correwith respeto the p-value? A.It is a less precise measure of test evinthrejection points. B.It is the largest level of significanwhithe null hypothesis is rejecte C.It ccomparerectly with the level of significanin reaching test conclusions. C is correct.When rectly comparing the p-value with the level of significance, it cusealternative to using rejection points to reaconclusions on hypothesis tests. If the p-value is smaller ththe specifielevel of significance, the null hypothesis is rejecte Otherwise, the null hypothesis is not rejecte对于一个分布而言,每个概率面积在横轴上都有一个对应点,例如significanlevel α对应的分位点就是临界值criticvalue(假设正态分布下α=95%,双尾检验,则右尾就会对应criticvalue=1.96)。而p-value也是一个概率面积,对应的点就是计算出来的检验统计量的那个点。所以当p-value<α时,就等价于检验统计量>criticvalue,也就是拒绝原假设。 能方便一下- 为何是smallest 而不是 largest 吗? 以下为我的理解 - p-value 是一个概率, i.e. 如果 p-value 为 0.06 - 则我们可以理解为有 6% 的可能性 the null hypothesis 为 true (base on the available samples). 而 level of significan只是一个 pre-finethresholof rejecting null hypothesis - i.e. if p-value = thresholthen we conclu there is statisticevinto support null hypothesis being true, otherwise, we rejenull hypothesis. e.g. when p-value 0.05 level of significan= 0.05 -- 则我们可以rejenull. Regaress p-value = 0.04, 0.03 etc -- 我们都可以rejenull, 那么level of significan岂不是一个upper boun 而p-value 也可以理解为 一个大值 where we coulrejea null? 请问我的理解错在了哪里? 谢谢

2024-03-25 10:15 1 · 回答

NO.PZ2017092702000133问题如下 Whiof the following statements is correwith respeto the p-value?A.It is a less precise measure of test evinthrejection points. B.It is the largest level of significanwhithe null hypothesis is rejecte C.It ccomparerectly with the level of significanin reaching test conclusions. C is correct.When rectly comparing the p-value with the level of significance, it cusealternative to using rejection points to reaconclusions on hypothesis tests. If the p-value is smaller ththe specifielevel of significance, the null hypothesis is rejecte Otherwise, the null hypothesis is not rejecte对于一个分布而言,每个概率面积在横轴上都有一个对应点,例如significanlevel α对应的分位点就是临界值criticvalue(假设正态分布下α=95%,双尾检验,则右尾就会对应criticvalue=1.96)。而p-value也是一个概率面积,对应的点就是计算出来的检验统计量的那个点。所以当p-value<α时,就等价于检验统计量>criticvalue,也就是拒绝原假设。 请问老师,a错在哪里?

2023-12-31 18:19 1 · 回答

NO.PZ2017092702000133问题如下Whiof the following statements is correwith respeto the p-value?A.It is a less precise measure of test evinthrejection points. B.It is the largest level of significanwhithe null hypothesis is rejecte C.It ccomparerectly with the level of significanin reaching test conclusions. C is correct.When rectly comparing the p-value with the level of significance, it cusealternative to using rejection points to reaconclusions on hypothesis tests. If the p-value is smaller ththe specifielevel of significance, the null hypothesis is rejecte Otherwise, the null hypothesis is not rejecte对于一个分布而言,每个概率面积在横轴上都有一个对应点,例如significanlevel α对应的分位点就是临界值criticvalue(假设正态分布下α=95%,双尾检验,则右尾就会对应criticvalue=1.96)。而p-value也是一个概率面积,对应的点就是计算出来的检验统计量的那个点。所以当p-value<α时,就等价于检验统计量>criticvalue,也就是拒绝原假设。 p value和test statistic 的关系?

2023-05-28 17:09 1 · 回答

NO.PZ2017092702000133 问题如下 Whiof the following statements is correwith respeto the p-value? A.It is a less precise measure of test evinthrejection points. B.It is the largest level of significanwhithe null hypothesis is rejecte C.It ccomparerectly with the level of significanin reaching test conclusions. C is correct.When rectly comparing the p-value with the level of significance, it cusealternative to using rejection points to reaconclusions on hypothesis tests. If the p-value is smaller ththe specifielevel of significance, the null hypothesis is rejecte Otherwise, the null hypothesis is not rejecte对于一个分布而言,每个概率面积在横轴上都有一个对应点,例如significanlevel α对应的分位点就是临界值criticvalue(假设正态分布下α=95%,双尾检验,则右尾就会对应criticvalue=1.96)。而p-value也是一个概率面积,对应的点就是计算出来的检验统计量的那个点。所以当p-value<α时,就等价于检验统计量>criticvalue,也就是拒绝原假设。 请老师指导,B哪里错了?

2022-12-29 21:05 1 · 回答

NO.PZ2017092702000133 问题如下 Whiof the following statements is correwith respeto the p-value? A.It is a less precise measure of test evinthrejection points. B.It is the largest level of significanwhithe null hypothesis is rejecte C.It ccomparerectly with the level of significanin reaching test conclusions. C is correct.When rectly comparing the p-value with the level of significance, it cusealternative to using rejection points to reaconclusions on hypothesis tests. If the p-value is smaller ththe specifielevel of significance, the null hypothesis is rejecte Otherwise, the null hypothesis is not rejecte对于一个分布而言,每个概率面积在横轴上都有一个对应点,例如significanlevel α对应的分位点就是临界值criticvalue(假设正态分布下α=95%,双尾检验,则右尾就会对应criticvalue=1.96)。而p-value也是一个概率面积,对应的点就是计算出来的检验统计量的那个点。所以当p-value<α时,就等价于检验统计量>criticvalue,也就是拒绝原假设。 能不能翻译一下每个?不怎么了解

2022-11-28 10:42 1 · 回答