p-value =
The probability that the a sample which is more extreme than the tested sample is drawn
under 1. the null hypothesis
and 2. normal distribution hypothesis with given population variance.
bzbz2019-04-30 22:21:08
好似話依個係成日比人誤解、誤用既stat概念之一
即使係專業人士都會被誤用
利申: 唔識
約德爾人2019-04-30 22:27:38
本身個因果邏輯就好奇怪
被人ban9係有原因既
山系野人2019-04-30 22:41:50
連大學教授都好難簡單咁解釋, 本身概念已經抽像
山系野人2019-04-30 22:59:05
This is one of the ways to explain P-Value. Since we assume that u=specified number and expected the sample mean should fall into, let say 99%. That 1% is the extreme value which is too far away from population mean.
If the sample mean is within this 1%, we will have sufficient evidence to say that u is not the specified number.
The concept is aligned with confidence interval. For instance, we have 99% confidence that the population mean will fall into this range but the true mean is not. We are skeptical that the mean is not what we assumed.
淫賤打椿機2019-04-30 23:30:49
咪就咁 P(getting the sample result given that null is true)
佢大嘅話 你咪信住個null係啱先囉
佢細嘅話 你咪唔好信囉
點為之大同細 就係你個level of significant
所以你先至會用p value嚟同佢嚟做比較