Correct me if I am wrong, The region of rejection is we picked up a significant level then use that to calculate rejection region while P value is the conditional probability of alternative hypothesis $Ha$ on $Ho$ is true then we compare with significant level?

Is significant level a conditional probability of Reject Ho on Ho is true?

  • $\begingroup$ $p$-value has nothing to do with $H_0$ being true. Check stats.stackexchange.com/questions/31/… for a start or this recent thread stats.stackexchange.com/questions/347653/… $\endgroup$
    – Tim
    May 24, 2018 at 11:58
  • $\begingroup$ Thanks. According to the link you refer, $P$($Samplemean$ ≥ 5ft 9inches|$Truevalue$=5ft7inches).Which the true value is $Ho$ is true which $u$ is population mean? I thought it is same as my understanding? $\endgroup$ May 24, 2018 at 12:33
  • $\begingroup$ $P(X|H_0) \ne P(H_0|X)$ $\endgroup$
    – Tim
    May 24, 2018 at 12:34
  • $\begingroup$ Thanks Tim. If I edit my question as this way, does it make more sense? $\endgroup$ May 24, 2018 at 12:42
  • $\begingroup$ re-read/digest the link, my understanding is: region of rejection is by positioning $Z$ test result in the confidence interval, we can see whether the null hypothesis is rejected or not. Same conclusion can be obtained from $P$ value test. After calculate $P$ value from alternativehy pothesis $Ha$ on Null hypothesis $Ho$, if value is greater than significant level then we don't reject $Ho$, but if value is smaller, we reject it. $\endgroup$ May 24, 2018 at 13:13

1 Answer 1


the region of rejection test is using Z value. It just simply reject the result without showing the strength of significance? This is why we introduce P value

  • $\begingroup$ To me, the rejection region shows the proportion of area under normal curve that accommodates values of Z beyond that of normal distribution. $\endgroup$
    – user10619
    Aug 9, 2018 at 14:24
  • $\begingroup$ It shows the area occupied by outliers (Z values more than the expected Z under the assumption of (say) alpha = .05 . The probability or alpha level for a given distribution of observed scores is usuailly calculated by simulation. If the calculated probability is less than or equal to say, assumed alpha (.05), we say that the effect-size is statistically significant at .05 . $\endgroup$
    – user10619
    Aug 16, 2018 at 11:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.