1
$\begingroup$

I'm studying the second course in statistics, and now I have a problem to understand $P$-values. Namely, one exercise is the following.

When testing hypothesis $H_0:\mu=\mu_0$ one gets the value of test statistic $z=1.7$. Determine the $P$-value for the alternative hypothesis $H_A:\mu>\mu_0$.

Is it that I have to calculate the $P$-value from $N(0,1)$-distribution or is there some other distribution in the background?

$\endgroup$
2
  • 1
    $\begingroup$ This sounds like homework so I will add the homework tag. You can delete it if I am mistaken. $\endgroup$
    – Peter Flom
    Nov 25, 2012 at 13:18
  • $\begingroup$ @PeterFlom It is indeed a homework. $\endgroup$ Nov 25, 2012 at 20:10

1 Answer 1

2
$\begingroup$

You are on the right track; think about what the z-statistic means. I am not sure what you mean by "count the p-value" but that may be just a linguistic problem.

$\endgroup$
4
  • $\begingroup$ So my problem is that in my lecture notes the example are from normal distribution. But as that is not said explicitly in the problem, should I assume that we test the hypothesis from $N(0,1)$-distribution? Probably I should try to find some lecture notes where P-values are defined more rigorously. $\endgroup$ Nov 25, 2012 at 16:19
  • $\begingroup$ If the statistic is a z-value you can assume their discussion a normal distribution. Z-values are derived from the standard normal distribution. $\endgroup$
    – John
    Nov 25, 2012 at 18:16
  • $\begingroup$ @JaakkoSeppälä Does the first sentence of the wikipedia article on p-values (which correctly defines the term) help you? $\endgroup$
    – Glen_b
    Nov 25, 2012 at 23:33
  • $\begingroup$ @Glen_b Yes it helps. John's comment was helpful as now I know that the values are derived from the standard normal distribution. $\endgroup$ Nov 27, 2012 at 12:25

Your Answer

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

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