Hi All, Can someone please comment on the reason behind taking twice the value of probability when doing two-sided test, reference to question under 3.2 section of chapter 09. I am able to understand the method of critical values calculation in two-sided test but unable to understand the logic behind multiplying by two. Many thanks. Sunil
If we didn't double it then we wouldn't be able to tell whether we'd reject H0 or not unless they told us it was a one or two sided test. As for one-sided, we would reject H0 if p-value<5%, whereas two-sided, we would reject H0 if p-value<2.5%. Hence, doubling it means it is consistent with a one-sided test (which has the full 5% on one side).