T2-Miller- Ch7 - Slide 15

[eldakrory]

Hello David
1- I noticed in slide 15 when using the C.I. approach you used the two sided value while it should be one side as H0 < or = 0. (and used one sided in the sig. approach)
2- How do we came up with the P value?

 

[David Harper CFA FRM]

Hi @eldakrory I just listened to it, I don't think i was emphasizing the CI. It's just in there to show the generic (2-sided) CI approach, in comparison to the significance approach. We could show a 1-sided CI, it's just unusual in my experience. The video means to focus on the 1-sided significance test. But, yes you are correct, the analogous confidence interval would be a 1-sided CI, which is not shown in the exhibit (but I don't think i even emphasize the CI in the video anyway, it's an XLS snippet from the learning XLS that aspires to compare methods; maybe I should add a 1-sided CI ...).

p-value calcs are in the learning XLS:

1 sided p value =T.DIST.RT(computed t,df) = T.DIST.RT(0.54, 9)
2 sided p value =T.DIST.2T(computed t,df) = T.DIST.2T(0.54, 9)