R10 Quants Q18 page 387

In question 18 to calculate confidence interval, book is using critical value of 1.96 for .025 which is according to 2 tail test for .05 level of significance, but based on information given in text, it is supposed to be 1 tail test which means for .05 level of significance critical value is close to 1.645. Please suggest, TIA

Where does it say anything about a one-tailed test???

Also, I am not sure how a confidence interval can be one-tailed?

You must be referring to this section of the question:

_ He also believes that for each 1 percent increase in pre-offer price adjustment, the initial return will increase by less than 0.5 percent, holding other variables constant. _

Bear in mind, that this is a hypothesis which we then test using our estimate, and for which we use a one-tailed test.

In question 18 you are constructing a 5% confidence interval around that estimate, so you want to know, the interval that contains 5% of values, so 2.5 above and 2.5 below.

What do you read in the text that leads you to believe that it’s supposed to be a 1-tail test?

I don’t see anything that would lead me to that conclusion.

Confidence intervals always have two tails.

thanks all for clarifying

One tailed: H0 > X or H0< X

Two tailed: H0 does not equal x

Since the questions is saying it’s less than, its a one tailed test.

I presume an interval will always have two limits (an upper and a lower limit). This rules out the possibility of it being a one-tailed test.

To be absolutely clear: confidence intervals (in the CFA curriculum) have two tails. Always.