Re: Normal distribution of ratio of means

Thank you.

1) So, you agree that the answer to my question is "yes," but you point
out that there is a more exact (and more complicated) way to determine
the confidence interval in such a case. Have I understood you right?

2) I will be able to take a look at Kendall and Stuart in a few days.
Do you know other sources where I could find a reference - not for the
more complicated way (based on Fieller's Theorem), but - for this "yes"
(that the 99%-confidence interval of a ratio of two means can be
determined as follows: ratio +/- (2.58 multiplied by the standard error
of the ratio))?

.

Loading