
T confidence interval
X1, X2,...Xn come from a random sample that is normally distributed with mu unknown and sigma^2 known. the confidence coefficient is .95 and the length of the confidence interval is less than .01*sigma. What is n?
I got a weird answer of 120<n<infinity. is this wrong?

(1) If sigma is known, you should be using a N(0,1) distribution and not a t.
(2) I guess this CI is for MOO, I mean mu, you need to say what we are estimating.
it could be one over mu, mu squared, sigma times moo.....
(3) If so, set
$\displaystyle 2{1.96\sigma\over \sqrt{n}}< {\sigma\over 100}$