Let (X1, X2,..., Xn) be a sample, where each Xi is a random variable of normal distribution with mean mu and variance sigma². Let us suppose that n = 20, and sigma² = 9. An experiment has yielded the results (X1, X2,..., X20), and we have calculated that the empirical mean x20 = 2.09.
1) Give a confidence interval with level of confidence 90% for mu.
2) How big would the sample have to be for the interval to be half as long?
3) Let us now suppose the variance is not known. Knowing that counting from i=1 to 20 (xi - x20)² = 14.6, give a confidence interval with level of confidence 90% for the value of mu.
4) We now suppose we know mu is known and is equal to 2. Give a confidence interval with a confidence level of 90% for the value of sigma².
5) Same question if we do not know the value of mu.
Please see attachment for proper format.
The response provides the confidence interval at 90% confidence, a new sample size based on the shorter interval, how the confidence interval would be found without the variance, and when sigma and mu variables are known.