In: Statistics and Probability
A simple random sample of size n is drawn from a population that is normally distributed. The sample mean, x overbar, is found to be 110, and the sample standard deviation, s, is found to be 10. (a) Construct a 96% confidence interval about mu if the sample size, n, is 12. (b) Construct a 96% confidence interval about mu if the sample size, n, is 26. (c) Construct a 99% confidence interval about mu if the sample size, n, is 12. (d) Could we have computed the confidence intervals in parts (a)-(c) if the population had not been normally distributed?
a)
sample mean, xbar = 110
sample standard deviation, s = 10
sample size, n = 12
degrees of freedom, n - 1 = 11
For 96% Confidence level, the t-value = 2.328
CI = (xbar - t*s/sqrt(n), xbar + t*s/sqrt(n))
CI = (110 - 2.328 * 10/sqrt(12) , 110 + 2.328 * 10/sqrt(12))
CI = (103.28 , 116.72)
b)
sample mean, xbar = 110
sample standard deviation, s = 10
sample size, n = 26
degrees of freedom, n - 1 = 25
For 96% Confidence level, the t-value = 2.167
CI = (xbar - t*s/sqrt(n), xbar + t*s/sqrt(n))
CI = (110 - 2.167 * 10/sqrt(26) , 110 + 2.167 * 10/sqrt(26))
CI = (105.75 , 114.25)
c)
sample mean, xbar = 110
sample standard deviation, s = 10
sample size, n = 12
degrees of freedom, n - 1 = 11
For 99% Confidence level, the t-value = 3.106
CI = (xbar - t*s/sqrt(n), xbar + t*s/sqrt(n))
CI = (110 - 3.106 * 10/sqrt(12) , 110 + 3.106 * 10/sqrt(12))
CI = (101.03 , 118.97)
d)
No, it is not possible to calculate the CI if population is not
normally distributed.