Which of the following statements about the twosample t proc

Which of the following statements about the two-sample t procedures for significance tests and confidence intervals for the difference of the means, mu1 - mu2, is/are FALSE? a. The two population variances must be equal. b. The two samples must be simple random samples. e. Although Normality is a requirement, with large sample sizes the central limit theorem guarantees that the results will be approximately correct for other population distributions. d. The two samples must be independent.

Solution

(a) is false because even if variances are not equal, we have a t-procedure to calculate test/CI

(b) is true

(c) is correct as sample mean is almost normal for large sample sizes.

(d) is false, because here we can use dependent sample t-procedures.