Question

1. Suppose I create a 99% con?dence interval for ?2 that does not contain zero. What does this imply about the results of a two-sided hypothesis test that there is no e?ect of X on Y? Assume the test is done at the 5% level. Explain.

2. Suppose I create a 99% con?dence interval for ?2 that contains zero. What does this imply about the results of a two-sided hypothesis test that there is no e?ect of X on Y? Assume the test is done at the 5% level. Explain

Answer 1

Zero is the null value of the parameter . If a  confidence interval includesthe null value, then there is no statistically meaningful or statistically significant difference between the groups.

(1)here the confidence interval does not contain zero , so there is significant effect of X on Y

we reject null hypothesis of there is no effct of X on Y

(2)here the confidence interval contain zero , so there is not significant effect of X on Y

we donot reject null hypothesis of there is no effct of X on Y