- Mr. A wants to determine how closely returns are related to growth. Specifically, he wants to test whether or not there is a one-to-one relationship between the two variables. He formulates the following competing hypothesis: H0 : b3 = 0 HA : b3 not equal 0 (given in the problem b3=1.16 & std error=.51) At the 0.05 significance level, he: A. rejects the null hypothesis, and concludes that a one-to-one relationship exists between returns and growth. B. rejects the null hypothesis, and concludes that a one-to-one relationship does not exist between returns and growth. C. does not reject the null hypothesis, and concludes that a one-to-one relationship exists between returns and growth. D. does not reject the null hypothesis, and concludes that a one-to-one relationship does not exist between returns and growth. Answer: C () Incorrect LOS: Study Session 3-1-C-b When conducting the following test H0 : b3 = 0 HA : b3 ¹ 0 the calculated test statistic is t=(1.16-1)/0.51=0.314. With a critical value of ±2.145, (obtained in the same manner as problem 9), the decision rule is to not reject the null hypothesis since –2.145<0.314<2.145; thus the sample data supports the hypothesis that b3 =1 and a one-to-one relation exists. =========================================================== I think the qn is wrong it should b3=1… anyone concur…
obviously question is wrong.
this is not a time series question. it is just ordinary correlation problem, which is from L1. Unit root problem only exists in time series regression, in which dependent variable and independent variable are in different time frame, which is not the case here. So the null should be h0: b3=0. and the test is equal to 2.3**, then the null can be rejected cuz the t@0.05<2? then should be A, right?