this question is from last year’s Windsor (schweser) exam: Forecast of Sales based on data from 180 months of sales Sales = 10.2 +5.6*DOL + 6.3*IP + 9.2*GDP Std Errors: Intercept =5.4 DOL= 3.5 IP =4.2 GDP= 5.3 Hypothesis 1: 2% Increase in DOL will result in an increase in sales of more than 11.2% Hypothesis 2: 1% Increase in industrial production (IP) will result in a 1% decrease in sales. He fails to reject each hypothesis at 99% confidence interval. Is he correct?
- no 2. no Don’t know if I’m seeing the real problem here, though.
By plugging values and at 99% for both he should reject hyp. So he is incorrect for both?? Any hidden calc required other than this…?
it looks like the standard errors were thrown in to confuse folks. Even if the coefficient is deemed insignificant it is still used to calculate the final answer.
Correct Answer: He should fail to reject the null for both. The part where i get tripped up is in setting up the hypothesis tests. For Hypothesis 1: 11.2/2 = 5.6 ,so 5.6-5.6/3.5 , making t stat =0, fail to reject since t critical =2.348 for one tail test. For Hypothesis 2: 6.3- (-1)/ 4.2 = 1.74…so fail to reject since t critical for one tail test =2.604