Consider the information in equation (12.17). The sample for this regression has 76 observations.

(a) Create a confidence interval for the true effect of the Corruption Perceptions Index (CPI) on gross national income (GNI) per capita. Does this confidence interval change the conclusion that the estimated effect of corruption on national income is surprisingly large? If yes, how? If no, why not?

(b) Test the null hypothesis that the Education Development Index (EDI) is unrelated to GNI per capita against the alternative that it has a positive effect, at the 5% significance level. Test it again at the 1% significance level.

What can we conclude about the apparent effects of education on GNI?

(c) The minimum and maximum values for the EDI in our sample are 537 and 995, respectively. What is the difference between the implied contributions to GNI at these two levels of EDI? Is it substantively large or small?

(d) Test the joint hypothesis that both coefficients are zero at the 5% significance level. The R2 for the regression in equation (12.17) is .7667.