Introductory Econometrics — DU B.A. (Hons.) Economics Previous Year Question Papers

1. State whether the following statements are True or False. Give reasons/ justifications for your answer. (i) A regression result based on OLS using 42 observations and 5 explanatory variables reported an R2 = 0.142. The Adjusted R2 should be negative. (ii) The slope coefficient of a regression obtained through OLS is 2.11. Since this value is now known and constant, the slope coefficient is not a random variable. (iii) In the presence of high multicollinearity, it becomes difficult to assess the individual statistical significance of partial regression coefficients due to inflated standard errors. (iv) Weighted Least Squares (WLS) produces efficient and unbiased estimates even in the presence of heteroskedasticity (v) A dummy variable regression, such as framework for testing whether the conditional mean of groups defined by . provides a differs across two (vi) The first difference transformation to eliminate autocorrelation assumes that the coefficient of autocorrelation must be –1. (6×3=18) 2. Suppose we have a regression equation relating consumption expenditures (y) to income (x) for 20 families in a certain region. Both y and x are measured in thousands of rupees. The estimated regression is as follows (i) Interpret the slope coefficient in the above regression. Test the overall significance of the regression at the 5% level of significance. Suppose y and x are measured in lakhs of rupees, instead of thousands of rupees, then how would the coefficients of the above regression be affected? Interpret the new slope coefficient thus obtained. (ii) Suppose the plot for estimated residuals with respect to x in the above regression indicates that the error variances are not constant but increase with the value of x. Will this cause a problem in OLS estimation? What will be the consequence of such a problem? (iii) In order to test the presence of the above problem, a formal test (Glejser test) was conducted, and the following result was obtained How would this help in confirming the problem identified in part ii above? State the null hypothesis of this test. (6×3=18) 3. A researcher aims to predict the Return on Assets (ROA) of Micro-Finance Institutions (MFI) using the following model: Where NoB is the number of borrowers, FM is the percent of female borrowers, and LO is the number of Loan Officers. Based on 54 MFIs, the following results were obtained using OLS: (i) Interpret the coefficients of NoB and FM. (ii) Based on the results above, how would you check for the presence of Multicollinearity in the regression? If found, will the aim of the researcher be difficult to achieve? (iii) The Variance Inflation Factor for NoB was reported to be 17.3. How do you think this number was obtained? (6×3=18) 4. A researcher studies monthly household water consumption (in kilolitres) in different regions. The following regression model is specified : Ci : Monthly water consumption HHSizei : Household size (number of members)