[This question paper contains 8 printed pages.] Your Roll NoOv.ssccrerses Sr. No. of Question Paper: 5105 H Unique Paper Code : 2922072401 Name of the Paper : Basics of Econometrics Name of the Course : Bachelor of Business administration (Financial Investment Analysis) [BBA (FIA)] Semester + IV Duration : 3 Hours Maximum Marks : 90 Instructions for Candidates i Ny Write your Roll No. on the top immediately on receipt of this question paper. Question 1 is compulsory. Attempt any four questions from question 2 - question 6. All questions carry equal weightage Use of simple calculator is allowed. PERO:5105 2 wn _ = uw 1. Explain the following with appropriate examples : (a) Explain the different functional form of regression models. (b) Dummy Variable Regression (c) Interaction Variable (d) Logistic regression (e) P-value 2. (a) Explain the problem of endogeneity? How can it be solved? (b) Consider the following model: Yi = a0 + aID1 + a2 D2 + 03 D3 + a4 DI *#X +B1 Xi + Ui Where Y = annual salary of college professor depending upon region of residence of professor D1 = dummy variable for region, | for North; 0 otherwise D2 = dummy variable for region, 1 for South; 0 otherwise D3 = dummy variable for region, | for East: 0 otherwise X*D1 = Interactive Dummy Xi = years of teaching experience Write the regression equation corresponding to each of the following: annual Salary of a professor g in North, Salary of a professor who is residing in South. Salary of a professor who is residing in East, Salary of a professor who is residing in West. How would you infer the significance of coefficient of D3 using the p-value? What impact does it have on the model? How would the model look like if we add the dummy variable for ‘WEST? Write the equation for it and explain the ‘dummy variable trap’ in this model’s context. Interpret the Coeff of interactive dummy. 3. A researcher, using a sample of 2000 individuals, is investigating how the probability of a respondent obtaining a bachelor’s degree from a three-year college is related to the respondent’s score on ASVXYZ. 30 percent of the respondents earned bachelor’s degrees. ASVXYZ ranged from 20 to 70, with mean value 52, and most scores were in the B10;5105 4 range 45 to 65. Defining a variable DEGR to be equal to | if the respondent has a bachelor’s degree (or higher degree) and 0 otherwise, the researcher fitted the OLS regression (standard errors in parentheses): DEGR = -0.863 +0.022 ASV XYZ (0.000) She also fitted the following logit regression: Z = -12.103 + 0.179 ASV XYZ (0.004) Where Z is the variable in the logit function. (a) Give an interpretation of the OLS regression, and explain why OLS is not a satisfactory Estimation method for this kind of model. (b) What does the slope coefficient in Logit model measures? Interpret the estimated logit model. From the estimated logit model, how would you obtain the expression for the probability of earning a bachelor’s degree? (c) What is the probability that an individual with a score of 40 on ASV XYZ will have a bachelor’s degree? What is the rate of change of probability at the score of 55? 5105 5 The relation between two variables age and income is investigated if it is same for men and women. You have the following relations: Income,= @, +0, Age, + Ue (1), i= 1,2. Income, = B, |B, Age, +A Fem,+v, (2), i-1,2... Where u, and v, are stochastic error terms. Fem, is female and takes the value | if the observation unit i is of female and zero for a male. Below is the regression output from OLS applied to equation (1) and (2): The estimation sample ie: 1 to 12237 Coefficient Std.Error t-value t~-prob Part.R°2 Constant 100878. 9431. 10.7 0.000 0.0082 age 10536. 313.8 33.6 0.000 0,0844 fem ~62966.0 4349. -14.8 0.000 0.0188 sigma 240096 RSS 7 0524297481014 R2 0,0962938 F(2,12234) = 651.8 [0.000]s« log-Likelihood -168964 DW 4.42 no. of observations 12237 no. of parameters 3 mean (ine) 72161 varine) 6. 37730010 P,v0.5105 6 ‘The estimation sample is: 1 to 12237 Coefficient Std.Errar t-value t-prob Part.R-2 Conetant, 125862, 9347, 13.5 0.000 0.0146 age 10373.8 316.3 32.8 0,000 0.0808 signa 242136 RSS 7.473283650+014 Re 0.0808074 F(1,12235) « 1076 [0.000] «+ log-1ikelihood -169068 DW aus no. of obvervations 12237 no. of parameters 2 mean (inc) 172161 varCinc) 6.3773e+010 Review the estimated coefficients. What is the precise interpretation of the coefficient X? Test the hypothesis that the relation between age and income is the same for men and women. Suppose the result looks like the one given below: The estimation sample is: 1 to 12237 Coerficient Std.Error t-value t-prob Part.R°2 Constant 190025. 1.27304004 -14.9 0.009 0.0179 age 13676.2 434.4 31.5 0,000 0.0749 fen 124028. 1.849e#004 = «6.710.000 0.0037 fenvoge ~6504.33 625.3 10.4 0,000 0,088 signa 239051 RSS 6990606162014 Re? 0.104216 F(3,12233) = 474.4 [0.000] + Jog=1ikelshood 168910 DW 1.13 no. of observations 12937 no. of paramaters 4 mean( inc) 172161 var(ine) 6. a773er010 5105 vi An interaction term; AGE * Female is included in the regression. Give an interpretation of the coefficient on this variable. Does this result add to our understanding of the gender differences in the income- age relationship? How would you now test the null hypothesis of no gender differences? 5. Using data for 500 individuals the following model of wage determination was estimated : Log (W) = 8, + B, D, + B, EDU, + B, (D, * EDU) + u, Where, W = Daily wages in rupees D = Dummy variable for gender, D=1 for females and 0 for males EDU = Years of education D*EDU = Interactive dummy The table below gives the estimated regression coefficients and their standard errors : Variables "| Siandard exrers c —— 0.1260 | D 0.23000 ——To.1780 EDU 0.0800 0.0090 | D*EDU 0.006 Oost | PEO:5105 8 (a) Write the regression relating Log (W) to EDU for males and females separately. (b) The returns to education are measured by the percentage increase in wages due to an extra year of education. Find the returns to education. for males and females. (c) Interpret the coefficient of interactive dummy. (d) Is the difference between returns to education for males and females statistically significant at 5% level of significance? 6. Explain the followings : (a) Explain the problem of Multicollinea ? How can it be solved? (b) Explain the problem of Autocorrelation? How can it be solved? (c) Akaike information criterion (AIC). (1000) |