You are given the time plots of the variables con, inv and inc. cont invt inct 3 Graph 1: Impulse Responses of VAR(2) system for con, inv and inc cont cont invt cont invt invt cont inct invt cont inct cont invt inct invt inct inct inct Table 1: Impulse Responses of VAR(2) system for con, inv and inc Responses to a one-standard error shock in con period con inv inc 1 0.00954 0.01359 0.00617 2 -0.00089 0.00460 0.00274 3 0.00289 0.00410 0.00102 4 0.00082 0.00046 0.00123 5 0.00037 0.00219 0.00053 6 0.00043 0.00048 0.00024 7 0.00021 0.00026 0.00029 8 0.00008 0.00040 0.00009 9 0.00010 0.00012 0.00006 10 0.00004 0.00007 0.00006 11 0.00002 0.00008 0.00002 12 0.00002 0.00002 0.00002 13 0.00001 0.00002 0.00001 14 0.00001 0.00001 0.00000 15 0.00000 0.00001 0.00000 16 0.00000 0.00000 0.00000 17 0.00000 0.00000 0.00000 18 0.00000 0.00000 0.00000 19 0.00000 0.00000 0.00000 20 0.00000 0.00000 0.00000 4 Responses to a one-standard error shock in inv period con inv inc 1 0.00000 0.04029 -0.00067 2 -0.00008 -0.01121 0.00183 3 0.00228 -0.00191 0.00173 4 -0.00003 0.00438 -0.00026 5 0.00022 0.00064 0.00024 6 0.00014 -0.00060 0.00033 7 0.00015 0.00045 0.00003 8 0.00004 0.00021 0.00004 9 0.00002 0.00002 0.00005 10 0.00003 0.00003 0.00001 11 0.00001 0.00003 0.00001 12 0.00000 0.00002 0.00001 13 0.00001 0.00000 0.00000 14 0.00000 0.00000 0.00000 15 0.00000 0.00001 0.00000 16 0.00000 0.00000 0.00000 17 0.00000 0.00000 0.00000 18 0.00000 0.00000 0.00000 19 0.00000 0.00000 0.00000 20 0.00000 0.00000 0.00000 Responses to a one-standard error shock in inc period con inv inc 1 0.00000 0.00000 0.00881 2 0.00255 0.00297 -0.00109 3 0.00220 0.00209 0.00122 4 -0.00081 0.00220 0.00090 5 0.00079 0.00043 0.00000 6 0.00031 -0.00022 0.00038 7 -0.00005 0.00080 0.00010 8 0.00014 0.00006 0.00002 9 0.00005 -0.00004 0.00009 10 0.00001 0.00015 0.00001 11 0.00003 0.00002 0.00001 12 0.00001 0.00000 0.00002 13 0.00000 0.00003 0.00000 14 0.00001 0.00000 0.00000 15 0.00000 0.00000 0.00000 16 0.00000 0.00000 0.00000 17 0.00000 0.00000 0.00000 18 0.00000 0.00000 0.00000 19 0.00000 0.00000 0.00000 20 0.00000 0.00000 0.00000 5 Graph 2: Impulse Responses 6 Graph 1 Graph 2 10 Graph 1 Graph 2 13 Question 4 Suppose that we examine the effect of a teaching method known as PSI on the performance of students in a course. The question is whether students exposed to the method scored higher on exams in the class. Data from students in two classes, one in which PSI was used and another in which a traditional teaching method was employed. You are given the following variables: Consider the following model: GRADEi = 0 + 1GPAi + 2PSIi + 3TUCEi + ui (a) You are given the following LPM estimation results: Pˆ(GRADE i = 1) = −1.58 + 0.52* GPAi + 0.30 * PSIi + 0.006 * TUCEi According to LPM, what is the chance of a student with a grade point average of 3, taught by traditional methods, and scoring 20 on the TUCE exam to get an A? Answer: Pˆ(GRADEi = 1) = −1.58 + 0.52*3 + 0.30 * 0 + 0.006 *20 = 0.10 This student would have about a 10% chance of getting an A. (b) According to LPM, what is the chance of a student with a grade point average of 3, taught by PSI method, and scoring 20 on the TUCE exam to get an A? 14 Pˆ(GRADEi = 1) = −1.58 + 0.52*3 + 0.30 *1 + 0.006 *20 = 0.40 Answer: This student would have about a 40% chance of getting an A. (c) What is the main disadvantage of LPM estimates? Answer: Negative probabilities can be predicted! (d) Below, you are some logit model estimation results. a. What is the chance of a student with a grade point average of 3, taught by traditional methods, and scoring 20 on the TUCE exam to get an A? b. What is the chance of a student with a grade point average of 3, taught by PSI method, and scoring 20 on the TUCE exam to get an A? c. Would getting into the PSI class increase the chances of getting an A or not? How much the difference? 15 Answers: a. The student who has a GPA of 3.0, is taught by traditional methods (PSI=0), and has a score of TUCE has approximately 7.27% chance of getting an A b. The student who has a GPA of 3.0, is taught by PSI method (PSI=1), and has a score of TUCE has approximately 35.3% chance of getting an A. c. The student would have about a 28.03% (=35.3-7.27) better chance. 16 **** Stata Commands ***** use logist.dta *LPM regress grade gpa psi tuce * What is the chance of a student with a grade point average of 3, taught by traditional methods, and scoring 20 on the TUCE exam to get an A? * According to these results, a student with a grade point of 3.0, taught by traditional methods, and scoring 20 on the TUCE exam would earn an A with probability of 0.1. So, this person would have about a 10% chance of getting an A. * logit estimation logit grade gpa tuce psi * margins margins, at(gpa = 3 psi = 0 tuce = 20) * How can you interpret the reported value for margin .0726861 * The student who has a GPA of 3.0, is taught by traditional methods (PSI=0), and has a score of TUCE has approximately 7.27% chance of getting an A margins, at(gpa = 3 psi = 1 tuce = 20) * How can you interpret the reported value for margin .0726861 * The student who has a GPA of 3.0, is taught by PSI method (PSI=1), and has a score of TUCE has approximately 35.3% chance of getting an A. * Would getting into the PSI class increase the chances of getting an A or not? How much the difference? * Yes, the student would have about a 28.03% (=35.3-7.27) better chance. 17 Question 5 Consider the following simple regression model: Yt = 0 + 1 X t + ut a) Write the likelihood function to derive the Maximum Likelihood (ML) estimators for 0 , 1 and 2 . Answer: b) Write the log likelihood function to derive the Maximum Likelihood (ML) estimators for 0 , 1 and 2 . Answer: c) Derive the Maximum Likelihood (ML) estimators for 0 , 1 and 2 . Are they unbiased estimators? Answer: 18 19