This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Introduction to Applied Econometrics Midterm 2 Do not turn page until told to do so! Name: ID & TA: Richard Walker, Winter 2010 Introduction to Applied Econometrics Richard Walker, Winter 2010 Midterm 2 Name: ID & TA: (50points) Answer all questions, using the spaces provided. There is room for doodling at the back. There is also a formula sheet and a table of sample correlation coefficients. I am interested in the determinants of earnings, for example education, gender and different kinds of intelligence (specifically numerical and verbal). Using the same dataset we’ve seen in class, I first run the following regression: LGEARN = β 1 + β 2 S + u (#1) where LGEARN is the natural log of EARNINGS (which in turn is hourly earnings in dollars) and S is years of schooling. The regression results are given in the relevant table at the end (regression #1). Note that I’ve deliberately omitted some of the results (indicated by ‘*’ ). 5pts 1. Use the regression results provided to determine the results of an Ftest (at the 5% significance level) of the explanatory power of regression #1. Explain what you’re doing. I then add two new regressors to my equation: LGEARN = β 1 + β 2 S + β 3 ASVAB02 + β 4 ASVAB03 + β 5 ASVAB04 + u (#2) where ASVAB02 is the individual’s score on an arithmetic reasoning test, ASVAB03 the individual’s score on a word knowledge test and ASVAB04 an individual’s score on a paragraph comprehension test. These variables are meant to capture the numerical (ASVAB02) and verbal (ASVAB03 & ASVAB04) dimensions of intelligence referred to above. The results are in the relevant table at the rear (regression #2). Again, some numbers have been withheld (indicated by ‘*’ )....
View
Full Document
 Spring '08
 Witte
 Macroeconomics, Econometrics, Regression Analysis, Natural logarithm, sample correlation, relevant table

Click to edit the document details