This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: TA session 9 Econ. 103, winter 2010 Wed., Mar. 3, 2010, 10:00 a.m. and 1:00 p.m. in PP2400E. Problem 2 First lets summarize the data and get an idea of what were looking at. sum Variable  Obs Mean Std. Dev. Min Max+ gradecat  657 2.245053 .9575967 1 5 female  657 .5022831 .5003757 1 ageyrs  657 16.15221 1.215236 14 18 white  648 .5185185 .5000429 1 black  648 .183642 .387491 1+ hispanic  648 .2268519 .4191195 1 other  648 .0709877 .2570026 1 beltpct  653 76.90658 23.67181 20 100 drivedrunk  642 .1308411 .3374895 1 ridedrunk  655 .3343511 .4721237 1 Most of these variables are dummy variables. The only ones that arent are gradecat and beltpct . Use the browse gradecat beltpct command to take a look at these exceptions. Even though they are not dummy variables, gradecat and beltpct are also discretevalued variables. 2.1 Generate dummy variables for gradecat The question asks you to generate dummy variables, specifically for the students who get Fs. One easy way to do this is just to write . gen getsFs=1 if gradecat == 5 . replace getsFs=0 if gradecat !=5 and this will do it. A fun and powerful command if you need to generate an entire set of dummy variables is . tabulate gradecat, gen(grade) which generates a set of dummy variables grade1 , grade2 , grade3 , grade4 , and grade5 , each corresponding to one of the categories in gradecat . For the purposes of this answer key Im going to assume we used both the above commands and have a dummy called getsFs as well as a set of dummies grade1 through grade5 . 1 2.2 Estimate the model: seat belt wearing= f(grades, race, sex, whether you have driven drunk, and whether you have rode with a drunk) Regression 1 . reg beltpct grade2 grade3 grade4 grade5 ageyrs white black hispanic female dri > vedrunk ridedrunk Source  SS df MS Number of obs = 629+ F( 11, 617) = 7.03 Model  39138.0031 11 3558.00028 Prob > F = 0.0000 Residual  312224.795 617 506.036945 Rsquared = 0.1114+ Adj Rsquared = 0.0955 Total  351362.798 628 559.494901 Root MSE = 22.495 beltpct  Coef. Std. Err. t P>t [95% Conf. Interval]+...
View
Full
Document
 Winter '07
 SandraBlack
 Econometrics

Click to edit the document details