So I'm reviewing over for my exam, and there is one problem in the book that I don't understand. It has to do with the dummy variable trap causing perfect multicollinearity. If anyone has the book by Stock and Watson, its 7.4 in chapter 7.
Anyways so we got a regression that predicts a persons wage going with wage=B0+B1College+B2Female+B3Age+B4Ntheast+B5Midwest+B6South where all of these are binary variables. To avoid perfect multicollinearity, west was omitted the equation.
The question says girl one is from south, girl two is from west, and girl 3 is from midwest...construct a 95% CI for the difference in expected earnings between girl one and girl two.
The answer the professor posted:
The expected difference between girl one and girl two is (X6girl1 - X6girl2) XB6=B6 Thus a 95% CI is B1+/- 1.96(SE). I mean I know how to calculate a CI...but where the hell does X6girl1-X6girl2 come from..i mean X6 is the variable for the region south, and girl 2 is from the west so why did he use X6?
why don't you answer it yourself, obama...you transcend math, after all
^ Told you affirmative action is junk.... how did this guy get into Harvard Law?? sheese....
The difference in earnings should just be the B6South coefficient. To get the expected earnings for the girl from the west, you input 0 for B4, B5, and B6. To get the expected earnings for the southern girl, well you obviously put 0 for the B4 and B5 dummy variables and then 1 for B6.
Ut iste voluptatem repellendus rerum. Culpa beatae aliquid eos et ut facere eum aut.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...