Suppose I am interested in modeling how wages are determined. My model of interest is log(wage) = β0 + β1educ + u where the wage is the hourly wage earned and Educ is total years of education a worker has. (a) Why do we expect β1 > 0? (b) Should years of experience working be included in a model of wage determination (why or why not)? (c) If years of experience working is excluded from my model, what is the likely effect on the bias of the OLS estimator of β1? (d) Interpret β1 in the above model. (e) If I inform you that years of education and years of work experience are correlated, does that mean that years of work experience should not be included in the model to eliminate collinearity (why or why not)? (f) My estimated regression is log wage = 0.284 + 0.073 · Educ. Interpret the coefficient on Educ from my estimated regression ? (g) R^2 = 0.304 in this model. What does this mean? (h) Would R^2 decrease if I added years of experience as a regressor in my model? (i) Would R^2 decrease if I added the individual's height as a regressor in my model? (j) How would your interpretation of β1 change if you used log(Educ) as a regressor instead of Educ?
Suppose I am interested in modeling how wages are determined. My model of interest is
log(wage) = β0 + β1educ + u
where the wage is the hourly wage earned and Educ is total years of education a worker has.
(a) Why do we expect β1 > 0?
(b) Should years of experience working be included in a model of wage determination (why or why not)?
(c) If years of experience working is excluded from my model, what is the likely effect on the bias of the OLS estimator of β1?
(d) Interpret β1 in the above model.
(e) If I inform you that years of education and years of work experience are
(f) My estimated regression is
log wage = 0.284 + 0.073 · Educ.
Interpret the coefficient on Educ from my estimated regression
? (g) R^2 = 0.304 in this model. What does this mean?
(h) Would R^2 decrease if I added years of experience as a regressor in my model?
(i) Would R^2 decrease if I added the individual's height as a regressor in my model?
(j) How would your interpretation of β1 change if you used log(Educ) as a regressor instead of Educ?
Trending now
This is a popular solution!
Step by step
Solved in 4 steps
