Please choose the best answer and briefly explain your reasoning. 1. Which of th
ID: 1104240 • Letter: P
Question
Please choose the best answer and briefly explain your reasoning.
1. Which of the following is true?
a) If two regressors are highly correlated (with a correlation coefficient 0.95), then the regression suffers from perfect collinearity.
b) With perfect collinearity, OLS does not work.
c) With multicollinearity, OLS does not work.
d) With perfect collinearity, OLS can still be unbiased, but standard errors tend to be large
2. Estimate a simple linear regression
Home _ price = + 1SQF + 2 Yearbuilt + 3Crime _rate + u using a random sample of houses in CA. If the error u captures other amenities in the neighborhood, and you believe that crime rate is negatively correlated with amenities, but is unrelated with square feet (SQT) and in which year a home was built (Yearbuilt). Which of the following statement is true?
a) Crime _ rate is exogenous.
b) Yearbuilt is endogenous.
c) OLS estimate of 3 will be biased, but OLS estimates of 1 and 2 will not be biased.
d) OLS estimates of all slopes will be biased
Explanation / Answer
1)Ans is D
With perfect collinearity, ols estimator remains unbiased and consisitent, only change is that they became more sensitive to small change in prices and standard error tends to be large.
2)D
Due to omission of relevant variable there is an exclusion error and ols estimafor becomes biased.