You are considering running a heteroskedastic-robust version of a regression pre

Question

You are considering running a heteroskedastic-robust version of a regression presented above. You are deciding between using OLS with robust standard errors vs. weighted least squares (WLS). How would each of these methods impact the coefficients and standard errors of your model compared to regular OLS? What would be one advantage and one disadvantage to using each method? After further consideration, you believe that an important explanatory variable has been omitted from the regression presented above, violating one of the Gauss-Markov assumptions. In this case, would WLS be preferred to OLS? Explain why or why not.

Explanation / Answer

2. Using weighted least squares is like giving importance to some variables and there can be a chance that high weighted variables might be unstable in out of time, which is at least not the case in OLS with robust standatd errors. Coeffients would be changed in case of weighted least squares, since the variable which needs more weightage in the problem will have a greater absolute value than that of OLS

Hence it is better to go with using OLS with robust standard errors. But sometimes the results from OLS and robust OLS are not much differing

Related Questions

Navigate

• Browse (All)
• Browse Y
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.