eQuizShow

Econometrics

Categorical Variables

Question:

How would you analyze gender descrimination with the following model?

wage = B0 + B1(educ) + B2(exper) + B3(tenure)


Answer:

You would use the following equation, and look at the null hypothesis that D0, D1, D2, and D3 are all equal to zero.

wage = B0 + D0*female + [B1*educ + D1*educ*female] + [B2*exper + D2*exper*female] + [B3*tenure + D3*tenure*female]

 

 



Question:

What are two ways to calculate the likelihood of a null hypothesis with multiple dummy variables?


Answer:

An F-test, which can be done in Stata, or a Chow test



Question:

How would you go about running a Chow test?


Answer:
  1. Split the data into two groups, male and female
  2. Find the SSR (sum of squared residuals) for the male group and the female group
  3. Find the SSR for the entire population
  4. Chow = [(SSR all - (SSR male + SSR female)) / (SSR male + SSR female)] * [(n - 2(k+1))/(k+1)]
  5. To calculate the statistic, this is an F distribution for k+1,n-2(k+1)


Question:
Answer:

Question:
Answer:

Diff-in-Diff

Question:

Test


Answer:

Test



Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer:

Panel Data

Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer:

IV

Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer:

General Review

Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer:

Question:
Answer: