Dselby86
commented
3 years ago We have conflicting Instructions. So choose 1 to complete
On Sat, Jun 19, 2021, 8:23 AM Dimetry-Adel @.***> wrote:
Hi everyone, Sorry for bothering you again,
On Canvas Announcements, there is a PDF file for Lab-05 Instructions containing more questions than the posted questions on the Course Schedule ( https://ds4ps.org/cpp-524-sum-2021/schedule/).
Kindly, which instructions we need to follow or we have to answer all questions from both instructions?
Course Schedule questions:
- Which of the three estimators you are using in each regression model (reflexive, post-test only, or pre-post with comparison)?
- What each estimator represents? (Explain in your own words what you believe each represents)
- Which estimate is most appropriate for your research question? (Select the one that you believe is the best measure for the pay discrimination component of the gender pay gap)
lab-05-instructions.pdf https://github.com/DS4PS/cpp-524-sum-2021/files/6680899/lab-05-instructions.pdf
PDF file questions:
-
Part I: The Naïve Model (A) Test the hypothesis that men get paid more than women using only the salary and gender variables. Run the following regression: Pay = b0 + b1*male What is the difference in pay? Is the relationship statistically significant? Interpret your results. (B) What sort of research design is this? Reflexive, post-test only, or pre-post with comparison (diff-in-diff)? (C) If you define pay discrimination as paying two people different amounts for the exact same work, do you feel this is a good estimate of pay discrimination? Why not? What are potential problems with the naïve model?
Part II: Improved Cross-Section Model (A) Improve upon your model from Part I by adding control variables to account for omitted variables. Interpret your results. (B) How did your results change, if any?
Part III: The Difference-In-Difference Approach (A) Use a difference-in-difference model to examine how the pay structure changes when a nonprofit has a male executive director and hires a new female Executive Director (ED). The “treatment” in this scenario is a change in the gender of the ED. (B) What is the counterfactual in this case? Is it a better estimate of the pay gap than cross-sectional, observational models from Part I and Part II? (C) Interpret the results of the model by answering the following questions: (C1) Are the nonprofits that switch from male to female EDs any different than those that hire another male ED? (C2) Do we see any secular trends (change in pay when a male replaces a male, i.e. the ‘control’ group)? (C3) Do we observe a (statistically-significant) pay gap for female EDs? How much is their pay different than what was expected? (D) What is the female ED counterfactual salary? What they would get paid if they were paid the same as male EDs?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/DS4PS/cpp-524-sum-2021/issues/25, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB4EHB7KBCJYBUG7LWPXATDTTSY7LANCNFSM467E4VUQ .
Dimetry-Adel
Hi everyone, Sorry for bothering you again,
On Canvas Announcements, there is a PDF file for Lab-05 Instructions containing more questions than the posted questions on the Course Schedule (https://ds4ps.org/cpp-524-sum-2021/schedule/).
Kindly, which instructions we need to follow or we have to answer all questions from both instructions?
Course Schedule questions:
lab-05-instructions.pdf
PDF file questions:
Part I: The Naïve Model (A) Test the hypothesis that men get paid more than women using only the salary and gender variables. Run the following regression: Pay = b0 + b1*male What is the difference in pay? Is the relationship statistically significant? Interpret your results. (B) What sort of research design is this? Reflexive, post-test only, or pre-post with comparison (diff-in-diff)? (C) If you define pay discrimination as paying two people different amounts for the exact same work, do you feel this is a good estimate of pay discrimination? Why not? What are potential problems with the naïve model?
Part II: Improved Cross-Section Model (A) Improve upon your model from Part I by adding control variables to account for omitted variables. Interpret your results. (B) How did your results change, if any?
Part III: The Difference-In-Difference Approach (A) Use a difference-in-difference model to examine how the pay structure changes when a nonprofit has a male executive director and hires a new female Executive Director (ED). The “treatment” in this scenario is a change in the gender of the ED. (B) What is the counterfactual in this case? Is it a better estimate of the pay gap than cross-sectional, observational models from Part I and Part II? (C) Interpret the results of the model by answering the following questions: (C1) Are the nonprofits that switch from male to female EDs any different than those that hire another male ED? (C2) Do we see any secular trends (change in pay when a male replaces a male, i.e. the ‘control’ group)? (C3) Do we observe a (statistically-significant) pay gap for female EDs? How much is their pay different than what was expected? (D) What is the female ED counterfactual salary? What they would get paid if they were paid the same as male EDs?
Thank you in advance.