G-492 Project Description
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
G-492 Project Description
Setup
This is a great opportunity to let your knowledge and skills shine.You must choose one of the following major companies to advise: Amazon,Apple,Anheuser-Busch, Dell,Delta,Netflix, Nike, Tesla, UBER, or Walmart. For the chosen company, you will provide clear, pointed, and well-supported analysis of one of the following causal relationships between company choice variables and outcomes (organized by category).Using this analysis,you will then make recommendations to meet your company's long-term strategic objectives.
1. The Bottom Line: The effect of price changes on profits
2. Sales and Marketing: The effect of changes in advertising budget on sales
3. Human Resources: The effect of salary changes on employment longevity
4. Customer Service: The effect of help desk employment levels on number of complaints
5. Production: The effect of machine maintenance rates on proportion of faulty units
Note that some of these relationships are vague and not company specific. However, once you have chosen a company and a task,I will tailor them (when necessary) to be relevant for that company. Then, I will provide each student a(fabricated) dataset containing sufficient information to conduct analysis for the question they have chosen to address. For the project, your recommendation should only take into account your analytical findings, and not those of other students
Each student submits a written analysis, concerning one of the above relationships.
1. Must contain 1800-2500 words of text.
2. Can have relevant and supporting graphs,tables,and figures. Be professional in
visualization.
3. Must contain a bibliography of works cited.
4. Using AI techniques to create the report is considered cheating. Write your own report.
Major Deliverables
1. Submit a pseudo code for cleaning data for the final project.
o 12 PM on Tuesday, March 7th.
2. Submit the final code cleaning your data for the final project.
o 12 PM on Friday, March 10th.
3. Final project first draft
o 11:59 PM on Friday, April 14.
o Late turn-ins are accepted by 12:00 PM on Sunday, April 16.
4. Peer-review of the first draft
o 12:59 PM on Friday,April 21.
o Late turn-ins are accepted by 12:00 PM on Sunday, April 23.
5. Final project submission
o 11:59 PM on Friday, April 28.
o No turn-ins are accepted.
Company research
These data are to be used in conjunction with your own company research to conduct your analysis. Note that your company research should not include any collection of raw data from other sources, and all direct technical analysis should be done using the data I provide. To be concrete,your company research should be used to determine things such as: relevant company history, strategic objectives, market and market competitors, organizational structure, etc. Then, treating the data I provided as actual data provided to you by the company, your individual project should combine the technical analysis with your company research.
Readers
Keep in mind that readers of your report are the clients who hired you-the decision makers of the company. Imagine they have hired you as a consultant and have given you the relevant data for analysis. Your clients had taken a class like G-492 more than 10 years ago, meaning that they might vaguely remember some terms, such as endogeneity, selection bias,fixed effects, instrumental variables, or control variables. You need to write a report assuming any intelligent clients of yours can understand.You will still have to use econometrics terms and math in your report but must be defined so that the readers can understand. Also, explain your logic in a language that they could understand.
Details on contents
Your job is to convince your clients that whatever you are claiming, assuming, or concluding is true. You must clearly explain what you did to get that number and why in manner that a smart individual with limited econometric knowledge will understand(see readers).You should combine all analytical skills that we have learned in class to create a convincing report to argue the causal effect of X on Y.
(1) The limitation of OLS regressions is that it requires additional assumption(s) to identify a causal effect. Describe whether, how, and why those assumption(s) is likely or not likely to be satisfied in your setup. Describe scenarios, cases, or environments could potentially cause the problem in OLS, using specific examples. Remember to do so in a language that people without business analytics backgrounds can understand.
(2)During analysis,you may try a number of different model specifications and econometrics tools but you would eventually settle on one (or a couple more) approach(s)to make causal inference and recommendations. Do not waste space to show what approaches you have tried and how you reached your final model specification unless it is absolutely needed to explain something you are trying to argue.
(4) Since the goal is to convince the readers that the estimates of your model represent the causal effect of interest, explain in plain words why your approach is suitable for causal inference. For example, clearly state what your final model specification is, why you have chosen such model, what and how econometrics techniques are implemented, and why they are needed.
(5) Use the measured effect to draw conclusions and formulate a recommendation in the context of your company's long-term strategic objectives. The purpose of the report is to offer a recommendation that is drawn based on empirical findings.
Visualization
Your writing and result visualizations should appear professional.Screen captures or messy tables and/or graphs will be considered as not professional.
A few notes concerning the project:
1. The companies you will be analyzing are generally large. Therefore, to keep the project requirements reasonable,you will be focusing on a narrow subset of your chosen company's business.
2. The number of observations may seem inflated in some cases. For example, you may have several hundred observations for employee offices when the company clearly has fewer than that. This won't materially affect your analysis or assessment, and it is only done to make the analysis more meaningful in this controlled environment.
3. Rubric is available on Canvas.
2023-04-15