Assignment 03 – Creating a data analysis report

Published

January 17, 2026

This is a group assignment. Continue with the research question you had chosen in class.

Use AI iteratively to help you create a report.

First run

Think in terms of two variables (\(x\) and \(y\)).
- Pick a single variable or a combined index of variables
- Pick a GDP variable
- Think about a causal link.
Create one carefully designed graph to illustrate the relationship (Graph 1)
At each step, explain your choices/decisions (ie why you chose a certain variable)
Create a Conclusion paragraph where you summarize your work and results in 80-100 words.
create another graph to show heterogeneity of country size groups using population (Graph 2)
Run a regression and interpret the coefficient. Make sure they are precise.
Use AI to get a well cratfted folder structure

Extended context window run 1

create a system prompt will preferences, expected output, style, etc
recreate a the report

Extended context window run 2

Once you have a draft, Upload slideshow for chapters 10 data analysis slides. and ask to follow it and revise report.
recreate a the report

Final repor

You have three versions. Use them to have a single one you like the most. Edit further if needed.

To submit

report of maximum 2 pages in .pdf format (including exhibits) (lastname_firstname_dawai_week03_report.pdf) (12p)
your code (or provide a link). (2p)
Text: What advice would you give to a fellow data analysis student on using AI to create a report? (How do various runs compare? What changes are decisive? How many iterations and useful?) (lastname_firstname_dawai_week03_advice.txt) (6p).

Upload your report to the student folder called Reports at moodle or similar service if applicable. This is for the next class — each group will read another groups report and discuss the good and bad aspects of the report.

AI use restrictions

For the report, use AI as your assistant, use it as input not as output. Don’t submit an AI-generated report without review and edit.
Do not use AI to help you generate the advice. We want to hear your personal examples and reflections, not AI-generated suggestions.

---
title: "Assignment 03 -- Creating a data analysis report"
date: "2026-01-17"
---

This is a **group** assignment.  Continue with the research question you had chosen in class. 

Use AI iteratively to *help* you create a report. 

### First run

  * Think in terms of two variables ($x$ and $y$). 
    * Pick a single variable or a combined index of variables
    * Pick a GDP variable
    * Think about a causal link. 
  * Create one carefully designed graph to illustrate the relationship (Graph 1)
  * At each step, explain your choices/decisions (ie why you chose a certain variable)
  * Create a Conclusion paragraph where you summarize your work and results in 80-100 words.
* create another graph to show heterogeneity of country size groups using population (Graph 2)
* Run a regression and interpret the coefficient. Make sure they are precise. 
* Use AI to get a well cratfted folder structure

### Extended context window run 1

* create a system prompt will preferences, expected output, style, etc
* recreate a the report

### Extended context window run 2

* Once you have a draft, Upload slideshow for chapters 10 [data analysis slides](https://gabors-data-analysis.com/chapters/).  and ask to follow it and revise report. 
* recreate a the report

### Final repor

* You have three versions. Use them to have a single one you like the most. Edit further if needed. 

### To submit

1. report of maximum 2 pages in .pdf format (including exhibits) (lastname_firstname_dawai_week03_report.pdf) (12p)
2. your code (or provide a link). (2p)
3. Text: What advice would you give to a fellow data analysis student on using AI to create a report? (How do various runs compare? What changes are decisive? How many iterations and useful?) (lastname_firstname_dawai_week03_advice.txt) (6p).
    
* Upload your report to the `student folder called Reports` at moodle or similar service if applicable. This is for the next class --- each group will read another groups report and discuss the good and bad aspects of the report.

### AI use restrictions

* For the report, use AI as your assistant, use it as *input* not as *output*. Don't submit an AI-generated report without review and edit. 
* Do not use AI to help you generate the advice. We want to hear your personal examples and reflections, not AI-generated suggestions.