Skill Name
Purpose
Describe what this skill does and when an economist should use it. Be specific about the problem it solves.
When to Use
- Use this skill when you need to…
- This is especially helpful for…
Instructions
Follow these steps to complete the task:
Step 1: Understand the Context
Before generating any code, ask the user:
- What is the research question?
- What data format are you working with?
- What software/language preference do you have?
Step 2: Generate the Output
Based on the context, generate [code/document/analysis] that:
- Follows best practices - Use standard conventions for [Stata/R/Python/LaTeX]
- Is reproducible - Include comments explaining each step
- Handles edge cases - Check for missing data, outliers, etc.
Step 3: Verify and Explain
After generating output:
- Explain what the code does
- Highlight any assumptions made
- Suggest next steps or improvements
Example Prompts
Users might invoke this skill with prompts like:
- “Run a difference-in-differences analysis on my treatment data”
- “Clean this dataset and create summary statistics”
- “Write the methodology section for my regression analysis”
Example Output
* Example Stata code this skill might generate
* ============================================
* Load data
use "data.dta", clear
* Summary statistics
summarize var1 var2 var3
* Main regression
regress y x1 x2 x3, robust
eststo model1
* Export results
esttab model1 using "results.tex", replace
Requirements
Software
- Stata 17+ / R 4.0+ / Python 3.10+
Packages
- For R:
tidyverse,fixest,modelsummary - For Python:
pandas,statsmodels,linearmodels - For Stata: Built-in commands
Best Practices
When using this skill, ensure:
- Data is properly formatted - Variables named clearly, no special characters
- Sample is defined - Clear inclusion/exclusion criteria
- Output is documented - All results are reproducible
Common Pitfalls
Avoid these mistakes:
- ❌ Running analysis without checking data quality first
- ❌ Ignoring missing values or outliers
- ❌ Not specifying robust standard errors when needed
References
- [Relevant documentation or textbook]
- [Academic paper on methodology]
- [Software package documentation]
Changelog
v1.0.0
- Initial release