| name | stata-regression |
| description | Design and write reproducible Stata regression workflows with diagnostics and publication-ready output. Use when Codex needs to specify OLS, fixed-effects, panel, logit/probit, or related regressions in Stata, choose clustering and controls, export tables with esttab or outreg2, or suggest diagnostics and robustness checks for empirical research. |
Stata Regression
Overview
Produce clear, reproducible Stata regression workflows for empirical research. Build estimation code, table-export code, and concise explanation of identification choices, diagnostics, and robustness checks.
Workflow
- Gather the specification before drafting code: dependent variable, main regressors, controls, fixed effects, panel structure, clustering level, sample restrictions, weights, and desired export format.
- Choose the estimator that matches the design. Use
regress for baseline OLS, reghdfe for multiple high-dimensional fixed effects, xtreg for panel models, and nonlinear estimators such as logit, probit, or poisson when the outcome requires them.
- Write a reproducible do-file skeleton with
version, data loading, quick validation checks, transformations, estimation blocks, stored models, and exported tables.
- Match standard errors to the research design. Explain why the chosen clustering or robust variance option fits the treatment variation and sampling structure.
- Add diagnostics and robustness suggestions after the main specification: alternative fixed effects, alternative samples, functional-form checks, balance or pre-trend checks when relevant, and table notes that document the design.
- Explain what each model estimates, what assumptions matter, and what the user should verify in the data before trusting the coefficients.
Output Expectations
- Write Stata code that can be pasted into a do-file with minimal cleanup.
- Keep model labels, variable labels, and export filenames explicit.
- Prefer publication-ready table workflows with
esttab or outreg2.
- State package dependencies when using community commands such as
reghdfe, esttab, or outreg2.
- Surface design risks such as bad clustering choices, omitted fixed effects, or unclear variable definitions.
References
Read references/stata-regression-recipes.md when choosing commands, table-export patterns, diagnostics, or robustness checks.
