How setzler/eventstudy works
eventStudy is a deprecated R package for estimating event-study treatment effects from long-format panel data. Its documented user workflow is to provide data and variable names to ES, which produces event-study estimates; users can then plot levels or treatment effects. Major documented capabilities include stacked controls, cohort-specific or pooled effects, anticipation handling, and adjustments for non-parallel trends. A built-in simulator is also documented. Source excerpts were unavailable, so internal call relationships beyond the README-described workflow are not asserted.
Main entry points: ES.R ES command · plot_functions.R Result plots · simulate.R Data simulator
Components 13
Analysis workflow
- ES command
eventStudy/R/ES.R
Treatment estimation
- Stacked controls
eventStudy/R/by_cohort_ES.R - Cohort effects
eventStudy/R/by_cohort_estimate_ATT.R - Pooled effects
eventStudy/R/estimate_ATT.R - Bootstrap inference
eventStudy/R/bootstrap_ses.R
Trend adjustments
- Anticipation handling
eventStudy/R/pre_process.R - Trend adjustments
eventStudy/R/wald_ES.R - Covariate balancing
eventStudy/R/ipw.R
Results and tools
- Event-study estimates
eventStudy/R/ES.R - Result plots
eventStudy/R/plot_functions.R - Data simulator
eventStudy/R/simulate.R
Other components
- Analyst
- Panel data
Connections 5
- Analyst → ES command: runs
README.md - Panel data → ES command: provides data
README.md - ES command → Event-study estimates: estimates effects
README.md - Event-study estimates → Result plots: plots results
README.md - Analyst → Data simulator: simulates data
README.md