rdrobust is a package for statistical inference in regression-discontinuity (RD) designs, a quasi-experimental method popular in social, behavioral, and natural sciences. It provides tools for point estimation, robust confidence intervals, bandwidth selection, and exploratory data analysis in Sharp, Fuzzy, and Kink RD settings. The package was authored by Sebastian Calonico.
Use Cases
- Estimating local average treatment effects at a cutoff based on the package's rdrobust() function.
- Selecting optimal bandwidths for RD analysis based on the rdbwselect() procedure.
- Conducting exploratory data analysis and creating RD plots based on the rdplot() tool.
- Applying robust inference methods to Sharp, Fuzzy, or Kink RD designs as described.
Strengths
- Provides a suite of specialized tools (rdrobust, rdbwselect, rdplot) for a specific and widely-used research design.
- Supports inference for three distinct RD settings: Sharp, Fuzzy, and Kink.
- Focuses on robust, data-driven statistical methods for causal inference.
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download.
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count and data scale are unknown, which may limit suitability assessment.
Provenance
- Source
- Sebastian Calonico
- Collection Method
- Software package for statistical analysis.
- Time Range
- null
- Freshness
- Last update date is unknown; freshness unverified.
- Geography
- null