Bucky's Archive for Data Analysis in the Social Sciences
Provides functions for various statistical techniques commonly used in the social sciences, including functions to compute clustered robust standard errors, combine results across multiply-imputed data sets, and simplify the addition of robust and clustered robust standard errors. The package was originally developed, in part, to assist porting of replication code from 'Stata' and attempts to replicate default options from 'Stata' where possible.
- NEWS for Bucky's Archive for Data Analysis in the Social Sciences
** Release 1.0.5
- Fixed bugs with predict method for robustified objects
- Added vcovHC method for polr objects (from MASS)
- Added predict method for mi.estimates objects based on lm or glm
** Release 1.0.4 (2017-11-30)
- Moved methods, sandwich, lmtest from "Depends" to "Imports"
- Fixed robustify examples ("coeftest" replaced by "summary")
- Added option for lazy evaluation to "mi.eval"
- Changed default for robust standard errors to HC0 to match Stata
- Added predict method for robustified objects based on lm or glm
- Improved compatibility between "mi.eval" and other packages
- Added compatibility with 'mice'
- Fixed compatibility with 'mitools'
- Added compatibility with 'stargazer' when using lm, glm, and a
few other types of models
** Release 1.0.3 (2017-11-20)
- Initial submission to CRAN
- Initial check-in to Github