R functions
standardize.R: scaling regression inputs by dividing by two standard deviations
bayesglm.R: generalized linear modeling with independent normal, t, or Cauchy prior distributions on the coefficients (default is Cauchy with center 0 and scale 10 for the intercept and 2.5 for all other coefficients)
regression.R: miscellaneous useful functions
The papers
Scaling regression inputs by dividing by two standard deviations, by Andrew Gelman
A default prior distribution for logistic and other regression models, by Andrew Gelman, Aleks Jakulin, Maria Grazia Pittau, and Yu-Sung Su
Back to Andrew Gelman's homepage.