Estimates models with increasing number of comparison subjects starting from 1:1 to using all available comparison group subjects.
psrange(df, treatvar, formula, nsteps = 10, nboot = 10, samples,
type = c("logistic", "ctree"), ...)
data frame with variables to pass to glm
vector representing treatment placement. Should be coded as 0s (for control) and 1s (for treatment).
formula for logistic regression model
number of steps to estimate from 1:1 to using all control records.
number of models to execute for each step.
the sample sizes to draw from control group for each step.
either logistic for Logistic regression (using glm
function) or ctree for Conditional Inference Trees (using the
ctree
function).
other parameters passed to glm.
a class of psrange that contains a summary data frame, a details data frame, and a list of each individual result from glm.