mlpsa.logistic: Estimates propensity scores using logistic regression.

View source: R/mlpsa.logistic.R

mlpsa.logisticR Documentation

Estimates propensity scores using logistic regression.

Description

This method will estimate a separate logistic regression model for each level 2 (or cluster).

Usage

mlpsa.logistic(vars, formula, level2, stepAIC = FALSE, ...)

Arguments

vars

data frame containing the variables to estimate the logistic regression

formula

the logistic regression formula to use

level2

the name of the column containing the level 2 specification

stepAIC

if true, the [MASS::stepAIC()] from the 'MASS' package will be used within each level.

...

currently unused.

Value

a list of glm classes for each level 2 or stepwise-selected model if stepAIC is true.

See Also

getPropensityScores


multilevelPSA documentation built on April 11, 2025, 6:18 p.m.