Unifying Estimation Results with Binary Dependent Variables

Calculate unified measures that quantify the effect of a covariate on a binary dependent variable (e.g., for meta-analyses). This can be particularly important if the estimation results are obtained with different models/estimators (e.g., linear probability model, logit, probit, ...) and/or with different transformations of the explanatory variable of interest (e.g., linear, quadratic, interval-coded, ...). The calculated unified measures are: (a) semi-elasticities of linear, quadratic, or interval-coded covariates and (b) effects of linear, quadratic, interval-coded, or categorical covariates when a linear or quadratic covariate changes between distinct intervals, the reference category of a categorical variable or the reference interval of an interval-coded variable needs to be changed, or some categories of a categorical covariate or some intervals of an interval-coded covariate need to be grouped together. Approximate standard errors of the unified measures are also calculated.


News

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("urbin")

0.1-6 by Arne Henningsen, 3 months ago


http://r-forge.r-project.org/projects/urbin/


Browse source code at https://github.com/cran/urbin


Authors: Arne Henningsen , Geraldine Henningsen


Documentation:   PDF Manual  


GPL (>= 2) license


Suggests sampleSelection, maxLik, mfx, mlogit, MASS, mvProbit


See at CRAN