Dr. Alexander W. Schmidt-Catran
Ados in the mlt-package (1.4 beta)
- Explained variance in multilevel models: mltrsq
This ado gives the Bosker/Snijders and
Bryk/Raudenbush R-squared values after hierarchical mixed models
(postestimation for xtmixed). Keywords: explained variance, multilevel, hierarchical mixed models, Bosker, Snijders, Bryk,
Raudenbush, r-square, r-squared, r2.
- Outlier detection in multilevel models: mltcooksd
calculates the influence statistics Cook's D and DFBETAs for higher-level units
after hierarchical mixed models (postestimation for xtmixed, xtmelogit and xtmepoisson). The purpose of this ado
is to detect outliers or influential cases at the higher level. Keywords: outliers, influential cases, regression diagnostics, multilevel, hierarchical
mixed models, Cook's D, DFBETAs, xtmixed, xtmepoisson, xtmelogit.
- Outlier detection in multilevel models: mltshowm
is an postestimation command
for mltcooksd. It can be used to obtain the estimation results of those models exculding the outliers.
- Scatter plots at higher levels: mltl2scatter
is an easy way to produce scatter plots at higher levels (with aggregated data). Together with mlt2stage it can be used
to plot estimated slopes against third variables. Keywords: scatter plot, aggregated data, two-stage plots, slopes as outcomes plots.
- Two-stage/Slopes as outcomes multilevel models:
is an easy way to estimate two-stage
(or slopes as outcomes) multilevel models. The program
estimates separate regressions (regress and logit) for each higher-level unit. mltl2scatter can be used to plot the
estimated coefficients against higher-level variables. Keywords: multilevel model, two-stage, slopes as outcomes, fixed effects.
© 2013 | Alexander Schmidt-Catran | All rights reserved |