dr Katarzyna Kopczewska
Class 05 / Outline:
-advanced panel models
Class 05 issues
Use the same dataset as in Class 04
TASKS
- Estimate basic models (pooled, FE, RE)
- Check for heteroskedasiticity and autocorrelation and other problems
- Try NEW models via .xtregar, .xtpcse, xtgls
- Find differences in models. Which one is the best. Try using postestimation commands.
Details:
. xtregar fits fixed affects and random effects models where the disturbance follows an AR(1) process:
where and
It can be FIXED or RANDOM effects
To be sure whether autocorrelation of effror term apperas update your STATA with . xtserial command.
When serial autocorrelation (in groups/series) and/or heteroskedasticity appear two types of models can help: .xtpcse or .xtgls
Testing for heteroskedasticity is given here:
. xtpcse: from STATA help file: “. xtpcse calculates panel-corrected standard error (PCSE) estimates for linear cross-sectional time-series models where the parameters are estimated by OLS or Prais-Winsten regression. When computing the standard errors and the variance-covariance estimates, . xtpcse assumes that the disturbances are, by default, heteroskedastic and contemporaneously correlated across panels. Use it when T > N
xtgls estimate panel models under various assumptions about heterogeneityof variances across panels and possible serial correlation.
Options:
noconstant suppress constant term
panels(iid)use i.i.d. error structure
panels(heteroskedastic)use heteroskedastic but uncorrelated error structure
panels(correlated)use heteroskedastic and correlated error structure
corr(independent)use independent autocorrelation structure
corr(ar1)use AR1 autocorrelation structure
corr(psar1)use panel-specific AR1 autocorrelation structure
Other useful stuff:
. xtline var1 # to produce line chart of given variable for every unit over time
. xtcsd # Testing for cross-sectional dependence in panel-data models INSTALL IT!
. xttest0 # Breusch and Pagan LM test for random effects, for use after xtreg, re,
. xttest2 # “calculates the Breusch-Pagan statistic for cross-sectional independence in the residuals of a fixed effect regression model or a GLS model estimated from cross-section time-series data. xtreg, fe estimates this model assuming independence of the errors”. INSTALL IT!
. xttest3 # “calculates a modified Wald statistic for groupwise heteroskedasticity in the residuals of a fixed effect regression model. It is for use after xtreg, fe or xtgls (with the default panels option)” INSTALL IT! This is new version of .xttest1, which was withdrawn