The runmixregls Stata command allows you to fit mixed-effects location scale models in Stata by calling Don Hedeker's MIXREGLS software bundled with the command.
These models extend standard two-level random-intercept models (xtreg, mle or mixed) by modelling the level-1 residual variance as function of covariates and an additional random effect.
In longitudinal data these models therefore allow you study within-individual variability as well as usual mean differences in behaviour over time.
The latest updates to MIXREGLS and runmixregls allow you to also apply this extension to two-level random slope models to allow for random individual specific time trends.
Other updates include allowing runmixregls to run on the M1 Mac in addition to the existing supported platforms, and a wider range of post-estimation prediction and residual options.
For more information see the runmixregls webpage http://www.bristol.ac.uk/cmm/software/runmixregls/.