Variable selection in joint modelling of the mean and variance for hierarchical data

Christiana Charalambous, Jianxin Pan, Mark Tranmer

Research output: Contribution to journalArticlepeer-review


We propose to extend the use of penalized likelihood variable selection to hierarchical generalized linear models (HGLMs) for jointly modelling the mean and variance structures. We assume a two-level hierarchical data structure, with subjects nested within groups. A generalized linear mixed model (GLMM) is fitted for the mean, with a structured dispersion in the form of a generalized linear model (GLM) for the between-group variation. To do variable selection, we use the smoothly clipped absolute deviation (SCAD) penalty, which simultaneously shrinks the coefficients of redundant variables to 0 and estimates the coefficients of the remaining important covariates. We run simulation studies and real data analysis for the joint mean–variance models, to assess the performance of the proposed procedure against a similar process which excludes variable selection. The results indicate that our method can successfully identify the zero/non-zero components in our models and can also significantly improve the efficiency of the resulting penalized estimates.
Original languageEnglish
Pages (from-to)1-27
Number of pages27
JournalStatistical Modelling
Issue number1
Publication statusPublished - 26 Nov 2014


  • Generalized linear mixed models
  • H-likelihood
  • Mean-covariance modelling
  • Multilevel data
  • Smoothly clipped absolute deviation penalty


Dive into the research topics of 'Variable selection in joint modelling of the mean and variance for hierarchical data'. Together they form a unique fingerprint.

Cite this