INFLUENCE ANALYSIS IN THE LINEAR MIXED MODEL

Creative Commons License

TÜRKAN S., TOKTAMIS O.

PAKISTAN JOURNAL OF STATISTICS, vol.28, no.3, pp.341-349, 2012 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 28 Issue: 3
  • Publication Date: 2012
  • Doi Number: 10.1002/bimj.201600188
  • Journal Name: PAKISTAN JOURNAL OF STATISTICS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.341-349
  • Hacettepe University Affiliated: Yes

Abstract

The linear mixed model is considerably sensitive to outliers and influential observations. It is known that outliers and influential observations affect substantially the results of analysis. So it is very important to be aware of these observations. Some diagnostics which are analogue of diagnostics in multiple linear regression were developed to detect outliers and influential observations in the linear mixed model. In this paper, the new diagnostic measure which is analogue of the Pena's influence statistic is developed for the linear mixed model.