Premium
Leverage and Influence in Autocorrelated Regression Models
Author(s) -
Puterman Martin L.
Publication year - 1988
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.2307/2347495
Subject(s) - leverage (statistics) , regression , regression analysis , econometrics , autocorrelation , statistics , linear regression , mathematics
SUMMARY This paper examines the effect of individual observations in regression models with AR (1) errors through use of the ‘hat matrix’ and other influential observation diagnostics. It shows both analytically and through examples that in many practical settings the first transformed observation can have a relatively large hat matrix diagonal component and a large influence on parameter estimates. This provides additional evidence for retaining the first observation when performing estimation in this setting.