Premium
ESTIMATION WITH HETEROSCEDASTIC AND CORRELATED ERRORS: A SPATIAL ANALYSIS OF INTRA‐URBAN MORTALITY DATA
Author(s) -
Haining Robert
Publication year - 1991
Publication title -
papers in regional science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.937
H-Index - 64
eISSN - 1435-5957
pISSN - 1056-8190
DOI - 10.1111/j.1435-5597.1991.tb01730.x
Subject(s) - statistics , heteroscedasticity , ordinary least squares , econometrics , regression analysis , estimation , regression , exploratory data analysis , mathematics , geography , economics , management
Regression analyses using data collected across contiguous a real units of varying size often fail to satisfy the statistical assumptions of ordinary least squares estimation, since errors arc neither homoscedaslic nor uncorrelated. This paper considers regression analysis in the case of mortality data collected across 87 contiguous Community Medicine Areas defined for the Creater Glasgow conurbation. Data properties are identified using exploratory techniques. A statistical model is specified, parameters estimated udner alternative assumptions about the errors, and the fit assessed using both confirmatory and exploratory methods. A pricipal objective of the study is to identify areas that have particularly high or low mortality rates.