作者: N. G. Cadigan
DOI: 10.1080/10705519509539992
关键词: Statistics 、 Data set 、 Generalized least squares 、 Mathematics 、 Least squares 、 Maximum likelihood 、 Measure (mathematics) 、 Structural equation modeling
摘要: Local‐influence diagnostics based on maximum likelihood (ML), generalized least squares (GLS), and unweighted (ULS) fit functions are developed for structural equation models (SEMs). The influence of observations, components variables is considered. illustrated with an example data set, comparisons made equivalent global measures influence. local the set ML GLS estimates very similar, but it much different from that ULS. observations also different. Although not possible to define a uniformly best measure influence, local‐influence here more versatile than global‐influence in assessing analysis SEMs.