作者: Thomas P. Hettmansperger , Joseph W. McKean , Simon J. Sheather
关键词:
摘要: A test for symmetry of the distribution errors in a linear model is proposed. It goodness-of-fit type based on discrepancy between two robust fits. The first fit appropriate under symmetric while second skewed as well distributions. proposed and asymptotically free. Besides deriving statistic’s null asymptotic distribution, its efficiency general class local alternative obtained which allows determination test’s relative with competitors.