P-values have little meaning in this context. Nor do regression coefficient
estimates and especially standard errors (they are biased).
Frank
-
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context:
http://r.789695.n4.nabble.com/leaps-regsubsets
Hi, does anyone know if there is a way to easily extract p-values from the
regsubsets() function?
Thanks,
James Stegen
p.s. this is a reposting due to me not putting in a useful subject heading...
--
James C. Stegen
NSF Postdoctoral Fellow in Bioinformatics
University of North Carolina
Chapel Hi
2 matches
Mail list logo