RE: Standards for "Skewness"

[Dale Glaser]

Even though in the context of structural equation modeling you may want to take a look at: West, S. G, Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables.  In R. H. Hoyle (Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 56-75). Thousand Oaks, CA: Sage. ...whereby highly nonormal variables are characterized as skewness = 3, kurtosis = 21 (p. 68); even though this chapter is frequently cited in the SEM domain when normality is examined, I'm not sure (as with any convention or cutoff value) how well this generalizes to other statistical techniques..........................dale glaser   -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Ronald B. Livingston Sent: Monday, December 20, 1999 11:25 AM To: [EMAIL PROTECTED] Subject: Standards for "Skewness" Hello:      Are there "standards" for describing the skew of a distribution?  For example, 0 to 1 = mild; 1 - 2 = moderate, etc.  I am aware of tests of significance for skew, but with large samples practically any skew is significant. Any references would be appreciated.   Sincerely,   Ron