Help with Satorra & Bentler

I have a problem regarding Satorra & Bentler:

Could anyone give a precise explanation of SB statistics and what it tends to do if there is excess kurtosis in the observed variables?


I am not an expert, so won't try to answer your question.
However, there are several excellent sources out there on the net that discuss this "corrected" chi-square and it's function.

Assuming you are on a campus somewhere, you can also use PSYCinfo or Medline and look for an article that, in layman's terms, discusses scaled chi-square values or SEM or CFA statistics for non-normal data (ie, under conditions of non-normality).

Also, any recent SEM text should have a mention of this, and any university library should have several.

good luck!