In my experience four or higher way interactions are are usually due to data entry errors. Before trying to interpret a fourway interaction, look at the boxplot for each cell.

Art Art@DrKendall.org Social Research Consultants University Park, MD USA (301) 864-5570

John.Antonakis@unil.ch wrote:

>Hi: > >Does anyone have any SPSS code to test a heterogenous slopes model? I estimated an ANCOVA model and the covariate interactes significantly with the independent variables (I have 3 IVs, and the four-way interaction of the 3IVs with the covariate is significant, as I expected). > >Apparently, the Johnson-Neyman procedure should be used in cases where the ANCOVA assumption for equality of regression slopes (on the covariate) are violated. > >Best, >John. > >********************************************************************* >John Antonakis, Ph.D. >Assistant Professor >University of Lausanne >522-BFSH-1 >CH-1015 Lausanne-Dorigny >Switzerland > >Tel: ++41 (0)21 692-3461 >Fax: ++41 (0)21 692-3495 > >******************************************************************** > > > >