I am interested in examining the main effects and interactions of a series of variables at both the between and within-person level. I originally conducted the analysis using the MIXED procedure in SPSS, but when I ran it in MPlus I was getting vastly different results. Ultimately, the problem seems to be that the variables were centred at the person and mean level in SPSS, whereas I used the "raw" variables in MPlus. This leads me to a couple of questions:
(1) Do I need to centre my variables in MPlus if I use the same ones at both the between and within-person levels?
(2) I have tried pre-centering all variables, calculating the interactions in SPSS, and then importing them into MPlus to see if I can recreate my results in SPSS. However, when I ran the model including the correlations between the variables only (i.e., not including the correlations between interactions), I get the following error (see below), and only the interaction terms seem to be affected (i.e., they have a bunch of stars instead of a variance). However, when I run the model excluding the correlations, it runs perfectly fine.
" One or more variables have a variance greater than the maximum allowed of 1000000. Check your data and format statement or rescale the variable(s) using the DEFINE command."