Mean-centering interaction in cross-lagged panel models

[HG Wells]

Hello everyone, I would be grateful and appreciate any suggestion about mean-centering a time-invariant moderator * time-varying predictor at wave 1 interaction in a cross-lagged panel model (3 waves).

I have mean-centered the time-invariant moderator in my data, before running the model, then specified the interaction by using lavaan's syntax (moderator:predictor).

Should I mean-center the time-varying predictor in my data, as well, or can I just enter the predictor un-centered, in lavaan?

Thank you

Emanuele

[Terrence Jorgensen]

Should I mean-center the time-varying predictor in my data, as well,

If you center it at any value , you should probably subtract that same value from all occasions, e.g., its grand mean across all occasions, or the lowest observed value across all occasions.

or can I just enter the predictor un-centered, in lavaan?

I don't see a problem with that, unless you are particularly interested in the estimated simple effect of the time-invariant moderator.  Because you refer to it as a moderator, I'm guessing it is not a focal predictor.

Terrence D. Jorgensen

Assistant Professor, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam

http://www.uva.nl/profile/t.d.jorgensen

[Emanuele Fino]

Thank you so much Terrence, this is very helpful and much appreciated.

--
You received this message because you are subscribed to a topic in the Google Groups "lavaan" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/lavaan/9DsYFNb-I1Q/unsubscribe.
To unsubscribe from this group and all its topics, send an email to lavaan+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/007fc21b-5dd7-4f1e-9c83-1b83b1646ad9n%40googlegroups.com.