[R] longitudinal analysis with latent construct

Carlos Santos Jr. carlosdennerjunior at yahoo.com.br
Wed Jun 17 06:30:54 CEST 2009


Folks,

I need to test a model that has one predictor (a construct with three indicators) influencing four other variables. Something like what I try to show below.

     X         --->   Y1, Y2, Y3, Y4
  /  |  \
 i1  i2  i3

Also, each variable was measured at 5 points in time. So, I'd like to model their change, in a longitudinal fashion (mixed model?).

I know it would be too much to find a script that does it all, but I thought that maybe you guys have a reference for me to read and learn the steps I'll have to follow to perform this analysis. Any guidance would be very welcome as I'm trying to migrate from EQS to R and thus have no experience with R yet.

Thanks in advance,
Carlos Santos Jr.


      Veja quais são os assuntos do momento no Yahoo! +Buscados
http://br.maisbuscados.yahoo.com




More information about the R-help mailing list