[R] SEM: multigroup model

3 views
Skip to first unread message

Amarnath Bose

unread,
May 23, 2013, 6:53:38 AM5/23/13
to r-h...@r-project.org
Dear R Gurus,

I am trying to run a multigroup SEM using Prof. John Fox's SEM package.

The two groups are "Ready to Eat" denoted by RTE and
"Ready to Cook" denoted by RTC.

I ran a omnibus CFA on the data of consumer perceptions & preferences and
am satisfied with what I got.

When I tried to do a multigroup SEM - my understanding is limited to the
SEM manual in CRAN - using the code below, I get the following message:
Error in summary.msemObjectiveML(sem.MG) :
no 'dimnames' attribute for array
Execution halted

The relevant part of my code follows:
mod.mg <- multigroupModel(sbmod.cfa,groups=c("RTC","RTE"))

sem.MG <- sem(mod.mg,data=srt,group="RTind",
formula = ~ inv + imp + tch + emo + loy + usg + sig + dif + ndif
+
vda + vdb + vdc + vdd + vde + vdf + vdg + vdh +
riskT + riskP + riskS + riskFi + riskFu + riskPs

)
summary(sem.MG)

I was expecting two sets of fit indices for RTE & RTC and want to do an
ANOVA
across the models; as well as possibly check for loading equivalence.

Can somebody please throw some light on where I am making a mistake?

Thanks
Amarnath Bose
--


*Amarnath Bose*
* **Associate Professor *
*Decision Sciences Department*
*Birla Institute of Management Technology
*
Tel: +91 120 2323001 - 10 Ext.: 398
Cell: +91 9873179813

[[alternative HTML version deleted]]

______________________________________________
R-h...@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

John Fox

unread,
May 23, 2013, 7:53:37 AM5/23/13
to Amarnath Bose, r-h...@r-project.org
Dear Amarnath Bose,

There's nothing obviously wrong with the commands that you report -- in
fact, your commands have the same structure as the multigroup SEM example in
?sem -- so the usual advice about including reproducible code producing the
error applies. If you like, you could send me your data and the complete R
script that you used.

Best,
John

-----------------------------------------------
John Fox
Senator McMaster Professor of Social Statistics
Department of Sociology
McMaster University
Hamilton, Ontario, Canada
Reply all
Reply to author
Forward
0 new messages