measEq.syntax() Error in TEST[[2]] : subscript out of bounds

[João Marôco]

Dear Terence,

First, thank you so much for the new measEq.syntax() function in semTools. This is a really life saver!

I am using the example code in https://www.rdocumentation.org/packages/semTools/versions/0.5-1/topics/measEq.syntax to test for invariance in my second order model, but I am puzzled by the fact that there's a out-of-bound subscript error. I wonder if for ordinal data (my itens are 5 point likert) some other parametrization is required?

This is my configural model:

modelo = '

Cp =~  SE1 + SE2 + SE3 + SE4 + SE5

Em =~ SE6 + SE7 + SE8 + SE9 + SE10

Co =~ SE11 + SE12 + SE13 + SE14 + SE15

# fator de 2a ordem

eng =~ Cp + Em + Co

'

and this is the code from the rdocumentation page:

## compare several invariance models

grupo="Country"

test.seq <- c("thresholds","loadings","regressions","intercepts","means","residuals")

meq.list <- list()

for (i in 0L:length(test.seq)) {

  if (i == 0L) {

    meq.label <- "configural"

    group.equal <- ""

    long.equal <- ""

  } else {

    meq.label <- test.seq[i]

    group.equal <- test.seq[1:i]

    long.equal <- test.seq[1:i]

  }

  meq.list[[meq.label]] <- measEq.syntax(configural.model = modelo,

                                         data = bd,

                                         ordered = c("SE1","SE2","SE3","SE4","SE5","SE6","SE7","SE8","SE9","SE10",

                                                     "SE11","SE12","SE13","SE14","SE15"),

                                         parameterization = "theta",

                                         ID.fac = "ul",

                                         ID.cat = "Wu.Estabrook.2016",

                                         group = grupo,

                                         group.equal = group.equal,

                                         long.equal = long.equal,

                                         return.fit = TRUE)

}

compareFit(meq.list)

I get this error:

> compareFit(meq.list)
Error in TEST[[2]] : subscript out of bounds

Any idea of what I should do differently?

Thanks in advance!

João

[Terrence Jorgensen]

I get this error:

> compareFit(meq.list)
Error in TEST[[2]] : subscript out of bounds

Any idea of what I should do differently?

Are you using the latest version of the software?

devtools::install_github("simsem/semTools/semTools")

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

[jpma...@gmail.com]

I believe I am.
There are warnings about non positive definite variances... I guess that's why there's no results and an out of bound error?

João Marôco
[Sent from mulher not that smart smartphone, with more errors than usual]

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at https://groups.google.com/group/lavaan.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/eff72b34-ca5d-4c45-b93b-0fa042e06258%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[jpma...@gmail.com]

Dear Terrence,

I installed both semTools and lavaan lastest versions from git_hub.

Still, have an error:

 

Warning messages:

1: In lav_model_vcov(lavmodel = lavmodel2, lavsamplestats = lavsamplestats,  :

  lavaan WARNING:

    Could not compute standard errors! The information matrix could

    not be inverted. This may be a symptom that the model is not

    identified.

2: In lav_test_satorra_bentler(lavobject = NULL, lavsamplestats = lavsamplestats,  :

  lavaan WARNING: could not invert information matrix

 

The models’ seems to be unidentified? I wonder if for ordinal items some other identification procedure (thresholds, or similar) needs to be done or if it’s just a problem with sample size (n=390, second order factor with 15 ordinal items, 8 groups) and polychoric correlation estimation…

I attach the rds database, if you care to try…

Thanks in advance!

João

 

From: lav...@googlegroups.com <lav...@googlegroups.com> On Behalf Of Terrence Jorgensen

Sent: 27 de junho de 2019 22:39
To: lavaan <lav...@googlegroups.com>

--

You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to

lavaan+un...@googlegroups.com.

To post to this group, send email to

lav...@googlegroups.com.
Visit this group at https://groups.google.com/group/lavaan.

To view this discussion on the web visit

https://groups.google.com/d/msgid/lavaan/eff72b34-ca5d-4c45-b93b-0fa042e06258%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Terrence Jorgensen]

or if it’s just a problem with sample size (n=390, second order factor with 15 ordinal items, 8 groups) and polychoric correlation estimation…

You have divided n=390 into 8 groups?!  That is a small sample even for a single-group model with DWLS estimation.  And dividing into 8 groups (at best, your smallest n = 390/8 < 50) means the samples would be too small even with multivariate normal indicators.

Sorry, but sometimes you just need to accept that you do not have the data needed to test a hypothesis with any kind of justified confidence.

[jpma...@gmail.com]

No, no... 390 per group... But, yes... For measurement invariance with WLSMV does not suffice. It works fine with ML tough...
Thanks for your insights,

João Marôco
[Sent from mulher not that smart smartphone, with more errors than usual]

-------- Original Message --------
From: Terrence Jorgensen <tjorge...@gmail.com>

Sent: Saturday, June 29, 2019 07:06 PM
To: lavaan <lav...@googlegroups.com>
Subject: Re: measEq.syntax() Error in TEST[[2]] : subscript out of bounds

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at https://groups.google.com/group/lavaan.

To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/59b1708d-3a2b-458f-b855-9b8a4c811b87%40googlegroups.com.