3 error messages in lavaan with transformed data

[hasna akub]

I built the model with one latent variable (WDI that was calculated by sum dam+ditch + aqued then standardized the result). for other variables, I used transformed data (look the attachment) because the original data is not normally distributed.

library(lavaan)

model <- '

# measurement model

WDI =~ ln.dams + ln.ditch +ln.aqued (should I use original or transformed variables to build latent variable???)

# regressions

ln.intro ~ WDI+asn.dev+ln.area+sqr.diversity

sqr.impact ~ ln.intro+WDI+asn.dev+sqr.diversity

sqr.diversity ~ ln.area

# residual correlations

WDI ~~ asn.dev 

WDI ~~ ln.area 

asn.dev ~~ ln.area'

fit.light <- sem(model, data = light)

summary(fit.light, standardized = TRUE)

 

however, when I built the model  I have 3 error messages which are

Warning messages:

1: In vnames(FLAT, "ov.x", warn = TRUE) :

  lavaan WARNING: model syntax contains variance/covariance/intercept formulas

  involving (an) exogenous variable(s): [asn.dev ln.area];

  Please use fixed.x=FALSE or leave them alone

2: In lavaan::lavaan(model = model, data = light, model.type = "sem",  :

  lavaan WARNING: some estimated variances are negative

3: In lavaan::lavaan(model = model, data = light, model.type = "sem",  :

  lavaan WARNING: covariance matrix of latent variables is not positive definite; use inspect(fit,"cov.lv") to investigate. 

so, I am wondering what mistake I had to have those error.

Should I use original data instead? 

 

[Jeremy Miles]

Transformations have nothing to do with these warnings. 

1. You have asn.dev ~~ ln.area, this is the cause of the first warning. As it says, leave it alone.

2. You have negative estimated variances, this suggests that your model is misspecified (although if they're only a teeny bit negative, don't worry.

3. The covariance matrix of your latent variables cannot occur with real data. This also suggests that your model is misspecified.

Hope that helps,

Jeremy

--
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 http://groups.google.com/group/lavaan.
For more options, visit https://groups.google.com/d/optout.

[hasna akub]

Thanks for help. 

I have a question; what do you mean by saying leave it alone. what should I do?

[yrosseel]

On 04/27/2014 04:53 PM, hasna akub wrote:
> *Thanks for help. *
> *I have a question; what do you mean by saying leave it alone. what
> should I do?*

Remove the line 'asn.dev ~~ ln.area' from your model syntax.

Yves.