CFA Using the Satorra-Bentler Scaled Chi-Square

[N.Plu.]

I am running a CFA using the Satorra-Bentler scaled chi-square adjustment due to non-normally distributed data. I have several questions about how the Satorra-Bentler adjustment functions. 1) When I run my model and request fit statistics with the S-B adjustment, it adds three residual variances to each manifest variable (as seen when I run semPaths). It groups the residual variance on each variable and shows a triangle error term type thing. These extra variances do not appear to correspond to anything in my data; however, it is definitely related to the S-B scale, because when I drop that specification, the variances are no longer there. Can anyone tell me what is going on with that? 2) When I run the model on two different groups, I only get one fitmeasures output. I have no idea how it is collapsing the fitmeasures into one, as it gives different chi-squares for the groups. Which chi-square is it using? If anyone can help with answers to these questions, I would greatly appreciate it. Thank you!

[Jeremy Miles]

Can you show us your code? The SB chi-square correction should not change any parameter estimates, just fit measures (based on chi-square, so not RMSEA) and standard errors.

Jeremy

On Mon, 10 Jul 2017 at 20:26 N.Plu. <npl...@gmail.com> wrote:

I am running a CFA using the Satorra-Bentler scaled chi-square adjustment due to non-normally distributed data. I have several questions about how the Satorra-Bentler adjustment functions. 1) When I run my model and request fit statistics with the S-B adjustment, it adds three residual variances to each manifest variable (as seen when I run semPaths). It groups the residual variance on each variable and shows a triangle error term type thing. These extra variances do not appear to correspond to anything in my data; however, it is definitely related to the S-B scale, because when I drop that specification, the variances are no longer there. Can anyone tell me what is going on with that?  2) When I run the model on two different groups, I only get one fitmeasures output. I have no idea how it is collapsing the fitmeasures into one, as it gives different chi-squares for the groups. Which chi-square is it using? If anyone can help with answers to these questions, I would greatly appreciate it. Thank you!

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

--

--

My employer has nothing to do with this email. 

[N.Plu.]

Here is my code. The different models are from different time points:

###Model 1###

model1 = '

NA=~ FNegAfw1+INegAfw1+FPosAfw1+BeInInw1+InPosAfw1+BeInDuw1
EC=~ Soothw1+AttFocw1+FrusTolw1

'

###Model 2###

model2 = '

NA=~ FNegAfw2+INegAfw2+FPosAfw2+BeInInw2+InPosAfw2+BeInDuw2
EC=~ Soothw2+AttFocw2+FrusTolw2

'

###Model 3###

model3 = '

NA=~ FNegAfw3+INegAfw3+FPosAfw3+BeInInw3+InPosAfw3+BeInDuw3
EC=~ Soothw3+AttFocw3+FrusTolw3

'

###########################################RESULTS#################################

fit1 = cfa(model1, data = imputeddata, orthogonal = FALSE, group = "gender", test = "Satorra-Bentler")
summary(fit1)
fitmeasures(fit1)

fit2 = cfa(model2, data = imputeddata, orthogonal = FALSE, group = "gender", test = "Satorra-Bentler")
summary(fit2)
fitmeasures(fit2)

fit3 = cfa(model3, data = imputeddata, orthogonal = FALSE, group = "gender", test = "Satorra-Bentler")
summary(fit3)
fitmeasures(fit3)

[Jeremy Miles]

Hi N.Plu

"These extra variances do not appear to correspond to anything in my data; however, it is definitely related to the S-B scale, because when I drop that specification, the variances are no longer there." 

If that's true, I think it's a bug.

Can you show us the code (and perhaps a bit of the output) where this happens?

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