Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.

# st: CFA and clustered estimator

 From Nicolas.Sommet@unil.ch To Subject st: CFA and clustered estimator Date Fri, 05 Apr 2013 19:57:35 +0200

```Dear Stata users,

I'm running a confirmatory factor analysis. I got a six-item
scale and I postulate a two-factor structure.
As my database consists of individuals (N = 81) nested in
groups (n = 46), I use "clustered sandwich estimator".

Here's the command I use*:
confa (x1: x1.1 x1.2 x1.3) (x2: x2.1 x2.2 x2.2), from(ones)
vce(cluster groupe)

My problem is that Stata does not generate value for
"Goodness of fit test" or "Test vs independance" (The Coefs,
SEs, Zs, ps and CIs seem normal though):
Goodness of fit test: LR = . ; Prob[chi2( .) > LR] = .
Test vs independence: LR = . ; Prob[chi2( .) > LR] = .

...same for the fit indices:
RMSEA  =      ., 90% CI= (     .,      .)
RMSR   =      .
TLI    =      .
CFI    =      .
AIC    =        .
BIC    =        .

Any idea of what's happening? And how to adress this issue?
FYI, cfa works good when clustered estimator is not
specified (i.e., confa (x1: x1.1 x1.2 x1.3) (x2: x2.1 x2.2
x2.2), from(ones))

Thanks for you consideration.
Nicolas Sommet

*I've learn how to do cfa reading: "Kolenikov, S. (2009).
Confirmatory factor analysis using confa. Stata Journal,
9(3), 329".
N.B.: I've tried to keep only groups comprising at least two
individuals (N = 67, n = 31), or to do boostraping
(bollenstine, reps(200) confaoptions( iter(20) )). I
obtained the same (non-)results.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/
```