Multigroup CFA 0 observations PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Heejung Chung posted on Wednesday, May 06, 2015 - 8:35 am
Hello. I am relatively new to Mplus so please bear with me . :-)

I am trying to run a MG CFA using survey data for 10 variables, 9 groups = groups here being countries. the number assigned to the groups being the labels used for countries in variable v3.

Variable:
Names are
v4v2 v5_1 v6_1 v7_1 v8_1 v11_1 v9v2 v10v2 v12v2 v13v2 country8 weight v3;
Missing are all (-9999) ;
Usevariables are
v4v2 v5_1 v6_1 v7_1 v8_1 v11_1 v9v2 v10v2 v12v2 v13v2 v3 ;
useobservations are (country8 EQ 1);
Grouping is v3 (2=WDE, 3=EDE, 4=GB, 13=S, 14=CZ, 16=PL, 25=E, 28=F, 32=DK);

However, I keep getting this error message .

*** ERROR
Group S has 0 observations.
*** ERROR
Group CZ has 0 observations.
*** ERROR
Group PL has 0 observations.
*** ERROR
Group E has 0 observations.
*** ERROR
Group F has 0 observations.
*** ERROR
Group DK has 0 observations.
*** WARNING
Data set contains unknown or missing values for GROUPING,
PATTERN, COHORT, CLUSTER and/or STRATIFICATION variables.
Number of cases with unknown or missing values: 4885

I've checked several times that the values are correct.
Many thanks in advance.
 Linda K. Muthen posted on Wednesday, May 06, 2015 - 10:51 am
Please send the output, data, and your license number to support@statmodel.com.
 Heejung Chung posted on Wednesday, May 06, 2015 - 12:34 pm
(sent via email)
I've also tried it by dropping all cases that are country8==0 and recoding the v3 variable into another country id variable in STATA, transferred to Mplus and tried it again without USEOB code.
In this case, it seem to have worked although modification indices were not given... another question.

However, there seems to be some changes in the data. This might be because as you can tell by the data file, the file was transferred from STATA via stata2mplus function and I am afraid something might have gone wrong in the process.

I have checked means etc after the transfer but that was fine.
I am also looking into more detail regarding the transfer.
 Heejung Chung posted on Wednesday, May 06, 2015 - 2:54 pm
Hi there.
Having checked again, it seems like there was a problem with the data being transferred where country8 EQ 1 has not selected all cases where country8 was coded as 1.

I am not sure exactly where it went wrong but now having dropped all cases in the STATA file before transferring and using the new data without the USEOBservation, it seem to work better.
 Bengt O. Muthen posted on Thursday, May 07, 2015 - 1:53 pm
Good.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: