Path analysis query PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 Michael Rolt posted on Wednesday, April 25, 2012 - 3:53 pm
Hi there,

This is my first time I'm using mplus so please excuse my basic level of understanding!

I am conducting a path analysis on a dataset containing both male and female participants. I would like to conduct a separate analysis on females alone, and am trying to use the Useobservations command to do so.

My input is...

Variable:
Names = gender thinid PASTAS Drive Diet Food Others fitmag beautmag thinshs muscular;

Usevariables = gender thinid PASTAS Diet Food Others beautmag;

Grouping is gender (1 = Female 2 = Male);
Useobservations = gender EQ 1;

Model:

EatDis by Diet Food Others;
EatDis on beautmag PASTAS thinid;
PASTAS on thinid beautmag;
thinid on beautmag;


Model Indirect:
Eatdis ind thinid;
Eatdis ind beautmag;
PASTAS ind beautmag;

Output:
Cinterval sampstat standardized(stdyx);

However, I am getting the following error message and no other results.

*** ERROR
Group MALE has 0 observations.

Any help would be gratefully received!

Regards,

Michael
 Linda K. Muthen posted on Wednesday, April 25, 2012 - 5:09 pm
You can't use the GROUPING option when you are not using both groups. Remove the GROUPING option.
 Michael Rolt posted on Wednesday, April 25, 2012 - 7:05 pm
Thanks Linda,

I removed the GROUPING option, reran the report with the following input...

Variable:
Names = gender thinid PASTAS Drive Diet Food
Others fitmag beautmag thinshs muscular;
Usevariables = gender thinid PASTAS Diet Food Others beautmag;
Useobservations = gender EQ 1;

Model:

EatDis by Diet Food Others;
EatDis on beautmag PASTAS thinid;
PASTAS on thinid beautmag;
thinid on beautmag;


Model Indirect:
Eatdis ind thinid;
Eatdis ind beautmag;
PASTAS ind beautmag;

Output:
Cinterval sampstat standardized (stdyx);

...but now have the following warnings.

*** WARNING in MODEL command
Variable is uncorrelated with all other variables: GENDER
*** WARNING in MODEL command
At least one variable is uncorrelated with all other variables in the model.
Check that this is what is intended.

Do I need to define my Gender as an IV in the Variable command section?

Michael
 Linda K. Muthen posted on Wednesday, April 25, 2012 - 7:29 pm
Remove gender from the USEVARIABLES list. It is not an analysis variable.
 Michael Rolt posted on Thursday, April 26, 2012 - 4:20 am
Works perfectly! Thanks for your patience Linda, it's all very new to me right now!

Michael
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: