TripleR Question

8 views
Skip to first unread message

Seth GMU

unread,
Sep 25, 2017, 5:22:57 AM9/25/17
to PERSOC - Personality and Social Relationships
Hi everyone

I was hoping someone might be able to help me with a TripleR question.

I'm wondering if the program can produce correlations among target effects for different traits.

So, for example, I'd like the correlation between target efx for neuroticism and agreeableness.

I also need the correlations between the self-ratings and the target efx - so, for example, the correlation between self-rated neuroticism and the target efx for agreeableness.

I tried exporting target efx and computing correlations, but I used multiple indicators for the traits and was not sure how to handle this.

Thanks in advance for any help

Seth

Terrence Jorgensen

unread,
Dec 14, 2017, 10:59:09 AM12/14/17
to PERSOC - Personality and Social Relationships
I'm wondering if the program can produce correlations among target effects for different traits.  So, for example, I'd like the correlation between target efx for neuroticism and agreeableness. 

Yes, but TripleR only fits up to 2 variables.  So if you plan to do this for all Big-5 factors, you would have to run bivariate models for each pair to get results.

Nestler (in press) proposed multivariate SRM for an arbitrary number of variables using (RE)ML estimation 


He provides R syntax for fitting it here:


But I don't know whether correlations with self-ratings are already included in that procedure.

I also need the correlations between the self-ratings and the target efx - so, for example, the correlation between self-rated neuroticism and the target efx for agreeableness. I tried exporting target efx and computing correlations, but I used multiple indicators for the traits and was not sure how to handle this. 

You shouldn't need to export anything, if self-ratings are included on the diagonal of the matrix(es) submitted to RR().  Look at the ?selfCor help page, or read the instructional document linked on the ?TripleR help page.

Terrence D. Jorgensen
Postdoctoral Researcher, Methods and Statistics
Research Institute for Child Development and Education, the University of Amsterdam

Reply all
Reply to author
Forward
0 new messages