Is the Betas file the one for numerical values of edge weightings?

[Henry Whitfield]

Hello All,

Thanks again for any help you can give me. I've learnt a lot from you guys already. 

This image (attached) is the Betas file with the qgraph visual it corresponds to. 

I'm looking at comparing the subgroups edges in terms of correlatory direction (blue or red) and weighting (numerical value in this table?).

Is this Betas file the one to use for this? 

I'm also not sure what the difference is between the left (white) table and the right (grey) table. 

Is there a publication that describes what this Betas file contains?

Many thanks,

Henry

Henry J. Whitfield MSc
PhD Candidate
ACBS peer-reviewed ACT trainer, TIR Trainer
Mindfulness Training Ltd.

http://www.psychflex.co.uk
Tel: 020 7183 2485

Like us on facebook!

[Katie Gates]

Great questions. 

There's actually a published tutorial that should help here: Lane, S. T., & Gates, K. M. (2017). Automated selection of robust individual-level structural equation models for time series data. Structural Equation Modeling: A Multidisciplinary Journal, 24(5), 768-782. It's a little dated, yet is still helpful with this foundational information in it. 

To answer your question, the grey part would be the A matrix, or matrix of contemporaneous relations. You'll see that negative estimates here correspond to blue paths in the figure (see "Openness -> Negative.Emotions"). The contemporaneous estimates are depicted with solid lines here. 

The white matrix is the Phi matrix, or lagged relations. The diagonal has estimates of the autoregressive relations, or how a variable relates to itself at the next timepoint. The dashed lines are lagged relations.  

For all lines in these individual paths, line width corresponds to the beta estimate, with wider widths = larger absolute values. 

Hope this helps,
Katie

[Henry Whitfield]

Thanks so much Katie,

I'm comparing two psychotherapy treatments in which there is some (0.55) correlation between treatment and GIMME subgroup. 

I'm wondering if adding up the weighted values between treatments to compare treatment outcomes (relating to specific edges) would be a good idea? 

As the group and subgroup models don't include positivity/negativity of correlations, I wish to zoom in on this. 

I'm an ACT therapist, and positive correlations between 'negative emotions' and 'Openness to Experience' could be interpreted as clinically interesting - understanding if people stayed open in spite of emotional pain or not. 

If this sounds like a terrible idea please let me know.

Many thanks again,

Henry

[Katie Gates]

Good question. Yes, you can use the estimates for the path weights in secondary analysis. They are theoretically normally distributed, so no need to transform them to conduct analyses such as t-test or correlations with these estimates. 

So, it sounds like a plan of analysis might be to have to take the estimates for the path 'negative emotions' and 'Openness to experience" and see how that relates to an outcome variable (either by looking at mean differences in that estimate if the outcome variable is categorial, or a correlation). 

hope this helps, 
Katie