Comparing Global Network Strength across two time-series periods

[aitkenm]

Hello,

I have a question about using the Network Comparison Test package with longitudinal data.

I have estimated separate dynamic networks in two time periods using panelgvar in the psychonetrics package. I would now like to compare global network strength for the temporal networks in period 1 vs. period 2. Is this possible using the NCT package?

Thank you in advance!

[KarolineHuth]

Hi @aitkenm,

unfortunately, the NCT cannot compare two dynamic networks yet. We might include that in future versions.

[SachaEpskamp]

For psychonetrics models you can do a multi-group comparison in the same way you do homogeneity/invariance testing in SEM. See here for some more details (this is about cross-sectional data but the logic applies to panel models too):

https://static-content.springer.com/esm/art%3A10.1007%2Fs11336-021-09764-3/MediaObjects/11336_2021_9764_MOESM1_ESM.pdf

For mlVAR models you can look at this preprint:

https://psyarxiv.com/dhp8s

[aitkenm]

Thank you, @KarolineHuth for confirming.

@SachaEpskamp Thank you for pointing me to these resources. The preprint on mlVAR was especially helpful. Is there a way to also compare global network strength (across groups/time periods)?