Implementing CCA as a SEM

[Philipp Deibert]

Hello,

for a project I am trying to conduct Canonical Correlation Analysis (CCA) with incomplete data. I found this paper (https://link.springer.com/chapter/10.1007/978-3-030-01310-3_22) formulating CCA using SEM. However, I know nothing about SEM and how to specify models and am somewhat confused looking through resources for the last few days. I'd appreciate any pointers on how the proposed method could be implemented using lavaan.

Thank you for any help on this!

[Mike Cheung]

Hi Philipp,

You may try the cancorr function in the mulSEM package (https://github.com/mikewlcheung/mulsem). It handles missing data with FIML. The details are in Gu, Yung, and Cheung (2019).

Gu, F., Yung, Y.-F., & Cheung, M. W.-L. (2019). Four covariance structure models for canonical correlation analysis: A COSAN modeling approach. Multivariate Behavioral Research, 54(2), 192–223. https://doi.org/10.1080/00273171.2018.1512847

--

---------------------------------------------------------------------
 Mike W.L. Cheung               Phone: (65) 6516-3702
 Department of Psychology       Fax:   (65) 6773-1843
 National University of Singapore
 http://mikewlcheung.github.io/
---------------------------------------------------------------------

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lavaan/972581e1-6798-4a03-8e92-477e0caa8378n%40googlegroups.com.

[Philipp Deibert]

Thank you for your reply!

I will be sure to check it out. Having the code available as open source is especially helpful in understanding what is going on, so thank you!

[Philipp Deibert]

The code seems to work fine. However, I'd like to better understand what is going on. The OpenMx code confuses me. I am unsure how the COSAN model related to the more general description of SEMs in terms of measurement and structure models. Is there perhaps a path diagram available for this approach?