RI-CLPM convergence issues

[C]

Hi all,

I am attempting to run a RI-CLPM on five waves of data with two constructs. I have been benefiting greatly from Hamaker et al (2015), Mulder & Hamaker (2021) (especially the lavaan scripts), Mund et al (2021).

Although I can get a CLPM to run without issues, I'm facing convergence issues with a basic RI-CLPM (no constraints over time). I get the sense that there is something up with the within-centred variables in the first wave of data for one of my constructs, because the (incorrect) regression estimates that *do* appear show seriously bizarre looking numbers for those, and when I do the analysis with only Waves 2-5, I have no convergence issues. I also find the model runs when I set the variance of that random intercept to 0, and covariance between the two RIs to 0, but not sure if this is an appropriate solution.

Does anybody have some recommendations on how to address this, or further reading I could look to?

Many thanks.