Estimating networks on items on different response scales
I am wondering if there are any considerations on estimating mgm networks on items with different response scales. Say a group of items with likert ratings from 1-4 and another set of items with 1 to 5 ratings.
My intuition is that I can combine those to estimate a joint network and that this should be fine since the networks are based on correlations (or partial correlations). But wondering if there are any considerations I am missing. I am actually estimating this using bootnet and qgraph packages in r, but I saw that the JASP network module is based on these packages, and I can't find a direct forum for those.