More here.
9/26/20 edit:
I made a small error in creating the last figure in this blog post for correlations between party ID and ideology. Specifically, I mistakenly included “Not sure” responses for the ideology survey items (operationalized as continuous variables), which resulted in attenuated correlations. Below is the corrected figure:
While playing around with this data, I calculated another interesting figure– a correlation matrix for individual level ideology at different years (which shows high stability for ideology, but lower than that for partisanship):
Agadjanian Politics 