I’m thinking basically a combination of a regression analysis and ANOVA/F-test to first find out whether there is a correlation and later zooming in on what the correlation exactly is (using something akin to a post-hoc battery of t-tests).