One could suppose that for some variables in social sciences and psychology could be an optimal level as well as in ecology and biology, so we could find a nonlinear correlation between variables.
Looks like exists a bunch of such methods and even linear correlation could someway catch nonlinear dependencies https://goldinlocks.github.io/Comparing-Linear-and-Nonlinear-Correlations/
Probably better nonlinear correlation coefficients are Distance correlation, Cosine similarity and Maximal information Coefficient.
Does exist some software for computing these coefficients and their p-levels by easy-way, without some programming?