We end this very short paper on this point:
> We should not [...] prioritize numerical correlation over explanation [because it risks] render[s] our fields closer to pseudoscience. Computational cognitive science in principle
can and should do more than merely model data.
doi.org/10.5281/zeno...
