Scientific realism faces the problem of the underdetermination of theory by evidence generated by the empirical equivalence of rival theories. It could overcome it if predictivism were true– weak predictivism, for strong predictivism stumbles over an apparently insurmountable difficulty. In that case, it would contribute also to the refutation of skepticism about the external world, but not, as it has been held, because experience supports more strongly the “naive” realism of common sense than such skepticism, but because scientific realism entails metaphysical realism.
Comesaña, M. (2015). Predictive success and realism (scientific and metaphysical). Revista De Filosofía, 71, pág. 23–30. Retrieved from https://revistafilosofia.uchile.cl/index.php/RDF/article/view/37931