Generalization of scientific studies (especially in the "softer" sciences) is a hard problem. On good approach is combining multiple studies that are deliberate

Generalization of scientific studies (especially in the "softer" sciences) is a hard problem. On good approach is combining multiple studies that are deliberately designed to asses different aspects of generalization. Psychology in particular has a problem with many papers poorly assessing their own generalizibility and making much stronger/broader claims than are justified by the actual design of their study. This, combined with the widespread teaching of badly run "famous" studies such as Calhoun, drove me to abandon getting a degree in that field.