Given a particular set of data, there are an infinite number of theories and models that can be proposed to explain that data. After all, nobody would propose a theory or model that doesn’t at least explain or fit the existing data. With that in mind, how are we to choose between competing theories and models if all of them can explain existing data? The key lies in predictive power. The best theories and models not only make strong, testable predictions but also get those predictions correct...
Published on March 22, 2019 00:25