Explore the concept of how reliability scores scale type I errors. E.g. if discriminability is 1, type I error is as parameterized in a statistical test. If discriminability is, e.g., 0.7, how does that inflate the type I error of the statistical test? It could be a great and impactful simulation study
Explore the concept of how reliability scores scale type I errors. E.g. if discriminability is 1, type I error is as parameterized in a statistical test. If discriminability is, e.g., 0.7, how does that inflate the type I error of the statistical test? It could be a great and impactful simulation study