Paradoxes in Scientific Inference