2 ms·

What's the "right balance" between Type I and Type II errors when assessing the quality/potential of new research?