Oakes on Confidence Intervals versus Tests


The researcher armed with a confidence interval, but deprived of the false respectability of statistical significance, must work harder to convince himself and others of the importance of his findings. This can only be good.

Michael Oakes

register