The article presents a new method for controlling errors in scientific studies with multiple hypotheses. Instead of the traditional way of setting error rates first, the researchers suggest fixing the rejection region and then estimating the error rate. This approach is shown to be more accurate and powerful, especially for the positive false discovery rate (pFDR). By using this new method, researchers can increase their study's power by over eight times compared to the standard method. The researchers also introduce the concept of the q-value, which eliminates the need to pre-set error rates.