In the WHI’s press release, Anderson justified her statistical decisions this way: “Because breast cancer is so serious an event, we set the bar lower to monitor for it. We pre-specified that the change in cancer rates did not have to be that large to warrant stopping the trial. And the trial was stopped at the first clear indication of increased risk.” In other words: We set the bar low enough to monitor for nonsignificant results if we could squeak out any.
