LAUSR

I enjoyed Professor Mayo’s (2022) comment in Conservation Biology very much, and agree enthusiastically with most of it. I agree with her that error probabilities (or error rates) are essential to consider: if one does not give thought to what the data would be like under the assumption that one’s theory is false, one is likely reinforcing confirmation bias rather than establishing the validity of the theory. As Mayo argues, “stress-testing” models and results is crucial. I also agree with Mayo that banning “bright lines” is not helpful. Some applications really require a decision to be made or a conclusion to be drawn. Should a species get specific legal protection, or not? Should a pipeline be built through a particular area, or not? Should drilling be permitted in a particular area, or not? Should there be economic incentives for an agricultural management intervention that is purported to sequester carbon, or not? Should a jurisdiction prohibit harvesting wild foods from urban ecosystems for safety concerns, or not? Although such decisions are not made solely on the basis of p-values, banning bright lines may preclude making decisions in a principled, reproducible way that controls error probabilities. Conversely, no threshold for significance, no matter how small, suffices to prove an empirical claim. As Fisher (1935) wrote: