Virginia Institute of Marine Science
Environmental Toxicology and Chemistry
Both the no-observed-effect concentration and its null hypothesis significance testing foundation have drawn steady criticism since their inceptions [1–5]. Many in our field reasonably advocate regression to avoid conventional null hypothesis significance testing shortcomings; however, regression is compromised under commonly encountered conditions (Green, present Perspective’s Challenge). As the debate to favor null hypothesis significance testing or regression methods continues into the 21st century, a sensible strategy might be to take a moment to ask, Are there now other choices? Our goal is to sketch out 1 such choice.
Newman, Michael C. and Krull, Marcos, Regression or significance tests: What other choice is there?—An academic perspective (2016). Environmental Toxicology and Chemistry, 34(11), 2439-2440.