Statistical significance sets a convenient obstacle to unfounded claims. In my view, removing the obstacle (V. Amrhein et al. Nature 567, 305–307; 2019) could promote bias. Irrefutable nonsense would ...