Reproducibility in Preclinical in Vivo Research: Statistical Inferences.

The article addresses the crisis of irreproducible preclinical in vivo research, emphasizing the role of statistical inference issues, over-simplification of models, flawed interpretation, and the persistence of dichotomous null-hypothesis significance testing. Topics include the need to move beyond reliance on "statistical significance," embracing estimation statistics, Bayesian approaches, and highlighting the complexities of translational behavioral neuroscience for improved reproducibility.