Registered reports for cancer biology: the plan is peer-reviewed before the result

Registered reports, adopted by over 300 journals mostly in psychology and neuroscience, reduce positive-result bias dramatically (around 44% positive results versus over 90% in conventional articles in one analysis). Cancer biology journals have barely adopted the format. A concerted push by the major cancer journals to offer registered reports for confirmatory preclinical studies (the last experiment before an investigational new drug application, for instance) would change the incentive at the point where it matters most for translation.