Efficacy of statistical and design interventions for improving reproducibility in biomedical research
A registered-report simulation study found that statistical and design interventions for improving biomedical research reproducibility had mixed and measure-dependent effects.

