Journals should pivot to only publishing experiments as registered reports. This would improve the quality of experiments and decrease publication bias.