A new survey shows that scientific papers where the results can't be replicated are far more likely to be cited in later studies.
https://advances.sciencemag.org/content/7/21/eabd1705
https://advances.sciencemag.org/content/7/21/eabd1705