The 2015 Reproducibility Project measured a 36 percent replication rate for one sample of 97 findings, but researchers still debate how representative that sample is, how to interpret a "failed" replication, and whether reforms such as pre-registration are sufficient to change publication incentives.
What would resolve this Sustained tracking of replication rates across psychology's subfields after widespread adoption of pre-registration and open data, compared against pre-reform baselines.
OpenPsychology (Methodology / Meta-Science)Wikipedia: Replication crisis