Nonreplicable publications are cited more than replicable ones

Area of Science:

• Psychology
• Economics
• Bibliometrics
• Scientific Publishing

Background:

• Replication is crucial for scientific validity.
• Studies suggest experts can predict replication success.
• Non-replicable findings may still receive significant scholarly attention.

Purpose of the Study:

• To investigate citation patterns of replicable versus non-replicable studies.
• To assess the impact of replication failure on study citations.
• To explore reasons behind the publication of non-replicable research.

Main Methods:

• Analysis of publicly available citation data.
• Examination of published papers in top psychology, economics, and general interest journals.
• Tracking citations before and after replication failure publication.

Main Results:

• Papers failing replication receive more citations than those that replicate.
• Citation advantage for non-replicable papers persists post-replication failure.
• Only 12% of citations acknowledge replication failures.

Conclusions:

• A citation bias favors non-replicable findings in top journals.
• Publication standards might be lowered for 'interesting' results, potentially compromising reproducibility.
• The scientific community's attention to non-replicable research highlights a potential conflict between impact and rigor.