publication bias
-
The Reproducibility Crisis Seen From the Bench
Most replication failures are not caused by fraud but by ordinary bench realities: drifting protocols, variable reagents, small samples and selective reporting.
Most replication failures are not caused by fraud but by ordinary bench realities: drifting protocols, variable reagents, small samples and selective reporting.