Scott E. Maxwell, Michael Y. Lau, and George S. Howard argued that a statistically significant original psychology study followed by a nonsignificant single replication does not necessarily constitute a genuine failure to replicate, because the replication may have had low statistical power.