Science
Authors request retraction of influential 2002 procrastination study amid allegations of data fabrication
An analysis of the original data files from the widely cited paper on deadlines and self-control revealed multiple inconsistencies, including duplicate observations and implausible correlations.
The short version
- Authors Dan Ariely and Klaus Wertenbroch have requested that the journal Psychological Science retract their highly cited 2002 study, "Procrastination, Deadlines, and Performance: Self-Control by Precommitment."
- An analysis published on Data Colada identified four major data red flags in the original research files after a separate team failed to replicate the study's findings.
- The 2002 paper, which has over 2,100 citations, argued that evenly spaced external deadlines significantly improve performance compared to self-imposed or single end-of-term deadlines.
Key facts
- A forthcoming replication paper in Psychological Science by Kyle Hyndman and Alberto Bisin reported a failure to replicate Study 2 of the 2002 paper.[Hacker News]
- Following the replication failure, an analysis published on Data Colada reviewed the original 2006 data files and concluded the data had been tampered with.[Hacker News]
- The Data Colada analysis found 18 out of 20 participants in one experimental condition had exact 'twins' with duplicate performance metrics spaced precisely 10 rows apart.[Hacker News]
- Variables that should have been highly correlated—such as subjective ratings of task interest and liking, or performance across similar tasks—showed virtually no correlation in the original data, despite correlating strongly in the replication data.[Hacker News]
- Self-reported task duration estimates in the original study lacked typical human rounding, with only 11.7% ending in a round number compared to 85% in the replication study.[Hacker News]
- After the analysis was shared with them, Ariely and Wertenbroch contacted Psychological Science to request a retraction, a process that is currently ongoing.[Hacker News]
What remains uncertain
- Ariely previously denied a request to include an analysis of the 2006 data files in the replication paper, questioning whether the files provided to the researchers in 2006 represented the actual data.[Hacker News]