In a study of 1,292 chemistry journal articles retracted from 2001 through 2021, misconduct was the largest assigned reason group: 756 papers, or 58.5%. Errors and concerns not classified as misconduct accounted for another 338 papers, or 26%. The finding describes reasons recorded for retracted papers—not the prevalence of misconduct across chemistry research.
What the chemistry retraction study found
Yulia Sevryugina and Ryan Jimenez analyzed retraction records for chemistry journal articles in a study published in ACS Omega in 2023. Their dataset contained papers retracted during 2001–2021, based on Retraction Watch records available on May 24, 2022. The authors assigned each paper to a primary reason group for the overall comparison.
As an Amazon Associate I earn from qualifying purchases.
| Primary reason group | Articles | Share of 1,292 retracted papers |
|---|---|---|
| Misconduct-related | 756 | 58.5% |
| Errors and concerns unrelated to misconduct | 338 | 26% |
| Other or not assigned to either group above | 198 | 15.5%, calculated from the reported total and group counts |
The two named groups do not account for every record in the sample. The remaining 198 papers are the arithmetic difference between the total and the two reported counts; the figures do not establish a single reason for that remainder.
What counted as misconduct—and what counted as error?
Misconduct
The study grouped fraud and plagiarism as misconduct. Fraud included falsification or fabrication involving data, images, or results. Plagiarism included reuse of others’ work, self-plagiarism, and copyright claims. The detailed reason counts include 526 plagiarism cases and 272 fraud cases.
#1 Best Overall
Within the plagiarism count, the authors report 306 cases of self-plagiarism and 210 cases of plagiarism of others. Since the database could record more than one reason for a paper, the underlying reason counts may overlap; they should not be added together as though each represented a distinct paper.
Errors and concerns
The separate error-and-concern group included problems involving data, analyses, methods, results, text, or images, as well as results or data described as unreliable or irreproducible, and cases where data were not provided. A retraction in this group is not, by itself, evidence of misconduct.
Why the 58.5% figure needs careful reading
The denominator is 1,292 already-retracted chemistry articles, not all chemistry papers. The result supports the conclusion that misconduct was the most common assigned reason group in this retraction dataset. It does not mean that most chemistry research is fraudulent, or that retraction automatically establishes wrongdoing by an author.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →The authors also compare the retraction count with Web of Science totals for chemistry research and review articles published across 2001–2021, describing the share as about 0.06%. That is a comparison of retraction records with a publication denominator; it is not an estimate of misconduct prevalence. Because the analysis covers retractions through 2021, it also cannot show whether the distribution of reasons has changed since then.
How the finding compares with an earlier analysis
A separate 2019 analysis by François-Xavier Coudert covered 331 retracted papers from 2017–2018 across chemistry, materials science, and related fields. Chemistry World reported that 69% were attributed to plagiarism or data manipulation. Coudert’s primary article argued that retraction notices should explain the reasons so readers can distinguish honest error from misconduct.
That 69% should not be treated as a direct trend comparison with the 58.5% result: the studies cover different fields, years, and classification approaches. The earlier analysis nevertheless points to a practical issue—clear notices help readers understand why a paper was withdrawn.
Rank #4
How long did retraction take?
For the papers in their analyzed sample, Sevryugina and Jimenez report a median of 1.7 years from publication to retraction. They report a median peer-review time of 71 days overall and 43 days for fraudulent manuscripts. These are sample-level statistics, not predictions about how long a particular paper will take to be reviewed or retracted.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
