The published articles serve as a source of reference for the research community hence; data reliability of the published articles remains as one of the major concern for authors and journal editors. Citation of the articles affected by research misconduct leads to propagation of inaccurate data. Consequently, retraction (withdrawal or cancellation) of these aforesaid articles takes place to prevent propagation of fabricated data.
Over the past few years, the frequency of retractions has risen sharply as compared to the number of articles published. A study conducted by Moylan et al. to observe the number of retraction cases for Biomed Central Journals reported that 134 retractions occurred during January 2000 to December 2015 (Figure).
Reasons for Retraction
In order to analyze the major reasons for retraction of publications; a research of PubMed database was carried out in May 2012. Since the year 1977, 2,047 PubMed indexed articles were retracted. The study revealed that 67.4% of retractions were attributed to some sort of misconduct (fraud, duplication and plagiarism); followed by only 21.3% of retractions due to error (Figure).
Rationale of Retraction
According to the guidelines of the Committee on Publication Ethics (COPE), the editors of the journals can retract the publications in any of the following scenarios:
- Data unreliability: as a result of misconduct (data fabrication/ falsification) or honest error (miscalculation or experimental error)
- Redundant Publication: findings are already published elsewhere without proper clarification or cross reference
- Plagiarism: self-, direct-, or mosaic-plagiarism
- Unethical research: not approved by an ethical committee
- Disclosure of competing interest: if authors fail to disclose a major competing interest that can influence the article interpretation
The major problem associated with retraction of articles is that, the validity of the paper has now been brought into question. This is particularly problematic for the authors who have cited this paper and part of their paper’s argument or research plan is based on the retracted study. This leads to an extensive loss of time and effort. Additionally, the retracted articles in the journals database record affect both the researcher as well as the journal’s reputation..
Prevention of retraction
Statistical techniques (analytical techniques and graphics) and diverse tools (such as Anti-Fraud Data Analytics) besides text-matching software have increased the detection rates of fraud and plagiarism thus; leads to decreased retractions due to research misconduct. An online database called “Retraction Watch” was launched in August 2010 (http://retractionwatch.com/). In addition of maintaining the retraction records of the scientific papers; the database also increases the transparency of the retraction process.
The retraction surge can be further minimized by increasing the usage of checklists by authors and reviewers, an enhanced focus on ethics, following a stringent peer review process, open access for the published articles by the journals, and establishment of uniform guidelines for retractions and retraction notices.
In the absence of reliable scientific literature; the retraction rates can further escalate in coming years. Thus, commencement of higher number of unbiased and centralized database that can be used to track the retracted publications would be a considerable undertaking.