Dissertation
Citation behavior patterns and sentiments towards retracted articles: a comparison of pre- and post-retraction citations
Doctor of Philosophy (Ph.D.), Drexel University
Aug 2021
DOI:
https://doi.org/10.17918/00000893
Abstract
Research articles are retracted for a number of reasons; yet often persist in the scientific community as evidenced by their continued reference in research. The persistence of retracted papers cited in current research undermines the foundation of confidence in the scientific community (Schneider, 2017). Retraction Watch estimates that 500 - 600 papers are retracted per year (Retraction Watch, 2016). Evidence of continued citation post-retraction can be counted in numerous publication databases and references; but does not convey the reasoning why researchers continue to reference these articles after retraction. We can reveal citation behavior patterns of authors or even groups of authors within a specific domain, say, by performing a text analysis of the context-providing sentences, or "citances" (Nakov, Schwartz, & Hearst, 2004). This is a relatively emergent analysis technique (Small, 2011) that is only beginning to develop classification schemes for these behaviors and theorizing reasons for these behaviors. The citation contexts to two papers retracted for different reasons (one for "unintentional error" the other for "misconduct and fraud") were examined and citation behaviors and sentiment were compared pre-retraction and post-retraction. A closer examination of post-retraction citations revealed that for one retracted paper ("misconduct and fraud"), the hypotheses of the citing papers changed and authors were stating the retracted status as a statement of fact and the subsequent consequences. For the second retracted paper ("unintentional error"), authors either ignored the retracted status, were unaware of the retraction, or only cited the portion that had not been disputed despite the full retraction of the paper. Despite the intention of withdrawing a paper and marking it no longer acceptable to the scientific community, the research shows that in addition to continued citation to a retracted paper as a statement of fact, authors have either found "valid" reasons for citing a retracted paper or a work-around to selecting portions of a retraction paper to reference. By uncovering the nature of this citing behavior, this research intended to illuminate how retracted articles continue to influence ongoing research post-retraction that can provide a framework upon which this issue can be addressed.
Metrics
51 File views/ downloads
97 Record Views
Details
- Title
- Citation behavior patterns and sentiments towards retracted articles
- Creators
- Kathleen Joan Padova
- Contributors
- Erjia Yan (Advisor)
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Philosophy (Ph.D.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Number of pages
- x, 110 pages
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- Computer Science (Computing) (2013-2026); College of Computing and Informatics (2013-2026); Drexel University
- Other Identifier
- 991015684143804721