Prevalence and Trends in Global Retractions Explored Through a Topic Lens

TL;DR

The study tackles the problem of rising global retractions by analyzing them through a topic-lens framework using a dataset of over $55{,}000$ retractions from Amend linked to a Web of Science topic hierarchy. It defines and computes metrics such as retraction rate $r$, compound annual growth rate $\beta$, and the relative retraction rate $R^3$ to reveal cross-disciplinary and multi-scale patterns, showing that retractions grow much faster than publications and are concentrated in specific topics. Key findings show that high-growth topics (e.g., within Electrical Engineering & Computer Science and Human-Computer Interaction) exhibit disproportionately high retraction severity, with some micro-topics surpassing global averages by large margins (e.g., $R^3$ values well above 1 and notable examples like $R^3=74.32$ for mlncRNA). The work provides a topic-centered framework and actionable indicators for governance policies to improve research integrity across disciplines, highlighting the need for enhanced post-publication review and field-specific scrutiny to curb fraudulent and low-quality research.

Abstract

Scientific publications form the cornerstone of innovation and have maintained a stable growth trend over the years. However, in recent years, there has been a significant surge in retractions, driven largely by the proliferation of low-quality and fraudulent papers. This study aims to examine retractions and their evolving trends through a topic lens. Our analysis of global retraction data reveals that the numbers of retraction have remained alarmingly high in recent years, with the growth rate of retracted papers significantly outpacing that of overall global publications. While retractions are observed across various fields, their distribution is not uniform. In disciplines characterized by high retraction rates, certain topics may only encounter minor issues, whereas in fields with lower retraction rates, some topics can experience substantial challenges. Moreover, an unexpected surge in publications has been observed in specific topics that also display abnormally high retraction rates. This study underscores several indicators that can assist the scientific community in pinpointing key fields that require rigorous scrutiny for potential low-quality and fraudulent research. Ultimately, our findings could serve as a benchmark for examining scientific integrity across diverse topics and offer crucial insights for developing tailored governance policies to enhance research integrity in each field.

Figures (11)

• Figure 1: Annual Count of Retracted Papers and Retraction Notices. * Data as of August 31, 2025.
• Figure 2: Retraction Rate by Discipline. This chart illustrates the retraction rates (retracted papers per 10,000 publications) across various disciplines, along with the total number of retracted papers for each. The dashed line represents the average retraction rate in the Web of Science.
• Figure 3: Annual Publications, Retracted Papers, and Retraction Rates by Discipline. This displays trends in publications, retractions, and retraction rates in macro topics within the Web of Science (WoS) by year of publication. All retraction rates declined after 2022, likely due to time lag in post-publication review, meaning that not all retractions may have occurred yet. The dashed lines represent the average retraction rates for each discipline.
• Figure 4: Science Map of Retractions. Each circle represents a meso-topic, sized by the number of retracted papers it contains, with the distance between circles indicating topic similarity based on mutual citations. The top 10 topics are highlighted.
• Figure 5: Characterizing Retraction Severity Across Meso-Topics. (A) Percentage of retractions relative to publications across various meso topics, with the dashed line indicating the diagonal. Notable topics are highlighted where the proportion of retractions significantly exceed that of publications. (B) Retraction index $R^3$ for various meso topics across disciplines, emphasizing those with the highest indices. The dashed line represents overall retraction severity. An index greater than 1 indicates a serious retraction issue, while an index below 1 suggests a milder problem.