Errors, questionable practices, or misconduct? A bibliometric and altmetric review covering two decades of retractions in Latin America

This study investigates the temporal trends, reasons for retraction, self-retractions, time to retraction, fields of knowledge, institutions, citation patterns, and altmetric data of retracted articles in Latin America through a comprehensive cross-sectional analysis. Data were sourced from Retraction Watch database, Web of Science, Scopus, and SciELO, resulting in a dataset of 375 retracted articles and 9375 metadata entries. Over a 20-year period beginning in 2002, a significant rise was observed in both the absolute and relative number of retractions involving authors affiliated with Latin American institutions. A consistent upward trend was evident across the primary causes of retraction, including errors, questionable practices, and misconduct. Misconduct emerged as the leading cause, accounting for 44.8% of cases, followed by editorial errors (20.5%), author errors (16.5%), and questionable practices (8.8%). Citation analysis revealed that 76% of retracted articles continued to be cited, accumulating a total of 9662 citations, with articles in the Health and Biological Sciences fields receiving the highest average citations. Altmetric analysis identified 22,337 online mentions, with Health Sciences topics receiving the most attention, particularly from the general public. These findings underscore a persistent misalignment between formal retraction notices and the ongoing dissemination of retracted research within both academic and public domains, since despite retraction, these articles remain widely cited and discussed, especially on social media platforms. By focusing on data from Latin America, this study emphasizes the critical need for better alignment between retraction processes, citation practices and public engagement to mitigate the continued influence of retracted research.