Technology article retractions rise in the developing world
Authored by Metwaly Eldakar at Minia University in Egypt and Ahmed Shehata at Sultan Qaboos University in Oman, the study published in Scientometrics on 20 September analysed retractions of technology-related publications from 1998-2022 in 90 countries in the Arab world, Africa, Asia, Latin America and the Caribbean.
A retracted publication is the original article that has been withdrawn from the public domain. Although it may still be available in archives or online databases, it should not be relied upon as a source of information as it is considered unreliable.
Technology fields in this study refer to academic disciplines that involve the application of scientific knowledge to create new products, services or processes that aim to solve problems or meet human needs including engineering, computer science, information technology and other related fields.
According to the study, 7,529 authors published the retracted technology articles with affiliations to 1,300 institutes and universities from 72 countries. In addition, all 3,703 retracted articles received 34,167 citations, with an average number of citations of 9.22.
The study found that the first retracted article was published in 1998, growing to a total of 2,412 in 2017. “Divided into 12-year and 13-year periods, the retraction trend indicates that the first period, 1998–2009, included 749 retracted publications, and the second period (2010-2021), which included 2,954 retracted publications, represents a considerable increase in retracted publications in recent years.
“This increase in withdrawals is worrying because incorrect data can cause bias in new technologies,” the study’s authors note.
The study put forward several recommendations for higher educational institutions in developing countries to reduce retraction numbers by fostering a culture of integrity.
These measures include educating scholars about academic ethics through comprehensive training on ethical conduct, responsible research methodologies and the consequences of misconduct.
They also include enhancing collaborative research practices by nurturing trust and the careful selection of collaborators. The need for adequate research funding was also highlighted as well as the establishment of robust systems for evaluating research, prioritising quality and impact over quantity and implementing transparent peer reviews.
The study also called for the establishment or strengthening of research ethics committees comprised of experts from various disciplines who review research proposals and protocols to ensure adherence to ethical guidelines.
Collaboration with international partners, especially those with strong research integrity practices, would facilitate valuable knowledge exchange and mentorship opportunities, it said.
Publication of research in reputable and peer-reviewed journals that have rigorous review processes in place to uphold research integrity was also important, the study noted.
Reasons for retraction
Using the Retraction Watch Database, the study found that most reasons for retractions were policy breaches by the author at 44.3% and journal-publisher investigations which accounted for 40.9% of retractions.
Other reasons included fake peer reviews, duplication and plagiarism, problems in referencing and-or attributions and misconduct by the author and rogue editors.
The study also highlighted the problem of retracted articles continuing to receive citations. For example, from 1998-2009 retracted publications received 7,811 citations, while during 2010-2022 retracted articles received 26,356 citations.
The study suggested that this ongoing citation cycle may be the result of a lack of awareness about retracted articles by people who rely on unreliable sources to obtain accurate information about retractions, such as predatory journals.
Some retracted articles may be cited for historical context or lack of alternative sources, even if their content has been discredited, it notes. Authors may also cite their own retracted work for personal gain or to support their research.
The study found that high-impact factor journals (HIFJ) have a higher number of retracted articles than other journals without or with low-impact factors because high-impact publications follow a rigorous review process and the submissions are checked using plagiarism checking tools.
The study also indicated that articles published in HIFJ are often more widely read and subjected to greater scrutiny and the data presented in it is more likely to prompt replication studies, making it easier to detect fraudulent or unreliable results.
The study showed that the universities with the highest number of retracted articles were the Islamic Azad University in Iran, University of Malaya in Malaysia and Tsinghua University in China, with 280, 115 and 62 retracted publications, respectively.
The study also showed that China (14 institutes and 16.12% of retracted articles), Iran (one institute and 7.56% of retracted articles), Malaysia (one institute and 3.11% of retracted articles) and Saudi Arabia (one institute and 1% of retracted articles) accounted for the majority of institutes and percentage of retracted technology-related articles.
“The high retraction rate may be due to the amount and growth of technology research participation in these countries. Naturally, the frequency of errors may increase with growth in publication intensity,” the authors write.
The study demonstrated that from a regional perspective, Asia, Africa, South America and North America, with the participation of 31, 20, 10, and 11 countries, respectively, exhibit the highest rate of retracted technology-related articles.
In addition, China, Iran and India in Asia, Egypt, Algeria and South Africa in Africa, Brazil and Colombia in South America, and Mexico in North America produced the most retracted technology-related articles in the target regions.
The study characterised 10 clusters of collaboration networks among institutes that have published retracted technology-related articles with the largest of the networks having 10 institutes, most of them Chinese. This indicates that these institutes constitute the central producers of retracted technology-related articles.
Arresting the damage
Samir Khalaf Abd-El-Aal, research professor at the Biotechnology Research Institute of the National Research Center in Cairo, Egypt, told University World News it was “extremely important that universities keep a close eye on their publications in the field of technology as the impact surrounding the retracting of this type of publications is very damaging due to the pivotal role technology plays in driving economic growth and enhancing social welfare.
“Universities must make use of the study's recommendations by adapting them to their contexts and use them as a basis for formulating strategic action plans aimed at reducing articles retraction,” Abd-El-Aal said.
UNESCO Science Prize laureate Professor Atta-ur-Rahman told University World News the report highlights a “growing menace” in the scientific publishing sector.
“Many international publishers have been badly affected by the compromise of their peer review processes,” said Atta-ur-Rahman, who is the former coordinator-general of the Standing Committee on Scientific and Technological Cooperation of the 57 member states of the Organisation of Islamic Cooperation and former federal minister of science and technology in Pakistan.
“The penetration into the peer review system by external criminal elements has led to large-scale retractions of articles among many publishers.
“This requires a number of actions to be taken by international publishers and the integrity of the peer review process must be maintained at all costs,” Atta-ur-Rahman said.
“Therefore, publishers must put forward strict measures for the selection of high-quality and relevant reviewers taking into account their h-index and the relevance of his or her expertise,” he said.
He also said the reviewer should be based in a different country to the author.
“As AI tools are becoming increasingly available to scientists, publishers must introduce AI detection methods to ensure that research papers have not been ‘cooked’ by intelligent software,” Atta-ur-Rahman said.
“A list of authors whose papers have been retracted needs to be maintained by Retraction Watch and extra care taken when future submissions from them are evaluated.
“If the retraction is the result of detection of fraud, then the institutions where such authors are employed should take punitive action against them,” Atta-ur-Rahman said.