False Research Results—A Global Problem That Includes the Arab World

Science has the power to improve health, strengthen economies and shed light on the unknown throughout the universe. Most scientists pursue research honestly and with noble aims. But a small and growing number of research papers are being retracted by journals for a myriad of reasons, including falsified evidence, conflicts of interest and plagiarism, specialists in science fraud say.

Research from China is one of the biggest offenders. A new data analysis by Al-Fanar Media, however, shows the problem is also widespread in the Arab region. Egypt and Saudi Arabia, two of the Arab countries with the highest numbers of published scientific articles, have some of the region’s highest rates of retraction. Egypt’s retraction rate is more than twice that of the United States and almost three times that of the United Kingdom.

“Bad science has an impact,” says Ivan Oransky, a journalist and co-founder of Retraction Watch, a blog that has also created a public database that catalogues retracted research papers. “It has an effect in terms of people in clinical trials with fraudulent results, but it also leads to mistrust of science more generally.”

[Enjoying this article? Subscribe to our free newsletter.]

Retractions, and more importantly, knowing about them is a crucial part of the advancement of science, argues Oransky. Unless researchers or journal editors publicly withdraw or correct false results, inaccurate scientific conclusions can contaminate the wider body of quality research, he says.

Research Misconduct Can Lead to Retractions

Journal editors or authors can formally withdraw an article and its findings for a number of reasons, some of which include misconduct in research. The researcher may have faked data, parts of the article could have been plagiarized or other circumstances could have called the results into question.

“The reason I care about retractions is because while it may be an imperfect indicator, it’s one way to tell the correction of science actually happens,” says Oransky.

Using the Retraction Watch database, Al-Fanar Media was able to determine the number of retracted research papers for each country of the Arab League since 1996. Al-Fanar Media then compared the number of retractions to the number of papers published in each country in the same time period using the Scimago Journal & Country Rank database. This made it possible to calculate the percentage of a country’s research that has been retracted from 1996 to 2018.

Globally, the proportion of retracted papers is currently about 0.08 percent per year, estimates Oransky. That proportion has been growing in recent years.

The results from Al-Fanar Media’s analysis show that many countries in the Middle East and North Africa region have comparatively low rates of retractions. But others, including some countries that have a relatively high output of scientific publications, such as Egypt and Saudi Arabia, have comparatively high rates of retraction, although not nearly as high as China’s. To a certain degree, the retractions show that science works and that false results are being caught and reported, rather than covered up.

The Comoros, Djibouti, Mauritania, Somalia and Yemen had no retracted research papers at all during the time period examined. But this is probably because of an absence of research: Mauritania, for example, has only recently established its first medical school.

“Countries with no retractions means either no one is doing research, or no one is watching,” says Oranksy.

Indeed, the no-retraction countries in the Arab world have very low research output. The Comoros for example, produced just 157 research papers from 1996 to 2018—had it received just a single retraction, the archipelago nation would have had the highest retraction rate in the region.

Bahrain’s High Rate of Retractions

Bahrain has the Arab world’s highest rate of research retractions—0.1201 percent of papers published by researchers in the Gulf state between 1996 and 2018 ended up being retracted.

“It looks like Bahrain is really bad, but it’s not the full picture,” says Oransky.

The negative statistic may be the result of articles written by just a few sloppy or unethical researchers. Bahrain produces a relatively small amount of research; approximately 6,600 articles in the time period that Al-Fanar Media investigated. Bahrain only has eight retracted research papers, but that was enough to make it the region’s worst offender. Libya has a very similar story. Egypt, by comparison, produces the most research in the region and churned out 204,000 papers in the same timeframe.

The results are arguably more meaningful for the countries with more substantial research cultures, such as Algeria, Egypt, Morocco, Lebanon, Saudi Arabia and Tunisia. Algeria and Morocco’s rates are roughly similar to those of the United Kingdom and United States. But Egypt, Lebanon, Saudi Arabia and Tunisia have retraction rates that are more than double that of the United States.

Both inside and outside of the Arab region, the retraction of a scientific article does not end its negative effects. Researchers in the future may not cite retracted articles or use their findings to design experiments, but by the time a paper has been retracted, its results may have already been incorporated into a meta-analysis—a study which pools the experimental results of multiple papers.

That distortion of research is something that Purnima Madhivanan, associate professor in the Department of Health Promotion Sciences at the University of Arizona, wants to end. In an article published this month in JAMA Internal Medicine, she and a colleague suggested that meta-analyses be retrospectively purged of results from retracted research.

They recommend that meta-studies should be given a defined life cycle, meaning authors would have to periodically check and make sure the research they included in the analysis had not been retracted or corrected since the meta-analysis was published. If changes had occurred, then a reanalysis would take place.

“In our own analysis, we found that the pooled estimates changed in a third of the analyses when the falsified data papers were removed,” she says.

Improving Scientific Process

One limitation to Retraction Watch’s research: It only monitors retractions from English-language journals. But Oransky is open to including articles published in Arabic if librarians or scientists in the region want to get involved. “If someone wants to help us with it and let us know when a paper has been retracted, we’ll enter ones in Arabic. We’d be delighted with that.”

Madhivanan says the solution to papers that need to be retracted isn’t necessarily about figuring out which countries or regions are worse offenders than others. She argues it’s more important to improve the way in which retractions are reported and publicized. “It’s a matter of being transparent and holding your colleagues to a standard,” she says. For example, sometimes scientific journals can be vague in explaining why an article has been retracted.

While she praises Retraction Watch, she argues that it isn’t right to expect journalists like Oransky to take on the task of logging retractions alone and she is concerned Retraction Watch can’t keep up: “It is not adequate for the rate and speed at which papers are being published. I worry all the time.” Scientists should be more involved with the process and take more responsibility, she says.

Oransky agrees that the current systems the scientific world has in place for checking the accuracy of publications are inadequate.

“I don’t think science is broken, I think the scientific correction process needs to get better,” says Oransky.