Fake Studies, Real Prescriptions

It's Monday, March ninth, and welcome to HigherIQ. A study published in the Proceedings of the National Academy of Sciences found that fraudulent scientific papers are now being published at a faster rate than legitimate ones. Not by lone researchers faking data to get ahead. By organized networks that operate like criminal enterprises, mass-producing fake studies, selling authorship slots, and hijacking dead journals to launder credibility. Millions of dollars flow through these operations every year. And the papers they produce end up in the same databases that doctors use to prescribe drugs, that regulators use to approve treatments, and that journalists use when they write "according to a new study." Today, we're doing a deep dive into how the system that's supposed to be the foundation of scientific truth actually works, how it's being exploited, and what you can do to tell the difference between real research and something that was manufactured for profit.

When most people hear "scientific fraud," they picture a single researcher in a lab, fudging numbers to get a paper published. And that does happen. But a team at Northwestern University, led by complex systems researcher Luis Amaral, spent years analyzing massive datasets of publications, retractions, and editorial records. What they found was something much bigger. There are organized networks, spanning multiple countries, that systematically fabricate research and sell it. Amaral called them "essentially criminal organizations, acting together to fake the process of science."

The operation works like this. At the center are paper mills, which function like factories for academic manuscripts. A paper mill produces fake studies, complete with fabricated data, invented methodologies, and conclusions that sound plausible enough to pass a surface-level review. Then there are brokers, middlemen who connect the paper mills to people who need publications on their resume. In many countries, career advancement in medicine, academia, and government requires a certain number of published papers. If you don't have time to do real research, or if your research didn't produce interesting results, you can buy your way onto a paper. First-author slots cost more. Fourth-author slots cost less. Prices run from hundreds to thousands of dollars per slot. The broker also finds compromised journals willing to publish the work, often with editors who are in on the arrangement and will push the paper through a sham peer-review process. The whole system exists because of a structural incentive problem in academia. In many countries, and increasingly in the U.S., career advancement depends on publication count. Hiring committees, tenure boards, and grant agencies use the number of papers a researcher has published as a proxy for productivity. When publishing becomes a currency rather than a byproduct of good science, a market for fake publications naturally emerges to meet the demand.

One of the most striking details in the Northwestern study is journal hijacking. When a legitimate academic journal stops publishing, its web domain sometimes lapses. Fraud networks buy the domain and start publishing under the journal's name, borrowing its reputation. The study highlighted a journal called HIV Nursing, which was formerly published by a professional nursing organization in the U.K. When the journal went defunct, someone purchased the domain and began publishing thousands of papers on topics completely unrelated to nursing. All of them were indexed in Scopus, one of the world's largest academic databases. To anyone searching that database, the papers looked like they came from a credible, established journal.

The fraud clusters in specific scientific subfields that lack robust oversight, and the networks are resilient. When a compromised journal gets caught and removed from a database, the operation simply moves to a different journal or hijacks another defunct one. The researchers found that the system is growing, not shrinking. The publication of fraudulent papers is outpacing the growth rate of legitimate science. And with generative AI now capable of producing convincing academic prose, the Northwestern team warned that the problem is about to get much worse. "If we're not prepared to deal with the fraud that's already occurring," researcher Reese Richardson said, "then we're certainly not prepared to deal with what generative AI can do to scientific literature."

A separate case, which went viral this month, shows how long this can go undetected. A Canadian medical journal affiliated with the Canadian Paediatric Society had been publishing clinical case studies for years. Doctors cited them in their own research. The cases were referenced in court proceedings. Then someone finally asked the author whether the cases were real. They weren't. Every case study was fictional, and the journal had never disclosed that to readers. For years, made-up patient stories were treated as medical evidence, shaping how doctors thought about real conditions in real children.

So why does this matter to someone who isn't a scientist? Because the downstream effects of published research touch almost every decision you make about your health, your diet, and your understanding of how the world works. When your doctor recommends a treatment, it's based on published studies. When a supplement company claims their product "is clinically proven," they're pointing to a study. When the FDA evaluates a drug, when a school district sets a screen time policy, when a news headline says "researchers found that coffee extends your life," there's a paper behind it. And right now, there's no easy way to know whether that paper came from a lab or a paper mill.

The consequences aren't hypothetical. The Canadian pediatrics case involved fictional patients whose made-up symptoms were cited as evidence in real medical research and legal proceedings. If a doctor adjusts how they treat a child based on a case study that never happened, the fraud has crossed from the academic world into the exam room. If a policy is shaped by research that was fabricated for profit, the people affected by that policy are living with the consequences of a lie. The lead researcher, Amaral, put it bluntly: "If we do not create awareness around this problem, worse and worse behavior will become normalized. At some point, it will be too late, and scientific literature will become completely poisoned."

The good news is that spotting weak or suspicious research isn't as hard as it sounds, once you know what to look for. A few questions go a long way. First, where was it published? Peer-reviewed journals with established reputations, think Nature, Science, The Lancet, the New England Journal of Medicine, PNAS, are much harder to game than obscure or newly launched journals. If a finding sounds dramatic and the journal name is unfamiliar, that's worth noting. Second, how big was the study? A study of thirty people is a data point. A study of six hundred thousand is a pattern. Small studies produce dramatic results that often shrink or disappear on replication. Third, has it been replicated? A single study is a finding. Multiple studies saying the same thing is evidence. Fourth, who funded it? A study on the health benefits of chocolate funded by a chocolate company isn't automatically wrong, but it warrants more skepticism than an independent study. Fifth, does the conclusion sound too clean? Real science is messy. If a headline says "X causes Y" and the study is small, observational, and hasn't been replicated, be cautious. And sixth, check whether the study is observational or a randomized controlled trial. Observational studies, which look at existing data and find patterns, can show that two things are associated but can't prove one causes the other. Randomized trials, where researchers actually assign people to different groups and test an intervention, are much stronger evidence. Most of the health headlines you see are based on observational studies, which means the finding is a clue, not a conclusion.

This is something we think about constantly on this show. HigherIQ covers studies regularly, and every time we do, we check where it was published, how large the sample was, whether it's been replicated, what the limitations are, and whether the conclusion matches the strength of the evidence. When we cover a mouse study, we tell you it's a mouse study. When a finding comes from an observational analysis and not a randomized trial, we say so. We do that because the difference between "this study suggests" and "science proves" is the difference between being informed and being misled. Now you know why it matters even more than you might have thought.

Fake research is no longer the work of a few bad actors. It's an industry with brokers, supply chains, and revenue streams. The same databases that doctors and regulators rely on contain papers that were manufactured for profit and never tested against reality. The system that's supposed to separate what's true from what isn't is under pressure, and the tools to fake it are only getting better. The best defense isn't to distrust science. It's to get better at reading it. Ask where it was published, how many people were in the study, and whether anyone has found the same thing twice. Those three questions will filter out more junk than most people realize.

Stay informed, stay curious, and we'll see you tomorrow.