Paper Integrity

We catch what peer review misses. 🔬 AI-powered research verification for journalists, researchers & editors. Try it free →

[Paper Integrity @paperintegrity]

Image manipulation in biomedical papers is reaching epidemic levels.

Western blots spliced. Microscopy images duplicated. Graphs with mathematically impossible error bars.

The tools to detect this exist. The will to use them systematically does not.

[Paper Integrity @paperintegrity]

The reproducibility crisis isn't just about methods. It's about incentives.

No one gets a Nature paper for replicating someone else's work. No one gets tenure for confirming existing findings.

The system rewards novelty. Reliability is an afterthought.

[Paper Integrity @paperintegrity]

A retracted study doesn't just disappear. It lives on in citations, textbooks, and clinical guidelines.

The average retracted paper gets cited 30+ times AFTER retraction.

We built systems to publish science. We forgot to build systems to un-publish it.

[Paper Integrity @paperintegrity]

Fun fact: there's no global database of research misconduct findings.

A researcher found guilty of fraud in one country can move to another and start fresh.

We track criminal records across borders. We don't track scientific fraud.

[Paper Integrity @paperintegrity]

Citation cartels: groups of researchers who systematically cite each other to inflate their metrics.

It's gaming the system, but technically not fraud. So it goes unchecked.

When we measure researchers by numbers, we get researchers who optimize for numbers.

[Paper Integrity @paperintegrity]

When a clinical trial is found to be fraudulent, what happens to the patients who were enrolled based on earlier fraudulent results?

Usually: nothing. No notification. No follow-up.

The ethics of research fraud extend far beyond the lab.

[Paper Integrity @paperintegrity]

If your doctor prescribed a drug based on a study later retracted for fabrication — would anyone tell you?

Probably not.

Retracted research still shapes clinical guidelines long after the paper is pulled.

[Paper Integrity @paperintegrity]

“It’s just one bad apple.”

That one bad apple published 200+ papers, trained 30 PhD students, and shaped a subfield for a decade.

Research fraud isn’t isolated. It’s networked.

[Paper Integrity @paperintegrity]

The retraction count crossed 50,000 last year.

But for every retracted paper, ~10 problematic ones remain uncorrected.

Retraction isn’t the finish line. It’s barely the starting line of accountability.

[Paper Integrity @paperintegrity]

A university investigated itself for misconduct and found nothing wrong.

An outside analysis found 40+ manipulated images across 12 papers.

You can’t be the judge in your own trial.

[Paper Integrity @paperintegrity]

Most people think peer review catches fraud.

It doesn't. It was never designed to.

Peer review checks methodology — not whether the data is real. That’s why fabricated papers sail through top journals.

[Paper Integrity @paperintegrity]

Your paper's credibility depends on details reviewers don't have time to check manually: do the degrees of freedom match the model? Are the confidence intervals plausible? Let software handle the math.

[Paper Integrity @paperintegrity]

10,000+ papers retracted annually isn't a success story for peer review — it's evidence that peer review alone can't scale to modern publication volumes. We need layered defenses.

[Paper Integrity @paperintegrity]

The academic job market pressures researchers to publish constantly. That pressure creates shortcuts. We can't fix the job market overnight, but we can make shortcuts harder to hide.

[Paper Integrity @paperintegrity]

The academic job market pressures researchers to publish constantly. That pressure creates shortcuts. We can't fix the job market overnight, but we can make shortcuts harder to hide.

[Paper Integrity @paperintegrity]

Sample sizes matter. When a study claims n=1000 but the statistics suggest n=200, that's a red flag screaming for attention. Automated checks catch this instantly.

[Paper Integrity @paperintegrity]

Why do banks verify every transaction but journals don't verify every statistic? Fraud detection is normal in finance. It should be normal in research too.

[Paper Integrity @paperintegrity]

5 red flags in a research paper:

1. Results too clean — real data is messy
2. No raw data available
3. Author has multiple retractions
4. Stats don't match methods
5. Images appear in multiple papers

None prove fraud. All warrant a closer look.

[Paper Integrity @paperintegrity]

The image duplication problem in biomedical research is staggering. AI can now detect manipulated Western blots with 90%+ accuracy — but most journals still rely on manual review.

The tooling exists. The adoption doesn't.

[Paper Integrity @paperintegrity]

"But it was published in a top journal" is not the defense people think it is.

Nature, Science, and The Lancet have all retracted high-profile papers in the last 2 years. Prestige is not a proxy for integrity.