An index to quantify scientific debt

Imagine the world of science as a giant, bustling library where researchers are constantly writing new books. For decades, the most famous way to judge a writer's success has been the "h-index." Think of this like a "popularity contest" score: if a scientist has an h-index of 20, it means they have written 20 books that have each been borrowed and talked about by at least 20 other people. The higher the score, the more influential the writer is considered to be, and the more likely they are to get funding or a job. It's a forward-looking badge of honor, assuming that a high score means the writer will keep producing great work.

But what happens when one of those "great" books turns out to be a lie? What if the data was faked, or the results were just a mistake? In the real world, when a book is proven false, the library pulls it off the shelf and puts a big "RETRACTED" stamp on it. However, the old popularity score (the h-index) doesn't change; the bad book still counts toward the writer's fame. This creates a problem: a scientist could have a high score built partly on books that are now known to be nonsense. This paper introduces a new idea to fix that imbalance. It suggests we need a way to measure not just the good books, but also the "scientific debt" caused by the bad ones. It's like realizing that a bank account balance looks great only if you ignore the massive overdraft fees piling up in the background.

The New Score: The "z-index"

The author of this paper, writing from a future date in 2026, proposes a new metric called the z-index. If the h-index measures your scientific "achievement," the z-index measures your "debt."

Here is how it works: Every time a paper is retracted, it adds to a scientist's z-index. But it's not just a simple "one strike" system. The paper argues that a retraction in a super-famous, high-impact journal is a much bigger deal than a retraction in a smaller journal. So, the z-index adds up the "Impact Factor" (a score of how famous and influential a journal is) of every journal where a retracted paper was published.

Think of it like a game of musical chairs where the chairs have different point values. If you knock over a small, wobbly chair, you lose a few points. But if you knock over a giant, golden throne, you lose a massive amount. The z-index is the total weight of all the chairs you've knocked over.

The "Debt" vs. "Achievement" Balance

The paper suggests that to get a true picture of a scientist's value, we shouldn't just look at their h-index. Instead, we should look at the difference: h minus z.

• Positive Result: If your "good book" score (h) is higher than your "bad book" debt (z), you are still a net positive contributor.
• Negative Result: If your debt is higher than your achievements, your net value is negative. This suggests that, on balance, your work might be doing more harm than good to the scientific community.

The author uses a real-world example (from the perspective of the paper's 2026 date) involving a scientist named Ranga Dias. In this scenario, Dias has an h-index of 19, which sounds impressive. However, he has five retracted papers from very famous journals like Nature and Physical Review Letters. When you add up the "fame scores" of those journals, his z-index jumps to about 124. His final score (19 minus 124) is a massive -105. This negative number acts as a warning flag, suggesting that the misinformation he spread outweighs his valid contributions.

Time Matters: The "Delayed Penalty"

The paper also realizes that when a mistake is fixed matters. If a scientist realizes they made an honest mistake and fixes it quickly, the damage is small. But if they drag their feet for years, the bad information spreads further, and the "debt" grows.

To account for this, the author creates a "modified z-index" (called z-tilde). This version uses a special math function that acts like a penalty timer:

• If a paper is retracted quickly (within a year), the penalty is small.
• If the retraction takes a long time (like two years or more), the penalty doubles or even quadruples.

Using this "time-weighted" system, the author recalculates the scores. For Ranga Dias, the debt grows even larger (to about 165) because his retractions took many months to process. His final score drops to -146.

Interestingly, the paper also looks at two Nobel Prize winners, Gregg Semenza and Thomas Südhof. Even though they have huge h-indices (223 and 238), their accumulated debt from retracted papers is so high that their net scores change significantly. For Semenza, the h minus z-tilde value becomes a massive -227. However, Südhof's score remains positive at 176, meaning that despite his retractions, his valid achievements still outweigh his scientific debt. This suggests that even the most celebrated scientists can have a "net negative" impact if they allow too much misinformation to linger, though in Südhof's case, his legacy remains net positive.

The Library's Own Scorecard

The paper doesn't stop at individual scientists; it applies the same logic to the journals themselves. Just as scientists have scores, journals have them too. The author calculates the z-index for the top 20 journals in the world.

The results are startling. The two most famous journals, Nature and Science, have massive h-indices (1495 and 1433), showing they publish many influential papers. However, they also have enormous z-indices (3650 and 2809). When the "time penalty" is applied, their debt balloons to 8290 and 6350.

The paper points out that these journals are sitting on a mountain of "scientific debris." While they are famous for their achievements, they are also responsible for a huge amount of misinformation that took a long time to clean up. In contrast, the New England Journal of Medicine (NEJM) has a much lower debt score relative to its fame. The author suggests this might be because NEJM has stricter rules about sharing data, which helps catch errors faster.

Why This Matters

The paper concludes that science needs a way to measure "negative impact" just as well as it measures "positive impact." Currently, the system only counts the wins. By introducing the z-index, the author hopes to encourage scientists and journals to be more careful.

The main takeaway is that if journals could retract bad papers faster (ideally within a year), the "debt" scores would drop significantly, and most scientists and journals would have a positive net value again. The paper doesn't claim to solve the problem of scientific misconduct, but it offers a new tool—a "debt meter"—to make the problem visible and to encourage the scientific community to clean up its act. It's a call to stop ignoring the bad books in the library and start counting the cost of the lies.