Judge researchers by how sound and honest their work is, not by how many papers they publish or how famous the journals are
Steps to make it happen
- Universities and the groups that fund research review their rules for judging researchers.
- They stop giving scores based only on the number of papers or how famous the journal is.
- Researchers pick just a few studies they are truly confident in and submit those.
- Reviewers actually read those studies and talk about how sound and how important they are.
- Studies that redo other people's experiments to check them, and reports of results that did not work out, are also praised.
- People who share their data and methods with everyone, and people who train younger researchers, are also rated highly.
- Each institution makes public the yardstick it used, so anyone can see it.
- Careful and honest research becomes the normal way to be rewarded, instead of rushing to pile up numbers.
Water so far 💧 0
The proposal in detail
This proposal covers every situation where researchers are evaluated, such as hiring, promotion, and the review of research funding. The ones who carry it out are universities, research institutes, the organizations that hand out public research money, and the academic societies that help set the standards for evaluation. The basic rule is that numbers such as how many papers a person has, which journals they appeared in, and how many times they were cited are not the main basis for evaluation. The numbers are not banned. They are kept as one piece of reference information and are not built mechanically into a score. Instead, researchers narrow their work down to a few representative results, and for each one they explain in a short text what it showed and how they checked it. The people doing the evaluation actually read those results and look at whether the claims are supported by evidence, and whether the methods and data are open so that others can check them. The work that counts includes not only papers but also studies that try to reproduce other research, reports of results that did not turn out as expected, published data and software, training of younger researchers, and efforts to explain science to the public. Institutions make public their own evaluation standards and the results of using them. Because the right way to evaluate differs from field to field, the specific items and their weights are decided field by field with researchers themselves taking part, and they are reviewed regularly.
BenefitsLess pressure to cheat, and more solid knowledge builds up
If there is less need to compete on numbers, the very motive to pad a paper count or tidy up results to look better becomes smaller. The big advantage is that this does not only punish misconduct after the fact. It builds, from the ground up, an environment where misconduct is less likely to happen. Researchers can spend time on a single question, and it becomes easier to take on basic research that takes a long time to produce results. When the work of rechecking other people's research and reports of results that did not work out are rewarded, the waste of repeating the same failures in different places around the world shrinks, and the foundation of science as a whole gets stronger. If sharing data and methods is tied to evaluation, research also becomes more transparent, and society finds it easier to trust what science produces. Contributions that used to show up poorly in numbers, such as training young researchers and organizing data, are now recognized, so researchers who work in many different ways are rewarded. Researchers in regions that cannot pay the cost of publishing in famous journals can also compete on the substance of their work, and it is hoped that this will widen the group of people who carry out research around the world.
DrawbacksJudgments may become subjective, with more favoritism and more work
What people on the opposing side worry about most is that evaluation will become subjective and fairness will be lost. One reason evaluation by numbers spread in the first place was to prevent hiring based on connections and personal relationships, and deference to powerful researchers, and to apply the same standard to everyone. If the numbers are removed and the decision is left to the judgment of the people who read the work, the reviewers' tastes and assumptions, and unconscious bias about a person's university, gender, or nationality, can easily creep in. In societies where connections are concentrated among powerful researchers and institutions in particular, closed-off evaluation could even grow stronger. For young researchers at unknown institutions, the number of papers was sometimes the objective proof of their ability. Also, reading each result carefully takes much more time and effort from reviewers. That adds more burden for researchers who are already worn out from reviewing papers. There is also a concern that if there is no common yardstick for comparing researchers across countries and fields, it will become harder for talent to move internationally and for people to choose partners for joint research.
Learn the background
In research settings around the world, publishing many papers and getting into famous journals is the shortcut to getting a job and winning research funding. Numbers are easy to understand and easy to compare, so they have been widely used as a yardstick for evaluation. But as a result, some people split one study into small pieces to raise their paper count, or rush out only the eye-catching results. People have pointed out again and again, in many fields, that a lot of studies cannot be reproduced when other researchers run the same experiment. Research misconduct, such as falsified data and companies that make papers for others and sell them, is also not unrelated to this pressure to compete on numbers. Basic research that takes years to produce results, and reports of failed results, also tend to lose out in the numbers. On top of that, many of the highly rated journals lean toward particular languages and regions, and the cost of getting published in them is high, so researchers in regions with less money are put at a disadvantage from the start. Now that technology that writes text automatically has spread and it has become easy to produce huge numbers of papers that only look polished, the weakness of judging by numbers has become even more obvious.
Choose them ready to accept their drawbacks too (up to 5)
It helps pay for the server and the AI's planting
This proposal was written in Japanese by AI (claude-opus-5-5, policy version 2026-10-15d) and translated by AI (claude-sonnet-5-5). Nobody has checked the facts. Read the original (日本語) How the AI writes