Make animal researchers register their plans before starting, and always publish results, even unexpected ones
Steps to make it happen
- Researchers who use animals write their plan in a shared public register before they start an experiment.
- They decide in advance what they want to test, how many animals they will use, and how they will judge the results.
- The committee that approves experiments checks that the plan is registered before it gives permission.
- When an experiment ends, researchers publish the results within a set time, even if the results are not what they expected.
- If results cannot be published right away, researchers write the reason and the planned date in the register.
- Other researchers search the register to see whether the same experiment has already been done before they make their own plans.
- Groups that pay for research check at the next review whether earlier results were published.
- Fewer wasteful repeats of the same experiment happen, and using fewer animals becomes the normal way of working.
Water so far 💧 0
The proposal in detail
This proposal covers experiments that use animals with backbones, as well as octopuses, squid, and other animals that have been pointed out as possibly feeling pain. It includes not only basic research at universities and public institutions but also safety tests of drugs and chemicals done by companies. Before starting, researchers register the question they want to answer, the main result they will use to decide, the number of animals and the reason for that number, how animals are assigned to groups and how the people looking at results are kept from knowing which group is which, how severe the expected suffering is, and the point at which the experiment will be stopped if suffering becomes too great. Registration is recorded with a date, and if the plan is changed later, the history of changes and the reasons are kept. When the experiment ends, within a set period such as one year after it finishes, the researchers publish the results, whether they match the hypothesis or not, along with the number of animals actually used and a short summary that people who are not experts can read. Delaying publication because of a patent application or a company secret is allowed, but there is a limit on how long it can be put off, and the reason for the delay is written in the register. To keep researchers and facilities from being harassed, details such as a facility's location and personal contact information can be left out of what is made public. The register is not controlled by a single organization. Instead, the registration systems that already exist in different places are linked with a common format so that they can be searched from anywhere. The ethics committees that review experiments and the organizations that fund research are responsible for checking whether results were registered and published.
BenefitsLess needless repetition and more reliable research
If both plans and results are published, researchers can check whether the same question has already been tested. This cuts down on unnecessary repeated experiments and, as a result, keeps down the number of animals used. Because results showing no effect are shared too, the bias that makes a treatment look more effective than it is becomes smaller, and the decision about whether to move on to clinical trials in humans can be made with more solid evidence. This also reduces the burden on patients who join trials of drugs that do not work, and it reduces wasted research money. Setting the criteria for judgment in advance also makes unintended bias less likely, such as picking only the convenient results to report. Combining and analyzing many studies together becomes easier too, so more knowledge can be drawn from fewer experiments. In addition, when the content of animal experiments becomes visible to society, citizens can learn why an experiment is needed and can discuss it, which also helps support trust in research.
DrawbacksA heavy load on researchers and a risk to new discoveries
The concern likely to come most strongly from researchers is the burden of the procedures. For labs with little staff or money, and for research institutions in regions where administrative systems are not well developed, the work of registering and publishing could put real pressure on the research itself. Also, much research is exploratory, moving forward from clues that only become clear once you try, and fixing the question and the criteria in advance does not always fit. If research that did not go as planned comes to be looked down on, the seeds of unexpected discoveries could be crushed. There are also persistent worries that a registered hypothesis could become known to competitors, and that a company's development information could leak. The risk that information about a facility could lead to researchers becoming targets of extreme protests is not zero. From the viewpoint of animal protection, a criticism is also conceivable: registration and publication do not reduce the suffering in experiments themselves, and by putting the procedures in order they may instead lock in animal testing as something legitimate.
Learn the background
Animal experiments are still used in many fields to study how diseases work and to check the safety of new drugs. At the same time, it has long been pointed out that research with animals is biased: results that worked are easily turned into papers, while results that showed no effect are often never published and get buried. When results go unpublished, other researchers repeat the same experiment, and animals are used each time. Also, because mostly successful results are collected, a treatment can look more effective than it really is, and there have been repeated cases in which people believed this, moved on to clinical trials in humans, and failed. For clinical trials with people, registering the plan in advance and publishing the results is now widely required around the world, but for animal experiments it remains limited to some voluntary efforts. The principle of replacing animals with other methods, using fewer of them, and easing their suffering is accepted worldwide, but if it is not known what has already been tried, there is a limit to how much the effort to reduce numbers can achieve.
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