“Why would our best researchers agree to have their analysis plan peer reviewed before they have seen the data?” That was the question a UK deputy vice-chancellor for research put to me last month while her institution debated whether to recognise Registered Reports in promotion criteria. I hear versions of it in every research strategy conversation now.
Registered Reports are a publishing format that reverses the usual sequence. An author submits a Stage 1 manuscript containing the introduction, method, proposed analyses, and a detailed power analysis before collecting data. Reviewers assess the research question and study design while they can still change something. If the plan passes, the journal grants in-principle acceptance, a commitment to publish the final results provided the protocol is followed. Stage 2 later includes the results and discussion, with any analyses that departed from the plan reported clearly as exploratory.
The point is to reduce two long-standing problems. Publication bias means null results often never see print because editors and authors prefer positive findings. HARKing, or hypothesising after the results are known, lets researchers present unexpected findings as if they had been predicted all along. Registered Reports cannot remove all flexibility, but they move the decision about quality to the point where the research question and method are at stake, rather than the direction of the result.
The format arrived at a moment when psychology was confronting the scale of those problems. In 2015, the Open Science Collaboration published a replication of 100 psychology studies and found that 36 replications produced statistically significant results, compared with 97 of the original studies. For many research leaders that number made the case for structural change, not another round of methods workshops.
Where uptake actually stands
Cortex and Perspectives on Psychological Science were among the first journals to formalise Registered Reports in 2013. The approach then spread unevenly across neuroscience, social psychology, ecology, and cancer research. The Centre for Open Science now maintains a directory of journals that offer the format, a list that has passed 300 titles. That sounds substantial until you look at how many Registered Reports those journals actually publish.
Most journals that offer the format receive few Stage 1 submissions each year. Authors worry that committing to a plan before data collection will punish exploratory work, that a failed Stage 1 review will cost months, or that a journal with lower name recognition will not impress a hiring committee. The result is a familiar mismatch: an option exists on paper, but researchers do not use it unless a funder or a supervisor makes it feel safe.
What funders are now asking for
Funders have moved more cautiously than journals, but a handful have stopped treating Registered Reports as merely an eligible publication type. Cancer Research UK’s Registered Reports funding partnership is one of the clearest examples: it supports studies through the Registered Reports route and ties funding decisions to the two-stage review process. A team receives provisional funding for a study and then works through Stage 1 peer review; in-principle acceptance unlocks the full award.
Wellcome Open Research, the open research platform for Wellcome-funded teams, has made Registered Reports an available article type so that researchers can publish the protocol and the completed study in one funded workflow. That matters because funder support reduces the risk authors otherwise carry: reviewers can ask for changes to the design before data exist, which is expensive if the grant has no room for that stage.
Public funder mandates are still far from universal. No major United States federal research agency requires Registered Reports across its entire grant portfolio, and the stronger federal push there has centred on data management and sharing requirements. UK and European funders have been more willing to pilot ring-fenced calls and to treat in-principle acceptance as a signal of research quality in grant review. The policy picture is uneven, but the direction has shifted from encouragement to selective mandate.
The numbers behind a slow revolution
Peer Community in Registered Reports, launched in 2021, has become a useful middle layer for researchers who do not want to bet on a single journal. The platform reviews Stage 1 protocols independently and recommends accepted work to a network of participating journals. That changes the arithmetic for an early-career researcher: the protocol can be reviewed once, and journal selection can happen after in-principle acceptance rather than before data collection.
Uptake data still show a small share of the published literature. Registered Reports are common enough in journals like Royal Society Open Science and Nature Human Behaviour to be visible, but they remain rare across biomedicine as a whole. Where they appear, adopters often describe the experience as slower at the start and faster at the end, because the final decision no longer turns on whether the results are exciting.
This slow build is part of a broader transparency shift. Researchers who follow Registered Reports now are operating inside the same policy environment that has produced stricter data sharing mandates and closer scrutiny of trial reporting. A protocol-first workflow is not a separate movement; it sits next to those rules and often makes compliance easier because data collection and analysis decisions are already documented.
Early-career researchers, hiring committees, and a different currency
For doctoral candidates and postdoctoral researchers, a Stage 1 in-principle acceptance can function as an insurance policy against the null-result drawer. A journal has already committed to consider the final paper, which means the research record is not held hostage by a non-significant finding. That is the argument research integrity officers make when they introduce the format to graduate schools.
Hiring committees and promotion panels have been slower to value it. Many rubrics still count journal placements and citation metrics, with little formal credit for a Registered Report protocol or a publication that reports a null result carefully. That creates a real penalty for the very researchers who are trying to reform the system. A department that adds in-principle acceptance to its promotion guidance does more to increase uptake than any number of reproducibility workshops.
The pressure is not theoretical. Postdoctoral researchers already face intense publication pressure, and adding a longer pre-data review can feel risky if the benefits are invisible to a future employer. Administrators can change that by naming Registered Reports in annual review templates and listing them in faculty activity reports. They can also ask search committees to read a Stage 1 protocol as evidence of methodological judgement.
What research offices can do this month
Most research offices already know which journals their faculty publish in. A practical first step is to identify ten journals in the institution’s most common disciplines that already offer Registered Reports and send that list to research groups in the next grant support newsletter. The Centre for Open Science directory makes the filtering quick, and publication teams usually maintain the same data locally.
Then pick one funding round where the institution has influence over the application process and add a line to the internal review criteria that asks reviewers to consider a submitted Registered Report protocol or in-principle acceptance as evidence of research quality. This does not require a new policy or a senate vote, but it changes what local reviewers notice.
Finally, give the administrators who manage research outputs a place in the conversation. They see which grant reports stall and which protocols sit unreviewed; they also know which faculty need practical help with power calculations. The universities that make Registered Reports work are the ones where someone in the research office owns the process by name, follows a protocol from Stage 1 to Stage 2, and helps the researcher reset when a study does not go as planned.
Photo by Krists Luhaers on Unsplash